Copyright is owned by the Author of the thesis.  Permission is given for 
a copy to be downloaded by an individual for the purpose of research and 
private study only.  The thesis may not be reproduced elsewhere without 
the permission of the Author. 
 
SOCIAL IDENTITY THEORY AND INTERGROUP 
RELATIONS IN GENDER DOMINATED OCCUPATIONS 
A thesis presented in partial fulfilment of the requirements for the degree of 
Doctor of Philosophy 
in Psychology 
at Massey University, Albany, New Zealand 
Christina Margaret Baird 
200 1 
1 
ABSTRACT 
Previous research has found that men and women have quite different 
experiences of working in opposite gender dominated occupations. The effect of 
societal status on the processes that occur in gender dominated occupations often results 
in negative outcomes for women and positive outcomes for men. The study aimed to 
explore the attitudes and beliefs of individuals working in gender dominated 
occupations. It focussed specifically on how people who work in gender dominated 
occupations react to their group's position as a numerical majority or minority and the 
groups' attitudes towards their situation. Social Identity Theory (SIT, Tajfel & Turner, 
1979) was adopted as a guiding framework for the research. 
The sample for the study consisted of male and female volunteers working in 
gender dominated occupations. 110 male and female nurses and 97 men and women 
working in two male dominated occupations (engineers and prison officers) 
participated. To achieve the aims of the study specific areas of SIT were measured. 
The areas included identification with the gender group, how prominent gender and 
occupation were in the self�oncept, and whether gender was used as a basis for 
categorisations of others. Also measured were perceptions of the status of the groups 
and beliefs about how fair and open to change the intergroup situation was. In addition, 
perceived acceptance of the ingroup and acceptance towards the outgroup, beliefs about 
changing the groups' position and the support or rejection of outgroups' beliefs were 
measured. The following specific measures were used: the Spontaneous Self�oncept, 
the Gender Salience Scale, and llinkle, Taylor, Fox-Cardamone and Crook's (1989) 
measure of group identification. SIT makes specific predictions about how beliefs 
about the intergroup situation and identification with the group are related to social 
change beliefs. 
Results showed that gender affected choice of social change beliefs for 
achieving positive distinctiveness, with men being higher on social mobility beliefs, and 
women higher in social creativity and social competition beliefs. Engineers were higher 
in social mobility beliefs and social creativity beliefs than nurses. Status also affected 
social change beliefs with low status groups being more likely to choose social 
competition strategies than high status groups. 
Women showed less support for social competition and social creativity beliefs 
in the outgroup than men. Women showed more support for social mobility beliefs in 
the outgroup than men. Nurses showed less support for social mobility beliefs in the 
out group than engineers. Nurses had lower social competition (maintaining status) 
beliefs than did engineers. Status also affected support of the outgroup's social change 
beliefs. The low status group showed higher support of social mobility beliefs in the 
outgroup than the high status group, and higher social competition beliefs than the equal 
and high status groups. The equal group showed less support for outgroup social 
competition than did the high status group. 
The results of regression analysis showed that gender was the best predictor of 
ingroup social mobility beliefs and ingroup social competition beliefs. Gender also was 
the best predictor of attitudes towards outgroup social mobility beliefs and ingroup 
social competition (to maintain status) beliefs. Gender identification was the best 
predictor of ingroup social creativity beliefs, and support for social creativity beliefs in 
the outgroup. Legitimacy beliefs were the best predictor of support or rejection of the 
outgroup's social competition beliefs. 
The results of this study highlight the importance of using an approach that 
explores the different variables that predict each social change belief rather than 
focussing on the relationship between identification and differentiation as previous 
studies have done. 
ACKNOWLEDGEMENTS 
I would like to take this opportunity to acknowledge the people that provided 
assistance and support throughout my thesis. I would l ike to express my appreciation 
for the advice and patient supervision provided by Drs Dave Clarke, Hillary Bennett and 
Robbie Sutton. 
I would also like to thank the organisations involved for allowing me to conduct 
research on their staff or members, and to all the individuals who took the time from 
their busy schedules to fill in the long questionnaire and share their experiences of 
gender dominated occupations that formed the basis of this research. 
Permission to conduct this research was granted from the Massey University 
Human Ethics Committee. 
Finally, I wish to thank Mike for his financial support and his patience 
throughout my thesis years. 
TABLE OF CONTENTS 
Abstract ..•••••.•••..•••••.•••....•••.•••••..•••••.•••..••..••••...•••••••••....•••••••.•...•.••••...•••...•..• 11 
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  iv 
Table of Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  v 
List of Tables 
Chapter 1 :  
1 . 1 
1 .2 
Chapter 2 :  
2 . 1  
2.2 
2.3 
2.4 
2.5 
Chapter 3 
3 . 1 
3 .2 
3 .3 
3 .4 
3 .5 
3.6 
3 .7  
3 .8  
....................................................................................................... Vlll 
Introduction 
Background ... . ...... .... ......... .......... ... . . .......... . ......... .... ....... .... ........ . . .  1 
Overview of the Current Study ... ..................... ... ..... ... ................... 5 
Gender Dominated Occupations and Tokenism 
Introduction ................ ... .......... .. . . . ...... . ............ ........... ..... ..... .......... 7 
Tokenism .. .. . . ........ ... . . ...... . .... ........... ...... .............. .......... ... . ........... . 7 
Male Dominated Occupations-
Engineers and Prison Officers . . ....... ............ . . . ....... . . . .... .... ... . . ....... . 1 9  
Female Dominated Occupations- Nurses ................... ... . . . . . ........ . . .. 25 
Chapter Summary .. . . ....... ... ..... .... ........................ ......... . . .. ...... .... .... 28 
Social Identity Theory 
Overview of SIT . . .... ... . ....... . . .. . ......... ............ . .. ........ . . ....... .. .... ....... 29 
S IT and Gender Relations .. ....... ... ...... . ... . ............. ....... .. .. .. . ..... .. . . . .. 32 
Social Categorisations . ..... . . . . .... . . ...... . ...... .... . . . . ......... . . ... . . . ...... .... ... .  35  
Social Identity . . . . ...... . .. ....... . . . ........... . ...... . .. . ..... ... . ....... .... ... . ...... . .... 38 
Salience of Identities . . ... . .. . ....... . ........ . ... ...... . ... ....... ...... . ....... . ... . . . .. 47 
Psychological Distinctiveness . . . .... .... . . .......... . . . . ....... . . . ......... ... .. . . ... 57 
Social Comparison Processes ....... . ...... . ............. .. . .... ..... . .. .. . ......... 59 
Beliefs about the Intergroup Situation ......... . . ... . .... . .... ... ... ...... .... . .. 6 1  
v 
3.9 
3 . 1 0  
3 . 1 1 
Chapter 4 
4. 1 
4.2 
4.3 
4.4 
4.5 
4.6 
Chapter 5 
5 . 1  
5 .2 
5 .3 
5 .4 
Chapter 6 
6. 1 
6.2 
6.3 
6.4 
6.5 
VI 
Social Change Belief Systems 
- Achieving Positive Distinctiveness ................. ... , '" . . . . . . . . . . . . . . . . . . . .  72 
Social Change Belief Systems 
-Maintaining Positive Distinctiveness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  80 
Chapter Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  86 
Research Themes and Hypotheses 
Introduction .................................................................................... 88 
Gender Identity ......... ... . . . ................. . ........... .................................. 88 
Identity Salience ............................................................................. 88 
Beliefs about the Intergroup Situation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  90 
Social Change Belief Systems 
- Achieving Positive Distinctiveness ..................................... ....... 9 1  
Social Change Belief Systems 
-Maintaining Positive Distinctiveness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  93 
Method 
Overview ........................................................................................ 96 
Participants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  97 
Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  99 
Design and Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  1 22 
Results 
Gender Identity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  1 25 
Identity Salience . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  1 3 1  
Beliefs about the Intergroup Situation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  1 39 
Social Change Belief Systems 
- Achieving Positive Distinctiveness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  1 46 
Social Change Belief Systems 
-Maintaining Positive Distinctiveness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  1 66 
Chapter 7 
7. 1 
7.2 
7.3 
7.4 
7.5 
7.6 
References 
Appendix A 
Appendix B 
Appendix C 
vu 
Discussion 
Gender Identity .............................................................................. 1 86 
Identity Salience ............................................................................. 1 92 
Beliefs about the Jntergroup Situation ......................................... .. 1 97 
Social Change Belief Systems -
Achieving Positive Distinctiveness ............................................... 203 
Social Change Belief Systems -
Maintaining Positive Distinctiveness ............................................ 2 1 3  
Conclusion ............................. .... .................................................... 221  
....................................................................................................... 223 
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  239 
Questionnaire One - Female Nurses .............................................. 239 
Questionnaire Two _ Male Engineers ............................................ 262 
Social Change Belief Items ........................................................... 280 
Items used in the Main Study ......................................................... 287 
Male Jngroup Measure ................................................................... 287 
Male Outgroup Measure ................................................................ 289 
Female Jngroup Measure ............................................................... 292 
Female Outgroup Measure ............................................................. 294 
Vlll 
LIST OF TABLES 
Table 1 Predicted Relationships between Identification and Social Change 
Bel iefs ............................................................... ............................ 93 
Table 2 Predicted Relationships between Acceptance and Social Change 
Beliefs ............................................................................................ 95 
Table 3 Internal Consistency Alphas of the Gender Identification Scale (Main 
Study) . ............................................................................................ 1 04 
Table 4 Factor Loadings for the Gender Identification Scale 
(Main Study) .................................................................................. 1 06 
Table 5 Internal Consistency Alphas for the Gender Salience Scale 
(Main Study) .................................................................................. 1 08 
Table 6 Internal Consistency Alphas of Social Change Beliefs 
(Pi lot Study) ......... ...................... .................................................... 1 1 3 
Table 7 Ingroup Social Change Beliefs Internal Consistency Alphas 
(Main Study) . ........ ......................................................................... 1 1 5 
Table 8 Outgroup Social Change Beliefs Internal Consistency Alphas 
(Main Study) .............................................................................. .... 1 1 5 
Table 9 Factor Loading for Social Change Beliefs Male Ingroup Measure 
(main study) ................................................................................... 1 1 8 
Table 1 0  Factor Loading for Social Change Beliefs Male Outgroup Measure 
(main study) ............................................................................... .... 1 1 9 
Table 1 1  Factor Loading for Social Change Beliefs Female Ingroup Measure 
(main study) ................................................................................. . .  1 20 
Table 1 2  Factor Loading for Social Change Beliefs Female Outgroup Measure 
(main study) ................................................................................... 1 2 1  
Table 1 3  Outgroup Social Change Beliefs Internal Consistency Alphas for 
Extracted Factors (Main Study) ..................................................... 1 22 
Table 14 Ingroup Social Change Beliefs Internal Consistency Alphas for 
Extracted Factors (Main Study) ..................................................... 1 22 
IX 
Table 1 5  Mean Gender Identification Scores (for each sample) ................... 1 27 
Table 1 6  Mean Gender Identification Scores for Each Status Group ........... 1 28 
Table 1 7  Mean Gender Identification Scores for Majority and Minority 
Groups ............................................................................................ 1 29 
Table 1 8  Correlations between Gender Identification and other variables 
(total sample) ................................................................................. 1 30 
Table 1 9  Free Answers on Gender Identification ......................................... 1 3 1  
Table 20 Mean Identity Salience Scores for each Sample ............................ 1 33 
Table 2 1  Mean Identity Salience Scores for Each Status Group .................. 1 32 
Table 22 Mean Identity Salience for Minority and Majority Groups ........... 1 34 
Table 23 Correlations between Gender Salience and Occupational 
Salience .......................................................................................... 1 34 
Table 24 Paired Sample T -Test results for Gender and Occupational 
Salience .......................................................................................... 1 35 
Table 25 Gender Salience Means for Outgroup Present and Outgroup 
Absent ....................... ..................................................................... 1 35 
Table 26 Correlations between Gender Salience and Other Variables ......... 1 3  7 
Table 27 Free Answers on Gender Salience ................................................ 1 38 
Table 28 Distribution of Respondents in their Beliefs about the Intergroup 
Situation ......................................................................................... 1 40 
Table 29 Beliefs about the Intergroup Situation held by Majority and 
Minority Group .............................................................................. 1 4 1  
Table 30 Perceptions of Legitimacy and Stability for Each Status Group .... 1 42 
Table 3 1  Mean Acceptance Scores for Each Sample ................................... 1 43 
Table 32 Mean Acceptance Scores for Each Status Group .......................... 1 44 
x 
Table 33 Mean Acceptance Scores for Minority and Majority Groups ........ 1 47 
Table 34 Mean Ingroup Social Change Beliefs for Each Sample ................. 1 48 
Table 35 Mean Ingroup Social Change Beliefs for Each Status Group ........ 1 50 
Table 36 Mean Ingroup Social Change Beliefs for Minority and Majority 
Groups ......... ................................................................................... 1 5 1  
Table 37 Correlations between Gender Identification Scales and Ingroup Social 
Change Beliefs (total sample) .......................................... .............. 1 52 
Table 38 Correlations between Gender Identification Scales and Ingroup 
Social Change Beliefs (men and women) ...................................... 1 54 
Table 39 Correlations between Gender Identification Scales and Ingroup 
Social Change Beliefs for Each Status Group .............. ... .............. 1 56 
Table 40 Correlations between Gender Identification Scales and Ingroup Social 
Change Beliefs for Majority and Minority Groups ........................ 1 57 
Table 4 1  Mean Ingroup Social Change Beliefs fro Legitimacy and Stability 
Beliefs ........ . . .. .. ..... .. ............ .. .. .. ........... .. .. .......... .. .... .. .................... 1 59 
Table 42 Correlations between Acceptance and Ingroup Social Change 
Beliefs (total sample) .. ........ ..... .. ............... .............. ....................... 1 60 
Table 43 Correlations between Acceptance and Ingroup Social Change 
Beliefs (men and women) ....... ....................................................... 1 60 
Table 44 Correlations between Acceptance and Ingroup Social Change 
Beliefs for Each Status Group ....................................................... 1 62 
Table 45 Correlations between Acceptance and Ingroup Social Change 
Beliefs for Majority and Minority Groups ..................................... 1 63 
Table 46 Standardized Regression Coefficients between Ingroup Social 
Change Beliefs and Predictor Variables . ........................... . ........... 1 64 
Table 47 Strategies used to Improve the General Situation for the Ingroup . 1 65 
Table 48 Reactions of the Outgroup to Social Change Strategies ................ 1 66 
Table 49 Mean Outgroup Social Change Beliefs for each Sample ...............  1 68 
Table 50 
Table 5 1  
Table 52 
Table 53 
Table 54 
Table 55  
Table 56 
Table 57  
Table 58 
Table 59 
Table 60 
Table 6 1  
Table 62 
Table 63 
Table 64 
Mean Outgroup Social Change Beliefs for Each Status Group ..... 1 70 
Mean Outgroup Social Change Beliefs for the Minority and 
Majority Groups ...... ....................................................................... 1 73 
Correlations between Gender Identification Scales and Outgroup 
Social Change Beliefs (total sample) .... ......................................... 1 72 
Correlations between Gender Identification Scales and Outgroup 
Social Change Beliefs (men and women) ..................................... 1 73 
Correlations between Gender Identification Scales and Outgroup 
Social Change Beliefs for each status group .................................. 1 76 
Correlations between Gender Identification Scales and Outgroup 
Social Change Beliefs for Majority and Minority Groups ............. 1 77 
Mean Social Change Beliefs for Legitimacy and Stability Beliefs 
1 8 1  
Correlations between Acceptance and Outgroup Social Change 
Beliefs (total sample) ............... ...................................................... 1 78 
Correlations between Acceptance and Outgroup Social Change 
Beliefs (men and women) .............................................................. 1 80 
Correlations between Acceptance and Out group Social Change 
Beliefs for each Status Group ....................................................... 1 8 1  
Correlations between Acceptance and Outgroup Social Change 
Beliefs (majority and minority groups) .......................................... 1 82 
Standardized Regression Coefficients between Outgroup Social 
Change Beliefs and Predictor Variables ........................................ 1 84 
Social Change Strategies Used by the Outgroup . ..........................  1 85 
Best Predictors of Ingroup Social Change Beliefs ......................... 2 1 2  
Best predictors of Out group Social Change Beliefs ...................... 220 
Xl 
CHAPTER ONE 
1 :  1 Background 
Comedian Paul Reiser highlights the complex intergroup relations that exist 
between the genders: "When you're a couple, each person represents their gender. 
You're the flag bearer for the whole team. And if any member of your team, anywhere 
in the world, past or present, does something to offend, you have to answer for it" 
(Reiser, 1 994, p. 1 74). Although men and women interact and co-operate as individuals 
at work and at home with their families, at times they are aware of difference between 
the genders and identify with their own gender. Viewing the genders as two different 
and often conflictual groups is encouraged by the media, where the conflict between 
genders often forms the basis of stand up and television comedy. Popular authors such 
as Gray ( 1 993) and Pease and Pease ( 1 998) also overemphasize the difference between 
the genders as groups and ignore the differences that exist within the gender groups. 
Gender relations are not straightforward as there is a degree of competition between the 
genders as groups, but there is also a high level of interdependence between the gender 
groups and between individual members of the groups (del Boca & Ashmore, 1 986). 
A large number of men and women work predominately with the opposite 
gender. Women's involvement in the paid workforce in New Zealand has continued to 
increase since women first began entering the workforce. However, the distribution of 
women and men across occupations and industries has remained segregated (Statistics 
NZ, 1 999). Detailed analyses of gender segregation have not been published from the 
last New Zealand census in 1 996. Statistics from the 1 99 1  New Zealand Census show 
different profiles for male and female dominated occupations. Women were 
concentrated in a narrower range of industries and occupations than were men. The 
proportion of women working in female dominated occupations (defined as less than 
32% male) increased between 1 97 1  and 1 99 1 .  In addition between 1 98 1  and 1 99 1  the 
1 0  occupations employing the largest numbers of men maintained the same degree of 
numerical male domination while the 1 0  occupations employing the largest numbers of 
2 
women became less numerically female dominated (NZ Department of Statistics, 1 993). 
Although the same detailed analyses were not published from the last census in 1 996, 
the trends have continued with the statistics showing that "women work in a narrower 
range of occupations than men and occupational segregation by sex remains a feature of -
the workforce" (Statistics NZ, 1 999, p. 95). For example, over 90 percent of early 
childhood workers, nurses, and secretaries are women, and over 90% of carpenters and 
mechanics are men. The 1 0  occupations employing the largest number of women and 
the 1 0  employing the largest number of men have only two occupations in common. 
These two occupations were supply and distribution managers, and salespersons and 
demonstrators (Statistics NZ, 1 999). 
According to McClean and Kalin ( 1 994) occupational segregation by gender is 
extensive, persistent, and slow to change. Occupational segregation is particularly 
important because of the implications that it has for wages, as women tend to be over­
represented in lower paying jobs. The traditional gender pay gap remains in New 
Zealand. A recent study shows that females earn between 1 8- 1 9% less than comparable 
males ("Qualifications play" 2000). Ryan ( 1 994) identifies three consequences of 
gender based segregation of the labour market. Firstly, individuals' perceptions of the 
managerial role and the characteristics of managers can become based on masculine 
notions of hierarchy and control. It also creates occupational hierarchies, where women 
are over-represented in the jobs at the bottom. These are usually repetitive, boring, and 
low skil led jobs offering poor wages and conditions. Thirdly, occupations such as 
childcare that reproduce domestic work are undervalued and seen as unskil led. Bond 
and Kemp ( 1 99 1 )  provided support for Ryan's statements, finding that respondents in 
New Zealand perceived female dominated occupations as lower in value than male 
dominated occupations. 
Research suggests that men in female dominated occupations have different 
experiences to women in male dominated occupations. Researchers have found that 
men find it easier to be accepted into female dominated jobs and that men's minority 
position can assist in their career progression (C. Williams, 1 992). Benokraitis and 
Feagin ( 1 986) even suggest that men are valued in non-traditional jobs simply because 
3 
they are not women. Men also find that "female jobs" can be stepping-stones to jobs in 
management. Although the New Zealand Ministry of Women's Affairs monitors 
women's representation in the workforce, there is no governmental body responsible for 
monitoring males' entry into female dominated occupations. There is also less research 
available on men in female dominated occupations than on women in male dominated 
occupations. Many companies and labour organisations do not keep statistics on the 
gender demographics of their workforces. There is also little research into the gender­
based stereotypes of men, men's gender roles and men's views of gender relations. 
Researchers on the effects of tokenism in the workplace have assumed that the same 
processes operate for any under-represented group regardless of their societal status 
(e.g., Kanter, 1 977). This assumption has been criticised for not taking into 
consideration the effect of the status differences between the genders that are brought 
into the workplace from society. For example, Zimmer ( 1 988) states that women's 
treatment in the workforce is due to "notions of inferiority rather than scarcity" (p. 72). 
Cockburn ( 1 99 1 )  identifies institutions and culture as two factors that obstruct 
sex equality in organizations. Institutions involve structures, procedures, and rules that 
disadvantage women. Cultures arise in discourse and interaction, and influence what 
people feel, think, and do. The men Cockburn studied created a masculine focused 
culture around the work they did that made women feel like they did not belong. There 
is some evidence that the type of discrimination confronting minorities in the workplace 
is changing. Overtly unequal treatment of women or racial minorities is no longer 
socially acceptable and has been replaced by more subtle or covert discrimination 
(Benokraitis & Feagin, 1 986; Swim, Aiken, Hall & Hunter, 1 995). 
The change to subtle or covert discrimination represents a change from 
institutional discrimination to cultural discrimination. Legislation and methods of 
complaint have been put into place to ensure at least an outward show of equality for 
minorities in the workplace. In New Zealand, this includes the ratifying of the 
Convention for the Elimination of Discrimination against Women in 1 984, the 
formation of the Equal Employment Opportunities Trust in 1 99 1 ,  and the introduction 
of the Human Rights Act in 1 993. The move to more subtle and covert discrimination 
4 
means that minorities are most likely to experience cultural and interactional forms of 
discrimination (Swim et al., 1995). Thus, the study of attitudes toward gender relations 
is crucial to understanding discrimination in the workplace. 
Occupations that are gender dominated become gendered, or gender stereotyped, 
which means the attributes or skills needed for the job are seen as either feminine or 
masculine (Liff & Aikenhead, 1992). This creates additional pressure for gender 
minorities in these occupations, as they may be perceived as gender role deviants. 
Minority women in male dominated occupations have been referred to as double 
deviants because they differ from the dominant gender in the workplace and they do not 
conform to occupational norms (Yoder, 1994). 
The current study aims to explore the attitudes and beliefs of individuals 
working in gender-dominated occupations. Social Identity Theory (SIT) (Tajfel 1975, 
1978, Tajfel & Turner, 1979) will be used as a guiding framework for the present 
research. SIT is particularly suitable because it takes into account the reactions and 
attitudes of both the majority and minority groups in an intergroup situation. The scope 
of SIT is very broad, covering identity, categorizations, actions, and discrimination 
between groups. SIT was a departure from the previous theories on intergroup relations 
such as Realistic Conflict Theory (Sherif, 1966, cited in Tajfel & Turner, 1979). SIT 
did not focus solely on individual causes of prejudice and intergroup action, and 
claimed not to impose individual processes onto intergroup relations as previous 
theories had done. In contrast, SIT considered how the group contributed to and 
constitutes part of the individual, and how these in turn influence the groups' actions 
and beliefs. SIT was designed to consider groups' situations where the groups did not 
have equal status, and makes specific predictions of what individuals in a group of 
lower status will do in reaction to their situation. SIT also includes predictions about 
how advantaged groups react to threats to their status, and can predict when the 
advantaged group will accept the change in position of the lower status group and when 
it will not (Finchilescu & de la Rey, 1991). SIT takes more factors into account than 
just numerical proportions when assessing groups' status, and so is ideal as a framework 
for studying men and women in gender dominated occupations. SIT has been 
previously applied to the study of gender relations (Breinlinger & Kelly, 1994, J. 
Williams, 1984) and has been successfully applied to organizational settings (e.g., 
Haslam, Powell & Turner, 2000). 
1.2 Overview of the Current Study 
Large numbers of people work in a gender dominated environment. 
5 
The present study used Social Identity Theory (SIT, Tajfel & Turner, 1979), as a 
framework to explore the relations between the gender groups in three gender 
dominated occupations: nursing, engineering and prison officers. The current study 
explored the attitudes of individuals' who work in gender dominated occupations 
towards their group's position as a numerical majority or minority. The particular focus 
was on the social change beliefs that individuals hold about their group's situation. The 
current study also explored whether status in society affects the attitudes of the majority 
and minority groups. Studying minority groups' attitudes to their situation, and 
majority groups' attitudes to other groups will be useful for organisations when trying to 
increase the numbers of a minority group in a gender-dominated occupation, or when 
trying to improve relations between different groups. These are areas of SIT that 
researchers have not previously studied, therefore this study will contribute to 
knowledge about relations between natural groups in natural settings. The study also 
extends SIT by providing information about the beliefs and attitudes of groups who see 
their status as equal, and about groups' support or rejection of each others beliefs. 
To achieve these aims specific constructs of SIT were measured. These 
included identification with the gender group, how prominent gender and occupation 
were in the self-concept and whether gender was used as a basis for categorisations of 
others. Also measured were perceptions of status, beliefs about how fair, open to 
change and accepting the groups were; beliefs about changing the groups' position and 
the attitudes of the groups to the outgroups' beliefs. 
The current study begins with a chapter proving a background of the experiences 
of men and women who work in gender dominated occupations. The three specific 
occupations that were studied are introduced in this chapter. Chapter 3 introduces SIT: 
particular attention is given to the research that has been conducted on gender groups. 
Chapter 4 outlines the research themes and hypotheses of the current study. Chapter 5 
outlines the methodology that was used, and Chapter 6 presents the results of the study. 
Chapter 7 is the discussion of the results and conclusions. 
6 
CHAPTER 2: GENDER DOMINATED OCCUPATIONS 
AND TOKENISM 
2.1  Introduction 
Chapter 2 aims to give a background on the experiences of minority genders in 
gender dominated occupations. Firstly, the issues and processes of tokenism and 
discrimination will be introduced; this includes an overview of previous research. 
Lastly, the three specific gender dominated occupations sampled, engineering, nursing 
and prison officers, will be described. 
7 
'Tokenism' as a term is widely used but less often adequately defined. Kanter 
( 1 977) defines tokens as those few members (under 1 5%) of a social type in an 
environment dominated by another social type who control the environment and its 
culture. She calls these category members tokens as they are treated as symbols rather 
than individuals. Tokenism does indicate more than just numerical under­
representation, but is a situation where one group is in control. Kanter ( 1 977) also 
described tokens as being readily identifiable as group members. Gender and race are 
often readily identifiable social categories facilitating tokenism. The dominant group 
also holds certain assumptions about the attributes, behaviour and suitable roles of the 
tokens. Lone category members are not always tokens. It is possible that their presence 
in the situation is taken for granted and seen as unremarkable then they do not 
experience tokenism processes. Tokenism is also often used to refer to the situation that 
exists when minority group members are hired or promoted solely because of their 
minority group membership or to ensure that government or organizational quotas are 
met (Zimmer, 1 988). 
2:2 Tokenism 
Gender dominated occupations have been approached for study in three main 
ways. Field research has looked at the general principles and processes of tokenism. 
8 
Laboratory based research has manipulated token situations, stereotypes and attitudes. 
A third group of studies has examined the experiences of individuals in specific 
occupations, including experiences of gender based discrimination. Many of the studies 
of gender based discrimination are carried out in gender dominated occupations as it is 
often assumed to be more prevalent in these occupations. Researchers in all areas have 
approached tokenism from a wide range of perspectives and disciplines including 
women's studies, sociology, and psychology. 
Principles and Processes of Tokenism 
The first group of research studies was initiated by the work of Kanter ( 1 977). 
Previously Kanter ( 1 976) had begun examining how some of the attributes that were 
previously seen as sex differences may actually be attributed to the structural position of 
the genders in the work place. Kanter identified four types of group based on the 
relative proportions in the group. 'Uniform' groups have only one type of member. 
' Skewed' groups have a ratio of around 85% of one type of member and 1 5% of a 
different type. 'Tilted' groups are those with a ratio of around 65% to 35%. Lastly, 
'balanced' groups have from around 50-60% of one group and 40 to 50% of another. 
These terms are now widely used to describe the different proportions in token 
situations such as those in gender dominated occupations. Token dynamics are 
strongest in the skewed groups. As wel l  as being numerically uncommon, tokens are 
also usually readily identified by ascribed characteristics such as gender and ethnicity. 
Kanter ( 1 977) identified three main processes that tokens experience, namely, 
'visibility' ,  'polarization' ,  and 'assimilation' .  Visibility imposed performance pressures 
on tokens as they felt that people noticed them, and often people remembered the names 
of the tokens more easily. Visibility also made the tokens feel as if they were 
representing their category to the dominant group and that the worth of the whole 
category hung on their performance. The women in Kanter's study also felt that they 
were being evaluated on two measures: "how as women they carried out the sales role 
and how as salesworkers they lived up to images of womanhood" (p. 973, author's 
9 
emphasis). The visibility of a token took the fonn of attention to the discrepant 
characteristics of the token rather than to their achievements and abilities. For example, 
colleagues noticed physical appearance more often than credentials. The additional 
attention created perfonnance pressures for the token and although the tokens felt that 
they had to work extra hard to prove themselves and have their achievements noticed, 
they felt that they could not work so hard that they made the dominant group look 
inferior. 
Polarization is the process by which the dominant group exaggerated their 
commonality and the difference of the token. According to Kanter, "the presence of a 
token makes dominants more aware of what they have in common at the same time that 
it threatens that commonality" (p. 975). The dominants emphasized and exaggerated the 
aspects of their culture that the tokens did not share. For example, one of the 
behaviours of the dominants that emphasized the tokens' difference was to question or 
apologize for certain behaviour. The men that Kanter ( 1 977) studied would often 
apologize for swearing or ask if they could tell a 'dirty joke' .  This served to reinforce 
the dominant group's own understanding of their culture and at the same time set 
boundaries on the behaviour that was expected from the token. Tokens were also 
excluded from infonnal networks because of their difference. Often their loyalty to the 
dominant group was tested. For example, the dominant group required tokens to laugh 
at jokes at the expense of their own group or to side with the dominant group against 
their own group. 
The last process Kanter ( 1 977) identified was assimilation. The dominant group 
distorted the characteristics of the tokens, to fit with their pre-existing stereotypes or 
generalizations about the tokens'  category. There were several aspects of this process. 
As the tokens were uncommon in their role, people they came across assumed them to 
be in the role more common for their gender. For example many of the women in 
Kanter's study were assumed to be the secretary, or in infonnal settings the wife of a 
businessman. In addition, the dominant group often gave tokens assignments more 
suitable to the alternative role, such as secretarial duties. This occurred even when the 
dominant group was aware of a token's true position. Another aspect of assimilation 
1 0  
was 'stereotyped role induction' .  Members of the dominant group Kanter observed tried 
to induce the tokens to take on stereotypical roles, with which the dominant group felt 
more comfortable interacting. Some tokens were treated as a 'mother' ,  used by the 
dominant members as a l istener for their problems and as a source of nurture and 
support. Other tokens were seen as sexually desirable or available and their role was to 
play 'the seductress' .  If the token placed in this role favoured one dominant member 
over others this caused jealousy among them. The 'pet' was a token who was treated as 
a symbol or mascot. Pets were an audience for displays of the dominant culture, and 
were often treated as 'precious or precocious' when they showed any occupational 
competence. Women who resisted being inducted into any of these roles were typically 
forced into the ' iron maiden' role. These women were seen by the dominant to be tough 
and often labelled as feminists. 
Kanter ( 1 977) focused mainly on numerical proportions and assumed that these 
had the most effect on tokens' experience. She also mentioned some other variables 
that will affect the tokenism processes. For example token status was heightened if the 
tokens' deviant characteristics were physically obvious, as is gender and if the tokens' 
group was new to the setting. The opportunity and power structure of the organization 
and sex roles also affected the tokens. She did briefly mention the effect of societal 
status on tokenism, but downplayed its relevance, claiming that "if the token's master 
status is higher than that of the situational dominants, some of the content of the 
interaction may change while the dynamics remain the same" (p. 986). Kanter claimed 
that the importance of these aspects is heightened if members of the dominant group are 
accustomed to interacting with the tokens in roles different to those they occupy in their 
current situation. 
Zimmer ( 1 988) criticizes Kanter for focusing too narrowly on the effects of 
numerical under-representation, and not taking into account the effect that societal 
sexism has on the organizational processes. Kanter ( 1 977) sees increasing the 
numerical representation of the group as the solution to tokenism processes. Zimmer 
claims that social attributes are at least as important as numbers and gender-neutral 
theories of organizational behaviour such as Kanter's ( 1 977) mask real ity. 
11 
One of the societal attributes that can influence group members in token situations 
is that although organizational structures are assumed to be gender neutral they are not. 
Acker (1990) stated that an organization's idea of a worker is usually a white male with 
total commitment to the organization. Cockbum (1991) also states that organizations are 
not gender neutral in their approach to workers as they do not take into account the 
differing responsibilities that men and women often take on in a gendered society. She 
claims that "until the symbolic man-as-citizen has his mind on the cooker, his eye on a 
toddler and a hand on granddad's wheelchair no constitution will guarantee social 
equality" (p. 95). Ryan (1994) suggests that discriminatory behaviour at work often 
results from the attitudes held by the dominant group that women are secondary earners 
with less attachment to work and less interest in advancement than men. 
Fairhurst and Snavely (1983a) highlight the importance of the power distribution 
between groups on tokenism processes. Tokens generally have less power in an 
organization than the dominant members. Fairhurst and Snavely found (1983b) that male 
tokens did not experience as much social isolation as the female tokens in Kanter's (1977) 
study. They suggest that this may be because male tokens are not as powerless as female 
tokens. Male tokens have a higher societal status than female tokens and this aids them in 
the workplace. Fairhurst and Snavely (1 983a) explain that this gives the male tokens a 
power base that balances the power women have from being a majority and enables them 
to resist their token status. 
In addition, men may find it easier to be accepted by female colleagues in female 
dominated occupations than women in male dominated occupations (Benokraitis & Feagin. 
1986). EHemers, Doosje, Van Knippenberg and Wilke (1992) studied the relationship 
between status and numerical minority or majority. Being a numerical minority or 
majority does not automatically determine status. Ellemers et al. (1992) emphasize the fact 
that status, power, and numerical size are different characteristics and that they should not 
be used interchangeably. They found that a high status minority was viewed more 
favourably than a low status minority. However, when the ingroup was a majority there 
was no difference in positive feeling towards the group in the low or high status conditions. 
Their research suggests that because males in female dominated occupations 
1 2  
are a high status minority they, may identify more with their group than women in male 
dominated occupations. 
Yoder ( 1 994) also discusses the different aspects that can affect tokenism 
processes. She claims that there were actually four factors at work in the situation that 
Kanter ( 1 977) studied, and Kanter predominately focused on one of these: numeric 
imbalance. The other three aspects that Y oder mentions are gender status, occupational 
gender-inappropriateness, and changes in the positions of the genders in the occupation 
as a whole. 
Benokraitis and Feagin ( 1 986) identify some characteristics of tokenism in their 
discussion of subtle discrimination. They suggest that tokenism is used as an outward 
show of equality; it creates the impression that no group is totally excluded while it 
enables the dominant group to control the entry of other groups. They describe tokens 
as marginal, alienated members of the workgroup. Tokens are often hired and retained 
only as long as it is cost effective. In this situation turnover is inexpensive, so as soon 
as the minority starts requiring equal treatment or promotion they can be replaced by 
other token members. They also describe situations in which the minority members are 
placed in highly visible niches that have some power and status as symbols to signify 
that the organization is not discriminatory, which may become a role trap for the token 
and may limit their upward mobility. Tokenism can also be used as a threat to members 
of under-represented groups, as it can be used to make them doubt their abilities and 
position. For example, the members of the dominant group may say that a token only 
received a promotion because of their token status rather than their abilities. 
Benokraitis and Feagin ( 1 986) also found that sabotaging tokens' work was an 
effective form of discrimination in both traditionally female jobs and traditionally male 
jobs. In female dominated jobs, men may use sabotage to take control and take 
advantage of women's inferior job status. In traditionally male jobs, sabotage controls 
women's entry because women are seen as "potentially threatening the 'old gang' 
cohesion, camaraderie, and esprit de corps" (p. 1 09). 
Another interesting example of covert discrimination that Benokraitis and 
Feagin discuss is the selection of women or other tokens who are brought into the 
1 3  
system and used to control the entry and promotion of other women. The selection of 
tokens decreases the tokens' identification with other women as they think they 
achieved their position through hard work. It gives the impression that the organization 
is gender neutral and does promote the minority group. Benokraitis and Feagin claim 
that letting some tokens succeed in the system also causes the other tokens to attribute 
their own positions to their lack of ability rather than structures within the organisation. 
The acceptance of some tokens discourages group challenges to the system. 
Swim, Aiken, Hall and Hunter ( 1 995) studied what they termed modern sexism. 
They found that individuals who were high in modern sexism were more likely to 
attribute sex segregation to individualistic causes. A denial of continuing 
discrimination, antagonism towards demands for equal treatment, and resentment of 
positive discrimination characterize modern sexism. They found that respondents who 
scored highly on the modern sexism measure also overestimated the percentage of 
women in male dominated occupations, which was related to their disbelief in the 
existence of continuing discrimination. 
Other types of discrimination that contribute to the gender segregation of 
occupations are access and treatment discrimjnation, which involve discriminatory 
sel ection by employers, differential treatment, direct discrimination and harassment and 
gender stereotypes and expectations. Socialization also contributes to gender 
segregation. Gender appropriate role models direct individuals into gendered interests, 
traits and attitudes. Individuals also match their gendered self-image to a gendered 
occupational image that can contribute to the choice of occupation (McLean & KalID, 
1 994). 
Experimental Research 
Laboratory experim ents have usually studied the effects of different numbers in 
the group on different attributes and attitudes. Crocker and McGraw ( 1 984) carried out 
laboratory experiments designed to test whether behaviour in groups were, as claimed 
by Kanter ( 1 977), caused by the structure and availability of power in the organization 
1 4  
or by gender differences. Crocker and McGraw used groups of differing proportions of 
males and females. They discovered that being the solo member of a gender did affect 
performance in a small group. The processes were similar for males and females but the 
outcomes were different depending on the status of the token. They found that gender 
salience is high for solos and the other members of their group. Gender was most 
salient for solo females. Males were more likely to be identified as group leaders in all 
groups. Group members were more satisfied with the group conditions when their 
group included a solo male than when it included a solo female. Crocker and McGraw 
conclude that the problems in gender dominated occupations may be caused by the 
violation of traditional gender roles, rather than just minority status. 
Yoder ( 1 994) also studied the different processes involved in tokenism focusing 
on perceptions of occupational deviance by measuring students rating of vignettes. She 
found that token numbers alone did not produce the tokenism processes identified by 
Kanter ( 1 977). Subordinated gender status also contributed regardless of the gender 
appropriateness or prestige of the occupation. She also found contrary to her 
expectations that women in male defined occupations were not rated lower by 
respondents on work related qualities than traditionally occupied women, although their 
personal qualities were rated lower. The work atmosphere of female tokens was rated 
lower than that of male tokens and female non-tokens. The research by Crocker and 
McGraw ( 1 984) and Yoder highlight that numbers alone do not create tokenism 
processes, interactions and the culture of the workplace also have an impact on tokenism 
processes. 
The tokenism processes that minorities experience can affect tokens' own self­
concepts. Macke ( 1 98 1 )  studied the impact of sex related characteristics such as 
physical attractiveness and occupational competencies such as intelligence on the self­
esteem of individuals in different gender dominated occupations. She found that sex­
related characteristics were important determinants of self-esteem for individuals 
working in occupations dominated by either gender. The relative importance of 
occupational competencies was higher for those working in same gender dominated 
occupations. She explained her findings by claiming that characteristics that others in 
1 5  
the social environment emphasize become more salient. Colleagues of tokens 
emphasize tokens' gender more than their occupational competency, so gender becomes 
more salient for tokens' self-concept. 
Another experiment designed to test differences between genders in token 
situations was conducted by McClean and Kalin ( 1 994). They focused on the matching 
of a gendered self-image to an occupation. They found that both men and women in 
male dominated occupations have more masculine self-concepts than those in female­
dominated occupations. They found that respondents saw stereotypically male 
occupations as high in dominance and low in affiliation. They also found that the 
individuals in the gender traditional occupations in their study showed a higher level of 
similarity between their self-concept and their relevant occupational stereotypes than 
those in the gender non-traditional fields. They also compared students '  self-concept 
with their occupational stereotypes, finding that individuals entering female dominated 
occupations were higher in affiliation, and women were overall more affiliative than 
men were. The students ' self-concepts were most similar to the field they were 
studying. McClean and Kalin's research suggests that women in male dominated 
occupations may not identify highly with their gender group because their self-concepts 
may be more masculine than other women's. 
McCauley and Thangavelu ( 1 99 1 )  also studied stereotyping and gender 
dominated occupations. They measured the strength of sex occupation stereotyping, 
using a sample of students and non-student adults. They specifically examined the link 
between occupational and personality-trait stereotyping. They found substantial gender 
stereotypes linked to occupation and personality traits. They found individuals '  general 
perceptions of which occupations are gender dominated was accurate, but that the 
respondents tended to underestimate the percentage of women in both male and female 
dominated occupations. Respondents who saw a strong link between gender and 
occupation also saw a strong link between gender and personality traits. 
Metcalfe ( 1 987) studied the self-concepts of men and women in management. 
She found that men and women were equally ambitious, controlling, and forceful. 
Women also saw themselves as more intellectual than men which was in fact 
16 
demonstrated practically by the women holding higher academic qualifications than the 
men. The women managers had often experienced resentment and suspicion from male 
colleagues. Like Kanter (1977), Metcalfe found that the women saw themselves as 
highly visible, and as having to prove themselves. Women did not feel comfortable in 
the male dominated managerial environment and felt alienated by the emphasis on 
competitiveness and aggression. The women experienced conflict between their gender 
identity and the role that they were fulfilling. The conflict was expressed as a pressure 
"to separate themselves from their identity and reflect male gender-stereotyped 
behaviour as embodied in organizational cuJtures" (p. 218). 
Research into Specific Occupations 
There has been a great deal of research focusing on the experiences of 
individuals working in specific gender dominated occupations, a brief overview of some 
of the research is provided here. 
Morgan, Schor and Martin (1993) studied gender inequality in banking, finding 
that men advanced faster to middle management than women. Men had more work 
experience outside of banking than did women, and tended to have more mixed career 
paths with experience in more than one area within banking. They discovered that 
within banking, men were more likely to work in lending which was higher paid and 
lead more directly to management, whereas women more often worked in customer 
service positions. 
Similar resuJts were found in a study of women in business. Schneer and 
Reitman (1994) studied the career progress of women in business, an area which has 
traditionally been male dominated. They found that although women and men had 
similar experiences at the start of their careers, their mid career experience was not the 
same. They found the mid career environment was not supportive of women and that 
women had lower job satisfaction, were less appreciated by their bosses, experienced 
more discrimination and earned less than men. 
1 7  
An unsupportive environment may be an important factor in token processes. 
O'Farrell  and Harlan ( 1 982) compared male co-worker hostility towards women and job 
satisfaction in non-traditional jobs, defined as those predominately held by men and 
considered atypical for women, versus traditional female jobs. They also found that the 
women in non-traditional jobs were more satisfied with their pay and work than the 
women in the traditional jobs, and that the women in traditional jobs were more satisfied 
with their supervisors. They found no difference in satisfaction with co-workers. Both 
groups identified job security, pay, and work content as the most important job 
characteristics, and social interactions as least important. They also surveyed the women 
in the non-traditional jobs about the reactions of their male co-workers. They found that 
1 8% strongly approved of women's presence, 44% approved, 1 1 % had no opinion, 9% 
disapproved and 1 8% strongly disapproved. Thirty percent of the women in non­
traditional jobs also reported that male co-workers gave them 'a hard time' and often 
excluded them from on the job training opportunities. Women in non-traditional jobs 
who had experienced harassment were less satisfied with every aspect of their job except 
pay. The women in non-traditional jobs who had not experienced harassment had higher 
overall job satisfaction than did the women in the traditional jobs. 
Women who enter male dominated occupations may gain status from their 
occupation, while men may lose status by joining a female dominated occupation. 
Wharton and Baron ( 1 99 1 )  studied the benefits and problems for women in male 
dominated workplaces compared to women in predominately female workplaces. They 
found that women in male dominated occupations were the most satisfied with their 
jobs� which was despite the fact that they were more likely to report discrimination. The 
least satisfied women worked in female tilted ( 1 5-30% male) settings. They explain 
these findings by suggesting that well-being is influenced by a combination of the 
positive aspects (such as higher pay) and negative aspects (such as discrimination) of the 
total situation. Wharton and Baron found that women' s  well-being was threatened most 
when males were a workplace minority, which is counter to the predictions of gender­
neutral theories of tokenism processes. Wharton and Baron emphasize that power and 
1 8  
status in society as wel l  as numbers in the workplace must be considered in analyses of 
gender dominated occupations. They also highlight the importance of gender mix in the 
workplace as an important determinant of job satisfaction. 
Wharton and Baron ( 1 987) also studied the effect of gender segregation on 
men's well-being. They found that the men in the mixed gender setting experienced 
lower job satisfaction and self-esteem than the men in the male dominated or the female 
dominated work setting. They concluded that token males might actually interact better 
with females in the female dominated setting than in the mixed setting. Their findings 
support theories such as SIT, that claim that increasing numbers of a new group 
increases the threat to the original group. They also discovered that the gender mix in 
the workplace affected women's well-being less than it affected men's well-being. 
They suggest that men have more to lose from gender integration than women do which 
may make them more sensitive to co-worker gender than women. 
Other studies of men in female dominated occupations have been conducted. In 
a study of four female-dominated occupations, C. Williams ( 1 992) found that men were 
mostly accepted into the occupations she studied, which were nursing, librarianship, 
social work and teaching. The men in her study experienced very little discrimination or 
prejudice from colleagues, although they did experience prejudice from the public, 
clients and people they met socially. She found that "unlike women who enter 
traditionally male professions, men's movement into these jobs is perceived by the 
'outside world' as a step down in status" (p. 262). 
C. Will iams ( 1 992) found that the men in her study often encountered 
preferential hiring, as there were so few men in their occupations that they were sought 
after. The women colleagues were often eager to accept men into their profession and 
yet they also resented the speed with which the men often advanced. Men were also 
advantaged when compared to female tokens in that even though they worked in a 
female dominated occupation, their superiors and supervisors were most often men. 
Men's difference is also often viewed in female dominated occupations as something 
positive, which gives men more of an incentive to group together than women tokens 
have (C. Williams, 1 992). For female tokens their gender is often seen as a negative 
thing, to be hidden, and this discourages bonding between the token women. 
1 9  
C. Wil liams ( 1 992) found that men were often encouraged to enter specialties 
that others assumed were more appropriate for them as men. Men were often 
encouraged into the less hands-on areas of their occupations such as management, 
teaching and administration. Benokraitis and F eagin ( 1 986) identify this as an aspect of 
subtle discrimination termed 'containment' .  Tokenism generally refers to the exclusion 
of minority members, or quantitative exclusion. Containment is exclusion of a 
qualitative nature. It is the process used to direct minorities to certain areas of an 
occupation considered more appropriate for their gender. 
Overall C. Wil liams found that men were advantaged by their minority status, 
and she emphasizes the need to use different solutions for enhancing men's experience 
of female dominated occupations than are used to assist women in male dominated 
occupations. 
2.3 Male Dominated Occupations -
Engineers and Prison Officers 
Engineers 
Since the pioneering work of Kanter ( 1 977) there have been numerous studies of 
women who work in male dominated occupations. Although there are similarities in the 
situations of women in the different male dominated occupations, there may also be 
aspects of their situations that are unique. 
The present profile of women in engineering is very similar to women in other 
male dominated occupations as previously discussed. Entry into the profession for 
women has become easier over the last 20 years; and increasing numbers of women are 
entering the profession. However, women are stil l  clustered in some fields of 
engineering and are under-represented at the more senior levels. In New Zealand at the 
last published census women made up 3 .6% of professional engineers (Institute of 
Professional Engineers New Zealand, 1 999). The largest number of female engineers 
20 
work in the chemical and civil engineering fields. Although women stil l  only make up 
22. 1 % of chemical engineers and 1 .7% of civil engineers in New Zealand (IPENZ, 
1 999). 
Studies have been conducted to try to discover the factors that contribute to the 
under-representation of women in engineering. Jagacinski, LeBold, and Linden ( 1 987) 
found that in the United States, women engineers reported lower salaries and less 
supervisory responsibility than did men after only five years in the workforce. 
Jagacinski et al . found one factor that accounted for most of the difference between the 
men and the women in their sample. The women were more likely to have taken a 
career break than the men were. However, even controlling for this factor did not 
account for all of the difference between the salary and supervisory responsibility of the 
men and women. They found no educational differences between the men and the 
women, and they could not explain the difference between the men and the women. 
They suggest that less obvious factors such as "stereotypes and cultural values" (p. 246) 
may affect the progress of the women engineers. 
Bailyn ( 1 987) conducted a study with matched pairs of male and female 
engineers, and like Jagacinski et al. ( 1 987) found that women were paid less and had 
less supervisory responsibility than comparable men. She found that women were lower 
in self-confidence than men were and that they put less value on technical expertise, 
than men. She suggests that technical work is gender alien for women. 
In historically gender-dominated occupations such as engineering, not only are 
women under-represented numerically but they may also be disadvantaged by the l ink 
between the occupation and the gender role. Robinson and McIlwee ( 1 989) compared 
engineers in four different specialties, aerospace, high tech, mechanical and electrical. 
Initially they thought that there would be less discrimination against women in the high 
tech and electrical areas, as these are the least traditional areas of engineering and have 
the highest numbers of women. They found that the women engineers working in 
aerospace and mechanical engineering had more experiences of overt discrimination and 
harassment from colleagues and sometimes from their managers, than women engineers 
in high tech and electrical areas. Surprisingly they found that the women did better in 
2 1  
these very male dominated environments, in tenns of moving into senior positions in the 
organization. They explain these results by suggesting a number of factors that 
contribute to women's advancement in these organizations. Firstly, in aerospace the 
companies were mostly working on contract to the government and therefore must have 
affinnative action programmes in place, which work in favour of the women in the 
organization. In addition, these types of organization tend to be very bureaucratic and 
fonnal which, "means that the 'rules of the game' are clear. The newcomer in a non­
traditional field can quickly learn the rules, and can rely on fonnal structures and 
qualifications as she seeks to advance" (p. 46 1 ). In the areas of high tech and electrical 
engineering, the organizations tend to be less traditional. The rules for promotion are 
less clear and rely more on "personal reputations and peer evaluations" (p. 46 1 ). The 
criteria for advancement rely on male defined styles of interaction. They also found that 
assertiveness and self-confidence were problems for the women engineers trying to 
work in these less structured environments. 
Robinson and McIlwee ( 1 989) suggest that structures and procedures within an 
organization are important because they define power relations. An individual 's ability 
to progress in an environment is affected therefore by the structures and by the 
socialised gender attributes that may assist or hinder their ability to work with the 
structures. 
In 1 99 1 ,  Robinson and McIlwee again studied engineers. They focussed 
particularly on the culture of engineering. They found that the culture of engineering 
was strongly identified with the male gender role. They argue that an occupational 
culture of this sort is maintained through the day-to-day interactions with others, and 
that confonnity to the culture is the best way to fit into the strong group environment. 
They discovered that the culture was expressed through a strong ideology of obsession 
with tools and other hands-on abi lities; which often excluded women, and prevented 
them from fitting in. 
Robinson and McIlwee ( 1 99 1 )  claim that women's odds of success improve as 
competence becomes less male defined. The tool-oriented hands-on aspect of the work 
is the most male identified. The other aspects that are important to the job for example, 
22 
research, writing, calculating, and administering are less male identified. As an 
engineer moves into more managerial positions, the work they do is less identified with 
the male gender role. The position of female engineers is therefore similar to male 
nurses. For male nurses the fact that management positions are less female identified 
than the hands-on nursing, works in their favour and enables them to rise to managerial 
positions quickly. However for the female engineers there are stil l  gender role conflicts 
with the managerial role. Evetts ( 1 993) described some of the difficulties that female 
engineers have moving into senior positions. She suggests that the manager ideal is sti l l  
male defined, and that authority is sti l l  seen as a masculine attribute. She also mentions 
the problems women have with their family life influencing the time available to them 
to put into their careers. Evetts also suggests that what is defined as the competence 
necessary for a managerial position is stil l  defined by the men holding those positions. 
These men can use criteria such as being solely committed to the job with no family 
commitments, which means that men in positions of power hold on to the rewards and 
exclude women. 
Prison Officers 
The position of female prison officers is similar to that of female engineers, 
although there are some more specific problems that are related to the physical aspects 
of the job. In the prisons involved in the present study 2 1  % of the prison officers were 
women. Corrections officers' work consists ofthree types of duties (Crouch, 1 985): 
routine housekeeping and 10gistical duties, establishing and maintaining personal 
authority in relations with inmates and handling confrontations and physical violence. 
Research outlines some of the difficulties female prison officers have 
encountered since being allowed to work in male prisons. In the United States, this 
occurred in the 1 970's. Before that female prison officers were only permitted to work 
in women's prisons (Crouch, 1 985). Jurik ( 1 985) relates the move to permit female 
officers in men's prisons to the shift away from custodial prisons to an emphasis on 
rehabilitative prisons. An emphasis on rehabilitation changes the skills and abilities that 
23 
are needed in a prison officer. Jurik sees the previous custodial prisons, which focussed 
on social control as organisations where the traits perceived as masculine are valued. 
Rehabilitative prisons need staff that are more service oriented which creates an 
atmosphere that can place a greater value on traits that are perceived as feminine. 
Jurik's ( 1 985) study identified five major areas that create problems for female 
prison officers trying to integrate into the prison service. These are tokenism, 
conflicting organizational goals, external environmental conditions, informal 
organisational structures, and inadequate strategies for institutionalising correctional 
reforms. Some of these areas are specific to the prisons that she studied, and some of 
the areas are more general. Tokenism is common to women in all male dominated 
environments. Conflicting organisational goals is the conflict between a security 
emphasis and a rehabilitative emphasis that had not been successfully resolved in the 
prisons Jurik studied. External conditions refer to such problems as overcrowding and 
lack of staff due to high turnover. The female prison officers also often missed out on 
formal training because of staff shortages, which made it more difficult for them starting 
out on the job. Like the female engineers, female prison officers also seem to have most 
difficulty when the structures for promotion are informal, which creates an environment 
where it is who you know, and how well you fit in, that leads to promotion rather than 
competence or skil l .  
Jurik ( 1 985) also found that the range of work that women were assigned in the 
prison was l imited, which in turn limited their chances for promotion, as they had not 
had the necessary experience in all areas. The types of assignments that the women 
were given were wholly dependent on the attitudes of their supervisors who assigned the 
tasks. The l imitation to certain roles and areas discouraged female officers to support 
each other as "these allocation patterns served to encourage ' intra-sex' competition for 
the valued assignments" (p. 385). Jurik emphasises the fact that it is the interaction of 
gender role socialisation and working conditions, that affects the women in male 
dominated occupations. She claims that previous researchers have focussed either on 
organizational dynamics or on individual attitudes not a combination of the two. 
24 
Crouch ( 1 985) identified three structural factors that played a part in women's 
integration within the prisons: the proportion of women in the environment, (as 
discussed by Kanter, 1 977), the paramili tary hierarchy, and the amount of security in the 
prison, minimum to maximum. What he terms the paramilitary hierarchy is the clear 
rank structure that exists, women tend to be inexperienced and at the bottom of the 
organization. These organizations tend to be rigid and traditional which limits the role 
of women. The different security levels in a prison determines the role of prison 
officers similarly to the emphasis identified by Jurik ( 1 985) the less secure the prison 
the greater the emphasis on rehabilitation. 
Crouch ( 1 985) summarises some of the positive effects that women officers have 
been previously found to have on the prison, these include normalising the prison world, 
creating a more relaxed atmosphere and promoting better dress and behaviour among 
the inmates. Women correctional officers are similar in their situation to female police 
officers, as a lot of the resistance to their presence is due to their perceived physical 
inadequacies in violent situations. However, previous research indicates that women are 
less confrontational than men and can sometimes diffuse potentially violent situations 
with more success than do men (Bell, 1 992). In fact, the women in Crouch's study 
found dealing with their work colleagues more difficult than dealing with the inmates. 
Crouch ( 1 985) also describes a set of norms that are perceived by male prison 
officers to define masculinity. The norms included aspects such as maintaining 
toughness and avoiding femininity, confidence and self-reliance, aggression and status. 
Women entering the prison workforce are a threat to the masculine occupational self­
image. In fact, Crouch identifies women as "a status threat, if they perform well ,  
masculinity is somehow devalued" (p. 540). Women prove the greatest threat to men 
with a lower class background or with traditional sex role definitions. The men who are 
most threatened will show the greatest resistance to women officers. He also claims that 
a vocal minority can sometimes introduce collective norms, which is especially true if 
those male officers who are accepting of women are not willing to admit it. 
Crouch ( 1 985) also highlights the lack of role models of behaviour that women 
have. As there are so few experienced women in the prisons, new officers only have 
25 
men to provide informal instruction. Crouch makes specific recommendations for the 
questions that future research should address including: the effect that the number of 
women in an institution has on the difficulties they face on the job, the degree to which 
male officers treat women in categorical ways, and how women officers adapt to, or 
handle, resistance from male officers. 
Zuppan ( 1 986) criticises previous researchers that have studied the effect women 
have on prisons, for concentrating on factors that are solely related to their gender for 
example, normalisation, and a more relaxed atmosphere. She claims that the researchers 
have ignored the skills and attitudes that women possess. She studied male and female 
officers to see if their abilities and attitudes to each other and to the inmates differed. 
She found that in fact there was very little difference between the male and female 
officers. Women rated safety for individuals more important than male officers did and 
women officers reported more job tension than men did. In other attitudes towards the 
inmates and the prison, there were no differences, which could be explained by 
recruitment processes that may only select women who have similar attitudes to existing 
officers. 
2.4 Female Dominated Occupations - Nurses 
The situation of men who work in a female dominated occupation is less clear 
than those of females who work in a male dominated occupation. In New Zealand, men 
make up 5% of registered nurses (Kiwicareers, 1 998). Ott ( 1 989) found that male 
nurses actually benefited from their situation. Ott also found that the female nurses 
resisted the presence of male nurses less than the male police resisted female officers. 
Floge and Meril l  ( 1 986) found that male nurses were not as disadvantaged by their 
numerical status as female physicians. They also found that male nurses had more 
egalitarian interactions with physicians than did female nurses. They mention the total 
environment in which nurses work, and how as they progress up the hospital hierarchy 
their position changes. There are more males than females in the hospital administrative 
structure. Waddell ( 1 995) states that less than 1 0% of nurses (in the United Kingdom) 
26 
are male but that men hold half of the management posts. He goes on to describe how 
he feels about being called a male nurse, and claims that nurses are always referred to as 
'she' ,  and expected to be female. He also mentions how his colleagues always ask 
questions about why male nurses chose nursing, but do not seem to question the 
females' career choice. 
Heikes ( 1 99 1 )  states that socio-cultural factors must be included in any 
consideration of tokenism as well as numerical factors. Although male nurses can be 
considered as a minority group numerically, they are still members of the majority group 
in society, which affects their position. Snavely and Fairhurst ( 1 984) describe male 
nurses as high status tokens because of their status in society. They found that male 
nursing students did not display the characteristics of numerical tokens as introduced by 
Kanter ( 1 977). 
There are some indications that male nurses tend to end up in the less 'caring' 
nursing roles such as the lCU and psychiatric units. Heikes ( 1 99 1 )  reports a case where 
a male nurse had to file a discrimination suit to get a job in the maternity area. Male 
nurses are also frequently asked to carry out such duties as lifting heavy patients, or 
dealing with difficult patients. Male nurses also have trouble as they do not fit into 
stereotypical gender roles, and so many are assumed homosexual . C. Williams ( 1 992) 
conducted a study on men working in professions where men are the numerical 
minority. She found that the male nurses experienced some of the same problems as 
women but with the opposite result. She termed the phenomenon "the glass escalator" 
to contrast it with the "glass ceiling" that previous researchers have found that women 
experience (Morrison, White, & Van Velsor, 1 987). She found that many male nurses 
encountered a preference for hiring and promoting men. Men were also often 
encouraged into certain areas of these occupations, which set them up for moves into 
management or administrative positions. For example, a nurse who wanted to pursue 
further studies in child and family health was encouraged to concentrate on adult 
nursing. She also discovered that the men developed good formal and informal 
relationships with their superiors, who were usually men. C. Williams ( 1 992) found 
that unlike women, the men in her study encountered the most discrimination and 
negative attitudes from the public (clients and outsiders), as their job was seen as 
inconsistent with male gender roles. She also found that men felt they had to protect 
themselves against sexual abuse suspicions particularly if they were working in areas 
such as paediatrics. 
27 
Buchan ( 1 995) suggests that in the next few years the advantages that male 
nurses once had will decrease. The number of male nurses in the United Kingdom 
increased by 7.4% between 1 99 1  and 1 994 and the increase is larger than the actual 
increase in the overall numbers of nurses. The increase in numbers of male nurses was 
accompanied by a decrease in the number of management positions available, and a 
move towards community nursing and away from institutional care, which is a 
traditional area for male nurses. 
One researcher (Villeneuve, 1 994) has suggested that some of the dynamics of 
males in nursing, such as the way they seem to gather in administrative and management 
positions, may be explained by the differences in career development between male and 
female nurses. He states that men in Ontario, Canada are more likely to work full time 
than women in nursing: 8 1 .2 % of men worked full time compared to 54.9 % of the 
women. He also reviewed relevant literature and found that male nursing managers 
tended to be more likely to have degrees than female nurses, who tended to collect more 
qualifications, but of lesser quality such as certificates. Men also had clearer ideas 
about their career paths even when they were still in school . He concludes that more 
efforts need to be made to attract and retain men to nursing. 
There has also been some research that has been carried out into male gender 
roles and their violation. Some researchers have assumed that violation of gender roles, 
such as working in a traditionally female occupation, should result in increased stress or 
lowered well-being. The research however has not found this to be true (Thoreson, 
Shaughnessy, Cook & More, 1 993). Other theorists suggest that men lose social status 
when they join female dominated occupations and they have lower self-esteem (Macke, 
1 98 1 ). Wharton and Baron ( 1 987) claim that it is not the numerical balance that is 
important in tokenism processes. However the "quality and quantity of interactions in 
the workplace" (p. 575.) may have great influence on tokens. For example, Benokraitis 
28 
and Feagin ( 1 986) describe the experiences of a male nurse who felt that the doctors 
singled him out as the most competent nurse. In contrast, female nurses often complain 
that doctors treat them like domestic help not professionals. 
2.5 Chapter Summary 
Gender equalities in the workplace remain; many occupations are sti l l  segregated 
by gender. Individuals who are tokens are very aware of their category membership and 
feel highly visible, they find that their category membership is overemphasised and that 
they are treated according to the roles the dominant group expects them to be fulfilling. 
The research into tokenism processes has often focused on the numerical proportions of 
the groups, without adequately exploring the different experiences of male and female 
tokens. Men still have a higher status in society than women, which may contribute to 
men being advantaged by their token status, while women are disadvantaged. As 
attitudes to gender equality change, procedures and rules to aid equality are more 
prevalent. Discrimination has become covert, which makes interactions between gender 
groups in the workplace increasingly important to the token's experience of working in 
a gender dominated environment. The colleagues of the tokens can either be supportive 
to their entry into the occupation or see them as threatening their position, which can 
affect the token's job satisfaction. Research has also shown that tokens can experience 
conflict between their gender and the characteristics needed for their occupation. 
Women in engineering often have trouble integrating into the very technical 
atmosphere created by their male colleagues. Female engineers are more successful in 
situations that have formal rules, and that do not rely heavily on interactions with 
colleagues and supervisors. Female prison officers are assigned limited tasks within the 
prison and also have trouble integrating into the very masculine culture surrounding the 
work. Male nurses integrate more easily into the female dominated environment than 
female engineers or prison officers. They are also often encouraged into specific areas 
and treated according to their gender category, which in contrast to female engineers 
works to the male nurses advantage. 
29 
CHAPTER THREE: SOCIAL IDENTITY THEORY 
SIT is the theory that was adopted in the present study as a theoretical 
framework to study gender groups in nursing, engineering and among prison officers. In 
this chapter an overview of SIT is given, then previous research using SIT to study 
gender relations is reviewed. The basic principles or areas of SIT are introduced, the 
research in each area is briefly described and then the area is related to gender relations. 
The principles of SIT include social categorisations, identification, social comparisons, 
beliefs about the intergroup situation, and social change beliefs. 
3: 1 Overview of Social Identity Theory 
SIT is an ideal framework to investigate gender relations. Social Identity Theory 
was developed in Europe in the late seventies by Tajfel and Turner (Taj fel, 1 974, 1 975 ; 
Tajfel & Turner, 1 979). It was developed as a reaction to the earlier reductionist 
theories of intergroup relations, which had their origins in the USA, for example 
Sherifs Realistic Conflict Theory (Taylor & Moghaddam, 1 994). These theories tended 
to concentrate on individuals ' behaviour in groups, or tried to understand group 
behaviour using the principles developed to understand individuals'  behaviour. SIT on 
the other hand explains how the group can contribute to an individual 's identity, and 
focuses on the "group in the individual" (Hogg & Abrams, 1 988, p. 3). S IT is 
concerned with the unequal status relations between groups being "originally developed 
to account for phenomena such as discrimination against, hostility towards or 
persecution of members of outgroups" (Emler & Hopkins, 1 990, p. 1 1 3). The theory 
makes specific and testable predictions about the effects of belonging to a group of low 
status, and the reactions of groups of high status to threatened status, which makes it 
particularly suitable for studying gender relations in gender dominated occupations. S IT 
conceptualises social interactions as fal ling on a continuum with solely interpersonal 
interactions at one end and intergroup relations at the other. 
30 
Social Identity Theory had its origins in the information processing principles of 
research on categorization. The theory originated from work done by Tajfel and 
colleagues studying at the categorization of lines of differing lengths (e.g., Tajfel & 
Wilkes, 1 963). They discovered that categorisation of the lines into groups made people 
overestimate the similarities within the groups and the differences between the groups. 
They termed this 'the accentuation principle ' .  Tajfel believed that this effect occurred 
in social categorizations - of people into groups for example as well  as in cognitive 
processing of visual stimuli . Social categorizations are the "cognitive tools, that 
segment, classify, and order the social environment" (Tajfel & Turner, 1 979, p. 40). 
People will accentuate the differences and similarities of characteristics related to the 
categorisation, and the accentuation is stronger when the categories are important, 
relevant, or highly valued. 
Minimal group experiments were commonly used to demonstrate and develop 
the principles of S IT. Researchers designed minimal group experiments to remove pre­
existing intergroup biases from the experimental situation, in order that any effects 
found could only be explained by the categorization of participants into groups (e.g., 
Tajfel, Billig, Bundy, & Flament, 1 97 1 ). They performed these experiments by 
randomly allocating individuals to groups. The individuals were told that the groups 
were based on criteria such as which modern artists respondents l iked best, or whether 
they under or over estimated the length of lines. The groups therefore had no history of 
contact or conflict and no face-to-face interaction between members, and no opportunity 
for future intergroup interaction. The popular way of measuring the differentiation 
between groups in the minimal group experiments was with the "Taj fel Matrices" 
(Bourhis, Sachdev, & Gagnon, 1 992). These are a series of tables in which participants 
reward points (which represent money) either to two ingroup members, two outgroup 
members or to a member of each. The choices that the respondents make can be 
categorized into five allocation strategies. The first is 'fairness' ;  in which the groups are 
allocated points evenly. 'Maximum joint profit' is the option that will lead to the most 
3 1  
money for all .  Thirdly 'maximum ingroup profit',  is a strategy which focuses on 
getting the most points for the ingroup regardless of the points given to the outgroup. 
The 'maximum differentiation' strategy is one where the respondent will chose options 
that increase the difference between the ingroup and'outgroup even at the expense of 
ingroup profit. Lastly, the strategy of favouritism is the combined use of maximum 
ingroup profit and maximum differentiation. 
When SIT was first being formulated and tested it was necessary to use 
laboratory created groups (Tajfel et al . ,  1 971 ; Turner, Sachdev, & Hogg, 1983) rather 
than real life groups. This was so that minimal groups could be used, to eliminate 
confounds such as an existing history, or other factors that could influence the 
perceptions of the groups. The initial use of minimal groups has led to a large 
proportion of the research conducted within the SIT paradigm being laboratory 
experiments. Skevington ( 1 989) suggests that the focus on laboratory experiments has 
had the result of restricting the use of alternative methodologies from SIT research. 
Griffin ( 1989) criticizes the laboratory experiment approach as it has led to SIT research 
that is remote from any social context. For example Tajfel ( 1 982) saw social identity to 
be variable across situations and affected by context, but most researchers treat it as 
stable. In addition Bourhis ( 1992) mentions that factors such as unequal status relations 
or power are often neglected in laboratory created groups. Not all experiments have 
included manipulations of status and many that do focus on only the low status group 
neglecting the potential SIT holds for examining the beliefs of high status groups. 
As the theory has developed studies have expanded into using natural groups 
(not experimental) such as gender (Hogg & Turner, 1987), ethnicity (Finchilescu and de 
lay Rey, 1 99 1 ), political affiliation (KelIy, 1988) and attitudes to the Gulf war (Heskin 
& Power, 1994). SIT has also been used to look at the relations between groups in the 
workplace (Haslam, 2000). 
The move away from the laboratory paradigm and into field studies has created 
some problems in terms of the methodology used. While the original studies were 
concerned with manipulating and controlling conditions, the field research focuses on 
measuring and observing factors, which has lead to inconsistency in the measures used 
32 
by researchers. Finchilescu and de la Rey ( 1 99 1 ) mention some of these difficulties 
when discussing the methods they used to study attitudes towards ethnic groups (whites 
and blacks) in South Africa. Finchilescu and de la Rey claimed that the measure they 
used to assess status, legitimacy and stability, had face validity but they could not 
measure criterion validity as there was no established measure of legitimacy and stability 
in the Tajfellian sense that they could compare their measure to. 
Researchers have therefore used many different methods of measuring key 
aspects of SIT such as legitimacy and stability in field research, which makes it hard to 
compare results from different studies. The different methodologies used could be 
contributing to any discrepancies in the findings. Relatively few studies have used a 
combination of laboratory created and real life groups, which may be because this is a 
difficult situation to create and control . Jackson, Sullivan, Harnish and Hodge ( 1 996) 
used both laboratory groups and a real life group (smokers and non-smokers) and 
Lalonde and Silverman ( 1 994) used faculty membership as well as laboratory created 
groups. There are factors that can influence group relations that may not be present in 
artificial groups. Jackson et al. ( 1 996) conducted a series of experiments using a 
minimal group, and two real life groups - cigarette smokers and men and women. They 
found differences between the real and laboratory groups and suggest "research is 
needed to determine when and why real-life groups differ from laboratory created 
groups" (p. 253). 
The diversity of research into SIT has resulted in a very broad range of group 
attributes being studied, very little of this has studied the social change beliefs of real 
groups. SIT has evolved considerably over the years, as this study focussed on one of the 
original aspects, the initial statements of the theory have been used where possible to 
guide the current research. 
3.2 SIT and Gender Relations 
Tajfel and Turner ( 1 979) conceptualised intergroup and interpersonal relations as 
existing at opposing ends of a continuum. In contrast Abrams ( 1 989) describes how 
individuals can experience both interpersonal and intergroup aspects of relations at the 
same time. This is particularly true in gender relations where "one of the clearest 
criteria for distinctive group memberships (sex) can also be the basis of the most 
33 
intimate human relationships" (p. 64). Skevington and Baker ( 1 989) suggest that 
intergroup and interpersonal gender relations should be seen as "existing simultaneously, 
dynamically interacting to transform each other" (p. 1 99). 
Although SIT seems to be very suitable for considering gender relations from an 
intergroup perspective, surprisingly little research has done so. The initial work on this 
issue was by Wil1iams and Giles ( 1 978), who reviewed gender relations from a SIT 
perspective. Hogg and Abrams ( 1988) discuss gender relations as a concrete example 
when explaining SIT. These reviews look at gender relations in a theoretical manner 
rather than undertaking empirical studies of gender relations. Some empirical research 
has been conducted (Skevington and Baker, 1 989) although these studies have set out to 
look at gender relations from an intergroup perspective, they have focused more on 
measuring women's gender identity, and whether SIT explains women's behaviour as 
well as it does men's (1. Williams, 1984). 
Condor ( 1 989) identifies some problems with the SIT research on women. She 
states that the original theory was dynamic and flexible but the reliance of researchers on 
traditional research methods has prevented researchers exploring the theory to its full 
potential. She points out three main areas where this has occurred. The first area she 
claims relies on the 'objective existence' of women as a group (e.g., Williams & Giles, 
1 978), referred to as 'gender dualism' .  The research focuses on the membership of 
women in the gender group without considering differences among women or 
differences in how they perceive their identity as women. The research in this area has 
failed to really explore the meanings of gender categorizations, or whether they are 
affected by context or other factors. Social change is often assumed to be about the 
change in status between men and women, rather than alternatives, such as the removal 
from usage of the categories men and women (Condor, 1 989; Hogg & Abrams, 1 988). 
The second area of research reviewed by Condor is studies from an individual 
differences perspective - which assume that constructs such as gender identification, 
salience and self-stereotyping are relatively stable (Abrams, Thomas and Hogg, 1 990; 
Gurin & Townsend, 1 986). Researchers in this area most frequently use measures based 
on standard attitude and personality scales, which leads to research that focuses on the 
quantity or strength of the construct, and neglects to explore alternative aspects of 
gender relations, for example how identification is expressed, or research on the 
meaning and use of the gender categories. 
34 
The final area of research studies reviewed by Condor is research which begins 
to consider the subject's perceptions of category membership, such as the research on 
gender salience (Oakes, & Turner, 1 986). These studies focus on people's use of gender 
categories as social perceptions or self-definitions. However this approach also focuses 
on the quantity of the constructs being measured. These three areas of research also pay 
l ittle attention to the interactions or relations between men and women when they are 
acting as groups. In addition the majority of research focuses more heavily on women 
and their category membership to the exclusion of men and their category memberships. 
Other criticisms have been made of SIT's approach to women. Some theorists 
suggest that there are differences in how men and women think about groups and that 
S IT does not consider this. J. Williams ( 1 984) has suggested that S IT is a theory of 
intergroup relations between men. The initial minimal group experiments that showed 
categorization created differentiation between groups were conducted with all male 
respondents (e.g. , Tajfel et aI , 1 97 1 ). J. Williams claims that the principles of 
comparison, distinctiveness and differentiation, which form the basis of SIT, are more 
characteristic of male group behaviour than female behaviour. She also mentions that 
the theory neglects other forms of intergroup interactions such as attachment and 
affiliation to others in the group and also to other groups. She suggests that these 
communal processes are more common among women than among men. J. Williams 
claims that the communal processes favoured by women contribute to their social 
position and help to maintain the unequal power relations between men and women. 
Bourhis ( 1 992) claims that J. Williams ( 1 984) takes a dispositional approach to 
the different orientations of men and women regarding group behaviour. A 
dispositional approach implies that group behaviours are internalised aspects of men and 
women. Bourhis ( 1 992) conducted an experiment to test whether women are disposed 
towards communal processes or whether it is being of low status that disposes groups to 
communal processes. He found that men also used communal processes when they 
35 
were in low status positions. Bourhis concludes that relative status positions cause more 
differences between males and females in groups than dispositional characteristics. 
Overviews of gender relations in a broad sense have been written, but currently 
very little research has been conducted into relations between the genders from a SIT 
perspective. There has been little research on SIT and gender relations since Breinlinger 
and Kelly (1994). 
3:3 Social Categorizations 
Categorization is a basic cognitive process that individuals use to cope with the 
large amount of information that they encounter on a day-to-<iay basis. People 
categorize both social and non-social stimuli (Tajfel, 1974). Categorization is a "process 
of bringing together social objects or events in groups which are equivalent with regard 
to an individuals actions, intentions, attitudes, and systems of beliefs" (Tajfel, 1 974, p. 
69). 
The minimal group experiments that Tajfel and his colleagues (Tajfel et al. 
197 1 )  initially worked on were designed to show that categorization alone was enough 
to create competition between groups. Later Billig and TajfeJ ( 1973) questioned 
whether these groups were truly minimal. In the initial experiments, there were 
unintended aspects of similarity and dissimilarity between the groups that could 
contribute to competition between the groups. The researchers allocated respondents to 
groups on the basis that they were either under estimators, or over estimators on a 
perceptual task, or in the second experiment on their preference of modem painters. The 
researchers suspected that individuals could have used these criteria to create self­
referent categorizations and to may have created perceived similarity between group 
members (Tajfel, Billig, Bundy and Flament, 1971). Tajfel and Billig (1973) conducted 
a minimal group experiment using random trivial categorizations; the participants were 
assigned to groups randomly by number. The random trivial categorization could not be 
used as a basis of subjective similarity. They concluded from the findings of 
discrimination measured on the Tajfel matrices, that it is solely categorization that 
causes the inter group competition, not the similarity of group members. Tajfel and 
36 
Turner ( 1 979) state that categorization does accentuate similarities within groups and 
differences between groups. This however only applies on dimensions salient in the 
specific context, relevant to the categorization and of importance and value to the group. 
Abrams and Hogg ( 1 990a) suggest that categorizations are flexible and that the 
one that best accounts for the similarities and differences between people will be used. 
Turner, Sachdev and Hogg ( 1 983) state that people will define themselves according to 
the available social categorizations if that wil l  help make sense of their current social 
context. Individuals' behaviours and reactions to others are affected by the 
categorizations they make, in their social environment (Abrams, 1 990). 
Individuals categorize themselves and others as part of groups. Categorization 
distinguishes between ingroups of which the individual is a member and outgroups. 
Including the self in an ingroup is called self-categorization; it underlies social identity, 
as social identity is the internalisation of the categorization (Skevington and Baker, 
1 989). Categorizations become self-referent, and lead to the learning of group norms, 
which are then applied to the individual's own behaviour. Behaviour becomes more 
normative as category membership becomes more salient (Hogg, 1 992). In this way, 
social categorizations help create and define an individual 's place in society. 
Tajfel ( 1 978) used a broad definition of what constitutes a group. He claimed 
that individuals are considered a group when they categorize themselves as such, and 
when they are consensually categorised as a group by others. Turner, Sachdev, and 
Hogg ( 1 983) point out that identification with a group can occur as a result of other 
people defining the individual as part of that category. If there is an external consensus 
among others about the individual ' s  category membership it can sometimes lead to "a 
subjective acceptance of the group membership despite the fact that at least initially 
there may be little interdependence and attraction between members" (p. 227). 
Identification may occur because categorization by others introduces recognition and 
acceptance of a common fate between the individual and other ingroup members. Being 
treated in a similar way to others may lead to this awareness. 
37  
Gender and Social Categorisations 
SIT states that individuals categorize themselves and others as parts of groups, 
which contributes to their social identities, and facilitates social comparisons. Social 
comparison in turn leads to an exaggeration of differences between groups and the 
similarities within groups. It is widely assumed that categorization by gender is 
common. J .  Williams ( 1 984) states, "it is uncontentious to state that sex is a basis for 
social differentiation and stratification" (p. 3 1 1 ). Del Boca and Ashmore ( 1 986) raise 
the question of whether men and women are actually groups, claiming that they are 
fundamental social categories but that does not necessarily make them groups. If 
categorizations into male and female are as common as del Boca and Ashmore ( 1 986) 
suggest, then men and women are functioning as groups. Although there are qualities of 
groups identified by del Boca and Ashmore ( 1 986) that the genders do not share, for 
example some qualities that groups have are interdependence among group members, 
interaction, structure, and definition as a group by members and others. Men and 
women do not exhibit these qualities, not all men or all women are interdependent, not 
all men or all women define themselves as members of the group, and not all men are 
treated in the same way by others. There is also a great deal of interdependence between 
men and women. 
Token situations may heighten the sense of sharing a common fate with others 
similarly treated. For example the research by Kanter ( 1 977) on women in a male 
dominated occupation suggests that women in this type of setting do feel categorized by 
others and as if they are representing their category. Kanter called the process 
'polarization' but what she describes is really the accentuation effect in SIT, the 
exaggeration of within group similarities and between group differences. 
38  
Social Categorizations Summary 
To summarise, the concept of social categorization came from the studies that 
Tajfel and Wilkes ( 1 963) performed in the area of information processing. They applied 
their findings to the social environment. When individuals make social categorizations 
of themselves and others, they emphasize the differences between groups and the 
similarities within groups. The experiments conducted showed that categorization into 
groups is enough to produce discrimination and ingroup favouritism. Individuals make 
self-categorizations, that is place themselves as a member of a category this is the basis 
of social identity. 
The genders are widely used as.a basis for categorisations, although the 
consensus of the meaning of the categorizations has been questioned. Working in a 
token situation may increase the awareness of category membership, as the groups tend 
to overemphasise similarities within groups and differences between groups. 
3:4 Social Identity 
The self-concept can be seen as a collection of self-images (Abrams & Hogg, 
1 990). It is made up of social identifications and personal identifications. Social 
identifications are those gained from categories or groups and personal identifications 
are individual descriptions (Hogg and Abrams, 1 988). Hogg and Abrams ( 1 988) 
suggest that individuals do not experience their total self-concept at any one time, but 
only experience a series of self-images that are dependent on the situation. Tajfel and 
Turner ( 1 979) define social identity as consisting of "those aspects of an individual ' s  
self-image that derive from the social categories to which he perceives himself as 
belonging" (p. 40). S IT states that individuals strive for a positive social identity. They 
gain positive social identity through membership in positively valued groups. 
Membership in groups with little value may result in a negative social identity. A less 
valued group may be a group with little power, status, or access to resources. Positive 
and negative evaluations are the result of social comparison processes between the 
ingroup and outgroups. 
39 
Tajfel ( 1 974) saw the social identities of individuals as dynamic and that 
individuals hold more than one identity at a time; gender identity, for example, can be 
seen as just one of an individual 's  possible social identities. This introduces new 
questions about the relationships between different social identities that individuals can 
hold at the same time. 
Tajfel and Turner ( 1 979) believed that group membership becomes internalised 
as part of the individual's self-concept. Members of the outgroup who consistently 
define individuals as part of a group can contribute to this process of internalisation into 
the individual 's  self-concept. Taylor and Moggaddam ( 1 994) suggest that it is possible 
for individuals who do not necessarily share any emotional ties or evaluation of a group 
to be influenced by the fact that they are identified as a part of a certain group. 
Gurin and Townsend ( 1 986) define social identity as a member's awareness of 
group membership and their feelings about their membership. They state that group 
identification had three aspects: the perception of intergroup similarities, the awareness 
of a common fate and the centrality of group identification in the self-concept of the 
individual . Turner, Sachdev, and Hogg ( 1 983) define identification as the process by 
which a social categorization becomes part of the self-concept. Brown, Condor, 
Mathews, Wade and Williams ( 1 986) identified three major aspects of group 
identification: awareness of membership in the group, evaluation of that membership 
and the affect or emotion associated with membership. 
Social identity is therefore the part of the self-concept that is related to an 
individual 's group or category membership. Social identity can be influenced by the 
categorizations and perceptions of others. Social identity is made up of awareness of 
group membership, the value attributed to that membership and the affect towards 
membership. According to SIT, social identity influences other processes, such as 
differentiation, self-esteem, identity salience, and the actions taken by the groups. 
40 
Identification, Differentiation and Self-Esteem 
SIT links identification with intergroup differentiation, predicting that how 
strongly an individual identifies with the group will affect how much they differentiate 
between their group and the outgroup. SIT postulated a complicated relationship 
between identification, differentiation, and self-esteem. The need for a positive self­
concept is a fundamental assumption upon which Social Identity Theory rests. Tajfel 
and Turner ( 1 979) state that "individuals strive to maintain or enhance their self-esteem, 
they strive for a positive self-concept" (p. 40). SIT states that individuals' group 
membership contributes to their self-esteem. Therefore, the strength of i dentification 
with the ingroup and the importance of the group membership in the self-concept will 
affect how much the group' s  status contributes to the group member's self-esteem 
(Locke, McClear & Knight, 1 996). Identification with a positively valued group 
therefore increases self-esteem, while belonging to a negatively valued group lowers 
self-esteem. Abrams and Hogg ( 1 990) state that if self-esteem is low the group member 
has various options depending on their level of identification. If the individual has low 
identification with the i ngroup, they can leave the group, so long as the groups have 
open boundaries. If the individual identifies highly with the ingroup, they wil l  seek 
ways to raise the value of their group, which can be through differentiation strategies 
such as ingroup bias. 
The predicted relationship between self-esteem, differentiation and identification 
has received mixed support from research. Kelly ( 1 988) carried out a study on members 
of political groups and found that the level of ingroup identification predicted i ntergroup 
differentiation. Other researchers have not shown such definite links. Skevington 
( 1 98 1 )  in her study of groups of nurses found that the high status group (registered 
nurses) showed the expected relationship between differentiation and identification but 
that the results from the low status group (enrolled nurses) were inconclusive. Brown 
and Williams ( 1 984) conducted a study in a bread factory and found mixed results with 
some groups showing the expected relationship and others showing the reverse. Brown, 
Condor, Mathews, Wade and Will iams ( 1 986) found in their study set in a paper factory, 
4 1  
that the relationship between differentiation and identification was inconsistent with only 
two out of the five groups showing the expected correlation and those only weakly. 
Oaker and Brown ( 1 986) found the reverse of the expected relationship: they 
found that there were lower levels of differentiation between different groups of nurses 
when identification was higher. Hinkle, Taylor, Fox-Cardamone and Crook ( 1989) 
conducted a study into identification and differentiation using students in small decision 
making groups. They divided identification into different subscales (emotional factors, 
cognitive factors and individual needs versus the group factors) and found that the 
relationship between identification and differentiation differed among the identification 
subscales. 
The relationships between identification and ingroup favouritism (or bias) and 
identification and differentiation have been criticised (Brown, 1995; Hinkle & Brown, 
1 990; Taylor & Moghaddam, 1 994) because of mixed results in research studies trying 
to show the relationship between these aspects. Turner ( 1 999) answers these criticisms 
by stating that often the researchers who criticize SIT have focused too narrowly on one 
or two aspects of the theory, and have not really taken into account the full picture that 
SIT provides. SIT never actually proposes a direct link (demonstrated statistically by a 
correlation) between identification and ingroup bias or between identification and 
differentiation. The total theory proposes mediating factors such as salience of identity, 
the structure of the intergroup relations, the relevance of the comparison dimension and 
the relevance of the outgroup. SIT also states that differentiation is only one option an 
individual can choose to achieve positive distinctiveness, depending on their beliefs an 
individual can also use social mobility and social creativity. Researchers have not paid 
as much attention to these group actions as they have to discrimination and bias. 
The inconclusive results from past research suggest the relationship between 
differentiation and identification may be more complicated than SIT first assumed. It 
could be that factors such as the basis of the group influence the relationship. Groups 
that are ideological and have l ittle actual personal contact between members such as 
those studied by Kelly ( 1 988) may be quite different to natural groups that have a great 
deal of personal contact, such as those studied by Oaker and Brown ( 1986). Oaker and 
42 
Brown ( 1 986) did find that the level of bias was affected by the amount of intergroup 
contact between the groups in their study, although this was only true for the low status 
group. 
Methods of measuring differentiation are inconsistent across studies, which may 
cause some of the inconsistency of results. Hinkle et al . ( 1 989) measured differentiation 
by requiring the participants to evaluate the in group and an imagined outgroup on 
instrumental and socio-emotional aspects of the group process. Hinkle et al . then used 
the difference between the ingroup and outgroup rating as a measure of differentiation. 
Other researchers have measured differentiation through interviews with the participants 
(Brown et aI . ,  1 986; Oaker & Brown, 1 986). 
Gender Identity 
There is stil l  some debate about the use of certain terms that are necessary for the 
study of gender relations. Some researchers use the terms 'gender' and ' sex' 
interchangeably (e.g., Kanter, 1 976). Nevertheless, some consensus is emerging: 
'gender' is more widely used to indicate the socially constructed categorization and 
'sex ' is used to indicate the categorization that occurs due solely to biological structures 
or physiological differences (Nicholson, 1 994). As sex and gender refer to different 
categorisations an individual can possess both a sex identity, which is based on the 
awareness of physiological or biological aspects of categorisation, and a gender identity 
that covers the social aspects of being a man or woman. The present research will only 
concern itself with gender identity. The term gender will be used to refer to the social 
categorizations and sex will be used to indicate biological or physiological 
categorizations. 
According to SIT, a social identity is the part of an individual's self-concept 
derived from group membership (Tajfel & Turner, 1 979). A fuller definition is given by 
Abrams ( 1 989) who defines gender identity as "a basic phenomenological sense of 
one's maleness or femaleness - which cannot be measured against the yardsticks of 
social consensus (for example, global sex-stereotypes), biology (sex) or distributional 
43 
data (psychometrics). It is essentially a matter of social comparison in specific 
contexts" (p. 59). Other researchers have described gender identity as " a  global sense 
or one's maleness or femaleness that is acquired early in l ife and is a group level defined 
concept" (Lorenzi-Cioldi & Doise, 1 990, p. 77). Gender identity has also been 
described as " the part of the self-concept that derives from membership in a gender 
category" (Gurin & Markus, 1 989, p. 1 53), or as " the sense in which people recognize 
their membership in one of the two gender groups" (Duveen & Lloyd, 1 986, p. 222). 
Gurin and Townsend ( 1 986) have studied the properties of gender identity and their 
relationship to gender consciousness. They define gender consciousness as a person's 
ideology about the position of his or her gender in society. They state that gender 
identity contains aspects of similarity, common fate, and centrality of the identity to the 
sense of self. Common fate influenced gender consciousness more than similarity and 
centrality. Spelman ( 1 988) c laims that gender identity is a relational concept: women 
discover the specific characteristics and attributes of their gender identity through 
comparison with men. This idea is demonstrated in a study by Thoreson, Shaughnessy, 
Cook, and Moore ( 1 993) who studied perceptions of the male gender role, which may 
indicate some of the categorisations that contribute to men's gender identity. They 
found that conceptualisations of the male role consist of the idea that men should seek 
status, should be tough, and should not engage in traditionally feminine activities. 
Thoreson et al . elaborate this theme and name it the 'anti-femininity' theme which 
conceptualises the male gender role as based on the view that women are adversaries. 
Their research indicates that part of what makes the male gender role is not being 
woman. Within the S IT framework there has been very little work done on the gender 
identity of men. Work on the development of gender identity such as that conducted by 
Abrams ( 1 989) has included the development of male gender identity but studies of 
adults have largely focused on women (e.g., Condor, 1 986; Gurin and Towsend, 1 986). 
The diversity within the gender groups and gender identity's interaction with 
other identities makes it difficult to specify exactly what group members are identifying 
with, when they identify with a gender group. Hogg and Abrams ( 1 988) suggest that 
gender identification may not constitute identification with all members of one's own 
44 
gender but may refer to a specific sub-group, of which the individual is a part. Lips and 
Colwill ( 1 988) describe a study into gender identity that found homemaker and working 
women saw themselves as very different, but they both disassociated from groups of 
women who felt deprived and identified with privileged women. They also found that 
both sets of women were more l ikely to compare themselves to other women than to 
men. 
Breakwell ( 1 979) questions whether gender categorizations have the same 
meanings for everyone. It is assumed that the categorizations into male and female are 
generally shared and yet Breakwell emphasizes that there are no external criteria of 
womanhood. Researchers such as Young ( 1 994) and Nicholson ( 1 994) have stated that 
women cannot be treated as a unified group because there are too many different women 
with too many different identities to include them all in one grouping. S IT (Tajfel and 
Turner, 1 979) does see identities as dynamic and interacting with each other, which is 
why salience is an important concept in S IT. 
Smith ( 1 988) criticised research that assumes that women can separate their 
gender identity from their racial identity. She states that the different identities interact 
and affect each other, for example identifying as a white women puts an individual in 
quite a different position in relation to other groups as identifying as a black women. 
She also points out that the dominant group often has the opportunity to assert their 
definition of the other group's identity, which occurs especially in relation to differences 
and similarities between groups. The example she gives is that white women can 
express solidarity with black women when it is to the white women's advantage. 
However white women can also emphasize the differences between white and non­
white women when that suits white women's purposes. 
Researchers have explored many different variables that affect the development 
of gender identity. These have included environmental gender typing (Katz, 1 986), 
socialization and social constructs (Sherif, 1 982), class, religion and race (Spelman, 
1 988), and intergroup relations (Abrams, 1 989). Sex role orientation can be a cause and 
a consequence of gender identity as it conditions the meaning of being a woman 
(Condor, 1 989). 
45 
SIT states that members of high status groups wil l  have higher group 
identification than members of low status groups as identifying with a high status group 
makes a positive contribution to the individuals self-concept. Therefore the following 
hypothesis is formulated for the present study, "Men will have higher gender 
identification than will women" (H I ). Researchers such as Williams and Giles ( 1 978) 
have assumed that women who accept the status quo have internalised their inferiority 
and have low self-esteem or negative self-concepts. Breinl inger and Kelly ( 1 994) state 
that this is the result of thinking of women with traditional beliefs at the opposite end of 
a bipolar scale to women with feminist or non-traditional beliefs, and assuming that the 
only way to a positive identity is through social competition. Condor ( 1 986) found that 
the reactions of women who identify highly with women are affected by their sex-role 
orientation. Women who identify strongly and hold a radical sex-role orientation are 
likely to choose social competition strategies, and women who identify strongly and 
hold a traditional sex-role orientation are less likely to choose social competition 
strategies. Traditional women can identify strongly with women (Condor, 1 986) and 
women who succeed in non-traditional jobs may actually, as a strategy to cope with their 
gender's  status in their workplaces, dissociate from their gender (J. Williams, 1 984). 
Other differences in gender identity between women with traditional beliefs and 
non-traditional beliefs have been researched. Gurin and Markus ( 1 989) found that 
women with a non-traditional sex role orientation thought a lot about being a woman, 
which may be due to the increased salience of gender identity for those who have more 
competitive beliefs. Women with a traditional sex-role orientation did not think a lot 
about being a woman. In previous studies into gender discrimination women have 
reported how they have become involved with feminist issues as a result of experiencing 
gender discrimination (Gomick, 1 983). These studies indicate that gender may become 
more salient and identification may increase for those people who experience gender 
discrimination. Therefore in the present study, "Gender discrimination will be 
positively correlated with self-concept gender salience and gender identification " (H2). 
Some researchers have explored the relationship between gender identity and 
other social identities. Marshall and Wetherell ( 1 989) suggest that the relationship 
46 
between women and their occupational identity has become "problematized" but the 
relationship between men and their occupational identity is "normalized". They 
conclude that men's occupational identity is more integrated with their gender identity, 
whereas women's occupational and gender identities are often separated and can be a 
source of tension. Marshall ( 1 991 )  found that strong identification with women and a 
traditional sex role attitude correlated with a weak occupational identity. 
Breakwell ( 1 979) also enlarges the discussion of gender identity by introducing 
the issue of internal and external criteria for group membership. Internal criteria are the 
opinions held by the individual, external criteria are the standards and norms the group 
upholds. People have opinions about how well they represent the group, for example, 
they can see themselves as not meeting up to the external criteria of the group standards 
for membership. Not meeting up to group standards can make group members feel 
marginalized, which is particularly relevant for individuals working in gender 
dominated occupations, as carrying out occupational roles associated with the other 
gender may cause them to doubt their ability to fulfil the external criteria of group 
membership for their gender. The dominated group can influence tokens so that they 
doubt their ability to fulfil the external criteria of gender membership, which is a source 
of discrimination in male dominated occupations. For example McIlwee and Robinson 
( 1992) found the association of engineering with the male gender role created conflict 
for the female engineering students in their study. They found that the women 
engineers perceived that the attention that was given to them because of their scarcity 
undermined their credibility as an engineer because the attention focussed on their being 
women. On the other hand acceptance by colleagues as engineering students 
undermined the female students credibility as women. For example jokes about the lack 
of physical attractiveness in female engineers, demonstrated that the image is of an 
engineer or an attractive woman but not both. 
The research on stereotyping of occupations and self-image is particularly 
interesting when considered from a SIT perspective. Research by McClean and Kalin 
( 1 994) focused on the matching of a gendered self-image to an occupation. They found 
that both men and women in male dominated occupations have more masculine self-
concepts than those in female-dominated occupations. McClean and Kahn interpret 
these findings as suggesting that individuals choose an occupation that matches their 
self-concept. However S IT would argue that the occupation that individuals have 
chosen contributes to their self-concept in such a way as to cause the similarities, 
between self-concept and occupation. 
Social Identity Summary and Research Themes 
47 
The relationship between differentiation identification and self-esteem is 
problematic for SIT. The majority of the research does not support the predictions of 
the theory, although previous research has focused too narrowly on identification and 
differentiation. Research indicates that discrimination and sex-role orientation may 
affect gender identification. S IT states that identification will be higher in high status 
groups as group membership makes a more positive contribution to their self-concepts. 
Research into women in male dominated occupations suggests that women experience 
more conflict between their occupational and gender identity than men do. The 
relationship between occupational and gender identity has not been studied in men 
working in female dominated occupations. There is also a scarcity of research into male 
gender identification. 
3.5 Sal ience of Identities 
It is possible to have many different social category memberships or identities. 
Salience is the prominence of one category membership over other possible categories 
in the self-concept. Turner ( 1 999) states that identity salience may be a mediating factor 
between identification and bias. 
SIT also states that the salience of group membership affects other group 
processes. These include group identification, self-stereotyping (Abrams, Sparkes & 
Hogg, 1 985;  Hogg & Turner, 1 987) speech styles (Hogg, 1 985), and self-esteem (lames 
& Greenberg, 1 989). Other researchers support a distinctiveness approach, which 
suggests that salience will be higher for numerical minorities. 
Theories of Identity Salience 
48 
Over the issue of category salience Social Identity Theory came into conflict 
with Distinctiveness Theory (McGuire & Padawer-Singer, 1 976, Taylor, Fiske, Etcoff & 
Ruderman, 1 978). Distinctiveness theory postulated that it is a basic perceptual 
tendency for novelty to be most salient, so numerically uncommon characteristics will 
be most salient. There were two groups of theorists in this area. Taylor and colleagues 
(e.g., Taylor, Fiske, Etcoff, & Ruderman, 1 978, Taylor & Thompson, 1 982) have 
studied distinctiveness in relation to the salience of characteristics of a target to a 
perceiver (Biemat & Vescio, 1 994). In contrast McGuire and colleagues (e.g. , McGuire 
and McGuire, 1 98 1 )  have focused on the effect the distinctiveness of a characteristic has 
on its salience in the self-concept. Studies have provided support for distinctiveness 
theory using a variety of characteristics. Taylor et al . ( 1 978) carried out a series of 
studies in which students watched slides and listened to recordings of groups 
interacting. The groups presented to the respondents were either mixed gender or mixed 
ethnicity. Salience was manipulated by the proportions of the gender or ethnic group 
members. They found that numerical distinctiveness (the fewer members of the group 
there were) led to higher ratings of prominence and assertiveness of the targets. 
Respondents also rated targets higher on stereotyped gender roles the more distinctive 
the target was. 
The salience of group membership in the self-concept has been widely 
measured. McGuire and colleagues (e.g., McGuire & Padawer-Singer, 1 976) 
developed the Spontaneous Self-Concept measure, which simply asks participants to 
(either orally or in writing) "tell us about yourself'. The information provided is then 
scored to assess whether gender or other category memberships are mentioned and how 
soon it is mentioned in comparison to other characteristics. The Spontaneous Self­
Concept measure has the advantage of being able to assess the relative salience of 
49 
different identities in the self-concept. The positive self-concept was also accompanied 
in later studies by the negation self-concept; which is the answer to the probe "tell us 
what you are not". They developed the negation self-concept because distinctiveness 
theory states that the most salient characteristics are the ones that are not shared by the 
majority. They assumed that in order to identify what characteristics are not common 
individuals must also be aware of what they are not. 
The research on distinctiveness theory has also studied whether ethnicity in 
children's self-concepts was affected by the numbers of their own ethnic origin in their 
class (McGuire, McGuire, Child & Fujioka, 1 978). Personal descriptive traits have also 
been studied, such as how salient hair and eye colour were in samples of school children 
(McGuire & Padawer-Singer, 1 976). The results of these studies supported 
distinctiveness theory's assumption that the more distinctive the trait or the ethnic group 
the higher the salience in the self-concept. McGuire, McGuire, and Winton ( 1 979) 
studied gender salience. Gender was found to be most salient for girls whose household 
(siblings and parents) was predominantly male and most salient for boys whose 
household was predominantly female. Distinctiveness theory was also supported by a 
study by Abrams, Sparkes and Hogg ( 1 985), into the salience of gender in families. 
They also found that the gender salience increased as distinctiveness of the individual 's 
gender within the family increased. 
S IT (Taj fel and Turner, 1 979) states that the presence of an out group member 
heightens intergroup difference and intragroup similarity, which leads to group 
membership being more salient for both the ingroup and the outgroup. So according to 
S IT it is the presence or absence of outgroup members not their relative numbers that 
makes category membership more or less salient. The presence of a solo member of one 
group will increase the group salience for both that solo member and the members of the 
majority group. SIT's position was supported by Oakes and Turner ( 1 986), who found 
that novel category memberships were not automatically more salient than common 
categories, but that their salience depended more on the significance of the 
categorization for the task carried out by the perceiver. Abrams et al . ( 1 990) found that 
category salience was influenced by the presence of the other group but was not affected 
50 
by the numerical proportions, as distinctiveness theory would predict. The predictions 
of SIT lead to hypotheses 3 and 4: "Individuals who have a member of the opposite 
gender in their immediate workgroup will have a higher self-concept gender salience 
than those who do not work with a member of the opposite gender "(H3). " The number 
of same gender colleagues or the distinctiveness of category membership will not be 
related to gender salience " (H4). 
There has been a great deal of debate between the proponents of these two 
theories. Oakes and Turner ( 1 986) criticize the distinctiveness hypothesis for focusing 
too exclusively on the numerical and cognitive aspects of salience and neglecting the 
social implications of group membership that may influence salience. For example, 
McGuire et al. ( 1 976) studied salience of traits such as eye colour and height that are not 
normally used as a basis for categorizations of group membership. 
Each theory of identity salience has research supporting its position. However 
much of the research on salience done within the SIT framework has focused on 
manipulating the salience of category membership for example by differing the numbers 
of each gender in a group, rather than measuring it in real l ife situations, and considering 
the conditions or contexts that make it increase or decrease (Oakes, 1 987; see also 
McCann, Ostrom, Tyner, & Mitchell, 1 985 and Oakes & Turner, 1 986). Due to the 
focus on manipulating salience in experimental work, researchers have not studied 
salience in the self-concept of the low status group member. There are also 
methodological differences that decrease the comparability between studies. 
Methodological Problems 
Research conducted into the salience of a target (McCann, Ostrom, Tyner and 
Mitchell, 1 985 ;  Oakes & Turner, 1 986; Taylor, Fiske, Etcoff and Ruderman 1 978) has 
involved manipulations of salience, of gender, race and other categories such as faculty 
membership and University attendance, in laboratory experiments. The gender salience 
of a target has most commonly been manipulated by changing the number of men and 
women in a group situation. Other researchers (Hogg & Turner, 1 987) however have 
5 1  
manipulated the gender salience by creating a group debate, which introduces 
competition to the group situation. In the low salience condition (intrasex) two 
members of same sex argued with each other, whereas in the high salience condition 
(intersex) two members of one sex argued with two members of the other sex. These 
inconsistencies in the methods used to manipulate gender salience makes comparisons 
between studies more difficult. 
The number of people used in each group to create a numerical minority has also 
varied. Some studies use five to one, (e.g., Oakes and Turner, 1 986) others use one to 
two (e.g., Cota and Dion, 1 986). Distinctiveness theory predicts that higher gender 
salience would be found in studies with a larger majority to minority ratio. Very few 
studies have measured the strength of the gender salience manipUlation that has been 
performed. There are also problems with using purely numerical proportions to 
manipulate salience, as there is no way of telling how the social status of the participants 
is affecting the situation. 
Another problem arises when research that has used real life situations is 
compared with laboratory based studies, where gender salience is manipulated. There 
has been very little research on whether identity salience is constant even for continuing 
real life contexts such as a classroom or home. Gender salience may fluctuate according 
to changes in the context. Other situational factors may increase the salience of a 
certain identity, for example the attitudes or gender of the teacher may influence gender 
salience in their pupils. Other intergroup processes may also affect salience for example 
Hoelter ( 1 983) found that salience increased as the commitment to that identity 
increased. 
Gender Salience 
Aspects and Definitions of Gender Salience 
As the self-concept is made up of many different identities, when gender identity 
is prominent over other identities it is called gender salience. It is necessary to clarify 
the different factors that are referred to as gender salience. Firstly, there is the salience 
52 
of the gender of a target person to an observer or onlooker. The second aspect is how 
salient gender is in the self-concept of a person. Thirdly gender salience can also refer 
to how likely a person is to use gender as a categorisation when making decisions or 
holding attitudes about others (Greenwood & Glidden-Tracey, 1 996). This type of 
gender salience is based on Bern's ( 1 98 1 )  gender schema work. 
The two main theories of identity salience - S IT and 'Distinctiveness Theory' 
have also considered the salience of gender identities. The SIT position on gender 
salience is that the presence of groups increases the salience of gender in both groups. 
The SIT position has been supported by studies examining the gender salience of a 
target (Oakes and Turner, 1 986) and gender salience in the self-concept (Abrams, 
Thomas and Hogg, 1 990). Abrams, Thomas and Hogg ( 1 990) carried out a study on 
female students who had previously been found to have traditional sex role ideologies. 
The researchers showed the students a sexist advertisement, which they viewed in 
groups of differing proportions of men and women. Their findings supported a SIT 
view of gender salience. Self-concept gender salience was higher in mixed groups than 
the all female groups, but the ratio of men to women did not affect salience. Traditional 
sex role attitudes were also higher for those with higher gender salience. SIT suggests 
that women will have higher gender salience than men as category membership will be 
more salient for the group of lower status, which leads to the following hypothesis, 
" Women will have higher self-concept gender salience than men - regardless of 
occupation "(H5). 
Distinctiveness theory states that how scarce gender is in the environment 
increases gender salience. Distinctiveness theory's position has also been supported by 
research, for example Abrams, Sparkes and Hogg ( 1 985) found that the self-concept 
gender salience of girls was most affected by the number of male siblings, and that self­
concept gender salience of boys was most affected by their number of sisters. They 
found that females with larger numbers of brothers had lower academic aspirations, and 
that males with more sisters had less stereotyped perceptions of the academic abilities of 
the genders. McGuire, McGuire and Winton ( 1 979) also studied gender salience in 
families. They found that boys had higher gender salience when females were in the 
53 
majority, and vice versa. Their findings supported distinctiveness theory, but they also 
discovered that the absence of a father increased boys' gender salience. Distinctiveness 
theory states that gender salience will be determined by numbers in the workplace which 
leads to the following hypothesis: "The minority in the workplace (female engineers and 
male nurses) will have higher self-concept gender salience than the majority (male 
engineers and female nurses) " (H6). 
A problem arises in the debate between the two theories of identity salience when 
comparisons are made between the different aspects of gender salience. Most of the 
researchers involved in the debate between distinctiveness theory and SIT neglect to 
consider the fact that two separate aspects of gender salience are being considered. For 
example, Oakes and Turner ( 1 986) claim that their study challenges the distinctiveness 
theory. In their study they presented respondents with tapes and slides of groups of 
different sex compositions interacting. Half of their respondents were told to assess the 
performance of an individual in the group and the other half were told to examine the 
whole group. For respondents focusing on the individual, salience was higher when there 
was a sole member of one sex. For the respondents focusing on the whole group salience 
was highest when there was an even sex distribution, which demonstrated that task focus 
was more important than distinctiveness in influencing the gender salience of a target. 
However, they do not consider that gender salience in the self-concept, which is what 
McGuire et al. ( 1979) have focused on may be influenced by different processes to those 
that influence the gender salience of a target. 
Crocker and McGraw ( 1 984) conducted a laboratory experiment and measured 
both self-concept gender salience, and gender salience of a target. They set up groups of 
six individuals in which the minority gender was the sole member of their gender. The 
majority members then rated the salience of the minority members' gender. They found 
that the gender salience of the female tokens was higher in the self-concept and as rated by 
the majority group than for the male tokens. SIT explains these results more easily than 
distinctiveness theory as aspects of intergroup status have influenced the effect of token 
status. Distinctiveness theory predicts no difference between male and female tokens. In 
contrast a study by Cota and Dion ( 1986) supported distinctiveness theory. They 
54 
studied gender salience in the self-concept using the spontaneous self-concept measure, 
in laboratory groups of differing sex compositions. 
Hurtig and Pichevin ( 1 990) conducted an experiment to assess the relationship 
between the context and the salience of gender to an observer. They state that in natural 
settings people are simultaneously perceivers of their self and others, as well as being a 
target for others perceptions. They also criticize research for ignoring the social 
meaning of the categories that can have an influence on salience. Similarly to Crocker 
and McGraw ( 1 984) they found that gender salience was higher for women than for 
men, especially when the context was sex-typed. 
The third factor referred to as gender salience has only been studied separately 
from the other two aspects; it is gender salience as a basis for the categorization of 
others. Glidden-Tracey and Wagner ( 1 995) looked at the effect of gender salience as a 
basis for attitudes and decisions within counselling situations. To do this they 
developed a measure of gender salience. There were methodological problems with the 
first measure of gender salience, because the measure focused on real events in the 
respondent's l ife, it had low generalisability and low variability. The problems with the 
first measure led them to develop a second measure that focuses on gender salience in 
specific contexts such as friendships, domestic tasks, and work situations (Greenwood 
and Glidden-Tracey, 1 996). Serbin and Sprafkin ( 1 986) studied the third type of gender 
salience in children. They examined children's use of gender-based categorisations and 
their use of gender to make affiliation choices. They related these to the development of 
sex role awareness. 
SIT appears to be the more useful theory to study gender salience. As it is a 
complete theory, S IT can be used to integrate the three different factors of gender 
salience. S IT suggests that an outgroup member's gender identity will be more salient 
to an observer from the ingroup, as the outgroup members presence increases 
differentiation between the groups. The presence of an outgroup member also increases 
gender salience in the self-concept of each group. When gender is salient in the self­
concept both ingroup and outgroup members will act according to group norms, which 
could make them more likely to use gender to categorise others, the third factor of 
gender salience. S IT predicts therefore that when gender is salient in the self-concept 
gender is more likely to be salient for categorising others, which leads to hypothesis 7:  
"Self-concept gender salience will be correlated with gender salience in categorising 
others IJ (H7). 
55 
S IT also suggests other aspects that may influence salience apart from numerical 
distinctiveness, such as intergroup threat or competition. SIT predicts that the presence 
of two groups in a workplace increases the salience of group membership for both 
groups, as found in a study by Abrams, Thomas and Hogg ( 1 990). 
Gender Salience in Gender Dominated Occupations 
Very little of the salience research has looked at gender salience in the 
workplace. The debate over the distinctiveness theory has led to the majority of the 
research focusing on numerical proportions and how numbers affect salience. However 
there may be other aspects of gender dominated occupations that influence gender 
salience. For example a policewoman may find being physically challenged raises her 
gender salience. Unfortunately, these have not been studied. Macke ( 1 98 1 )  is one of 
the few researchers to look at what she call s  gender salience in gender dominated 
occupations, using both men and women in her sample. She found that those working 
in the opposite gender dominated occupation did have heightened gender salience. 
However the method she used to measure salience is not the same as in other studies of 
gender salience. She measured salience by measuring how large a contribution sex­
related characteristics make to self-esteem - which is not a direct measure of gender 
salience. She compared these with the impact job-related characteristics such as 
intelligence had on self-esteem. Macke found that job related characteristics played a 
larger part in the self-esteem of participants in same sex occupations than did gender 
related characteristics. 
Kanter' s  ( 1 977) tokenism research demonstrates the SIT point of view of gender 
salience. She describes how in the organization she studied, the presence of the token 
group (in her case women) served to increase the tokens' and the majority group's 
56 
awareness of their gender. One of the three tokenism processes identified by Kanter 
( 1 977) is visibility or the feeling tokens have that they stand out. Visibility seems to be 
an awareness of the salience of their gender to an observer. Kanter's work is one of the 
few studies of gender dominated occupations that mentions salience. Simeone ( 1 987) 
claims that people have a heightened awareness of the sex of a token, but neglects to 
suggest what effect, if any, this would have on the self-concept of the token. Much of 
the work on tokenism processes mentions the heightened visibility of the token, and the 
attention that others in their workplaces give to their gender. S IT predicts that when 
others in the environment focus on one particular identity that will increase its salience 
relative to other possible identities. The fol lowing hypotheses are formed: "For male 
nurses and female engineers there will be an inverse relationship between occupational 
and gender salience. For male engineers and female nurses there will be a positive 
relationship between occupational and gender salience " (H8). Male nurses and female 
engineers will have higher gender salience than occupational salience. Female nurses 
and male engineers will have higher occupational salience than gender salience" (H9). 
Salience of Identities Summary and Research Themes 
The methods used to study salience have been inconsistent. For example, 
salience of a target to an observer has mostly been studied in laboratory situations where 
salience has been manipUlated. The laboratory studies have often been compared to 
studies that have measured identity salience in the self-concept of a minority group 
member. The studies are measuring different constructs that may act in different ways. 
Alternative factors that could be influencing identity salience such as discrimination 
have not been explored adequately. 
SIT theory appears to have more support for its predictions about gender salience 
than distinctiveness theory, especially when studying salience in the self-concept. 
Kanter's  research ( 1 977) suggests that low status or token members of the workforce are 
more visible than are the high status members, because their 'category membership i s  
emphasized. S IT predicts higher gender salience in  the token group's self-concept than 
57 
for the high status group. Infonnation about how gender salience in the self-concept 
relates to gender salience when categorising others has not previously been researched. 
3.6 Psychological Distinctiveness 
One of the major assumptions underlying SIT is that individuals strive to 
maintain or achieve a positive social identity (Tajfel & Turner, 1 979). This is achieved 
through belonging to a group that is positively valued. The process of social 
comparison is how individuals assess the relative value of their own group membership 
compared with other groups. In order to achieve a positive identity, individuals must 
see the ingroup as positively differentiated or distinct from the outgroups with which it 
is compared. One way groups achieve positive distinctiveness is by accentuating the 
differences between groups on the characteristics that are favourable to their group 
(Brewer, 1 99 1 ). This requires mutual comparison and differentiation on a value or 
characteristic shared by the groups (Tajfel & Turner, 1 979). Once the group achieves 
distinctiveness, the group must maintain positive distinctiveness. One way in which a 
group can maintain positive distinctiveness is through the use of signs and symbols that 
advertise their distinctiveness to other groups (Taj fel, 1 974). 
Groups experience more pressure to be distinct as the similarity of groups 
increases. S IT claims that groups need to be different from each other to be positively 
valued. Status differences can reduce the comparability between groups, as they 
decrease the similarity. However, status differences do not reduce comparability if they 
are perceived as changeable (Taylor & Moghaddam, 1 994). If an outgroup threatens the 
psychological distinctiveness of a group, the group wil l  be met with hostility and the 
ingroup will attempt to maintain distinctiveness (Taj fel, 1 974). 
Jackson, Sullivan, Harnish and Hodge ( 1 996) studied some of the ways in which 
groups favour the ingroup in situations of intergroup comparisons. In their experiment, 
they manipulated the positive or negative value of the groups by changing personality 
descriptions of the groups. They discovered that a negative characteristic was rated less 
undesirable when it distinguished the ingroup from the outgroup. The ingroup was also 
rated higher on all other characteristics when the ingroup was negatively distinct. 
Jackson et al. described these as efforts to enhance the positive distinctiveness of the 
mgroup. 
58 
Brown ( 1 995) states that the research into psychological distinctiveness 
indicates different motives for derogation of an outgroup and favouritism towards an 
ingroup, although SIT claims the motive for both outgroup derogation and ingroup bias 
is achieving positive distinctiveness. Support for Brown's claim is provided by Struch 
and Schwartz ( 1 989) who found that ingroup favouritism was unrelated to aggression 
towards the outgroup. They found that the predictors of aggression such as conflict, and 
permeability of group boundaries were not good predictors of ingroup bias. 
Hinkle and Brown ( 1 990) point out that sometimes ingroup favouritism occurs 
on some dimensions of a measure and not on others and argue that SIT cannot 
adequately explain this. Turner ( 1 999) questions the measures of ingroup bias that have 
been used. The researcher usually selects the dimensions of in group bias and does not 
test whether they are actually relevant to the respondents. SIT states that the relevance 
of dimensions of ingroup bias, and their centrality to the group influences how 
important they will be in determining group action. 
Brown ( 1995) claims that the postulated relationship between the similarity of 
the groups and the amount of threat they produce is not supported by the SIT research. 
However Turner ( 1999) rightly argues that Brown is being too simplistic in his analysis 
of the relationship. Turner states that SIT never stated that that threat was only due to 
similarity. Similarity is only one dimension of social comparison. In addition the 
amount of threat a similar group causes wil l  be mediated by the stability, legitimacy and 
the relative status of the groups. 
Gender and Psychological Distinctiveness 
The process of social comparison is how individuals assess the relative value of 
their gender group membership compared with the outgroup. In order to achieve a 
positive identity from gender group membership the gender must be seen as positively 
distinct from the outgroup. SIT expects that the more similar the gender groups see 
each other as being the more of a threat they will perceive each other to be. As 
individuals interact so often with members of the opposite gender, it seems that they 
would be a relevant outgroup for social comparisons. However there is l ittle research 
that has looked into how gender groups choose the relevant groups for social 
comparisons. 
59 
SIT provides a good explanation for the findings of Wharton and Baron ( 1 987, 
1 99 1 ), who found that the individuals in gender dominated occupations were least 
satisfied when the numbers of men and women approached even. SIT can explain this 
as a minority outgroup does not threaten the dominant group, and the dominant group 
may still control entry into the occupation. However, as the numbers of the minority 
group increase they may cause more of a threat to the positive distinctiveness of the 
dominant group, increasing dissatisfaction for the dominant group. 
Psychological Distinctiveness Summary 
Individuals attempt to achieve and maintain a positive social identity, which 
occurs through membership in a group that is positively distinct from other comparable 
groups. The process through which the value of the ingroup relative to outgroups is 
assessed is social comparison. Similarity between groups causes threat to the 
psychological distinctiveness of the groups. For gender dominated occupations 
increasing numbers of the minority group may be perceived as more of a threat than a 
small minority group. 
3.7 Social Comparison Processes 
Tajfel ( 1974) saw social comparison as the link between social categorizations 
and social identity. Group membership therefore only takes on value and meaning 
when it is evaluated in relation to other groups. Tajfel wrote, "a group becomes a group 
in the sense of being perceived as having common characteristics or a common fate only 
60 
because other groups are present in the environment" (Tajfel 1 974, p. 72). Tajfe1 ( 1 975) 
also suggests that what a group is not can play an important part in defining a group' s  
identity. For example "very often we are what we are because, ' they' are not what we 
are" (p. 1 08). There must be an element of consensus about these categorizations and 
value judgments in other words they must be shared socially. Tajfel stated that all 
information is socially defined. 
SIT states that the ingroup does not compare itself to all available outgroups on 
all factors, but compares itself to relevant outgroups on mutually valued attributes and 
characteristics (Tajfel & Turner, 1 979). Criteria that have significance for the group and 
that define and differentiate the groups will be used most often for comparison. 
Likewise, similarity, proximity, and situational salience of outgroups affects a group's 
choice of outgroups for comparison. Turner, Brown, and Tajfel ( 1 979) state that the 
dimension, on which the groups are compared, must be important, relevant, or salient. 
When the comparison results in a positive evaluation for the ingroup, this creates 
pressure on the ingroup to maintain its position. However ifthe comparison leads to a 
negative evaluation for the ingroup this results in low prestige for the ingroup (Taj fel & 
Turner, 1 986). 
Hinkle and Brown ( 1 990) point out some problems with SIT's description of 
social comparisons. They suggest that it is impossible to know which social 
comparisons are important for groups' social identity because previous research has 
found more in-group favouritism on some comparisons than others, and favourable out­
group comparisons can and do occur. They also question whether social comparisons 
happen as spontaneously as Tajfe1 ( 1 974) claims. For example in a study of groups in a 
bread making factory Brown and Williams ( 1 984) found that the respondents in their 
study showed little evidence of making spontaneous comparisons between the groups. 
SIT focuses on comparisons between groups, but other researchers have claimed 
that individuals make more comparisons with other individuals within their group, than 
with other groups (Major, 1 994). Major ( 1 994) argues that this is especially common 
when individuals are assessing what they deserve, because of the greater availability and 
diagnosticity of the other ingroup members. 
6 1  
Social Comparison Processes Summary 
In order to assess the ingroup's positive distinctiveness social comparisons with 
other relevant groups are made. The outcome of social comparison is information about 
the status of the groups and the intergroup situation. 
3.8 Bel iefs about the Intergroup Situation 
The type of action a group takes to maintain or achieve positive distinctiveness 
will be influenced by the group's beliefs about the intergroup situation. Their belief 
about the status, permeability, stability, and legitimacy of the intergroup relations 
determines the actions the group will take. 
Status 
One of the outcomes of intergroup comparison is information about the relative 
status of the groups. Tajfel and Turner ( 1 979) state that status "reflects a group's 
relative position on some evaluative dimension of comparison" (p. 43 ,  emphasis added). 
Tajfel and Turner's definition implies that group status can only be determined by social 
comparison. The status of the group an individual belongs to determines whether that 
group will make a negative or positive contribution to the self-concept and the self­
esteem of the group members. 
There is some confusion about the use of the terms ' low status' ,  'disadvantaged',  
'minority,' and ' subordinate' .  These are generally used interchangeably as if their 
meanings are the same, which causes some confusion, as a low status group such as 
women can be the largest in number. Women, make up 5 1  % of the population 
(Statistics NZ, 1 999), but in many respects remain a disadvantaged group and are 
sometimes for this reason described as a minority group. A minority group may or may 
not be a disadvantaged group. Taylor and Moghaddam ( 1 994) claim that some of these 
terms suggest that the group's position is attributable to their own characteristics. They 
62 
suggest that using advantaged and disadvantaged should be the preferred tenns, as these 
do not logically imply internal or external causes for their positions. In the present 
research, high and low status will indicate the groups' perception of their relative 
advantages and disadvantages and majority and minority will describe numerical 
proportions. 
S IT states that a group of high status should be positively distinct from the 
out group and this should provide group members with a positive identity. Ellemers, 
Doosje, Van Knippenberg, and Wilke ( 1 992) found that participants in their research 
saw the high status group as more attractive than the low status group. Ellemers, van 
Knippenberg & Wilke ( 1 990) also found that low status groups were more dissatisfied 
with their group membership than high status groups. 
Ellemers et al . ( 1 992) studied the difference between low status and minority 
groups. They found that minority status does not immediately reduce the positive 
distinctiveness of a group. However there is an interaction between status and numbers. 
The respondents rated a high status minority the most attractive group; Ellemers et al . 
termed this an elite group. They found that the members of high status groups were 
more proud of their group membership, had higher identification with the ingroup and 
were less attracted to the outgroup than low status group members. The high status 
minority members were more proud of their group membership and more inclined to 
work as a group. However, when the ingroup was a majority there was no difference 
between the low and high status responses, a low status majority group was no more 
attractive than a low status minority group. In addition to numbers status effects can 
also be affected by favourable treatment of the groups by a third party. Commins and 
Lockwood ( 1 979) found that favourable treatment of a group reduced ingroup bias. 
Favourable treatment did not however change the perceived status ofthe groups. 
A group member will make comparisons between the group that he or she 
belongs to and other relevant and similar groups. Comparisons lead to evaluative 
infonnation about the group such as its relative status. Status is not merely determined 
by numbers but by other factors such as power and resources. Low status threatens the 
groups' positive distinctiveness and leads to action. High status makes a positive 
63 
contribution to group members identity, so members of high status groups identify more 
strongly with the ingroup than do members of low status groups. These assumptions 
from SIT lead to the following hypothesis for the current study: "Respondents who see 
themselves o/high status will have higher gender identification than respondents who 
see themselves o/low status"(H I O). 
Gender and Status 
The result of social comparison is information about the relative status of the 
gender groups. According to SIT, being a numerical minority or majority does not 
automatically determine status, but if a gender group perceives their position as low in 
status they will take action to achieve positive distinctiveness. SIT is particularly 
helpful when it comes to studying the differences between males and females in 
opposite gender occupations because the theory takes more into account than just 
numerical minority and majority when assessing status. 
Despite women making up 5 1  % of the population, they are often considered by 
researchers as having lower status than men have (e.g., Swan & Wyer, 1 997). They are 
assumed to be of low status due to the history of oppression by men, their lack of 
economic power and their under-representation at decision making levels. Research 
findings suggest that some women, especially younger women, now consider that 
women have achieved equality with men in society (Cockburn, 199 1 ,  Condor, 1 989). 
Women in male dominated occupations are assumed to be of lower status than men, due 
to the history of absence from the professions and to the fact that men remain in the 
decision-making positions of these professions. Men entering female dominated 
occupations are in a different position. Fairhurst and Snavely ( 1 983b) explain the 
position of men as a change in the status configuration and state that "the numerically 
derived organizational environment would favour females, and the sex based hierarchy 
of the larger society would continue to favour males " (p. 354). Fairhurst and Snavely 
state that societal status will affect males' perceptions of their status in female 
dominated occupations. Males' position in female dominated occupations is also 
64 
enhanced by the fact that men are often in decision making or supervisory positions in 
these occupations, men are more likely to have a same gender boss than women in male 
dominated occupations (c. Williams, 1 992). The fol lowing specific hypotheses are 
derived from the research outlined above: "The majority offemale engineers will see 
themselves as low status, and the majority of male engineers will see themselves as high 
status " (H 1 1 ). If numbers of each gender determine status within an occupation then 
"The majority of male nurses will see themselves as low status, and the majority of 
female nurses will see themselves as high status " (H I 2). On the other hand, if societal 
status plays a part in the status of a gender group within an occupation " The majority of 
female nurses will see themselves as low status and the majority of male nurses will see 
themselves of high status " CH 1 3). 
Permeability 
Permeability is a concept used in SIT to describe the boundaries between groups. 
If the groups have permeable boundaries members can freely join a different group, if 
the boundaries are impermeable group membership cannot be changed. If group 
members see the intergroup situation as permeable, they will adopt individual strategies 
for improvement (Lalonde & Silverman, 1 994). On the other hand, if members of a low 
status group perceive the group boundaries to be impermeable then they will adopt a 
strategy that involves change for the whole group. Permeable boundaries can help 
legitimise the dominant group's position, as permeable boundaries demonstrate that 
entry to the dominant group can be achieved through effort or ability, which discourages 
group action (Lalonde & Silverman, 1 994). As members of the gender groups do not 
commonly become members of the other group acceptance will be measured in the 
current study rather than true permeability, therefore permeability will be referred to as 
acceptance. The high status group is expected to protect its position therefore in the 
present study: "High status respondents will express lower acceptance of the outgroup 
than the low status respondents, and the low status respondents will perceive lower 
acceptance of the ingroup than will the high status respondents " (H1 4) .  
65 
SIT predicts that the penneability of group boundaries will affect identification 
with the ingroup, if the boundaries are penneable group members' identification will be 
lower (Tajfel & Turner, 1 979). Research by Ellemers et al . ( 1 990) supported this 
prediction. Members of groups with penneable boundaries identified less with the low 
status ingroup and more with the high status outgroup, especially group members with 
high ability. Ellemers et al . ( 1 992) found that when the intergroup situation was 
penneable the high status minorities showed relatively high levels of identification with 
the ingroup. When the conditions were penneable ingroup identification was stronger 
in minority groups than majority groups, and when the boundaries were impenneable 
there was no difference in identification. Their research leads to the fol lowing specific 
hypothesis: "Gender identification will be negatively correlated with perceived 
acceptance oJ the ingroup " (H 1 5). 
S IT also predicts that the penneability of group boundaries will affect which 
strategy is chosen for achieving or maintaining group distinctiveness. Research into the 
effects of penneability of group boundaries on the choice of strategies to deal with being 
a group of low status (Ellemers, van Knippenberg, & Wilke, 1 990; Jackson, Sullivan, 
Harnish, & Hodge, 1 996; Lalonde & Silvennan, 1 994) has tended to involve 
manipulating the penneability of group boundaries. These studies have usually taken 
place in laboratory groups. 
Struch and Schwartz ( 1 989) attempted to measure the perceived penneability of 
two natural ethnic groups. They used three items to measure penneability. The first 
item dealt with distinguishing members of the two groups from each other. The next 
two items asked respondents how easy it was to move from one group to the other. 
They found that perceived conflict decreased the penneability of group boundaries and 
increased aggression towards the outgroup. 
J ackson et al . ( 1 996) manipulated penneabi lity of group boundaries in real life 
groups by presenting infonnation to the respondents about the ease of stopping smoking. 
They found that the positively distinctive group members - non-smokers, saw 
themselves as more similar to ingroup members and rated themselves more desirable on 
all dimensions, but these ratings were unaffected by the permeability of group 
boundaries. 
Gender and Permeability 
66 
S IT theorists seem to have conflicting views over the permeability of the gender 
groups; theorists l ike Brown ( 1 995) claim that groups based on gender and ethnicity are 
impermeable. Hogg and Abrams ( 1 988) state that passing from a gender group is very 
difficult as the membership is externally or physically designated. However Hogg and 
Abrams ( 1 988) also suggest that social mobility beliefs are "highly characteristic of 
women seeking traditionally male-occupied positions in organisations" (p. 56). 
Women in these situations may see the boundaries as permeable, or may not realize that 
they are impermeable. Vaughan and Hogg ( 1 995) suggest that individual 's perceptions 
of the permeabi lity of group boundaries may not necessarily reflect reality. Women who 
have been successful in male dominated occupations often try to fit in by de­
emphasizing their gender and taking on the characteristics of the dominant group. 
Baker-Miller ( 1 976) found that "some women stil l  try to imitate the dominant group by 
gaining status and power at the expense of subordinates, often now other women" (p. 
1 35). Taking on the characteristics of the dominant group results in behaviour such as 
taking part in jokes at the expense of their own group, and may explain why some 
women who are successful in male dominated occupations do not support other women 
coming into the organisation, because they are not identifying with women but with the 
dominant group. S IT would classify these beliefs as social mobility, which is associated 
with low ingroup identification. 
It may be that although the boundaries between groups are not permeable in that 
members of one gender group commonly become members of the other, that it i s  
possible for one group to accept the members of  the other group to the extent that the 
categorization into groups is no longer salient. Acceptance is similar to the concept that 
Tajfel ( 1 975) termed assimilation, which is when a whole group is accepted into the 
outgroup. As members of the gender groups do not commonly become members of the 
other group acceptance will be measured in the current study rather than true 
permeability, therefore permeability will be referred to as acceptance in this study. 
67 
From the research on gender dominated occupations that has found that women 
are more accepting of men in female dominated occupations (Fairhurst & Snavely, 
1 983; Kanter, 1 977) the following hypothesis is derived: " Women will be more 
accepting of men in female dominated occupations. than men will be of women in male 
dominated occupations" (HI 6). 
Stability 
The stabil ity of the groups' positions, affects the choice of strategy to maintain 
or achieve positive distinctiveness. Stability is the perception of whether the groups' 
situation will change in the future. SIT predicts that if the low status group can perceive 
alternatives to the present situation they are likely to choose strategies to change their 
group's position (Tajfel & Turner, 1 979). ' Stable' and 'unstable' are terms now used to 
refer to the group positions but originally Tajfel ( 1 974) termed these 'secure' and 
'insecure ' .  When the low status group perceives the situation to be stable, they are 
l ikely to choose strategies that change the status of the group without confronting the 
out group (Tajfel & Turner 1 979). 
Ellemers et al . ( 1 990) found that an unstable group situation resulted in 
increased ingroup identification for the members of the low status group, but not for the 
high status group, which reflected the group's concern for improvement of group status 
as a whole. Ellemers et al . also found that members of a low status group under stable 
conditions favoured an alternative status criterion for evaluation, which is evidence of a 
social creativity strategy. 
Taylor and Moghaddam ( 1 994) state that although SIT assumes that perceived 
instability would lead to greater intergroup discrimination, research findings do not 
support this. Taylor and Moghaddam oversimplified the SIT prediction. S IT does not 
predict a simple relationship between stability and discrimination. Finchilescu and de la 
Rey ( 1 99 1 )  suggest that the effects of the legitimacy of status relations take precedence 
over the influence of stability. Tajfel ( 1 975) originally believed that legitimacy and 
stability would affect each other, which implies an interaction between legitimacy and 
stability. He stated that there would be greater discrimination from an illegitimately 
high status group when the status relations were unstable than stable. 
68 
Research on the stability of intergroup relations has focused on manipulating the 
status and stability of groups in laboratory settings. Finchilescu and de la Rey ( 1 99 1 )  
used natural groups i n  their research and measured status and stability on an intergroup 
perception ladder, which requires respondents to rank the different groups on 1 1  rungs. 
Their findings did not support their hypotheses that the low status group would show the 
least ingroup favouritism under the stable-legitimate condition, and the high status 
group would show the most discrimination under unstable-legitimate conditions. They 
attempted to explain the failure of the results to support their hypotheses, by questioning 
the accuracy of their measure of stability. Finchilescu and de la Rey ( 1 99 1 )  suggest that 
the notion of stability should be continuous rather than a dichotomous variable as 
conceptualised in SIT. They claim that groups have varying positions of stability in 
relation to each other and that the relations between stability and legitimacy are not 
clear. Although Turner ( 1 999) has since claimed that that stability and legitimacy will 
interact in their effect on bias and differentiation. Finchilescu and de la Rey also claim 
that "there may also be qualitative differences in the perceptions of stability and 
legitimacy when they reflect continuous life experiences, in contrast to a state induced 
for an experimental session" (p. 230). No research has explored the qualitative 
difference in these factors between real and created groups. 
Taylor and Moghaddam ( 1 994) state that research findings do not always show 
that instability leads to discrimination. However researchers often fail to take into 
account mediating factors; such as identification and that S IT originally viewed an 
interaction between legitimacy and stability (Tajfel, 1 975). The research findings could 
also be due to some of the methodological limitations in the measurement of stabi lity. 
In addition discrimination is not the only option for the group to achieve positive 
distinctiveness, a group can also use social creativity or group members can achieve 
positive distinctiveness through social mobility strategies. 
69 
The different beliefs about the intergroup situation affect each other. Ellemers, 
van Knippenberg and Wilke ( 1 990) found that respondents perceived the group's 
position as fairer when they perceived the intergroup situation as unchanging (stable) 
and the group boundaries as permeable. For this reason it is expected in the current 
study that "Stability beliefs will not be independent of legitimacy beliefs" (H 1 7). 
Gender and Stability 
Many change to the relations between the gender groups have occurred in the 
last century, such as the increasing numbers of women in paid employment. These 
changes indicate that the intergroup situation between men and women may be unstable. 
The SIT definition (Tajfel & Turner, 1 979) of stability is that alternatives to the current 
situation are imaginable. There is no research into the perceptions of men and women 
of the stability of gender relations. The history of change however does suggest that 
changes will easier to imagine. 
Legitimacy 
The legitimacy of the group's positions are how fair or justified they are 
perceived to be. If the status relations are i l legitimate, the groups are more likely to take 
action to improve or maintain the group's positive distinctiveness, than if they are 
legitimate. Tajfel ( 1 975) predicts that beliefs about legitimacy affect the actions of the 
high status group as wel l  as the low status group. If the high status group believes that 
they are legitimately of high status, they will be discriminatory towards the low status 
group. 
Tajfel ( 1 975) described three types of legitimacy: legitimacy as perceived by the 
ingroup, legitimacy as perceived by the outgroup and lastly legitimacy as measured 
against some objective standard, for example by rules and regulations. Tajfel ( 1 975) 
stated that action to achieve positive distinctiveness for the ingroup was more l ikely the 
70 
larger the discrepancy between the first two types of legitimacy. To date, no published 
research has tested this part of the theory. Tajfel ( 1 981 )  also expected an interaction 
between stability and legitimacy. He stated that unstable situations are more likely to be 
perceived as illegitimate, and if a situation is perceived as illegitimate, it can lead to 
perceptions of instability. 
Finchilescu and de la Rey ( 1 991 ) measured the legitimacy of the group relations 
in real life groups by measuring where respondents thought two ethnic groups should 
ideally stand on an intergroup perception ladder. Respondents who placed whites 
higher than blacks saw the relations as legitimate, respondents who placed blacks on the 
same or at a higher level than whites saw it as illegitimate. Finchilescu and de la Rey 
found that the effect of legitimacy was enhanced if the intergroup situation was also 
perceived as unstable. They found that members of the advantaged group (whites), who 
viewed the status relations as illegitimate gave less discriminatory responses than those 
who perceived them to be legitimate. 
El1emers, Wilke and van Knippenberg ( 1 993) found that low status group 
members accepted their position more when the situation was legitimate. They 
manipulated legitimacy in their experiment by assigning low status to the groups on the 
basis of poor performance (legitimate), or number of items completed (illegitimate). 
They also found that ingroup identification increased when the group members were 
illegitimately of low status. Legitimacy also affects group members' social change 
beliefs. In Ellemers et aI . ' s ( 1 993) study permeability and stability were the strongest 
predictors of participants' social change beliefs. Legitimacy however had more 
influence than permeability and stability on the respondents' evaluation of the group 
situation and identification with the group. 
Moghaddam and Perreault ( 1 992) carried out a study on ethnic groups and found 
increased individual action was associated with higher beliefs in the legitimacy of the 
situation. Turner and Brown ( 1 978) found legitimacy also affected other group 
processes such as ingroup bias and satisfaction. They found that members of a group 
with legitimately low status were more satisfied with their position. They also found as 
predicted by SIT that identification with the group increased if the group was 
illegitimately of low status. 
7 1  
Skevington ( 1 98 1 )  measured legitimacy by providing respondents with 
information about the working conditions of two groups of nurses (this covered aspects 
such as pay and holidays) they then rated their agreement with seven statements relating 
to the information. In contrast to predictions from SIT, she found no difference in 
perceptions of legitimacy between the high and low status group. In the current study, it 
is expected that "More high status respondents will perceive the situation to be stable 
and legitimate than low status respondents " (H I 8) 
Gender and Legitimacy 
The perceptions of legitimacy of the intergroup relations between the genders 
play a big role in determining the action if any that wil l  be taken by the groups. 
However there is very little research on individuals' perceptions of the legitimacy of the 
intergroup relations between genders. There has been very little research into the beliefs 
about intergroup situation among men and women. 
Beliefs about the Intergroup Situation 
Summary and Research Themes 
Low status threatens the groups' positive distinctiveness and leads to action. 
Being of high status makes a positive contribution to group members identity, so 
members of high status groups identify more strongly with the ingroup than members of 
low status groups. The status of men in society is assumed to be high because of their 
advantaged access to resources and power, women are assumed to have low status. 
Societal status will affect the experiences of minorities in gender dominated 
occupations. 
Permeability is the perception that intergroup boundaries can be crossed. 
Permeability affects the level of identification with the ingroup and the choice of action 
72 
the groups will take to maintain or achieve positive distinctiveness. Permeability ofthe 
gender groups is not possible in the sense that becoming a member of the other group is 
common but assimilation may be possible, so a measure of acceptance was used in the 
present study. 
The stability of the intergroup situation is the perception that changes in the 
intergroup situation are possible. Stability affects the choice of strategy to maintain or 
achieve positive distinctiveness. Social change beliefs will also be influenced by the 
legitimacy of the group's position. There is very little research on the stability or 
legitimacy of the intergroup situation between genders. SIT predicts an interaction 
between legitimacy and stability. High status group members are more likely to see the 
situation as legitimate as that is less threatening to the positive distinctiveness of their 
group. Research carried out mainly on laboratory groups has upheld some of SIT 
theories predictions about these factors. Permeability, stability, and legitimacy all affect 
social change beliefs. It is unclear which of these factors is most important in 
determining beliefs. 
3.9 Social Change Belief Systems 
-Achieving Positive Distinctiveness 
SIT proposes that membership in a low status group results in low self-esteem 
and a negative self-concept for members because of the loss of positive distinctiveness. 
The theory predicts that members will try to gain a positive self-concept. The type of 
action the group takes wil l  be determined by the group's social change belief systems. 
The belief systems are social mobility, social creativity, and social competition. The 
belief system that the low status group members will choose depends on their beliefs 
about the intergroup situation and is mediated by the level of ingroup identification 
(Turner, 1 999). Taylor and Moghaddam ( 1 994) also claim that SIT neglects the 
possibility for simultaneous use of more that one strategy to achieve or maintain positive 
distinctiveness, as Tajfel and Turner present the options as a choice between strategies. 
73 
Gender and Social Change Belief Systems 
Williams and Giles ( 1 978) used SIT to consider the way women have as a total 
group made efforts to change their situation, focusing on the feminist movement and the 
changes it has initiated. Breinlinger and Kelly ( 1 994) suggest that because SIT is one of 
the few theories that makes predictions about individual responses to belonging to a low 
status group it provides a useful framework for considering women's reactions to being 
a disadvantaged group. They used a Q-sort methodology to categorize women's 
responses to 'status inequality' into the social change beliefs systems from SIT. These 
included individual mobility, social creativity (including new dimension, changing the 
values, changing the outgroup) and social competition. They found that women's 
responses are more complex than SIT allows. They found a fourth belief that was a 
combination of an awareness of discrimination for some women, and an 
acknowledgement that some women succeed through individual strategies of hard work 
and taking opportunities. As there is so little research in this area it is difficult to 
conclude whether this demonstrates that SIT is too simplistic in its perception of social 
change beliefs or whether it may be due to the more qualitative methodology used in 
this study. Marshall ( 1 99 1 )  also looked at the relationship between identification and 
social change beliefs in a sample of women lawyers. The women in her sample who 
identified strongly with women fel l  into two categories, one group had a feminist 
perspective, which she categorised as social competition beliefs. The other group with 
strong gender identity had traditional values that had a negative impact on the opinions 
of women's success as lawyers. There were also three groups of women who 
disassociated from their gender group and focussed on occupational position, which 
Marshall termed a social mobility perspective according to S IT. 
74 
Social Mobility 
The social mobility belief system involves the group members of a low status 
group believing that efforts to improve their status individually by leaving their original 
group and joining the advantaged group are possible. Social mobility does not improve 
the status of the original group. Assimilation of a whole group into the outgroup is also 
possible. According to SIT the use of this social change belief system is dependent on 
the individual perceiving the group boundaries to be permeable. An individual may also 
disassociate themselves from a group psychologically if physical boundaries make 
social mobility impossible. In the current study it is expected that for the low status and 
equal status respondents "Acceptance of the ingroup will be positively related to 
ingroup social mobility beliefs "  (H 1 9). 
Social mobil ity usually involves high status members of the low status group 
who will increasingly identify with the advantaged group and disidentify with the 
original group, which weakens group bonds and decreases the likelihood of groups 
taking action as a group (Tajfel & Turner, 1 986). SIT predicts that identification with 
the group will be low for those with social mobility beliefs, the fol lowing hypothesis is 
formed for the high and low status groups: "Identification will be negatively related to 
social mobility beliefs "  (H20). 
Lalonde and Silverman ( 1 994) studied the effects of permeability on 
respondents' actions to improve the group or individual status. They manipulated the 
permeability of groups in a laboratory setting, so that there were three levels of 
permeability. These were open, closed and token in which the experimenter informed 
the low status group, unsophisticated decision makers, that the high status group, 
sophisticated decision makers, would accept 2% of unsophisticated members for entry. 
Their results supported the prediction from SIT that individual strategies such as social 
mobility were preferred when the boundaries were permeable. In addition they found 
that under the token condition the low status group members were less inclined to take 
collective or individual action to gain membership in the high status group, but showed 
75 
high levels of acceptance of the situation. These findings have quite important 
implications for situations in which there are token group members as they demonstrate 
that acceptance of tokens into the high status group reduces the l ikelihood of action 
against the dominant group. 
Moghaddam and Perreault ( 1 992) considered the choice between individual and 
collective mobility strategies among immigrants to Canada. They measured strategy 
use by presenting respondents with two statements that represented a collective 
orientation and two that represented an individual orientation. Their research supported 
the SIT prediction that individual action (a social mobility belief system) was related to 
belief in the legitimacy of the status relations. The SIT prediction leads to the following 
hypothesis: "Respondents of low status who see the situation as legitimate will have 
higher social mobility beliefs than those of low status who see the situation as 
illegitimate " (H21 ). Respondents of low status who see the intergroup situation as 
stable, will have higher social mobility beliefs than those of low status who see the 
situation as unstable " (H22). 
Jackson et al. ( 1 996) found in their minimal group experiment that permeabil ity 
did not influence the choice of social mobility belief systems. They inferred social 
mobility beliefs from individuals' level of perceived similarity to the ingroup. Jackson 
et al . also studied natural groups - smokers and non-smokers. They found that 
individuals in the negatively distinct group saw themselves as more similar to the 
jngroup when the situation was permeable. Their findings are contrary to the SIT 
prediction that identification with the outgroup will increase and identification with the 
ingroup wil l  decrease for those with social mobil ity beliefs in permeable situations. 
They explain their findings by suggesting that temporary group membership could be 
less threatening to positive distinctiveness than when group membership is permanent 
under impermeable conditions. 
Gender and Social Mobility 
Evidence of social mobility beliefs are found in much of the l iterature on 
tokenism� especially in descriptions of women who are successful through their own 
76 
merit and do not support others. Women with this social change belief try to succeed in 
male environments and do not perceive this in gender competitive terms. Women who 
use social mobility as a strategy tend to explain differences between the sexes as a 
matter of preference and regard the boundaries between sexes as symbolic. These 
characteristics allow the dominant group to explain their position as due to their 
personal characteristics (Hogg and Abrams, 1 988). 
Breinlinger and Kelly ( 1 994) found clear support for social mobility beliefs in 
their study on status inequalities among women. These beliefs were characterized by 
belief in individuals' abilities and the perception that inequalities between the genders 
did not exist. Social mobility beliefs were also associated with a belief that the 
intergroup situation was legitimate and an opposition to group action. 
As individuals with this belief system focus more on individual characteristics 
and the boundaries between groups are less important, they should also have lower 
gender salience, and perceive less gender discrimination than individuals who choose 
other strategies. 
Social Creativity 
If the disadvantaged group perceives the status relations as stable and or 
legitimate and impermeable then the preferred social change belief system is that of 
social creativity. The predictions of SIT lead to the fol lowing hypotheses for the low 
status and equal groups "Acceptance will be negatively related to ingroup social 
creativity beliefs "  (H23). "Respondents who see the intergroup situation as stable will 
have higher social creativity beliefs than respondents who see the intergroup situation 
as unstable " (H24). Social creativity encompasses three different types of response: 
77 
Redefining the Value of Characteristics 
Redefining the value of characteristics involves group members revaluing 
characteristics that were previously given a negative value in a positive way. Effort is 
required to make the outgroup accept the new value of the characteristic. Redefining the 
value of a characteristic is most successful when it is central or criterial to the low status 
group (Hogg & Abrams, 1 988). There are two stages necessary for the revaluing an 
existing characteristic: agreement among the ingroup and acceptance by the outgroup of 
the new value (Tajfel, 1 98 1 ). lackson et al. ( 1 996) conducted a minimal group 
experiment, in which the respondents were categorised into groups on the basis of 
whether they were dot underestimators or overestimators. They measured three things: 
similarity to the ingroup, the desirability of the groups distinguishing characteristics and 
ratings of the ingroup compared to the outgroup on those characteristics. They found 
that when the situation was impermeable the respondents tended to change the value of 
the dimension that caused their negative distinctiveness, and rated the ingroup more 
favourably on all other dimensions, which are examples of social creativity strategies. 
Using a New Dimension for Comparison 
A new dimension of comparison on which the low status group is better can be 
used as the basis for comparison. The classic experiment demonstrating this aspect of 
social creativity, was conducted by Lemaine ( 1 974). He found that a group of children 
given resources of a lesser quality than another group with which to build a hut 
emphasized the fact that they had built a garden around their hut, to the other group and 
to people judging the huts. Mumrnendey and Simon ( 1 989) found that although groups 
sti l l  tended towards ingroup favouritism on dimensions equally important to the in and 
outgroup, their laboratory groups tended to show outgroup favouritism when the 
dimensions were unimportant to the ingroup but important to the outgroup. 
78 
Changing the Outgroup 
A group can find new relevant comparison groups with which to compare itself. 
If the group can find other groups with even lower status with which to compare itself, it 
appears more favourable. Hogg and Abrams ( 1 988) state that the comparison must be 
to relevant outgroups, usually one that is similar in outlook and that holds a similar 
status to the ingroup. Individuals may also compare themselves with less advantaged 
members of the ingroup. 
Gender and Social Creativity 
Groups will often choose social creativity beliefs when the intergroup situation 
is stable. It consists of redefining the value of characteristics, using a new dimension of 
comparison and changing the outgroup. One example of social creativity is the 
exaggeration of existing differences between gender groups on dimensions valued by 
women. Breinlinger and Kelly ( 1 994) found examples of social creativity strategies 
among the women in their sample. They found that social creativity was characterised 
by women's traditional contributions to society being highly valued. Although there 
was awareness among the women with these beliefs that other women face 
discrimination. 
Social Competition 
When the status positions of the groups are unstable and/or i l legitimate and the 
boundaries are impermeable, the most l ikely social change belief system for the low 
status group is social competition. Social competition is a collective group effort to 
improve the status of the ingroup, which usually involves direct conflict with the 
dominant group. Action based on this belief system results in the most change 
occurring in the status of the disadvantaged group. SIT predicts that group members 
who see the group situation as stable will be less likely to choose social competition as a 
strategy for change, as they perceive no alternatives to the current situation. The 
79 
following hypotheses for the low and equal status respondents are drawn from the SIT 
predictions: "Acceptance will be negatively related to social competition beliefs "(H25). 
"Respondents who see the intergroup situation as illegitimate will have higher social 
competition beliefs than will respondents who see the situation as legitimate. " (H26), 
and Respondents who see the intergroup situation as unstable will have higher social 
competition beliefs than will respondents who see the situation as stable "(H27). 
Ellemers et al. ( 1 993) measured use of individual or collective strategies by 
whether participants choose to keep their individually earned points or donate them to 
the group. They found that collective group action was favoured under unstable 
intergroup conditions. Ellemers et al . ( 1 990) asked respondents whether they intended 
to try to achieve a higher status position for the in-group and if they wanted to compare 
their answers with those given by the high status group. They found that under unstable 
conditions collective action was the most popular social change strategy, and the low 
status group members identified strongly with the ingroup in this situation. S IT predicts 
that social competition is characterised by identification with the group "Identification 
will be positively related to social competition " (H28). 
Gender and Social Competition 
Social competition beliefs result in direct conflict between the gender groups. 
Individuals with these beliefs see the intergroup situation as unstable - they can imagine 
alternatives to the present situation. Breinlinger and Kelly ( 1 994) found that social 
competition beliefs among the women in their sample were characterized by anger at 
inequalities between the genders, a perception that the situation was i l legitimate, and 
disagreement with social mobility strategies. They also found some resistance to social 
competition beliefs with some respondents believing that radical social competition 
beliefs were inappropriate in the current "post-feminist" times. 
Women who choose social competition strategies are more likely to be seen as a 
threat by men, which means that they are more likely to experience discrimination 
against them as men try to maintain their positive distinctiveness. 
80 
Gender is likely to be more salient for those members of the low status and high 
status group who choose social competition. Hogg and Abrams ( 1 988) say that women 
who choose social competition (and perceive the intergroup situation to be i llegitimate) 
would find their gender more salient because they engage more often in "positive 
differentiation" with the other group. Discrimination and gender salience will be 
associated with social competition beliefs as they are more threatening to the outgroup's 
position than social mobility or social creativity strategies. 
Social Change Belief Systems -
Achieving Positive Distinctiveness Summary and Research Themes 
The perceived legitimacy and stability of the intergroup relations and the 
permeability of the groups boundaries will influence which of three belief systems low 
status groups choose to try and achieve positive distinctiveness. The level of 
identification with the ingroup will also affect this decision. The belief systems 
identified by SIT are social mobility that focuses on individual action, social creativity 
that changes the comparisons between groups, and social competition which is direct 
competition with the high status group. Social change belief systems have largely been 
ignored in research. The actions of the low status group will in tum produce reactions 
from the high status group. Little research has been conducted on the belief systems of 
the gender groups, and the variables that affect the choice of beliefs systems in gender 
groups need more careful study. 
3. 1 0  Social Change Bel ief Systems 
- Maintain ing Positive Distinctiveness 
According to SIT, when the high status group perceives the group's position as 
stable and unchanging, their social identity will be secure. If a low status group 
attempts to achieve positive distinctiveness, the positive distinctiveness of the high 
status group can be threatened. SIT states that the high status group will then make 
8 1  
attempts to maintain its positive distinctiveness (Tajfel & Turner, 1 979). The high 
status group can also choose between the belief systems of social mobility, social 
creativity and social competition. Perceptions of the intergroup situation, permeability, 
stability, and legitimacy will determine this choice (Tajfel, 1 975). The type of threat 
that the low status groups are causing to the high status group's distinctiveness, will also 
affect the actions of the high status group (Hogg & Abrams, 1 988). SIT researchers 
have neglected the reactions and beliefs of the high status groups, in both laboratory and 
field based studies. 
Social Mobility 
The use of this belief system is dependent on the individual perceiving the group 
boundaries to be permeable. The following hypothesis is formulated for all status 
groups "Acceptance will be positively related to outgroup social mobility beliefs (H29). 
The use of this belief system by the low status group is not very threatening to the high 
status group. The high status or advantaged group retains control over the permeability 
of its boundaries, and they can make it difficult to pass into their group so that social 
mobility is no threat to their dominant group (Abrams, 1 992). They tolerate a small 
number of new members (tokenism) and they can tighten the boundaries if the numbers 
of outgroup members wanting to join grows too large (Hogg & Abrams, 1 988). Social 
mobility is even beneficial for the dominant group as it decreases the likelihood of 
group action against it from the low status groups as it decreased identification with the 
group (Lalonde & Silverman, 1 994). The high status group therefore is expected to 
support social mobility beliefs in the outgroup, as it does not threaten their position. 
The fol lowing hypothesis is formulated: "For respondents of high status, identification 
will be positively related to outgroup social mobility beliefs " (H30). 
The use of this belief system by the low status group leaves the actual group 
positions unchanged. Although a few individuals may join the high status group, the 
high status group sees them as exceptional and does not see their characteristics as 
belonging to the low status group as a whole (Tajfel, 1 98 1 ). The fol lowing predictions 
are made based on SIT's claims about the relationship between identification and social 
mobility beliefs: "For respondents of low and equal status, identification will 
negatively related to outgroup social mobility JJ (H3 1 ). 
Social Creativity 
82 
lf the high status group perceives the status relations as stable and/or legitimate 
then the preferred belief system to maintain their position is that of social creativity. 
The belief systems of social creativity do not necessari ly achieve any change in the 
groups ' relative positions. Each of the types of social creativity will result in a slightly 
different reaction from the high status group. Hogg and Abrams ( 1 988) explain that a 
high status group may respond positively to some of the social creativity actions, used 
by the low status group. For example some social creativity actions reinforce the 
attributes that cause the low status group's inferior position and they will be supported 
by the high status group. They give the example of men admiring women for "their 
physical beauty, for their ability to care, cook and have babies" (p. 2 1 3). Reinforcing 
women's abilities in these areas does not produce change between the groups. They 
also mention some examples of social creativity strategies from the dominant group; 
men may emphasize the abilities women have for example, "women should take 
advantage of their capacity to care" (p. 2 1 3) that reinforce their disadvantaged position. 
Therefore in the current study "Respondents who see the intergroup situation as 
legitimate will have higher outgroup social creativity beliefs than will respondents who 
see the intergroup situation as illegitimate (H32). Respondents who see the intergroup 
situation as stable will have higher support for outgroup social creativity beliefs than 
will respondents who see the intergroup situation as unstable (H33). 
The groups can accept social creativity strategies into the existing arrangement 
as long as the difference between the groups favours the high status group in prestige 
and power (Hogg & Abrams, 1 988). As social creativity is not threatening to the high 
status group the following hypothesis for the high status group was formulated: 
"Identification will be positively related to outgroup social creativity beliefs JJ  (H34). 
The low status group however is not expected to support attempts by the high status 
group to maintain their status, so for the low status group "Identification will be 
negatively related to outgroup social creativity beliefs JJ (H35). 
Redefining the Value of Characteristics 
83 
Hogg and Abrams ( 1 988) state that the high status group will perceive attempts 
by the low status group to redefine the value of characteristics as a threat. As the 
perception of status relies on mutual comparison between groups, the low status group 
needs to convince the high status group of the new value before this belief system can 
be successful .  The dominant group can then try to ensure that the low status group is 
not too successful by challenging or repudiating the revaluation, or they may shift to 
other dimensions of comparison, in which the high status group remains superior. 
Using a New Dimension for Comparison 
Tajfel and Turner ( 1 986) state that introducing comparison on a new dimension 
can threaten the dominant group's positive distinctiveness. The high status group may 
tolerate some use of new dimensions, but only to a certain extent, once it becomes too 
much of a threat to their distinctiveness they will take action to maintain their position. 
The high status group can then reinforce the original dimensions, or create new ones on 
which they are once again superior (Hogg & Abrams, 1 988). Hinkle and Brown ( 1 990) 
suggest that the high status group can favour the low status group on some dimensions 
considered 'second rate' as a means of maintaining their positive distinctiveness by 
implying the others groups' distinctiveness is alternative or counter to norms. The main 
task for the low status group undertaking this belief system is to convince the high status 
group of the legitimacy of the new dimension of comparison. 
Changing the Outgroup 
Changing the Outgroup provides no threat to the dominant group who may 
actually encourage it. This belief system can help the high status group maintain the 
status quo as it can increase competition between low status groups. 
Social Competition 
84 
Social Competition results in the most change occurring in the status of the 
disadvantaged group, meaning it is the most threatening to the high status group, and 
will cause conflict and antagonism between the groups. For this reason it is expected 
that for all status groups: "Identification will be negatively related to outgroup social 
competition " (H36). In particular when the high status group sees their position as 
legitimate they will act in a discriminatory way towards the low status group (Tajfel , 
1 975), which leads to the following hypotheses for the high status group. "Respondents 
who see the situation as legitimate will have higher social competition (maintaining 
status) beliefs than respondents who see the intergroup situation as illegitimate (H37). 
Respondents who see the intergroup situation as unstable will have higher social 
competition (maintaining status) beliefs than respondents who see the situation as 
stable " (H38). If the high status group sees their position as illegitimate, their reactions 
are more complex. They may react with greater discrimination if the intergroup 
positions are stil l  stable, but if they are unstable, they may react with less discrimination. 
Finchilescu and de la Rey ( 1 99 1 )  found that the white (high status) respondents in their 
study who perceived the inter-racial situation as illegitimate discriminated less than did 
those who perceived it as legitimate. The fol lowing hypothesis is formulated for the 
high status group based on this research: "Respondents who see the intergroup situation 
as legitimate will have lower outgroup social competition beliefs than respondents who 
see the situation as illegitimate " (H39). "Respondents who see intergroup situation as 
stable will have lower outgroup social competition beliefs than respondents who see the 
situation as unstable " (H40). 
85 
A high status group's positive distinctiveness can also be threatened by a conflict 
of values; this can occur if the group's position is i llegitimate. In this situation, being a 
member of the group will no longer provide a positive self-concept for group members 
(Tajfel, 1 975). If the group boundaries are permeable, individuals may leave the high 
status group. They may actually join the low status group, as have some high-profile 
upper-class revolutionaries (Hogg & Abrams, 1 988). If the high status group member 
remains in their group the threat to the positive self-concept can only be resolved if the 
group finds a new justification for maintaining the status relations. 
In contrast to the prediction from SIT, Finchilescu and de La Rey ( 1 99 1 )  found that the 
white high status group members who perceived the situation to be legitimate and 
unstable were not more discriminatory than those high status members who perceived 
the situation to be legitimate and stable. 
Hogg and Abrams ( 1 988) outline some examples of social competition from the 
dominant gender group. When women are becoming successful, measures that reinforce 
the man's responsibility for household income (tax legislation) are often introduced, 
which reinforces traditional values. Sometimes men can accept the changing intergroup 
situation but acceptance can take different forms. When considering occupations 
sometimes the occupation can be actively opened to women, the job is then 
downgraded, the pay drops, men leave, and the occupation is feminised, which is 
referred to as gynopathy. Alternatively some attitudes of acceptance of women can be a 
patronizing gesture of a dominant group with a secure identity - they can afford to be 
generous, as their positive distinctiveness is not threatened. 
Social Change Belief Systems - Maintaining Positive 
Distinctiveness Summary and Research Themes 
The social change belief systems of the high status group is an under researched 
area of SIT. SIT predicts that the high status group will make efforts to maintain its 
positive distinctiveness in the face of efforts by the low status group to achieve positive 
86 
distinctiveness. The belief systems will be influenced by the permeability of group 
boundaries, and the legitimacy and stability of the intergroup relations. The high status 
group's reaction will also be affected by the amount of threat they perceive from the low 
status group, and levels of identification with the ingroup. The amount of perceived 
threat is dependent on the belief system chosen by the low status group. Social mobility 
will not be seen as a threat but social competition is threatening to the positive 
distinctiveness of the high status group. The various forms of social creativity produce 
differing amounts of threat. The efforts of gender groups to maintain status have never 
been studied. 
3. 1 1  Chapter Summary 
The current study focussed on specific areas of SIT. In the area of gender 
identification SIT states that identification is influenced by status beliefs and in turn 
affects the choice of social change beliefs. Little research has assessed the role of 
gender identity in the choice of social change beliefs. In the area of gender salience SIT 
appears to have more support for its view of gender salience than distinctiveness theory, 
especially when studying salience in the self-concept. Low status threatens the groups' 
positive distinctiveness and leads to action. As belonging to a group of high status 
makes a positive contribution to group members identity, members of high status groups 
identify more strongly with the ingroup than members of low status groups. The status 
of men in society is assumed to be high because of their advantaged access to resources 
and power, women are assumed to have low status. Perceptions of the intergroup 
situation include permeability or acceptance, legitimacy and stability. Very little 
research has studied the stability or legitimacy of the intergroup situation between 
genders. Legitimacy influences the social change beliefs of the groups to achieve or 
maintain positive distinctiveness. SIT predicts an interaction between legitimacy, 
stability and status. Research carried out mainly on laboratory groups has upheld some 
of SIT theories predictions about these factors. 
87 
Social change beliefs for achieving status are social mobility, which focuses on 
individual action, social creativity, which changes the comparisons between groups, and 
social competition, which consists of direct competition with the high status group. 
Social change belief systems are an area of SIT that has largely been ignored in research 
on gender relations. The actions of the low status group will in turn produce reactions 
from the high status group. The high status group's reactions will also be affected by 
the amount of threat they perceive from the low status group, and levels of identification 
with the ingroup. The amount of perceived threat is dependent on the belief system 
chosen by the low status group. Social mobility will not be seen as a threat but social 
competition is threatening to the positive distinctiveness of the high status group. The 
various forms of social creativity produce differing amounts of threat. 
88 
CHAPTER 4 :  RESEARCH THEMES AND HYPOTHESES 
4.1 Introduction 
The present study aims to research some of the issues identified in the previous 
chapter that researchers have neglected. The research will focus on gender 
identification, gender and occupational salience, status and beliefs about the intergroup 
situation and social change beliefs for achieving and maintaining positive 
distinctiveness. This chapter sets out the research themes and specific hypotheses in 
each of these areas for the present study. 
4.2 Gender Identity 
The first research theme for the current study is gender identification. The 
differences in gender identification between men and women and occupational groups 
will be studied. Relationships between gender identification and other factors such as 
discrimination, salience, status and beliefs about the intergroup situation will be 
examined. The relationship between identification and social change beliefs will also be 
examined. The following specific hypotheses were formed. 
H I  Men will have higher gender identification than will women. 
H2 Gender discrimination will be positively correlated with self-concept gender 
salience and gender identification. 
4.3 Identity Sal ience 
The second research theme for the current study is identity salience. 
Occupational and gender salience in the self-concept will be explored and the Gender 
Salience Scale will measure gender salience in categorising others. The differences in 
gender and occupational salience between men and women and occupations groups will 
89 
be studied. Relationships between identity salience and other factors such as 
discrimination, and numbers in the workgroup will be examined. SIrs predictions 
about salience as opposed to those of distinctiveness theory will be tested. Specific 
hypotheses were formed about the relationship between salience and other variables and 
for the theories. 
H3 Individuals who have a member of the opposite gender in their immediate 
workgroup will have higher self-concept gender salience than those who do 
not work with a member of the opposite gender. 
H4 The number of same gender colleagues or the distinctiveness of category 
membership will not be related to gender salience. 
H5 Women will have higher self-concept gender salience than men regardless of 
occupation. 
H6 The minority in the workplace (female engineers and male nurses) will have 
higher self-concept gender salience than the majority group (male engineers 
and female nurses). 
H7 Self-concept gender salience will be correlated with gender salience in 
categorising others. 
H8 For male nurses and female engineers there will be an inverse relationship 
between occupational and gender salience. For male engineers and female 
nurses there will be a positive relationship between occupational and gender 
salience. 
H9 Male nurses and female engineers will have higher gender salience than 
occupational salience. Female nurses and male engineers will have higher 
occupational salience than gender salience. 
4.4 Beliefs about the Intergroup Situation 
The third research theme for the current study involves status and other beliefs 
about the intergroup situation. Hypotheses regarding status, legitimacy, stability and 
acceptance towards the outgroup and ingroup are formulated below. Differences 
between men and women and occupations will be examined. Predictions about the 
relationships between status and beliefs about the intergroup situation and other 
variables will be studied. Status wil l  be associated with beliefs about the intergroup 
situation as predicted below. 
Status 
H I O  Respondents who see themselves of high status will have higher gender 
identification than respondents who see themselves of low status. 
H I I The majority of female engineers will see themselves as low status, and the 
majority of male engineers will see themselves as high status. 
H l 2  The majority of male nurses will see themselves as low status, and the 
majority of female nurses will see themselves as high status. 
H 1 3  The majority of female nurses will see themselves as low status and the 
majority of male nurses wil l see themselves of high status. 
Acceptance 
The specific hypotheses for acceptance and social change beliefs are summarised in 
Table 2 (p. 95). Specific hypotheses are also listed below. 
90 
H 1 4  High status respondents will express lower acceptance of the outgroup than 
the low status respondents, and the low status respondents will perceive lower 
acceptance of the ingroup than will the high status respondents. 
H I 5  Gender identification will be negatively correlated with perceived acceptance 
of the ingroup. 
H I 6  Women will be more accepting of men in female dominated occupations, than 
men will be of women in male dominated occupations. 
Legitimacy and Stability 
H 1 7  Stability beliefs will not be independent of legitimacy beliefs. 
H 1 8  More high status respondents will perceive the situation to be stable and 
legitimate than low status respondents. 
4.5 Social Change Bel ief Systems -
Achieving Positive Distinctiveness 
9 1  
The fourth research theme for the current study is social change beliefs for 
achieving positive distinctiveness. The differences in social change beliefs between 
men and women and occupations groups will be studied. Relationships between social 
change beliefs and other factors such as discrimination, salience, status and beliefs about 
the intergroup situation will be examined. No predictions can be made for the equal 
group, as SIT does not cover the actions of equal groups. The following specific 
hypotheses were formulated, and are summarised in Tables 1 (p. 93) and 2 (p. 95). 
Social Mobility 
H 1 9  F or respondents of low and equal status acceptance will be positively related to 
ingroup social mobility beliefs. 
92 
H2O For respondents of low and high status, identification wil l  be negatively related 
to social mobility beliefs. 
H2 1  Respondents of low status who see the intergroup situation as legitimate will 
have higher social mobility beliefs than those of low status who see the situation 
as illegitimate. 
H22 Respondents of low status who see the situation as stable, wil l  have higher social 
mobility beliefs than those of low status who see the situation as unstable. 
Social Creativity 
H23 Acceptance will be negatively related to ingroup social creativity beliefs. 
H24 Respondents who see the intergroup situation as stable wil l  have higher social 
creativity beliefs than will respondents who see the intergroup situation as 
unstable. 
Social Competition 
H25 Acceptance will be negatively related to social competition beliefs 
H26 Respondents who see the intergroup situation as illegitimate will have higher 
social competition beliefs than will respondents who see the situation as 
legitimate. 
H27 Respondents who see the intergroup situation as unstable will have 
higher social competition beliefs than will respondents who see the situation as 
stable. 
H28 Identification will be positively related to social competition. 
Table 1 
Predicted Relationships between Identification and Social Change Beliefs 
Social Change Beliefs Low 
Outgroup Social Mobility 
Outgroup Social Creativity 
Outgroup Social Competition 
Ingroup Social Mobility -CS) 
Ingroup Social Creativity + 
Ingroup Social Competition +(S) 
Note: (i) + == positive relationship, - == negative relationship. 
(ij) SIT predicted relationships are designated with (S) 
(iji) 0 == no predicted relationships 
Status 
Equal 
o 
o 
o 
o 
o 
4.6 Social Change Belief Systems -
Maintain ing Positive Distinctiveness 
93 
High 
+(S) 
+(S) 
-CS) 
+(S) 
The fifth research theme for the current study is social change beliefs for 
maintaining positive distinctiveness. This area includes the beliefs the groups have 
about how the out group should be achieving status. The differences in social change 
beliefs between men and women and occupations groups will be studied. Relationships 
between social change beliefs and other factors such as discrimination, salience, status 
and beliefs about the intergroup situation wil l  be examined. 
Social Mobility 
H29 Acceptance will be positively related to outgroup social mobility beliefs. 
H30 For respondents of high status, identification will be positively related to 
outgroup social mobility beliefs. 
H3 1 F or respondents of low and equal status identification will be negatively 
related to outgroup social mobility beliefs. 
Social Creativity 
H32 Respondents who see the intergroup situation as legitimate will have higher 
support for outgroup social creativity beliefs than will respondents who see 
the intergroup situation as illegitimate. 
94 
H33 Respondents who see the inter group situation as stable will have higher support 
for outgroup social creativity beliefs than will respondents who see the 
intergroup situation as unstable. 
H34 For respondents of high status identification will be positively related to 
outgroup social creativity beliefs. 
H35 For respondents of low status identification will be negatively related to 
outgroup social creativity beliefs. 
Social Competition 
H36 Identification wil l  be negatively related to outgroup social competition beliefs. 
H37 Respondents who see the situation as legitimate will have higher social 
competition (maintaining status) beliefs than respondents who see the situation 
as illegitimate. 
H38 Respondents who see the intergroup situation as unstable wil l  have higher social 
competition (maintaining status) beliefs than respondents who see the 
intergroup situation as unstable. 
H39 Respondents who see the intergroup situation as legitimate will have lower 
outgroup social competition beliefs than respondents who see the situation as 
illegitimate. 
95 
H40 Respondents who see the intergroup situation as stable will have lower support 
for outgroup social competition beliefs than will respondents who see the 
situation as unstable. 
Table 2 
Predicted Relationships between Acceptance and Social Change Beliefs 
Status 
Social Change Beliefs Low Egual High 
Outgroup Social Mobility + + +(S) 
Outgroup Social Creativity -CS) 
Outgroup Social Competition 
-CS) 
Ingroup Social Competition -CS) +(S) 
lngroup Social Mobility +(S) + 
Ingroup Social Creativity -CS) + 
Note: (i) + = positive relationship, - = negative relationship. 
(ii) SIT predicted relationships are designated with (S) 
(iii) 0 = no predicted relationships 
96 
CHAPTER 5: METHOD 
5.1  Overview 
In this chapter the methodology for the current study is outlined. To begin an 
overview of the study is provided. Then the participants in the study are described. The 
development of each measure is then described including the results from the prepilot, 
the pilot, and the two stages of the main study. 
The current study consisted of prepilot testing, a pilot study and a main study in 
two stages. In order to test the hypotheses of this study it was necessary to develop or 
adapt ways of measuring the constructs involved that would be suitable for use with 
gender groups in occupational settings. The questionnaire that was used in the two 
stages of the main study had seven sections (see Appendix A). The first section 
measured gender and occupational salience in the self-concept; this was adapted from 
The Spontaneous Self-concept measure (Abrams, Thomas, & Hogg, 1 990). Section two 
was a measure of gender identification (Hinkle, Taylor, Fox-Cardamone and Crook 
1 989). Section three of the questionnaire measured gender salience as a basis of 
categorisations of others on the Gender Salience Scale (Greenwood and Glidden-Tracey, 
1 996). Section four measured respondents' perceptions of the groups'  status, the 
legitimacy of the group's positions and the stability and permeability of the intergroup 
situation. The researcher developed this section based on previous measures used by 
Brown and Williams, ( 1 984) and Finchilescu and de lay Rey, ( 1 99 1 ). Sections five and 
six measure social change beliefs; these items were adapted from items used by 
Breinlinger and Kelly ( 1 994). The final section asked for demographic information so 
that the sample could be described. Space was also provided in the main study for 
respondents to comment on their choices and answers for each section. Free answers 
were intended to increase the relevance of the numerical data by providing insights into 
the respondents' experiences. To increase the relevance of the questionnaires to 
participants separate questionnaires were created for each gender group and for each 
occupation, by using wording appropriate for each gender and each occupation. 
97 
The development of the questionnaire began with the selection of suitable 
measures to base the questionnaire on. Before the questionnaire was piloted approval 
for all stages of the research was obtained from the Massey University Human Ethics 
Committee. A prepilot study was conducted on a sample of Massey University 
Students. After adj ustments were made to the questionnaire a pilot study was conducted 
on a sample of men and women working in gender dominated occupations. The 
main study consisted of two stages. Firstly data were collected from men and women 
working in a female dominated occupation. Following the completion of the female 
dominated occupation data collection, data were collected from men and women 
working in two male dominated occupations. 
5.2 Participants 
Prepi/ot Testing 
The prepilot sample consisted of 24 female Massey university students and 6 
male Massey university students. The average age of the female students was 32. 1 years 
and the average age of the male students was 30 years. 
Pilot Study 
Male Dominated Occupation 
The male dominated occupation sample for the pilot study consisted of police 
officers. Of the female officers, 1 2  (48% of questionnaires distributed) returned 
questionnaires and 30 (55% of questionnaires distributed) male officers returned the 
questionnaires. Male officers had a longer tenure (M = 1 0.5,  SD = 8.32) on average than 
female officers (M = 3 .2, SD = 2. 1 0), t (39) = 4.40, P < .00 1 .  Male officers (M = 38.8 ,  
SD = 8. 1 0) were also older on average than female officers (M = 30.2, SD = 4.37), t (39) 
= 3.46, p < .00 1 .  
98 
Female Dominated Occupation 
The female dominated occupation chosen for the pilot study was nursing. Of the 
female nurses 1 9, (32% of questionnaires distributed) returned questionnaires; of the 
male nurses 1 0  (33% of questionnaires distributed) returned questionnaires. Male 
nurses had a mean tenure of 2.7 (SD = 2.84) years, while female nurses had a mean 
tenure of 5 .7 (SD = 8 .03) years, t (25) = l . 1 3 , p  = > .05. Male nurses had a mean age of 
3 3 . 1  (SD = 5 .34) years, and female nurses had a mean age of 38 .6 (SD = 1 2 .54) years, t 
(26) = 1 .60, p > .05. 
Main Study 
The main study sample consisted of 207 volunteer respondents of whom 83 were 
men and 1 24 were women. 
Female Dominated Occupations 
The female dominated occupation sample consisted of 3 1  male and 79 female 
nurses, a questionnaire return rate of 26% for men and 28% for women. The majority of 
nursing respondents were New Zealanders of European descent (73%); the next largest 
group were from the United Kingdom ( 1 0%). The mean length of time that female 
nurses had been in their present job was 4.8 (SD = 5 .53) years; while the mean length of 
time male nurses had been in their present jobs was 3.5 (SD = 3 .0 1 )  years, t ( 1 03) = 
1 .56, P > .05. The average age for male nurses was 36.8 (SD = 8 .27) years; the average 
age for female nurses was 36.9 (SD = 9.97) years, t ( 1 06) = 1 .45, p > .05. 
Male Dominated Occupations 
The male dominated occupation sample consisted of engineers and prison 
officers. Of the male prison officers, 1 1  returned the survey and 3 female prison officers 
99 
answered the survey, representing a questionnaire return rate of 4% for male prison 
officers and 3% for female prison officers. The largest ethnic group was New 
Zealanders of European descent (7 1 %). The mean length of time male prison officers 
had been in their present jobs was 9.2 (SD =5.32) years; the mean length oftime for 
females was 9.5 (SD = 3.77) years, t ( 12) = . 1 0, p  > .05. Male prison officers had a 
mean age of 40.8 (SD = 5.35) years females had a mean age of 45.0 (SD = 3.46) years, I 
( 1 1 )  == 1 .26, P > .05 .  
Of the engineers, 4 1  males and 42 females returned questionnaires, representing 
a volunteer rate of 1 .6% of all engineers who receive the IPENZ e-zine. The largest 
ethnic group was New Zealanders of European descent (89%), the next largest group 
was Europeans (4%). On average the male engineers (M= 6.7, SD = 8.63) had been in 
their present jobs longer than female engineers (M= 2 .5, SD = 2. 1 5), t ( 8 1 )  = 3 . 1 , p  < 
.00 1 . Male engineers (M= 39.5, SD = 1 0.62) were also older on average than the 
female engineers (M= 29. 1 ,  SD = 7.28), t (79) = 5 . 1 3, p  < . 00 1 .  
5.2 MATERIALS 
Measure Development 
Prepilot Testing 
To assist with the development ofthe questionnaire it was piloted on a small 
group of male and female (30) Massey University students before the pilot study. The 
prepilot testing aimed to establish the internal reliability and face validity of the 
measures. It was also necessary to test the questionnaire layout and length. The pre­
pi lot testing also enabled items measuring social change beliefs to be tested for current 
relevance to the gender groups in New Zealand. The length of the questionnaire was 
found to be problematic for respondents at this stage so some items measuring social 
change beliefs were removed (see Appendix B for revisions to the social change belief 
scale) and the GSS was shortened. 
1 00 
Pilot Study 
The pilot study was designed to repeat the testing of the measures on a sample of 
individuals working in gender dominated occupations. The pilot study had several 
aims; the first aim was to retest the measures on the actual sample of interest (men and 
women in gender dominated occupations). It was also necessary to test whether 
applying the identification scale to gender groups affected the scales internal reliability, 
and whether condensing the GSS affected its internal reliability. The measures of 
legitimacy, status and stability also needed to be retested after changes were made 
following the prepilot. The measure of social change beliefs also needed retesting after 
changes had been made. The pilot study provided information on the internal reliability 
and factor construction of the scales. Internal consistency reliability was considered 
adequate at .75 .  Respondents in the pilot study also provided information about the ease 
of use of the questionnaire. Participants were from two small provincial cities not 
included in the region of the main study. Questionnaire packs included a questionnaire, 
information sheet, freepost envelope for return, and a letter from the organization saying 
they had given permission for the research to be carried out. Questionnaire packs were 
distributed randomly through the internal mail systems of the organizations. 
Materials 
The Spontaneous Self-Concept 
The first section was the Spontaneous Self-Concept measure (McGuire & 
Padawer-Singer, 1 976; McGuire, McGuire, Child & Fuj ioka, 1 978; McGuire, McGuire 
& Winton, 1 979). The Spontaneous Self-Concept measure is based on the assumption 
that invoking a free response from the respondent will provide a more authentic measure 
of the dimensions that people use when thinking about themselves than a measure where 
the researcher chooses the dimensions that the respondent then rates. The Spontaneous 
Self-Concept measure simply asks respondents to "tell us about yourself'. Respondents 
1 0 1  
have been given numbered lines to fi ll in (McGuire & Padawer-Singer, 1 976) or have 
had responses to the question tape recorded (McGuire et al . 1 978;  McGuire et al. 1 979). 
Prepi lot Testing and Pilot Study 
In the prepi lot testing, the following instructions were used: "Please write down 
five important pieces of information about yourself'. A score was then given to each 
mention of gender or occupation. This ranged from 1 not mentioned and therefore not 
salient, to 6 mentioned first, therefore highly salient. These are the instructions and 
scoring methods used by Abrams, Thomas, and Hogg ( 1 990). In the prepilot testing 
1 5% of respondents did not fill in this section. This indicated that the respondents were 
having problems filling in this question, so additional instructions were provided in the 
pilot study. 
In the pilot study the instruction "Please tell me about yourself' was added to the 
questionnaires to make the instructions clearer. This is also the instruction previously 
used by McGuire and colleagues (McGuire & Padawer-Singer, 1 976; McGuire, 
McGuire, Child & Fuj ioka, 1 978, McGuire, McGuire & Winton, 1 979). 
Only 4% of the respondents in the pilot study left this section blank, which indicated 
that the additional instruction did assist the respondents in filling in this section. 
Main Study: 
The main study instructions were the same as those used in the pilot study. The 
Spontaneous Self-Concept measure was presented first in the questionnaire; this was to 
reduce the extent to which information from the following sections affected the salience 
of gender and occupation. The spontaneity of the Spontaneous Self-concept Measure 
would be affected by the focus of the quantitative measures on gender and occupations. 
Some respondents still had difficulties with this section; thirteen respondents (6%) did 
not fill in this section. 
1 02 
Gender Identification 
The second section of the questionnaire measured gender identification. Brown, 
Condor, Mathews, Wade, and Williams ( 1 986) developed a measure of group 
identification based on the three aspects of group identification identified by Tajfel 
( 1 978), awareness of group membership ( cognitive), the value of the membership 
(evaluative) and affect towards membership (emotional). Their scale had ten items: five 
reflected positive attitudes towards the group, and five reflected negative attitudes. 
Brown et al. 's scale showed a Cronbach'  s alpha of . 7 1 .  To assess the validity of the 
scale they compared their respondents' scores on the identification scale to their 
interview responses. They concluded that the scale had adequate validity. Brown et al. 
also conducted a factor analysis on the scale they found three intercorrelated (non­
independent) factors. The identification scale can be adapted to measure identification 
with different groups, by including the name of the group in each item. Kelly ( 1 988) 
adapted Brown et al . 's scale in this manner to measure identification with political 
groups. The scale had adequate internal reliability (a.= 0.79) in Kelly's study. She also 
carried out a factor analysis that showed two factors, one consisting of the negative 
items, and one consisting of positive items. 
Hinkle, Taylor, Fox-Cardamone, and Crook ( 1 989) based their own 
identification scale on the scale developed by Brown et al. They retained six of the 
original items used by Brown et al. and included three new items. Their scale was also 
designed to measure the three aspects of identification as subscales. The new 
identification scale had higher internal consistency than the previous studies (a. = 0.85). 
A three factor principal components factor analysis (with oblique rotation) conducted by 
Hinkle et al. demonstrated clear loadings on the cognitive and emotional factors. They 
named the third factor individual group opposition as it represented opposition between 
individual needs and group opposition. Item 6 did not load clearly onto a factor. 
In the present study the scale developed by Hinkle et al . was used to measure 
gender identification. To measure gender identification the general term group used by 
1 03 
Hinkle et al. was replaced with men and women as appropriate. For example, the first 
item used by Hinkle et al. was " 1 identify with this group". For the present study, this 
became "I identify with women as a group", for the women's questionnaire, and "I 
identify with men as a group", for the men's questionnaire. The nine items are 
rated on a nine-point scale (from strongly disagree to strongly agree), which provides a 
score ranging from -36 (low gender identification) to 36 (high gender identification). 
Items 3, 6, 7, and 8 are reverse scored. There were no changes made to the scale 
between the three stages of testing. 
Prepi lot Testing and Pi lot Study 
Prepilot testing showed that the Gender Identification scale had below adequate 
reliability a = .66. Removing item three increased the reliability to a= .78 (N=28). The 
sample was not adequate for a factor analysis to be conducted. 
In the pilot study the gender identification scale again had a low Cronbach's 
alpha of .56 (N= 66). Removing item 3 - 'I feel held back by my group',  increased the 
alpha to a= .63. A factor analysis with varimax rotation resulted in a three-factor 
structure, with three items loading on each factor. The factor analysis of the pilot study 
data appeared to be more successful than the previous factor analysis by Hinkle et al . 
( 1 989). All items in the pilot study loaded onto one of the three factors. Hinkle et al. 
( 1 989) found that item 6 - ' I  do not fit in well  with my group' ,  did not load clearly onto 
a factor. The internal consistency reliability scores for the extracted factors were also 
higher than the reliability for the Hinkle et al. ( 1 989) subscales. The three factors 
extracted in the pilot study also fit in with the three aspects of identification that the 
scale was designed to cover - the knowledge of group membership, the value of that 
membership and the emotional ties to the group. These results need to be treated 
cautiously however due to the small numbers of respondents in the pilot study that was 
too low for reliable factor analysis. 
1 04 
Main Study 
In the main study the Gender Identification scale showed adequate internal 
reliability for the total sample (a = .79), and adequate internal reliability for men and 
women in the total sample. The Cronbach's alphas for each sample are shown in Table 
3 .  
Table 3 
Internal Consistency Alphas of the Gender Identification Scale (Main Study). 
Gender 
SamEle Combined Male Female 
Total Sample .79 .79 .78 
(202) (8 1 ) ( 1 2 1 ) 
Male .70 .72 . 7 1  
Dominated (96) (5 1 ) (45) 
Occupations 
Female .67 .59 .70 
Dominated ( \ 06) (30) (76) 
Occupations 
Note: n 's are in brackets 
A factor analysis was carried out on the gender identification scale. An oblique 
rotation was used because Hinkle et al. ( 1 989) claim the factors are correlated. Two 
factors emerged from the analysis as shown in Table 4. The two factors that Hinkle et 
al. saw as measuring emotional and cognitive aspects of group membership did not 
separate in this analysis. The second factor includes the items Hinkle et al. called 
individual/group opposition and item six that did not load with any factor in their 
original analysis. 
The first factor accounted for 48 % of the variance, and combined the two 
factors that Hinkle et al. called cognitive and emotional (a = .92, N = 207). Factor two 
accounted for 1 9% of the variance and included the items that Hinkle et al. termed 
individual group opposition and item 6 that did not load onto a factor in the Hinkle et al . 
study (a = .56, N = 207). This is a similar factor structure to that found by Kelly ( 1 988), 
1 05 
who found the negative and positive items loading onto two separate factors in her study 
of identification with political parties. 
Gender Salience Scale 
The third section of the questionnaire measured Gender Salience as a basis for 
categorising others. Glidden-Tracey and Wagner ( 1 995) initially measured Gender 
Salience on a scale that they had developed to measure respondents' reactions to and 
opinions about real events they had encountered. There were problems with the 
generalisability and low variability of this scale. Greenwood and Glidden-Tracey 
( 1 996) then went on to create the Gender Salience Scale (GSS). The GSS consists of 
what Greenwood and Glidden-Tracey call interpersonal contexts the respondents 
indicate how much they rely on gender as a basis for responding to the contexts. The 
GSS asks respondents to choose one statement from five that reflects the different levels 
of Gender Salience. Choosing from between specific statements was designed to create 
more accurate levels of Gender Salience than using a Likert response format as choosing 
between different statements ensures that respondents are sharing the same anchors. 
Greenwood and Glidden-Tracey ( 1 996) reported an adequate internal reliability, a = .78. 
They also checked the test-retest reliability of the GSS, r = .85, p < .00 1 .  To measure 
construct validity, scores on the GSS were compared with scores on the Attitudes 
Towards Women Scale (Spence & Helrnreich, 1 972, cited in Greenwood & Glidden 
Tracey, 1 996). The scores were related in the predicted direction, which indicates that 
attitudes towards women in society are linked to Gender Salience. 
The Gender Salience Scale used in this study consisted of 20 statements that 
represent different contexts in which gender may be a salient feature for judging others 
Table 4 
Factor Loading for the Identification Scale. (Main Study, All Samples). 
Item Factor A Factor B Communalities 
2 .  I am glad to be a woman/man. .91  -.04 .83 
1 .  1 identify with women/men as a group. .9 1  . 1 1 .83 
4. 1 think that women/men work well together. .85 .09 .73 
5 .  1 see myself as  important part of the group of working women/men .84 .0 1 . 7 1  
9. 1 feel strong ties to other women/men. .80 .0 1 .64 
7 .  1 consider my gender to be important. .76 -. 1 8  .62 
6. 1 do not fit in well with other women/men. -.07 .84 . 7 1  
8. I feel uneasy with other women/men. .03 .83 .69 
3 .  I feel held back because 1 am a woman/man. . 03 .50 .25 
(Oblique rotation, KMO= .877, N = 207) 
1 07 
behaviour. These included opinions about friends, health, household tasks, occupation 
and finances. Five statements that represent the five levels of gender salience follow 
each context. The respondents circle the statement that best reflects their opinion. The 
statement is given a score of one to five. The total score is a mean score for the whole 
scale, the maximum score was five, and the minimum was one. Greenwood and 
Glidden-Tracey ( 1 996) conceptualise the five levels of gender salience as follows: 
Level I :  Low Gender Salience: The individuals' actions and decisions are completely 
independent of gender. 
Level 2: Low-Moderate Gender Salience: Individuals actions and decisions are slightly 
determined by gender. 
Level 3 :  Moderate Gender Salience: Individuals actions and decisions are partially 
(roughly half the time) determined by gender. 
Level 4: High Moderate Gender Salience: Gender is a major factor determining 
individuals' actions and decisions. 
Level 5: High Gender Salience: Individuals actions and decisions are totally 
determined by gender. 
Prepi lot Testing and Pilot Study 
Prepilot testing showed that the full twenty item scale had high internal 
consistency reliability a = .84 (n = 27). Respondents however expressed concern over 
the length of the questionnaire. Some of the contexts (such as healthcare and birth 
control decisions) seemed irrelevant for the sample of men and women working in 
gender dominated occupations used in the present study, so were removed from the GSS 
for the pilot study. 
For the pilot study only 1 3  contexts from the original 20 were used. 
The items that focused on work, finances and household tasks were retained, as these 
items seemed most relevant to the sample of people working in gender dominated 
occupations. The pilot study showed a low Cronbach's alpha of .59 for the 1 3  retained 
contexts. 
1 08 
Main Study: 
The 1 3  contexts used for the pilot study were also used in the main study. The 
GSS demonstrated adequate internal reliability for the main study total sample (a = .79). 
The minimum gender salience score was 1 .2 and the maximum was 3 .5 .  The 
Cronbach's alphas for each sample are given in Table 5 .  
Table 5 
Internal Consistency Alphas for the Gender Salience Scale (Main Study) 
Gender 
SamQle Combined Male Female 
Total Sample . 79 . 79 .78 
(202) (8 1 )  ( 1 2 1 )  
Male Dominated .7 1 .73 .67 
Occupations (92) (50) (42) 
Female .6 1 . 6 1  . 6 1  
Dominated ( 1 00) (29) (7 1 )  
Occupations 
Note: n 's  are in brackets 
Beliefs about the Intergroup Situation 
Status: 
Brown and Williams ( 1 984) measured the status of groups in their study of 
workers at a bakery by asking the respondents to rank order the four different groups in 
the bakery, at the present. Finchilescu and de la Rey ( 1 99 1 )  used a similar measure in 
their study; they measured status on an intergroup perception ladder. An intergroup 
perception ladder requires respondents to indicate the position of the groups on a 
drawing of a ladder with 1 1  rungs. The respondents indicated where they believed the 
position of the groups to be at the present time. Skevington ( 1 98 1 )  measured status 
relations between nurses by asking the respondents to create a list of the advantages and 
disadvantages that both groups had. The ratio of advantages to disadvantages was an 
indication of the status relations. The current study used a combination of these 
techniques to measure status. 
Legitimacy and Stabi l ity: 
1 09 
Brown and Williams ( 1 984) measured the legitimacy of status relations by 
requiring respondents to rank order the groups in the organization, as to where the 
respondents felt their status ought to be. The difference between where the groups are 
and where they ought to be is an indication of the legitimacy of the status relations. In 
the same manner as Brown and Williams ( 1 984) Finchelescu and de la Rey ( 1 99 1 )  used 
the difference between the respondents' indication of where the groups are at the present 
time on the intergroup perception ladder and where they ought to be, as a measure of 
legitimacy. In the Finchelescu and de la Rey ( 1 99 1 )  study respondents also indicated 
where they thought the groups would be in the future. The difference between the 
respondents '  indication of where the groups are at the present time on the intergroup 
perception ladder and where they think they will be in the future was used as a measure 
of the stability of the intergroup situation. The present study used a combination of 
these techniques to measure legitimacy and stability. 
Acceptance: 
As members of one gender group cannot commonly change their group 
membership to �he other group a measure of acceptance of the groups was used instead 
of a measure of the actual permeability of group boundaries. The measure of acceptance 
asked about the ease of integration between the groups. 
1 1 0 
Prepilot Testing and Pi lot Study 
In the prepilot testing to measure status the respondents were asked to place the 
groups (men and women) on lines in order of who has the most advantages in the 
workplace at the present time. Legitimacy was measured by asking respondents to place 
the groups (men and women) in order of who they thought should have the most 
advantages. The respondents were asked to indicate who will have the most advantages 
in the future, the difference between this and where they placed the groups in the present 
indicated their perceptions of the stability of the intergroup situation. Only two l ines 
were provided for the respondents to place the groups on in the pre-pilot, as SIT (Tajfel 
& Turner, 1 979) claims groups are of high or low status. A large number of respondents 
(33%) requested an equal option or complained that an equal option was not available. 
The prepilot indicated that many respondents believed the intergroup situation to 
be equal. An equal option was added to the status measure for the pilot study, to ensure 
that information about those who do believe the situation is equal was not missed. 
Additional lines were added to the status legitimacy and stability measures, beside the 
lines used in the original measure. This enabled the groups to be placed equally on the 
lines. Many respondents ( 1 0%) did stil l not understand how to express equality on this 
measure and found it an unsatisfactory way of expressing equality. 
Main Study- Female Dominated Occupations 
In the main study, the measure of status consisted of the question: "Which of 
these groups do you think has the most advantages in your workplace, at the present 
time?" The respondents then chose between three response options: females have the 
most advantages, males have the most advantages, or males and females have equal 
advantages. Respondents ticked the box of the statement they agreed with. This 
enabled the respondents to be directed to answer different sections of the social change 
1 1 1  
beliefs measure designed specifically for their status beliefs. The measure of legitimacy 
and stability used the same format as the status question. 
A measure of acceptance of the groups was added to the main study. 
Acceptance was measured on two, nine-point scales. One asked the difficulty or ease 
that males experience in integrating with and being accepted by females in their 
occupation, and the other asked the difficulty or ease females experience being accepted 
by and integrating with males in their occupation. 
Main Study- Male Dominated Occupations 
The measure described above for the nursing sample was also used for the male 
dominated occupations. In addition, a continuous measure was used to measure status, 
legitimacy and stability. This was to assist with data analysis and to provide additional 
information about beliefs about the intergroup situation that are not provided by a 
categorical measure. The extra items asked the respondents to rank (on a nine point 
scale) the status of men and women in their profession (status), how fair or unfair the 
status was (legitimacy), and whether men's and women's future position would be 
worse, the same or improved (stability). Acceptance was measured in the same way as 
for the female dominated occupations, on two, nine-point scales. 
Social Change Beliefs 
Breinlinger and Kelly ( 1 994) studied women's social change beliefs using Q­
Sort methodology. They chose their items from a combination of pilot interviews, 
previous research and current events in newspaper and magazine articles. Judges who 
had some knowledge of SIT sorted these items into the various beliefs. Only statements 
that were unanimously chosen to represent beliefs were used. In the present study, the 
items from Breinlinger and Kelly's study were used to develop a scale to measure men 
and women's social change beliefs. Some items that initially seemed out of date or 
irrelevant for a working sample were discarded for example, "when I hear about women 
1 1 2 
who work, I feel good about staying home" (p. 1 6). The items for the male respondents 
were created by adapting the female items used by Breinlinger and Kelly ( 1 994). For 
example, an item measuring social competition was originally, "If companies and 
institutions gave women as many advantages as men then we would see more women in 
powerful jobs". For men this became "If companies and institutions gave men as many 
advantages as women then we would see more men in powerful jobs". 
Additional items were also developed from literature and pilot interviews. The 
social change beliefs of social mobility, social creativity (new dimension, changing the 
values, and changing the outgroup), and social competition were measured on nine point 
scales (from -4, strongly disagree to 4 strongly, agree). Appendix C shows the items 
used to measure each social change belief in the main study. 
Prepilot Testing and Pi lot Study 
For the female respondents in the pre-pilot testing and pilot study the measure of 
social change beliefs consisted of one measure of ingroup social change beliefs. For the 
male respondents in the pre-pilot and pilot the measure consisted of items measuring 
ingroup social change beliefs, with a few items measuring their acceptance of rejection 
of the females social change beliefs (outgroup measure). Forty-six male items and 
forty-six female items were tested those that contributed to low reliability or received 
negative comments from respondents were discarded (refer to Appendix B for discarded 
and retained items). The social change belief scale alphas for the pilot study are shown 
in Table 6. Thirty items were retained for the main study. 
Main Study - Female Dominated Occupations 
The measure used in the pilot study was found to be inadequate as the belief 
items treated women as low status and men as high status, but male or female 
respondents could see themselves of high, low or equal status. The measure used in the 
pilot study did not adequately measure the respondents' beliefs in response to this, so 
1 1 3 
changes were made to the measures. The measure of social change beliefs was organized 
into three sections, a respondent only answered one section depending on their 
perception of the status of their group. If they saw themselves of low status, they 
answered questions designed to measure their beliefs about improving their groups 
position (ingroup measure) these beliefs are referred to in this study as ingroup social 
competition (achieving), ingroup social mobility and ingroup social creativity. If the 
respondents saw themselves as of high status, they answered questions designed to 
measure their support or rejection of the beliefs of the low status group (outgroup 
measure), these beliefs are referred to in this study as outgroup social competition, 
outgroup social mobility and outgroup social creativity. The outgroup measure also 
included items measuring social competition beliefs for maintaining their group's 
status(referred to as ingroup social competition maintaining). If the respondent thought 
that the groups were equal, they answered some items from both the outgroup measure 
and the in group measure. 
Table 6 
Internal Consistency Alphas of Social Change Beliefs (Pilot Study) 
Social Change Beliefs 
Ingroup social competition 
Ingroup social mobility 
Ingroup social creativity 
Outgroup changing the outgroup 
Ingroup changing the outgroup 
Outgroup changing the values 
Ingroup changing the values 
Outgroup new dimension 
Ingroup new dimension 
Note: n 's are in brackets 
Female 
.8 1  
(30) 
. 8 1  
(29) 
.69 
(27) 
. 6 1  
(30) 
. 4 1  
(3 1 ) 
. 58  
(27) 
Gender 
Male 
. 82 
(36) 
.43 
(40) 
. 65 
(40) 
. 79 
(40) 
.02 
(40) 
. 69 
(40) 
.46 
(40) 
.49 
(40) 
. 76 
(40) 
1 1 4 
Main Study - Male Dominated Occupations 
The method of measuring the reactions of the high, equal and low status groups 
used for the female dominated occupation sample was found to be confusing for the 
respondents, and limited the analysis of the social change beliefs because of the large 
numbers of respondents who saw the situation as equal . For the male dominated sample 
all respondents filled in the outgroup measure and the ingroup measure, which enabled 
extra information to be collected from both groups about their support or rejection ofthe 
beliefs of the outgroup. 
Open-ended questions were also used to measure social change beliefs, for both 
the female and male dominated occupation samples. This enabled additional qualitative 
information to be collected. Two questions were asked with the outgroup measure of 
social change beliefs: 
1 .  What are the types of strategies that you notice the outgroup using to try to 
improve the situation for their gender group in your occupation? 
2. What are your reactions to these strategies? 
Three questions were asked with the ingroup measure of social change beliefs: 
1 .  What are the types of strategies that you use to improve the general situation for 
your gender group in your occupation? 
2. What sort of reactions are these met with by other members of the ingroup? 
3 .  What sort of  reactions are these met with by  the outgroup? 
The social change beliefs used in the main study were tested for internal reliability; 
many of the scales had low internal reliability as shown in Tables 7 and 8 .  
1 1 5 
Table 7 
IngrOUP Social Change Beliefs Internal Consistency Alphas (Main Study) 
Scale Gender 
Male Female 
Ingroup Social Mobility .79 .82 
(58) (54) 
I ngroup Social Competition (achieving status) .24 .80 
(58) (55) 
Ingroup Changing the Values .55 .55 
(55) (54) 
Ingroup New Dimension .85 .8 1 
(5 8) (55) 
Ingroup Changing the Outgroup .46 .28 
(58) (54) 
Ingroup Social Creativity .80 .74 
(54) (50) 
Note n 's  are in brackets 
Table 8 
Out group Social Change Beliefs Internal Consistency Alphas (Main Study) 
Scale Gender 
Male Female 
Outgroup Social Mobility .79 . 7 1  
(60) (57) 
Outgroup Social Competition . 59 .44 
(57) (55) 
Ingroup Social Competition (maintaining status) .67 .69 
(74) ( 1 04) 
Outgroup Changing the Values .42 .55 
(60) (57) 
Outgroup New Dimension .79 .80 
(59) (57) 
Outgroup Changing the Outgroup .68 .44 
(56) (43) 
Outgroup Social Creativity .77 .8 1 
(54) (43) 
Note: n's  are in brackets 
1 1 6 
A principal components (with oblique rotation) factor analysis on each ofthe 
four measures (male ingroup, male outgroup, female ingroup, female outgroup) was 
performed to isolate the clearest associations of beliefs. The factor analysis results are 
shown in Tables 9, 1 0, 1 1  and 1 2 . It was found that two of the types of Social 
Creativity, Changing the Values and New Dimension, were not separate factors for any 
of the groups. Items were discarded until strong factors with adequate (above .75) 
internal reliability emerged. Items were discarded if they did not load onto a factor, or if 
they loaded with items that were not theoretically related to each other. Some items 
were not included in the combined sample analysis as the items were only used for one 
sample, either the male dominated occupation or the female dominated occupation. As 
shown in Table 9 the male in group measure factor analysis resulted in three factors: 
Social Creativity (combining changing the values, new dimension and changing the 
outgroup), Social Mobility and Social Competition. Factor one accounted for 23. 1 9% 
of the variance, factor two for 20.50% and factor three accounted for 1 1 .64% of the 
variance. Nine items were discarded. Item 4 - 'When I see the kinds of jobs other men 
are doing I'm satisfied with my own achievements '  was originally designed to measure 
changing the outgroup. Item 1 3  ' I rarely consider how much my female colleagues get 
paid', was also originally included to measure changing the outgroup. These two items 
loaded negatively with the social competition items, so were reverse scored and 
included as part of the social competition factor. 
The male outgroup measure resulted in four factors as shown in Table 1 0  (p. 
1 1 9): Social Creativity (combining changing the values and new dimension) which 
accounted for 22. 1 2% of the variance, Social Mobility which accounted for 1 8 .75% of 
the variance, Social Competition which accounted for 1 0.90% of the variance and 
Changing the Outgroup which accounted for 8.5% of the variance. Ten items were 
discarded. 
The female ingroup measure resulted in four factors as shown in Table 1 1  (p. 
1 20): Social Competition (26.88% of the variance), Social Mobility ( 1 5 .56% of the 
variance), Social Creativity (9% of the variance) and Changing the Outgroup (8.29% of 
the variance). Eight items were discarded. Items 7, 1 6, 1 9  and 23 were originally 
1 1 7 
designed to measure social competition beliefs however when they were reverse scored 
they loaded onto the social mobility factor. The social mobility factor was characterized 
by agreement with the social mobility items and disagreement with four of the social 
competition items. 
The female outgroup measure showed three factors: Social Creativity 
(combining changing the values, new dimension and changing the outgroup) which 
accounted for 25.70% of the variance, Social Mobility ( 1 4.32% of the variance) and 
Social Competition ( 1 0.20% of the variance). Seven items were discarded. Item 1 3 -
'men should rarely consider how much female colleagues get paid' was designed to 
measure changing the outgroup however it loaded with social competition items, so it 
was reverse scored and retained as part of the social competition factor. The results of 
the factor analysis are shown in Table 1 2  (p. 1 2 1 ). 
Mean scores were used, based on the factors that emerged from the factor 
analysis where possible. Tables 1 3  and 1 4  show the internal consistency alphas for the 
extracted factors. 
Table 9 
Factor Loading for Social Change Beliefs Male IngrouQ Measure (main study N =56). 
Item Factor I Factor 2 Factor 3 Communalities 
1 2. Better at handling situations with the potential for violence. .83 - 1 . 5 -. 1 5  .68 
5 .  Men make better work colleagues than women. .79 -.26 .07 .66 
2 .  Men are tougher and stronger than women. .78 -.07 -. 1 1  .60 
23 .  General values and outlook are superior to women's. .77 - .25 . 1 7  .67 
1 6. Breadwinner is the most important in the family. .68 -.03 .08 .47 
27. More important to be clinical than emotional. .53 . 32  . 1 1 .44 
1 0. Better place if women were more like men in their values. .51  - .29 .48 .62 
28. Male nurses have no reason for complaint. .46 . 1 3  -.00 .24 
22. No excuse for other men not succeeding. .45 .4 1 -. 1 7  .45 
24. I try to compare myself with other men. .42 .20 .0 1 .23 
1 5 . It's important that people make the most of opportunities. - . 1 5  .90 -.04 .8 1  
25. It's up to each individual person -.02 .86 -.02 .75 
3 .  Drive and ambition are still important in our society. -.25 .79 .07 .63 
6. If you want something badly enough you can achieve it. .02 .75 .02 .56 
9. Most people j ust don't have enough drive and ambition. .34 .5 1 . 1 7  .44 
17 .  Men should keep on fighting until they have equality. -.07 . 1 2  .88 .74 
26. There is sti l l  a lot of prejudice and discrimination. .20 . 1 3  .77 .65 
14 .  Men should be angry about their low status in nursing. .06 .08 .76 .57 
1 3 . I rarely consider how much female colleagues get paid. . 1 5  .43 -.48 .50 
4 .  When I see the kinds of jobs other men are doing. .3 1 .2 1 -.43 .35 --
00 
Table 1 0  
Factor Loading for Social Change Beliefs Male OutgrouQ Measure (main study N = 55). 
Item Factor I Factor 2 Factor 3 Factor 4 Communalities 
1 1 . General values and outlook are superior to men's. .75 -.07 - . 1 6  -.22 .63 
1 7. Women are more versatile and flexible than men. .74 -.06 -.20 - . 1 7  .59 
1 5 .  Women make better work colleagues than men. .67 -. 1 5  -.08 . 1 5  .47 
20. Better place ifmen were more like women in their values. .67 -.04 . 1 5  -. 1 4  .53 
23 .  Men want women to stay in the home. .65 -.20 . 1 4 - .06 .5 1 
27. Women are more sensitive and caring than men. .59 .06 - . 1 1  -.29 .46 
6. Women are better at doing many things at once than men. .57 .37 -.08 . 1 2  .45 
1 9. A woman has a lower status in society than a man. .55 .0 1  .33 .48 .65 
8. It is more important to be intuitive than it is to be logical. .44 .28 . 1 5  .38 .4 1 
28 .  Men and women have equal opportunities. - .34 .80 .00 - . 1 1  .79 
4 .  If women want to get ahead in their careers. - . 1 2  .76 .07 .00 .59 
24. Now that there are so many women in senior positions. . 1 9  .65 . 1 3  -.05 . 5 1  
2 1 .  It is important now that women make the most of the freedoms they've got. .29 .62 . 1 6  - . 1 7  .58 
26. If you're good at your job, then you'll be promoted. - .33 .62 -.25 -. 1 1  .58 
9. It's up to each individual woman. -.09 .59 - .24 .06 .39 
22.  Compared with Muslim women. .20 .57 .36 -.34 .72 
7 .  Women should be angry about their low status in nursing. - . 1 8  -.02 .95 -.05 .88 
1 2 .  Women must keep on fighting. -. 1 7  -.0 1 .94 -.03 .87 
14 .  When women see how little some other women have. . 1 6  .08 . 1 5  -.82 .77 
1 0. When women see the kinds of jobs. .03 .30 .03 -.75 .74 
1 8. When female nurses/engineers hear about women who stay in the home. .20 .00 .04 -.69 .54 ...-...-
1.0 
Table 1 1  
Factor Loading for Social Change Beliefs Female IngrouI2 Measure (main study N = 5 1) 
Item Factor I Factor 2 Factor 3 
9. It's up to each individual woman. .82 . 3 1 - .03 
1 8. When I hear about women who stay in the home. .68 -.20 .05 
2. If companies and institutions provided child care facilities. .57 -.09 . 1 5  
1 2. Women must keep on fighting. .53 -.48 .06 
28. Men and women have equal opportunities. .05 .88 .04 
4. If women want to get ahead in their nursing .00 .88 .09 
26. If you're good at your job, then you'll be promoted. . 1 8  .86 .09 
1 6  (r). There is still a lot of prejudice and discrimination. -.25 .71  -.25 
7 (r). Women should be angry about their low status -. 1 9  .70 -. 1 1  
1 9  (r). As a woman I have a lower status in society than a man - .26 .64 -.08 
24. Now that there are so many women in senior positions. . 1 0  .6 1  .04 
23 (r). Men want women to stay in the home. .08 .36 -.26 
1 7. On the whole women are more versatile and flexible .04 . 1 4  .82 
1 1 .  General values and outlook are superior to men's. . 1 0  . 0 1  .75 
1 5 .  Better work colleagues than men. -. 1 4  -. 1 0  .72 
3.  Drive and ambition are overrated. -.34 .05 .70 
20. Better place if men were more like women in their values. - .01 -. 1 3  .62 
6 .  Women are better at doing many things. .20 .03 .6 1 
27 .  Women are more sensitive and caring than men. .27 . 1 0  .53 
25. Men have an unequal share of status. .33 -.06 .52 
22. Compared with Muslim women. . 1 2  .20 -.2 1 
8 .  I think it is more important to be intuitive. -.3 1 -. 1 6  .24 
2 1 .  Women make the most of the freedoms the:{ve got .23 .29 -.09 
Note: r indicates reverse scored items 
Factor 4 
-.07 
- .23 
.33 
.03 
- . 1 4  
-. 1 3  
-.08 
. 1 7  
.39 
.23 
-.28 
.35 
.08 
.29 
.0 1 
-. 1 7  
.03 
-.20 
-.26 
. 1 2 
-.76 
-.73 
-.5 1 
Communalities 
.77 
.62 
.46 
.56 
.80 
.77  
.76 
.78 
.73 
.58 
.48 
.37 
.63 
.6 1 
.54 
.55 
.44 
.52 
.50 
.45 
.69 
.44 
.69 
....... 
N o 
Table 1 2  
Factor Loading for Social Change Beliefs Female OutgrouQ Measure (main study N = 55) 
Item Factor I Factor 2 Factor 3 Communalities 
1 0. Better place if women were more like men in their values .74 -. 1 0  -. 1 1  .56 
1 6. Breadwinner is the most important in the family. .74 . 1 5  -. 1 6  .65 
24. Men should try to compare myself with other men .73 -.08 .42 .65 
2. Men are tougher and stronger than women. . 7 1  .05 -.20 .59 
5 .  Men make better work colleagues than women. .68 -.02 -.02 .46 
1 2. Men are better at handling situations with the potential for violence. .66 .04 . 1 3  .45 
23. Men's general values and outlook are superior .65 .03 - .06 .44 
7. It is more important to be logical than it is to be intuitive .65 . 1 9  .00 .49 
1 5 . It is important that people make the most of the opportunities. -.05 .85 -. 1 1  .75 
25. It's up to each individual person. -.29 .80 .22 .67 
6. If  you want something badly enough. -.02 .66 - . 39 .65 
3. Drive and ambition are still important in our workplaces. . 24 .65 -.07 .55 
1 8 .  Men and women have equal opportunities i n  nursing . 1 0  .55 . 1 3  .33 
22. There are so many men in senior positions. .04 .42 .03 . 1 8  
9 .  Most people just don't have enough drive and ambition .23 .33 - . 3 1 .32 
14 .  Men should be angry about their low status. . 1 6  .0 1 -.71 .56 
26. There is still a lot of prejudice and discrimination. .25 .00 -.62 .47 
1 3 . Men should rarely consider how much female colleagues get paid .28 .3 1 .61 .50 
1 7. Men must keep on fighting -.36 .24 -.50 .4 1 
1 .  If men in nursing/engineering had as any advantages as women .36 -.0 1 -.45 .36 
N -
Table 1 3  
Outgroup Social Change Beliefs Internal Consistency Alphas for Extracted 
Factors (Main Study) 
Scale Gender 
Male Female 
Outgroup Social Mobility .82 . 7 1 
(60) (57) 
Outgroup Social Competition .99 .55 
(59) (55) 
Outgroup Changing the Outgroup .78 
(59) 
Outgroup Social Creativity .8 1 .85 
(58) (57) 
Note: n 's  are in brackets 
Table 1 4  
Ingroup Social Change Beliefs Internal Consistency Alphas for Extracted 
Factors (Main Study) 
Scale Gender 
Male Female 
Ingroup Social Mobility .8 1 .87 
(57) (54) 
Ingroup Social Competition (achieving status) .73 .70 
(58) (55) 
Ingroup Changing the Outgroup .53 
(56) 
Ingroup Social Creativity .82 .82 
(57) (53) 
Note n's are in brackets 
5.3 Design and Procedure 
1 22 
Before questionnaires were distributed permission to survey members of the 
gender dominated occupations were initially obtained from the appropriate management 
or research staff in each organization. 
Nurses at five hospitals were surveyed; 285 female nursing questionnaire packs 
and 1 20 male nursing questionnaires packs were distributed. Included with the 
1 23 
questionnaire were a freepost return envelope and an information sheet. At two 
hospitals management distributed the questionnaires to charge nurses on each ward who 
then asked volunteers to fill in the questionnaires. At the remaining three hospitals, 
management e-mailed the charge nurses of the hospitals informing them of the study, 
and the questionnaires were delivered to each ward by the researcher. The charge nurses 
then sought volunteers to fill in the questionnaires. A thank you letter was sent to the 
charge nurse of each of these wards one week later -this included a reminder to the 
volunteers to return the questionnaires. The questionnaires were returned directly to the 
researcher in the freepost envelope provided with the questionnaires. 
At the prisons, managers were informed of the study and they distributed 
questionnaires to volunteers among their sections, 273 male questionnaires and 87 
female questionnaires were distributed. The questionnaires were returned directly to the 
researcher in the freepost envelope provided with the questionnaires. 
To recruit engineers for the study a request for volunteers to fill in a 
questionnaire among engineers was published in their professional institute's e-zine. 
Volunteers then contacted the research by e-mail requesting a questionnaire. The 
questionnaires were mailed or e-mailed to the respondents who returned the 
questionnaires by e-mail or in the freepost envelope provided. 
The questionnaires were coded by the researcher and analysed with the SPSS 
statistical program. To test for differences between the groups in the study, analysis of 
variance was used. Two-way ANOVAs could not be used as gender and occupation 
were not independent, X2 ( 1 ,  N = 207) = 1 3 .87, p < .00 1 .  Chi Square analyses were used 
for the categorical variables - status, legitimacy and stability. To test for relationships 
between variables regression analyses and correlations were used. The statistics were 
considered significant when p < .05, two-tailed tests of significance were used. For 
hypotheses were direction was predicted one-tailed tests of significance were used. All 
tests were performed on the full sample, and differences between occupations and 
genders were also tested. 
The answers to the free response and open-ended questions were content 
analysed by the researcher. Content analysis according to Miles and Huberman ( 1 994) 
is a process of searching for categories and themes from the data. As the current study was 
conducted using SIT, the categories for the content anaJysis were derived from the concepts and 
categories of SIT. 
Measures Used in the Main Study 
Measure Method of Scoring Questionnaire Section Item Numbers 
Nurses Engineers 
Spontaneous Self I'{) ranking Section A Section A -
Concept Measure (6= highly-salient) 
Gender Identification: Total (range -36-+36) Section B Section B 1 -9 
Gender Identification - Factor Score Section B Section B 2, 1 ,4,5,9,7 
Emotional/COWlitive 
Gender Identification Factor Score Section B Section B 6,8,3 
Individual/Group 
Opposition 
Gender Salience Scale Total/number of items Section C Section C 1 - 1 3  
(ranl\e 1 -5) 
Status High, (3) Low ( I ), Section D Section 0 Nurses I Engineers: 1 ,4,5 
Equal (2) 
Legitimacy Legitimate (0), Section 0 Section 0 Nurses: 2 Engineers: 2, 6,7 
lllegi timate (I) 
Stability Stable (0), Unstable ( I )  Section 0 Section 0 Nurses: 3 Engineers: 3,8,9 
Acceptance Total (raO!!e-4-+4) Section 0 Section 0 Nurses: 4a and 4b EnWneers: 10, 1 1  
Men Women Men Women Men Women 
Social Mobility Total/nwnber of items E & G  F & G  F F 1 5,25,3,6,9 28,4,26, 16,7, 1 9, 
(ingroup measure) (range -4 -+4) 24,23 
Social Competition Totallnwnber of items E & G  F & G  F F 1 7,26, 1 4, 1 3,4 9, 1 8,2, 1 2  
(ingroup measure- (range -4 -+4) 
achieving) 
Social Creativity Total/number of items E & G  F & G  F F 1 2,5,23, 16,27, 1 7, 1 1 , 1 5,3,20,6, 
(ingroup measure) (range-4 -+4) 10,28,22.24 27,25 
Social Mobility Total/number of items F & G  E & G  E E 28,4,24,2 1 ,26,9, 1 5,25,6,3,1 8,22,9 
(outgroup measure) (range-4 -+4) 22 
Socia[ Competition Total/number of items F & G  E & G  E E 7, 1 2  14,26, 1 3, [ 7, [ 
(outgroup measure) (range-4 -+4) 
Social Creativity Total/number of items F & G  E & G  E E 1 1 , 1 7, 1 5,20,23, 10, 16,24,2,5, 1 2, 
(outgroup measure) (range-4 -+4) 27,6, 1 9,8 23,7 
Social Competition Total/nwnber of items F & G  E & G  E E 1 ,2,3,4,5 1 ,2,3,4,5 
(maintaining) (range-4 -+4) 
1 24 
1 25 
CHAPTER 6 RESULTS 
Chapter 6 presents the results from the two stages of the main study. The results 
section is organised according to the main areas of SIT that were examined in this 
study. The first section presents the results of the gender identification measure. The 
second section covers the results of the identity salience scales. Thirdly, the results of 
the beliefs about the intergroup situation are described. Sections 4 and 5 cover social 
change beliefs for achieving and maintaining status. 
6.1  Gender Identity 
Total Sample 
Table 1 5  shows the means and standard deviations for the gender identification 
scale as a full scale and for the subscales extracted by factor analysis, of 
cognitive/emotional aspects and individual/ group opposition. One-way ANOVA's 
were performed to test for differences in gender identification between the different 
groups in the study. 
Gender Identification and Gender Groups 
Differences in gender identification between the genders are shown in Table 1 5 . 
As predicted (hypothesis 1 )  for the full scale men scored significantly higher than 
women in gender identification, F ( 1 , 200) = 7.72, p = < .0 1 .  Men also scored 
significantly higher on the cognitive/emotional subscale than women, F ( 1  ,205) = 7.48 , 
p < .0 1 .  No significant differences between the genders were found on the individual 
group opposition subscale. 
Gender Identification and Occupations 
The male dominated occupation sample had a higher mean score on the full 
scale than the female dominated occupation, F ( 1 , 200) = 201 .34, p < .00 1 .  
1 26 
The male dominated occupation sample also had a higher mean score on the 
cognitive/emotional subscale than the female dominated sample, F ( 1  ,200) = 442.89, P 
< .00 1 .  The female dominated sample scored higher on individual group opposition 
than the male dominated sample, F ( 1 ,200) = 5 .03 , p  < .05 . A high score on this 
subscale represents low conflict between individual and group needs which contributes 
to high identification. The female dominated occupation experienced less opposition 
between individual and group needs than the male dominated sample. 
Within the female dominated sample, there was no significant difference in 
gender identification between men and women. Within the male dominated sample 
men scored higher than women on the individual group opposition subscale, F ( 1 ,95) = 
5 .45 ,p  < .05, showing that men experienced less opposition between individual and 
group needs than women. 
Gender Identification and Status 
Table 1 6  shows the mean gender identification scores for the total scale and each 
subscale for each group of respondents, according to their perception of their group's 
status. Contrary to hypothesis 10 that high status groups would have higher gender 
identification there was no significant differences between the low equal and high 
groups in gender identification on the full scale. On the cognitive/ emotional subscale 
the low status group had higher gender identification (M = 0.56, SD = 2.09) than the 
equal group (M = -0.6 1 ,  SD = 1 .86), F (2, 202) = 7. 1 2, P < .00 1 .  On the 
individual/group opposition subscale, the low status group had a lower score CM = 0.94, 
SD = 1 .78) than both the equal (M = 2. l 5, SD = 1 .47) and the high status group (M = 
1 .77, SD = 1 .25), F(2, 20 1 )  = 1 1 .53, p < .00 1 .  The low status group experienced more 
opposition between individual and group needs than the equal or high status groups. 
Table 1 5  
Mean Gender Identification Scores (for each sample) 
Gender Identification Female Dominated Male Domjnated Occupations 
Scale OccuEations 
Female Male F Female Male F 
n =  79 3 1  45 52 
Full Scale -0.50 -0.40 0.24 1 .46 1 .45 0.00 
(0.90) (0.8 1 )  ( I ,  1 04) ( 1 .06) ( 1 .06) ( \ , 94) 
CognitivelEmotional - 1 .73 - 1 .60 0.44 1 .65 1 .32 1 .7 \  
Subscale (0.95) (0.79) ( I ,  \ 08) ( \ .22) ( 1 .27) ( \ , 95) 
Individual/Group 2 .00 1 .90 0 . 1 0  1 .07 1 .82 5 .45* 
Opposition Subscale ( \ .47) ( 1 .72) ( 1 ,  1 07) ( 1 . 84) ( 1 .30) ( 1 , 95) 
* p < .05, ** p< .0 1 ,  ***  P < .00 1 
Note: (i) standard deviations and degrees of freedom are in brackets, 
(ii) n's may vary depending upon missing values. 
Combined Sample 
Female Male F 
1 24 83 
0.23 0.77 7 .72** 
( \ .35) ( 1 .32) ( \ , 200) 
-0.50 0.23 7 .48** 
( 1 .94) ( 1 . 80) ( \ , 205) 
1 .66 1 .85 0.70 
( 1 .66) ( 1 .46) ( 1 , 204) 
Occupations (dominated by) 
Female Male 
1 1 0 1 09 
-.047 1 .46 
(0.87) ( \ .05) 
- \ .69 \ .47 
(0.9 \ )  ( \ .25) 
1 .97 \ .47 
( 1 .54) ( 1 .6 1  ) 
F 
20 1 .34*** 
( 1 , 200) 
442.89*** 
( 1 , 205) 
5.03* 
( 1 , 204) 
>-' 
N -.) 
Table 1 6  
Mean Gender Identification Scores for Each Status GrouQ 
Gender Identification 
Scale Low (L) 
n= 57  
Full Scale 0.72 
( 1 .5 1  ) 
CognitivelEmotional Subscale 0.56 
(2.09) 
Individual/Group Opposition Subscale 0.94 
( I .  78) 
***  p < .00 1 
Note: (i) standard deviations and degrees of freedom are in brackets, 
(ii) n's may vary depending upon missing values. 
Status 
Equal (E) High (H) 
99 49 
0.34 0.37 
( 1 .39) ( 1 . 1 2) 
-0.6 1 -0.24 
( 1 .86) ( 1 .59) 
2. 1 5  1 .77 
( 1 .47) ( 1 .25) 
(iii) For equal variances the Bonferroni Post Hoc test was used, for unequal variances the Dunnett T3 Post Hoc test was used. 
Post Hoc 
F 
1 .48 
(2, 1 97) 
7. 1 2***  L>E 
(2, 202) 
1 1 .53 ***  L<E 
(2,20 1 )  L<H 
-
IV 00 
1 29 
Gender Identification and Majority and Minority Groups 
Respondents were divided into majority and minority groups. The minority 
group was made up of males in female dominated occupations and females in male 
dominated occupations. The majority group was made up of females in female 
dominated occupations and males in male dominated occupations. Table 1 7  shows the 
means and standard deviations on the gender identification fuJl scale and subscales for 
the minority and majority groups. The minority group was higher in gender 
identification on the full scale, F ( 1 ,200) = 4.73, p < .05, and on the cognitive/emotional 
subscale, F ( 1 ,200) = 9.72, p < .0 1  than the majority group. The majority group was 
higher on the individual group opposition subscale than the minority group, F ( 1 ,200) = 
5.2 1 , p  < .05. The majority group had less conflict between individual and group needs 
than the minority group. 
Table 1 7  
Mean Gender Identification Scores for Majority and Minority Groups 
cremder ldentificatioD Group 
Scale Minority Majority 
n =  76 13 1 
Full Scale 0.71  0.29 
( 1 .33) ( 1 .36) 
CognitivelEmotional Subscale 0.32 -.52 
( 1 .93) ( 1 .85) 
Individual/Group Opposition 1 .4 1 1 .93 
Subscale ( 1 .82) (1.40) 
* p < .05, ** P < .01  
Note: (i) standard deviations and degrees of freedom are in brackets. 
(ii) n's may vary depending upon missing values. 
Gender Identification and Other Variables 
F 
4.73* 
( 1 , 200) 
9.72** 
(1 , 205) 
5 .21  * 
( 1 , 204) 
Table 1 8  shows the correlations between gender identification as a total scale 
and subscale and the other variables. Contrary to Hypothesis 2, gender identification 
was not related to gender discrimination, r = -.07,p > .05 on the full scale. Individual 
group/opposition was negatively related to discrimination, r = -.24, p < .00 1 . High 
individual group opposition was associated with experiencing gender discrimination. 
Table 1 8  
Correlations between Gender Identification and other Variables (total sample 
N= 207) 
Variable 
1 .  Gender Identification F u1l Scale 
2. Gender Identification Cognitive Emotional Subscale .92··· 
3. Gender Identification lndividual Group Opposition .36··· 
4. Discrimination -.07 
S. Age -.05 
6. Gender Salience Spontaneous Self Concept -.09 
7. Acceptance of the ingJOup -.09 
8. Acceptance towards the outgroup -.07 
• p < .05, •• p < .0 I, ".p < .00 I 
Note: u's may vary due to missing values. 
2 3 
-.03 
.02 -.24·· 
-.07 .03 
-.09 - 00  
-.2 1 ·· .25··· 
.03 -. 1 7· 
1 30 
Table 1 9  shows the free answer responses on gender identification for the female 
dominated and male dominated samples respectively. A large number of the male 
dominated sample (23%) mentioned how gender was unimportant in the workplace, 
however no members of the female dominated sample mentioned this. The two largest 
groups of responses for men in female dominated occupations were mentioning the 
types of things that affect gender identification (33%), particularly the effect of the 
working environment. Secondly was an emphasis on individuals rather than gender 
groups (33%), particularly identifying more with some individuals within the gender 
group than others. The largest category of remarks for the women in the female 
dominated sample was their lack of identification with their gender group (33%), this 
focused on aspects such as the context and differences in identification with different 
groups of women. For the males in male dominated occupations the things that affect 
gender identification, positive remarks about men and the emphasis on the 
unimportance of gender were mentioned equally (23% each). The females in the male 
dominated occupation sample made the most remarks about the negative aspects of 
working with other women (43%). 
Table 1 9  
Free Answers on Gender Identification 
Female Dominated Male Dominated 
Occupations Occupations 
Gender Male Female Male Female 
n = I S  n = 24 n = 22 n = 2 1  
Lack of Identification with gender 0% 33% 1 8% 0% 
group 
Emphasis on individuals 33% 20% 9% 1 0% 
Discrimination 27% 20% 0% 1 0% 
Things that affect Identification 33% 27% 23% 1 9% 
Positive remarks about gender 0% 27% 23% 24% 
Negative remarks about gender 0% 1 3% 9% 43% 
Remarks about outgroup 1 3% 0% 5% 0% 
Remarks about the unimportance of 0% 0% 23% 1 4% 
gender 
Miscellaneous 7% 1 3% 9% 0% 
6.2 Identity Salience 
1 3 1  
Table 20 shows the means and standard deviations for gender salience, on the 
Spontaneous Self Concept Scale, occupational salience on the Spontaneous Self 
Concept Scale and the Gender Salience Scale. One-way ANOVA's were performed to 
test for differences in gender salience between the different groups in the study. 
1 32 
Identity Salience, Gender Groups and Occupations 
As shown in Table 20 no significant differences were found in identity salience 
between the men and women in the sample, or between men and women in each 
occupation. Hypothesis 5 that women would have higher self-concept gender salience 
than men was not supported. There were also no differences in identity salience between 
occupations. 
Identity Salience and Status Groups 
As shown in Table 2 1  no differences in gender or occupational salience were 
found between the status groups. 
Table 2 1  
Mean Identity Salience Scores for Each Status Group 
Salience Scale 
n= 
Gender- Spontaneous Self­
Concept 
Occupations - Spontaneous Self­
Concept 
Low (L) 
57 
\ .94 
( 1 .92) 
2 .92 
(2. 1 5) 
Status 
Equal (E) 
99 
1 .85 
( 1 .85) 
2 .64 
( 1 .99) 
Gender Salience Scale 2.35 2.28 
(0.47) (0.44) 
Note: (i) standard deviations and degrees of freedom are in brackets, 
(ii) n ' s  may vary depending upon missing values. 
High (H) 
49 
1 .66 
( 1 .60) 
3 .07 
( 1 .97) 
2.32 
(0.48) 
Identity Salience and Majority and Minority Groups 
F 
0.3 1 
(2, 1 88) 
0.46 
(2, 1 88) 
0.59 
(2, 202) 
Comparing the mean salience scores for the minority and majority groups' 
results in one significant difference as shown in Table 22 (p. 1 36). Hypothesis 6 is  
only supported on the Gender Salience Scale on which the majority group had a higher 
mean score than the minority group, F ( 1 , 205) = 5 .27, p < .05 . The relationship 
between gender salience and occupational salience was explored. Contrary to 
predictions (Hypothesis 8) there were no correlations between occupational and gender 
Table 20 
Mean Identity Salience Scores (for each sample) 
Salience Female Dominated Male Dominated Occupations Combined Sample Occupations (dominated by) 
Scale Occu2ations 
Female Male F Female Male F Female Male F Female Male F 
n= 79 3 1  45 52 1 24 83 1 1 0 97 
Gender Spontaneous 2 .05 2.00 0.02 1 .63 1 .53 0.08 1 .90 1 .7 1  0 .52 2 .04 1 .58  3 . 1 9  
Self Concept ( 1 .96) ( 1 .96) ( 1 , 1 0 1 )  ( 1 .63) ( 1 .56) ( 1 ,88) ( 1 .85) ( 1 .72) ( 1 , 1 9 1 )  ( 1 .95) ( 1 . 59) ( 1 , 1 9 1 )  
Occupation 3.0 1 2.75 0.35 2.77 2.57 0.20 2.92 2 .64 0.90 2.94 2.67 0.89 
Spontaneous Self ( 1 .99) (2.08) ( I , I O ! )  2 . 1  I 2.00 ( 1 ,88) (2.03) (2.02) ( I ,  1 9 1 )  (2.0 1 )  (2 .04) ( 1 , 1 9 1 )  
Concept 
Gender Salience Scale 2.40 2.25 2.45 2.20 2.33 1 .90 2.32 2.30 0. 1 5  2.35 2.27 1 .70 
{0.442 (0.45) ( 1 , 1 0 1 )  0.42 0.50 ( 1 ,95) {0.442 (0.482 ( 1 , 2052 {0.45) (0.462 {1 ,205) 
Note (i) standard deviations and degrees of freedom are in brackets 
(i i) n's may vary due to missing values 
1 34 
salience for the full sample (Table 26� p. 1 39) or for the majority or minority groups as 
shown in Table 23. 
Table 24 shows the relationship between gender and occupational salience, for 
the majority and minority groups. As predicted (Hypothesis 9) the majority group also 
had significantly higher occupational salience than gender salience, t = -4.27, p < .0 1 .  
Contrary to predictions the minority group also had significantly higher occupational 
salience than gender salience, t= -3 .40, p < .0 1 .  
Table 22 
Mean Identity Salience for Minority and Majority Groups 
Salience Group 
Scale Minority Majority 
n =  76 
Gender Spontaneous 1 .77 
Self Concept ( I .  77) 
Occupation Spontaneous 2.76 
Self Concept (2 .09) 
Gender Salience Scale 2.22 
(0.43) 
* p < .05 
Note: (i) Standard deviations and degrees of freedom are in brackets. 
(ii) Note: n's may vary due to missing values. 
Table 23 
1 3 1  
1 . 85 
( 1 . 83) 
2.84 
(2.00) 
2.37 
(0.46) 
Correlations between Gender Salience and Occupational Salience 
Group 
Minority Majority 
n =  7 1  1 22 
Salience Scale 2 3 2 
I .  Gender Salience SSC 
2.  Occupational Salience .2 1 . 1 0  
3 .  Gender Salience Scale -.02 . 1 1 .09 -.0 1 
Note: n 's  may vary due to missing values. 
F 
0.08 
( I ,  1 9 1 )  
0.08 
( I ,  1 9 1 )  
5.27* 
(1 , 205) 
3 
1 35 
Table 24 
Paired Sample T -Test Results for Gender and Occupational Salience (total 
sample) 
Spontaneous Self Concept 
Scale 
n=  
Gender Salience 
Occupational Salience 
··· p < .OOI 
Minority 
7 1  
1 .77 
( 1 .77) 
2.76 
(2.09) 
-3.40***  
(70) 
Group 
Majority 
1 22 
1 .85 
( 1 .83) 
2.84 
(2.00) 
-4.27***  
(1 2 1) 
Note: standard deviations and degrees of freedom are in brackets 
Gender Salience, SIT and Distinctiveness Theory 
In order to test hypothesis 3 the sample was grouped into those individuals who 
worked with members of the outgroup (outgroup present) and those who did not work 
with any members of the outgroup (outgroup absent). Table 25 shows that there was no 
differences in gender salience (Self Concept and GSS) between those respondents who 
work with a member of the outgroup and those that do not. 
Table 25 
Gender Salience Means for OutgrOUP Present and Outgroup Absent. 
Salience 
Scale 
n= 
Gender Spontaneous Self Concept 
Gender Salience Scale 
* p < .05, ** p < .0 1 ,  *** P < .00 1 
Presence of Out group 
Out group Outgroup 
Absent Present 
22 1 79 
1 .50 1 .86 
( 1 .54) ( \ .83) 
2.42 
(0.54) 
2.3 1 
(0.44) 
Note (i) standard deviations and degrees of freedom are in brackets 
(ii) n 's  may vary due to missing values 
F 
0.72 
( 1 , 96) 
0.3 1 
(I, I 03) 
1 36 
Table 26 presents correlations between the salience scales and other variables. 
Hypothesis 4 that the number of same gender colleague would be related to gender 
salience was not supported, r = .0S , p  >.OS . The only significant correlation between 
gender salience and the other variables measured, is a positive relationship between the 
Gender Salience Scale and age, r =. 1 7, p < .05. The older respondents were more likely 
use gender as a basis for attitudes about others. The number of same gender colleagues 
was not related to gender salience, r = .OS, p >.OS . Hypothesis 7 that Gender Salience 
in the self-concept would be related to the Gender Salience Scale was not supported, r = 
.0S, p >  .05. 
Table 26 
Correlations between Gender Salience and other Variables (total sam12le N= 207) 
Variable 2 3 4 5 6 7 8 9 10 1 1  
I .  Gender Salience Spontaneous 
Self Concept 
2. Occupational Salience . 1 4  
Spontaneous Self Concept 
3 .  Gender Salience Scale .05 .03 
4. Acceptance to [ngroup -. 1 4  -.05 .05 
5.  Acceptance to Outgroup .03 -.08 .04 . 1 3  
6. Percent of Ingroup in .05 .05 . 10 .30" -.36" 
Workgroup 
7. Percent of Outgroup in - 03 -.05 -.09 -.3 1 " .36" 
Workgroup 
8. Gender Identification Full -.09 -.08 -.04 -.09 -.07 -.20" . 1 9" 
Scale 
9. Gender Identification -. 1 0  -.07 -.03 -.2 1 "  .00 -.28" .27" 
Cognitive Emotional Subscale 
10. Gender Identification .03 -.04 .0 1  .25" -. 17' . 1 80 -. 1 7' 
Individual Group Opposition 
1 1 . Discrimination .07 -.08 -.03 -. 1 2 -.09 .00 - .00 -.07 .02 -.240.' 
1 2. Ase .08 .06 . 1 7' .06 -.22" . 1 90' -. 1 9'· -.05 -,07 .03 -.03 
• p < .05, ··  p < .0 1 ,  • •• p <.OOI 
Note: n's may vary due to missing values. 
....... 
W -...l 
1 38 
Table 27 summarises the information gathered from the free response section on 
gender salience. In the female dominated occupation sample the majority of men 
mentioned the causes of differences between genders (38%), the category included 
comments about the physical differences between the genders, or socialization 
differences. The second largest group of comments from the men in the female 
dominated occupation talked about the distinction between how they think it should be 
and what actually happens in reality (38%). The largest group of comments from the 
women in the female dominated occupations mentioned other reasons or causes for 
attitudes such as culture and religion (44%). 
Within the male dominated occupation sample the largest category of comments 
from men mentioned the other causes or reasons for attitudes (36%). A larger number 
of men in male dominated occupations mentioned the upholding of tradition (32%) as a 
reason for gender differences than men in the female dominated occupation sample 
(0%). Women in the male dominated occupation sample most often mentioned 
alternative reasons for attitudes (38%). 
Table 27 
Free Answers on Gender Salience 
Female Dominated Male Dominated 
Occupations Occupations 
Gender Male Female Male Female 
n = 8  n = 1 6  n = 22 n = 1 3  
Causes of Gender Differences 38% 38% 1 8% 8% 
Belief in Different Gender Roles 0% 1 3% 1 8% 8% 
Distinction between Reality and 38% 1 9% 5% 8% 
Beliefs 
Upholding of Tradition 0% 25% 32% 3 1%  
Challenging Tradition 0% 0% 5% 23% 
Alternative bases for attitudes 1 3% 44% 36% 38% 
Miscellaneous 0% 0% 32% 1 5% 
6.3 Beliefs about the Intergroup Situation 
Status 
1 39 
Chi square tests were performed on status to test the distribution of beliefs. For 
the overall sample the distributions of perceptions of ingroup status, '1: (2, N = 205) = 
2 1 . 1 1 ,  p < .00 I were unequal, more respondents believed their status to be equal than 
high or low. 
Differences between men and women and those working in male and female 
dominated occupations were also explored. As shown in Table 28 status beliefs were 
not independent of occupation, X2 (2, N= 205) = 1 6.04, p < .00 1 .  The perceptions of 
the status of the groups were not independent of gender, X2 (2, N =  205) = 23.79, p < 
.00 1 .  Hypotheses 1 2  and 1 3  were not supported, men and women in female dominated 
occupations did not have significantly different beliefs about the status of the groups, X2 
(2, N= 205) = 1 .67, p > .05. Men and women in male dominated occupations had 
different distributions of beliefs about the status of the groups, X2 (2, N = 97) = 42.68, p 
< .0 1 .  Hypothesis 1 1  was supported, the majority of female engineers (7 1 %) saw 
themselves of low status, X2 (2, N = 45) = 3 1 .60, p < .00 1 . The majority of male 
engineers saw themselves as of equal (42%) or high status (48%), X2 (2, N = 52) = 
1 3 .42, p < .00 1 .  
Table 29 shows the beliefs about the intergroup situation for the majority and 
minority groups. More of the minority group members saw themselves as of low status 
than majority group members, X2 ( 1 ,  N = 205) = 33 .92, p < .0 1 
Table 28 
Distribution of Respondents in their Beliefs about the Intergroup Situation 
Measure Female Dominated Occu2ations Male Dominated Occu2ations 
Percent of Percent of i Percent of Percent i 
Females Males females of males 
n= 77 3 1  45 52 
Low Status 1 7  23 1 .67 7 1  1 0  42.68** 
Equal Status 65 52 24 42 
High Status 1 8  26 4 48 
Legitimate 69 47 4.72* 27 42 2 .59 
Illegitimate 3 1  53 73 58 
Stable 77 80 0. 1 1  6 1  77 2.54 
Unstable 23 20 39 23 
* P < .05, ** p< .0 1 ,  
Note: n ' s  may vary due to missing values 
Combined Sam21e 
Percent of Percent i 
Females of males 
1 22 83 
37 14 23 .79**  
50 46 
1 3  40 
53 44 1 .73 
47 56 
7 1  78 1 .08 
29 22 
Occu2ations (dominated b�2 
Percent of 
Female 
1 08 
1 9  
6 1  
20 
63 
37 
78 
22 
Percent 
of Male 
97 
38 
34 
28 
35 
65 
70 
30 
...-
.+>. o 
t 
1 6.04* 
1 5 .60** 
1 .82 
Table 29 
Beliefs about the Intergroup Situation held by Majority and Minority Group 
Scale Group 
Percent Minority Percent Majority X2 
n= 76 1 29 
Low Status 5 1  1 4  33 .92**  
Equal Status 36 56 
High Status 1 3  30 
Legitimate 35 58 1 0.5 1 * *  
I l legitimate 65 42 
Stable 69 77 1 .50 
Unstable 3 1  23 
** P < .0 1 ,  note n's may vary due to missing values 
Legitimacy and Stability 
The distributions of the beliefs of legitimacy were not unequally distributed. 
The Chi square shows unequal distributions of the perceptions of stability of the 
intergroup relations, X2 ( 1 ,  N= 1 99) = 45.35, p < .00 1 ,  more respondents believed the 
intergroup situation to be stable than believed it to be unstable. As predicted 
(hypothesis 1 7) stability beliefs were not independent of legitimacy beliefs X2 ( 1 ,  N = 
1 98) = 28.33, p < .00 1 .  
1 4 1  
Legitimacy beliefs were not independent of occupation, X2 ( 1 ,  N = 202) = 1 6.60, 
p < .0 1 .  The majority of nurses perceived the situation as legitimate, and the majority of 
engineers perceived the situation as illegitimate. Within female dominated occupations 
legitimacy beliefs were dependent on gender, X2 ( 1 ,  N = 202) = 4.72, p < .05 . The 
majority of women in female dominated occupations saw the situation as legitimate and 
the majority of men saw the situation as i llegitimate. 
A larger percentage of the minority group saw the intergroup situation as 
i l legitimate than the majority group, X2 ( 1 ,  N = 202) = 1 0.5 1 ,  P < .0 1 .  
Table 30 shows the distribution of the beliefs of legitimacy and stability of the 
different status groups. As predicted (Hypothesis 1 8) status is not independent of 
legitimacy beliefs, X2 (2, N = 202) = 1 75 .54, p < .0 1 ,  or stabil ity beliefs, X2 (2, 202) = 
29.86, p < .0 1 
Table 30 
Perceptions of Legitimacy and Stability for Each Status Group 
Status 
Scale Percent of Percent Percent X2 
Low of Equal of High 
n =  56 97 49 
Legitimate 2 98 8 1 75 .54** 
Il legitimate 98 2 92 
Stable 60 92 55 29.86** 
Unstable 40 8 45 
** p < .O I  
Acceptance 
1 42 
Table 3 1  shows the means and standard deviations of acceptance towards the 
ingroup and acceptance of the outgroup for each sample. No significant difference in 
perceived acceptance was found between men and women. Respondents in male 
dominated occupations felt significantly less accepted than those in female dominated 
occupations, F ( 1 ,  1 83) = 9.66, p < .0 1 .  Men in female dominated occupations were 
more accepting of the outgroup than women in female dominated occupations, F ( 1 ,  
1 03) = 7. l I , p  < .0 1 .  
Women in male dominated occupations felt significantly less accepted than men 
in male dominated occupations, F( 1 ,  79) = 1 4.72, p < .0 1 .  Men in male dominated 
occupations were less accepting towards the outgroup than women in male dominated 
occupations, F( 1 ,  77) = 3 1 .29, p < .0 1 .  
Mean acceptance scores were also analysed by ingroup status the results are 
shown in Table 32. There were significant differences in acceptance of the ingroup 
between low, equal and high respondents F (2, 1 8 1 ) = 1 3 .0 I , p  < 0 1 .  As predicted 
(hypothesis 1 4), the low status group perceived significantly lower acceptance (M = -
0.43) of the ingroup than the equal CM = 1 .22) and the high status CM = 1 .29) groups. 
There was no significant difference between the equal group and the high status group. 
Table 3 1  
Mean Acceptance Scores (for each sample) 
Scale Female Dominated OccuEations Male Dominated OccuEations 
Female Male F Female 
n=  74 3 1  45 
Acceptance Towards 1 .38 0.74 2.2 1 -0.56 
I ngroup ( 1 .73) (2.56) ( 1 , 1 02) (2. 1 6) 
Acceptance of Outgroup 0.70 1 .90 7 . 1 1 ** 1 .86 
{2 .07) (2. 1 82 {1 , 1 032 {2.05) 
** p< .0 1 ,  ***  p < .00 ]  
Note (i) standard deviations and degrees of freedom are in brackets 
(ii) n 's  may vary due to missing values 
Male F 
36 
1 . 1 9  1 4.72***  
( 1 . 88) ( 1 ,79) 
-0.74 3 1 .29***  
(2.062 {1 ,77) 
Combined SamE1e OccuEations (dominated bi:2 
Female Male F Female Male F 
1 1 8 67 1 05 8 1  
0.64 0.99 1 .07 1 . 1 9 0.22 9.66** 
(2. 1 2) (2.2 1 ) ( 1 , 1 83) (2.02) (2.2 1 )  ( 1 , 1 83) 
1 . 1 4  0.50 3.33 1 .06 0.7 1 1 .05 
(2. 1 3) {2 .49) {I , 1 82) (2 . 1 7) (2.42) {1 , 1 82) 
Table 32 
Mean Acceptance Scores for Each Status Group 
Status 
Scale Low (L) Equal (E) 
n= 54 89 
Acceptance Towards lngroup -0.43 1 .22 
(2.25) ( 1 .82) 
Acceptance of Outgroup 1 .30 1 .05 
(2.34) (2. 1 8) 
* p < .05, ** p < .00 1 
Note (i) standard deviations and degrees of freedom are in brackets 
(ii) n 's  may vary due to missing values 
High (H) 
4 1  
1 .29 
(2. 1 4) 
0.20 
(2 .26) 
Post 
F Hoc 
1 3 .0 1 ***  L < E 
(2, 1 8 1 )  L < H 
3 .04* L > H 
(2, 1 80) 
(iii) For equal variances the Bonferroni Post Hoc test was used, for unequal variances the Dunnett T3 Post Hoc test was used. 
1 45 
Significant differences were found in the acceptance of the outgroup between the 
respondents of low, equal and high status, F (2, 1 80) = 3.04, P < .05. As predicted the 
low status group (M = 1 .30) was more accepting to the outgroup than the high status 
group (M = 0.20). 
Acceptance was also examined for the minority and majority groups the results 
are shown in Table 33. There were significant differences in acceptance between those 
in the minority group and those in the majority group. The minority group felt less 
accepted than the majority group, F ( 1 , 1 83) = 1 9.22,p < .0 1 .  The minority group was 
more accepting of the outgroup than the majority group, F ( 1 , 1 82) = 26.20,p < .0 1 .  
Hypothesis 1 6  that minority group females (female engineers) would feel less accepted 
than minority group males (male nurses) was supported. The minority women felt less 
accepted (M= -0.56) than minority group males (M= 0.74), F ( 1 ,74) = 5.70,p < .05. 
Male majority group members (male engineers) were less accepting of the outgroup (M 
= -0.74) than the female majority group members (female nurses) (M= 0.70), F ( 1 ,  1 07) 
= 1 1 .60,p < .00 1 .  
Correlations between acceptance of the ingroup, acceptance of the outgroup and 
other variables were analysed. The results are shown in Table 26 (p. 1 37). Acceptance 
of the ingroup was positively related to the number of the ingroup in the immediate 
workgroup, r = .30,p < .0 1 .  The higher the numbers of the ingroup the more accepted 
the group felt. Acceptance towards the outgroup was related positively to the number of 
the outgroup in the workplace, r = . 36,p < .0 1 .  The higher the number of outgroup 
members in the workplace the more accepting the group was to them. As predicted 
(hypothesis 1 5) acceptance was related to gender identification. Acceptance towards 
the outgroup was negatively related to the individual group opposition subscale of 
gender identification, r = -. 1 7,p  < .05, and negatively to age, r = -.22,p < .0 1 .  Low 
individual group opposition was related to low acceptance towards the outgroup. 
Acceptance towards the outgroup was lower in the older respondents. Acceptance of 
the ingroup was negatively related to the cognitive/emotional aspects of gender 
identification, r = -.2 1 , p  < . 0 1 ,  high gender identification (cognitive/emotional aspects) 
was related to less perceived acceptance of the ingroup. Acceptance of the ingroup was 
also positively related to the individual/group opposition subscale of identification, r = 
1 46 
.25 , p  < .00 1 .  High opposition between individual and group needs was related to low 
perceived acceptance of the ingroup. 
6.4 Social Change Bel ief Systems -
Achieving Positive Distinctiveness 
Total Sample 
Table 34 shows the means and standard deviations for the ingroup social change 
beliefs. One-way ANOVA's were performed to test for differences in social change 
beliefs between the different groups in the study. 
Social Change Belief Systems and Gender 
Table 34 shows the means and standard deviations for the social change belief 
scales, for achieving positive distinctiveness. There were significant differences 
between men and women on social change beliefs. From the overall sample women 
had lower social mobility beliefs than men, F ( 1 ,  1 78) = 1 9. 1 9, p  < .00 1 ,  and higher 
social competition beliefs than men, F ( 1 ,  1 76) = 1 0 1 .50, P < .00 1 .  Women also had 
higher social creativity scores than men, F ( 1 ,  1 79) = 1 6.07, p < .00 1 .  
Social Change Belief Systems and Occupations 
There were significant differences between male and female dominated 
occupations and between men and women on social change beliefs. Respondents in 
female dominated occupations scored lower on social mobil ity beliefs than those in 
male dominated occupations, F ( 1 , 1 78) = 5.93 , p  < .05. Respondents in female 
dominated occupations had lower social creativity scores than those in male dominated 
occupations, F = ( 1 ,  1 79) 4.58, p < .05. 
Table 33 
Mean Acceptance Scores for Minority and Majority Groups 
Scale Minorit� Majori� 
N Female Male F Female Male F Minority Majority F 
45 3 1  74 36 76 1 09 
Acceptance Towards -0.56 0.74 5 .70* 1 .38 1 . 1 9 0.27 -0.03 1 .32 1 9.22***  
I ngroup (2. 1 6) (2.56) ( 1 ,74) ( 1 .74) ( 1 .88) ( 1 , 1 07) (2 .40) ( 1 .78) ( 1 , 1 83) 
Acceptance of Outgroup 1 .86 1 .90 .0 1  0.70 -0.74 1 1 .60***  1 .88 0.24 26.20***  
(2.052 {2. 1 82 {1 ,73) {2 .072 {2.062 ( I ,  1 072 {2.09) {2 . 1 72 {1 , 1 832 
* p < .05, **  p< .0 1 ,  *** P < .00 1 
Note: (i) standard deviations and degrees of freedom are in brackets 
(i i) n 's  may vary due to missing values 
Table 34 
Mean Ingroup Social Change Beliefs for each Sample 
Scale Female Dominated OccuEations Male Dominated OccuEations Combined SamEle OccuEations (dominated by) 
Female Male F Female Male F Female Male F Female Male F 
n= 6 1  22 45 52 1 06 74 83 97 
Ingroup Social Mobility 0.64 0.55 0.06 0.33 1 .9 1  30.23***  0.5 1 1 . 50 1 9. 1 9***  0.6 1 1 . 1 8  5.93 * 
( 1 .24) (2.07) ( 1 ,8 1 ) ( I . 79) (0.95) ( 1 ,95) ( 1 .50) ( 1 .50) ( 1 , 1 78) ( 1 .49) ( 1 .60) ( 1 , 1 78) 
Ingroup Social Competition 0.78 - 1 . 1 1 1 1 .60***  1 . 5 1  - 1 .78 1 95 .9 1 *** 1 . 1 0  - 1 .59 1 0 1 .50*** 0.27 -0.26 2.57 
(2.36) ( 1 .78) ( 1 ,79) ( 1 . 1 8) ( 1 . 1 3) ( 1 ,95) ( 1 .97) ( 1 .38) ( I ,  1 76) (2.37) (2.0 1 )  ( 1 , 1 76) 
Ingroup Social Creativity -0.66 - 1 .90 1 0.42**  0.03 -0.98 1 4 .6 1 *** -0.37 - 1 .25 1 6.07*** -0.99 -0.5 I 4.58* 
(1 .6 1 ) ( 1 .34) ( 1 ,8 I) ( 1 .40) ( 1 .21) ( 1 ,95) ( 1 .56) ( 1 .3 1) ( I ,  1 78) (1 .63) ( 1 .39) ( I , I 78) 
* p < .05, * *  p< .0 1 ,  ***  p < .00 1 
Note: (i) standard deviations and degrees of freedom are in brackets 
(ii) n 's  may vary due to missing values 
1 49 
Social Change Belief Systems and Status 
Differences in mean social change beliefs were also examined for each of the 
status groups. These results are shown in Table 35 .  There were significant differences 
in all three social change beliefs between the low equal and high groups. 
There were significant differences between the status groups on social mobility 
beliefs, F (2, 1 76) = 1 6.54, P < .0 1 .  The low status group (M = 0.03) had significantly 
lower social mobility beliefs than both the equal (M = 1 .24) and the high status groups 
(M = 1 .67). Differences in social competition beliefs were significant, F (2, 1 74) = 
1 3 .8 1 , p  < .0 1 .  The low status group (M = 0.98) had significantly higher social 
competition beliefs than both the equal (M = -0.2 1 )  and the high status groups CM = -
1 .46). The equal group had significantly higher social competition beliefs than the high 
status group. There was also significant differences in social creativity beliefs, F (2, 
1 76) = 5 .6 1 , p  < .0 1 .  The low status group (M = -0.20) had significantly higher social 
creativity beliefs than both the equal (M = -0.93) and the high status group (M = - 1 . 1 8). 
Social Change Belief Systems and Minority and Majority Groups 
Table 36 shows the means and standard deviations for the minority and majority 
groups. There were no significant differences between the majority and minority group 
on social creativity beliefs. The minority group had lower social mobility beliefs than 
the majority group, F ( l ,  1 78) = 1 2.02, p < .0 1 ,  and higher social competition beliefs 
than the majority group, F ( 1 ,  1 76) = 1 0.44, p < .0 1 .  
Table 35  
Mean Ingroup Social Change Beliefs for Each Status Group 
Scale Low (L) Equal (E) 
n= 55 97 
Ingroup Social Mobility 0.03 1 .24 
( 1 .67) ( 1 .45) 
Ingroup Social Competition 0.98 -0.2 1  
( 1 .67) (2 .40) 
Ingroup Social Creativity -0.20 -0.93 
( 1 .46) ( 1 .58) 
** p< . 0 1 ,  * * *  P < .00 1 
Note: (i) standard deviations and degrees of freedom are in brackets 
(ii) n 's  may vary due to missing values 
Status Post Hoc 
High (H) F 
27 
1 .67 1 6.54*** L<E 
(0.76) (2, 1 76) L<H 
- 1 .46 1 3 . 8 1 ***  L>E 
( 1 .  1 3) (2, 1 74) L>H 
E>H 
- 1 . 1 8  5 . 6 1  **  L>E 
( 1 .07) (2, 1 76) L>H 
(iii) For equal variances the Bonferroni Post Hoc test was used, for unequal variances the Dunnett T3 Post Hoc test was used. 
VI o 
Table 36 
Mean Ingroup Social Change Beliefs for Minority and Majority Groups 
Minorit� GrouE Majori� GrouE 
Scale Female Male F Female Male F Minority Majority F 
n= 45 22 6 1  52 67 1 1 3 
Ingroup Social Mobility 0.33 0.55 0. 1 9  0.64 1 .9 1  36.48***  -.40 1 .22 1 2.02*** 
( 1 .  79) (2.07) ( I ,  65) ( 1 .24) (0.95) ( 1 , 1 1 1 ) ( 1 .87) ( \ .28) ( 1 , 1 78) 
Ingroup Social Competition 1 .5 1  - 1 . 1  I 5 1 .3 1  *** 0.78 - 1 .78 50.92*** 0.65 -0.42 1 0.44*** 
( 1 . 1 8) ( 1 .78) ( 1 ,65) (2.36) ( 1 . 1 3) ( 1 , 1 09) ( 1 .86) (2.28) ( 1 , 1 76) 
Ingroup Social Creativity 0.04 - \ .90 28.95 *** -0.66 -0.98 1 . 36 -0.60 -0.8 1  0.79 
***  p < .00 1 
(1 .40) ( 1 .34) (1 ,65) ( 1 .6 1 )  (1 .2 1) (1 , 1 1 1) (1 .65) (1 .44) (I� 1 78} 
Note: (i) standard deviations and degrees of freedom are in brackets 
(ii) n 's  may vary due to missing values 
-
\.J) 
-
Social Change Belief Systems and Identification 
Total Sample 
1 52 
As shown in Table 37 a positive correlation was found between gender 
identification as a full scale and in group social mobility, r = .2 1 ,  p < .05, and ingroup 
social creativity, r = .30, p < .0 1 .  High gender identification was related to high social 
mobility and high social creativity beliefs. Hypothesis 28 that gender identification 
would be related to ingroup social competition beliefs was not supported on the full 
scale, r = -.09, p >.05. The individual group opposition subscale was negatively 
correlated with ingroup social competition (achieving status), r = -.27, p < 0 1 .  Low 
individual group opposition was related to low beliefs in social competition to achieve 
status 
Table 37 
Correlations between Gender Identification Scales and Social Change Beliefs (total 
sample N = 207). 
Variable 2 3 
I .  Gender Identification Full Scale 
2. Gender Identification CognitivefEmotional e .92***  
3 .  Gender Identification Individual/Group Opposition .36*** -.03 
4. Ingroup Social Mobility .2 1 * . I 5 . 1 3  
5 .  Ingroup Social Creativity .30**  .44* *  -. 1 3  
6. Ingroup Social Competition (achieving status) -.09 .02 -.27* *  
* p < .05, * *  p < .0 1 ,  ***  p < .00 1 ,  
Gender Groups 
Table 38 shows the correlations between gender identification and social change 
beliefs for men and women. For the male sample gender identification was positively 
1 53 
related to social mobility r = .35 ,  p < .0 1 ,  high gender identification was associated with 
high social mobility beliefs .  Gender identification was also positively correlated with 
social creativity, r = .29, P < .05, high gender identification was associated with high 
social creativity beliefs. Identification was negatively correlated with social 
competition, r = .25, P < .05, high gender identification was associated with low social 
competition beliefs. 
For the females gender identification was positively related to ingroup social 
creativity, r = .28, p < .0 1 ,  high gender identification was associated with high social 
creativity beliefs. Social mobility was correlated with individual/group opposition, r = 
.32, P < .00 1 ,  individual/group opposition was associated with low social mobility 
beliefs. The cognitive/emotional subscale was positively related to social competition, r 
= .23, p < .05, high identification on cognitive/emotional aspects was associated with 
high social competition beliefs. 
Status 
Table 39 shows the correlations between gender identification and social change 
beliefs for each of the status groups. For the low status group gender identification was 
positively related to ingroup social creativity, r = .28, p < .05. High gender 
identification was associated with high social creativity beliefs. Individual/group 
opposition was negatively related to social competition r = -.32, p < .05, low opposition 
between individual and group needs was associated with low social competition beliefs. 
Hypothesis 20 that gender identification would be negatively related to social mobility 
was not supported, r = .08 , p >.05. 
For the respondents who saw themselves of equal status gender identification 
was positively related to social mobility, r = .32, p < .0 1 ,  and social creativity, r = .2 1 , p  
< .05. High gender identification was associated with high social mobility and high 
social creativity beliefs. 
Table 38 
Correlations between Gender Identification Scales and IngrouQ Social Change Beliefs (men and women} 
Variable 
n=  
1 .  
2 .  
3 .  
4 .  
5 .  
6 .  
Gender Identification full scale 
Gender Identification CognitivelEmotional 
Gender Identification Individual/Group Opposition 
Ingroup Social Mobility 
Ingroup Social Creativity 
IngrouE Social ComEetition {achieving status} 
• p < .05, •• p < .01,  •••  P < .00 1 ,  
Male 
75 
2 
.93** 
.48***  . 1 4  
.35**  .38*** 
.29* .42** 
-.25 * -. 1 6  
Female 
1 05 
3 2 
.9 1 *** 
.29***  -. 1 3  
. 1 3  -.0 1 -. 1 4  
-. 1 8  .28** .35**  
-.20 . 1 7  .23* 
3 
.32***  
-. 1 5  
-. 1 5  
For the high status group hypothesis 20 that gender identification would be 
negatively related to social mobility beliefs was not supported. Individual/group 
opposition was however positively related to social mobility beliefs, r = .46, p < .05. 
Opposition between the individual and group was associated with low social 
competition beliefs for the high status group. 
Majority and Minority Groups 
1 55 
Table 40 shows the relationship between gender identification and social change 
bel iefs for the majority and minority groups. For the minority group gender 
identification was positively related to ingroup social creativity, r = .47, p < .00 1 and 
in group social competition, r = .40 ,p < .00 1 .  High gender identification was related to 
high social creativity and social competition beliefs. Individual/group Opposition was 
positively correlated with ingroup social mobility, r = .28, p < .05. High 
individual/group opposition was related to low social mobility beliefs. 
For the majority group gender identification was positively correlated with 
social mobil ity, r = .46, P < .00 1 .  High gender identification was related to high social 
mobility beliefs. Gender identification was negatively related to social competition 
(achieving status), r = -.39, p < .00 1 .  High gender identification was related to low 
social competition beliefs. 
Table 39 
Correlations between Gender Identification Scales and Ingroup Social Change Beliefs for Each Status Group 
Variable Low Status Equal Status High Status 
n ==  55 97 27 
2 3 2 3 2 3 
Gender Identification Full scale 
Gender Identification Cognitive/Emotional .92** .93* *  .93 ***  
Gender Identification Individual/Group Opposition .39** .00 .45***  . 1 0  .4 1 **  . 1 1  
I ngroup Social Mobility .08 .06 .07 .32* *  .29** .20 .27 .2 1 .46* 
Ingroup Social Creativity .28* .35** -. 1 4  .2 1 * .27* *  -.07 . 1 5  .29 -.24 
Ingroup Social Competition (achieving status) . 1 1  .2 1 -.32* -. 1 5  - . 1 3  -.0 1 .04 .0 1  - . 1 1  
* P < .05, * *  p< .0 1 ,  *** P < .00 1 
Note: n's may vary due to missing values 
Table 40 
Correlations between Gender Identifications Scales and Ingroup Social Change Beliefs for Majority and Minority Groups. 
Variable 
I .  Gender Identification Full Scale 
2. Gender I dentification Cognitive/Emotional 
3 .  Gender Identification Individual/Group Opposition 
3 .  Ingroup Social Mobility 
4 .  Ingroup Social Creativity 
5 .  Ingroup Social Competition (achieving status) 
* *  p< .0 1 ,  ***  P < .00 1 
.89***  
.34* *  
- .04 
.47***  
.40* *  
Minority 
67 
2 
- . 1 3  
- . 1 6  
.62*** 
.59***  
3 
.94***  
.44***  
.28* .46***  
-.2 1 .03 
-.27* -.35 *** 
Majority 
1 1 3 
2 
. 1 2  
.48***  
.07 
-.39***  
3 
. 1 3  
-. 1 1  
-.03 
...... 
VI -.) 
1 58 
Social Change Belief Systems, Legitimacy and Stability 
Social change beliefs were also compared according to the respondents' beliefs 
about legitimacy and stability. The results for this analysis are shown in Table 4 1 . As 
predicted (Hypothesis 2 1 )  respondents who saw the intergroup situation as legitimate 
had higher social mobil ity bel iefs than those who saw the situation as illegitimate, F = 
7. l 4 , p  < .0 1 .  Hypothesis 26 that respondents who see the intergroup situation as 
il legitimate would have higher social competition beliefs than those who saw the 
situation as legitimate was not supported, F = 0.46, P > .05. 
Hypotheses 22, 24 and 27 that there would be differences in social change 
beliefs among respondents who saw the intergroup situation as stable and unstable were 
not supported. There were no differences in social change beliefs between respondents 
who saw the situation as stable or unstable. 
Social Change Belief Systems and Acceptance 
Table 42 shows the correlations between acceptance of the ingroup and 
acceptance towards the outgroup and social change beliefs, for the total sample. As 
predicted (Hypothesis 1 9) acceptance of the ingroup was positively related to ingroup 
social mobility, r = .25 , p  < .0 1 .  High in group social mobility beliefs were related to 
perceived acceptance of the ingroup. Acceptance towards the out group was negatively 
related to social mobility beliefs, r = -. l 7, p  < .05. High acceptance of the outgroup was 
associated with low ingroup social mobility beliefs. As predicted (hypothesis 23) 
ingroup social creativity was negatively related to acceptance of the ingroup, r = -. 1 8, p  
< .05. Low acceptance of the ingroup is associated with high belief in social creativity. 
As predicted (hypothesis 25) ingroup social competition (achieving status) was 
negatively related to acceptance towards the out group, r = -. 1 9, P < .05. High social 
competition beliefs were associated with less acceptance of the outgroup. 
Table 4 1  
Mean Ingroup Social Change Beliefs for Legitimacy and Stability Beliefs 
Scale Legitimac� Beliefs Stability Beliefs 
Legitimate Illegitimate F Stable Unstable F 
n= 94 82 1 32 4 1  
Ingroup Social Mobility 1 .25 0.63 7. 1 4** 1 .00 0.73 0.96 
( 1 .44) ( 1 .62) ( 1 , 1 74) ( 1 .6 1  ) ( 1 .4 1 )  ( 1 , 1 7 1 )  
lngroup Social Competition -0. 1 7  0.06 0.46 -0. 1 4  0. 1 I 0.4 1 
(2.40) ( 1 .90) ( 1 , 1 72) (2.2 I )  (2.04) ( I ,  I 7 1 )  
Ingroup Social Creativity -0.9 1 -0.55 2.37 -0.69 -0.97 1 .05 
( 1 .60) ( l .43) ( 1 , 1 74) ( 1 .57) ( 1 .42) ( 1 , 1 73) 
•• p< .01 
Note: (i) standard deviations and degrees of freedom are in brackets 
(ii) n 's  may vary due to missing values 
1 60 
Table 42 
Correlations between Acceptance and Ingroup Social Change Beliefs eN =207) 
Variable 2 
I .  Acceptance of the ingroup 
2. Acceptance towards the outgroup . 1 3  
3 .  Ingroup Social Mobility .25** -. 1 7* 
4. Ingroup Social Creativity -. 1 8* .09 
5. Ingroup Social Competition (achieving status) -. 1 4  -. 1 9*  
* p < .05, **  p< .0 1 ,  
Note: n 's  may vary due to missing values 
Table 43 shows the correlations between acceptance of the ingroup and acceptance 
towards the outgroup and social change beliefs for each gender group. For the men in the 
sample acceptance towards the outgroup was negatively related to ingroup social mobility, 
r = - .27, p < .05 . High beliefs in social mobility was associated with low acceptance of the 
outgroup. For the women in the sample acceptance of the ingroup was positively related to 
social mobility, r = .42, p < .00 1 .  Perceived acceptance of the ingroup was associated with 
high social mobility beliefs. 
Table 43 
Correlations between Acceptance and Ingroup Social Change Beliefs (men and women) 
n= 
Variable 
I .  Acceptance of the ingroup 
2. Acceptance towards the outgroup 
3 .  Ingroup Social Mobility 
4.  [ngroup Social Creativity 
5. Ingroup Social Competition (achieving status) 
* p < .05, ** p< .0 1 ,  * * *  P < .00 1  
Note: n's may vary due to missing values 
.2 1 
- .08 
- . 1 2  
-. 1 7  
Male 
59 
2 
-.27* 
-. 1 4  
. 1 4  
.09 
Female 
1 1 8 
.42*** 
-. 1 8 
-.06 
2 
-.02 
. 1 4  
.09 
Table 44 shows the correlations between acceptance of the ingroup and outgroup 
and social change beliefs for each status group. For the low status group acceptance of 
the ingroup was positively related to ingroup social mobility, r = .35, p < .0 1 and 
1 6 1 
negatively related to ingroup social creativity, r = -.3 1 , p  < .05. For the low status 
respondents high acceptance of the ingroup was related to high social mobil ity beliefs 
and low social competition beliefs. For the respondents who saw themselves as of equal 
status acceptance towards the outgroup was positively related to perceived acceptance 
of the ingroup, r = .56, p < .05. For the high status respondents acceptance of the 
ingroup or out group was not associated with social change beliefs. 
Table 45 shows the correlations between acceptance of the ingroup and 
acceptance towards the outgroup for the minority group and the majority group. 
For the minority group acceptance ofthe ingroup is negatively related to ingroup social 
creativity, r = - .38, p < .0 1 and social competition, r = - .37 , p  < .0 1 .  High acceptance of 
the ingroup is associated with low social creativity and social competition beliefs. 
Among the majority group acceptance towards the outgroup is positively related to 
acceptance of the ingroup, r = .39, p < .00 1 .  
Table 44 
Correlations between Acceptance and Ingroup Social Change Beliefs for Each Status Group 
Variable 
n= 
I .  Acceptance of the ingroup 
2. Acceptance towards the outgroup -. 1 9  
3 .  Ingroup Social Mobility .35** 
4. Ingroup Social Creativity -.3 1 * 
5. lngroup Social Competition (achieving status) -.22 
* p < .05, ** p< .0 1 , *** p < .00 1 
Note: n's may vary due to missing values 
Low Status 
53 
- . I 5 
.09 
. 1 5  
Equal Status 
87 
.56*** 
-.02 
.02 
.05 
High Status 
1 9  
. 1 0  
-.09 . 1 9  -.24 
.04 -.23 .26 
. 1 2  -. 1 9  .32 
1 63 
Table 45 
Correlations between Acceptance and Ingroup Social Change Beliefs for Majority and 
Minority Groups. 
Variable 
1 .  Acceptance of the ingroup 
2. Acceptance towards the outgroup 
3. Ingroup Social Mobility 
4 .  Ingroup Social Creativity 
5. Ingroup Social Competition (achieving 
status) 
** p< .0 1 ,  * * *  p < .00 1 
Note: n's may vary due to missing values 
Minority 
67 
2 
. 1 5  
.22 -. 1 3  
-.38** .04 
-.37** -.08 
Majority 
93 
.39***  
2 
. 1 4  - .04 
.04 .09 
. 1 6  . 1 9  
In order to examine the relative importance of the variables determining each 
social change belief a stepwise regression analysis was performed, the results are shown 
in Table 46. For the total sample the results indicated that the best predictor of in group 
social mobility beliefs was gender, � = .2 1 ,  P < .0 1 .  Women were lower in social 
mobility beliefs than men. The second best predictor of ingroup social mobility beliefs 
was discrimination, � = -.20, p < .05. Few experiences of discrimination were 
associated with high social mobility beliefs. The best predictor of in group social 
creativity was gender identification, � = .27, P < .0 1 .  High gender identification was 
associated with high social creativity beliefs. The second best predictor of social 
creativity beliefs was gender, � = -.25 , p < . 0 1 . Women were higher in social creativity 
beliefs than men. Gender was also the best predictor of social competition beliefs � = -
.5 1 , p  < .00 1 ,  women had higher social competition beliefs than men. The second best 
predictor of social competition beliefs was discrimination, � = . l 7, p  < .05, more 
experiences of discrimination was associated with higher social competition beliefs. 
Table 46 
Standardized Regression Coefficients between Ingroup Social Change Beliefs and Predictor Variables (Total Sample N= 207) 
Predictor Variables 
Social Change Identification Legitimacy Stability Acceptance Acceptance Discriminatio Gender 
Beliefs: lngroup Out�roup n 
Ingroup Social . 1 5* -. 1 6  .00 . 1 7* -. 1 3  -.20* .2 1 **  
Mobility (.03) (.03) (.00) ( .03) ( .02) (.04) (.05) 
Ingroup Social .27** .09 -. 1 3  -.07 .06 . 1 6  -.25 ** 
Creativity (.08) ( .0 1 )  (.02) (.00) (.00) (.02) ( .07) 
Ingroup Social .03 .06 .00 -.02 .08 . 1 7* -.5 1  ***  
Competition (.00) (.00) ( .00) (.00) (.0 1 )  (.03) (.27) 
* P < .05, * *  p< .0 1 ,  *** P < .00 1 
Note: partial ? are shown in  brackets. 
.27 7.94**  
( l , 1 39) 
.2 1 1 0. 1 9* *  
( 1 , 1 39) 
.35 5 1 .67***  
( 1 , 1 39) 
1 65 
Table 47 shows the results of the open-ended questions on which strategies are 
used to improve the general situation for the in group in the workplace. For the female 
dominated sample, strategies that were categorized as social competition were the most 
commonly mentioned by both men and women (30.7% and 42% respectively). 
Examples of social competition strategies included encouraging male nurses to apply 
for promotions or positions and campaigning for increased childcare facilities. Both 
men and women also mentioned strategies that were intended to improve the situation 
of nursing as a whole, which included campaigning for higher pay. These strategies 
were seen to improve the situation of both men and women. Men also mentioned use of 
social mobility strategies that focused on individual improvement, more than did 
women. A common social mobility strategy mentioned by both men and women was 
studying for extra qualifications. 
For the respondents in male dominated occupations both men and women 
supported social competition strategies for the profession which included efforts to raise 
the profile of engineering as a whole. The most common strategy for women was social 
mobility, which was characterized by demphasizing gender and focusing on doing the 
best job possible. Men in male dominated occupations most commonly did not actually 
use strategies to improve the situation for their group. 
Table 47 
Strategies Used to ImQrove the General Situation for the IngrouQ 
Social Change Strategies Female Dominated 
OccuEations 
Male Female 
n =  1 3  30 
Social Competition for gender 23% 20% 
Social Competition for profession 7.7% 36.7% 
Social Support 1 5 .4% 20% 
Social Creativity 7.7% 1 6.67% 
Social Mobility 30.7% 1 0% 
None 7.7% 1 0% 
Other 7 .7% 1 0% 
Male Dominated 
OccuEations 
Male Female 
20 24 
0% 8.3% 
20% 29.2% 
1 0% 1 2.5% 
1 5% 0% 
20% 37.5% 
30% 8.3% 
30% 1 2 .5% 
1 66 
Table 48 shows how each gender feels the outgroup reacts to their strategies for 
improving their status. The majority of male and females in female dominated 
occupations felt that the outgroup was supportive of their strategies. In male dominated 
occupations, the majority of women felt that men were supportive of their strategies. 
More women felt that their strategies were reacted to with competition than men, this 
included portraying women who joined women's groups as "hairy-legged man haters" 
or lesbians. 
Table 48 
Reactions of the Out group to Social Change Strategies 
Reactions to Female Dominated Male Dominated Occupations 
Strategies 
n =  
Support 
Competition 
Indifferent 
Other 
Don't Know 
Occul2ations 
Male Female Male 
1 3  22 1 2  
53 .8% 54.5% 25% 
38.5% 23% 0% 
0% 22.7% 1 6.7% 
0% 0% 4 1 .7% 
1 5 .4% 0% 25% 
6.5 Social Change Bel ief Systems -
Maintaining Positive Distinctiveness 
Total Sample 
Female 
1 9  
63.2% 
26% 
1 5 .8% 
5.3% 
0% 
Table 49 shows the means and standard deviations for the outgroup social 
change beliefs. One-way ANOVA's were performed to test for differences in social 
change beliefs between the different groups in the study. 
1 67 
Social Change Beliefs and Gender 
Table 49 shows the means and standard deviations for the social change beliefs 
for maintaining positive distinctiveness. There were significant differences between 
male and female dominated occupations and between men and women on social change 
beliefs. From the overall sample women showed higher support for social mobility 
beliefs in the outgroup than men, F ( 1 , 1 79) = 1 4.40, p < .0 1 ,  and higher social 
competition (maintaining status) beliefs than men, F ( 1 ,  1 8 1 )  = 9.43, p < .0 1 .  Women 
showed lower support for social competition beliefs in the outgroup than men, F ( 1 , 1 62) 
= 4.92, p < .05, and lower support for social creativity beliefs in the outgroup than men, 
F ( 1 , 1 79) = 1 1 .60, p < .0 1 .  
Social Change Beliefs and Occupations 
Respondents in female dominated occupations showed lower support for social 
mobility beliefs in the outgroup than respondents in male dominated occupations, F 
( 1 , 1 79) = 4.3 1 ,  p < .05. Respondents in female dominated occupations also had lower 
social competition (maintaining) beliefs than those in male dominated occupations, F = 
( 1 ,  1 62) 40.38, p < .0 1 .  
Social Change Beliefs and Status 
Differences in mean social change beliefs were also examined for each of the 
status groups. These results are shown in Table 50. There were significant differences 
between the status groups in their support of social mobility beliefs in the outgroup, F 
(2, 1 77) = 4.69, p < .05. The low status group (M = 1 .50) showed higher support of 
social mobility beliefs in the outgroup, than the high status group (M = 0.40). Status 
groups were also different in their support of social competition in the outgroup, F (2, 
1 60) = 6.79, p < .0 1 .  The equal status group (M = - 1 .40) showed lower support for 
Table 49 
Mean Out grOUP Social Change Beliefs for Each Sample 
Outgroup Social Change Female Dominated OccuEations Male Dominated OccuEations 
Beliefs Female Male F Female Male F 
n==  63 23 45 52 
Outgroup Social Mobility 1 . 1 4  -0.56 1 4.46***  1 .67 0.83 8.97** 
(2.00) ( 1 .27) ( 1 , 84) ( 1 .0 1 )  ( 1 .60) ( 1 ,93) 
Outgroup Social Competition - 1 .27 0.6 1 9.23** - 1 .2 1  -0.84 1 .06 
( 1 .62) (2.48) ( 1 ,69) ( 1 .34) ( 1 .97) ( 1 , 9 1 )  
Outgroup Social Creativity - 1 .65 -0.98 3.92* - 1 .6 1  -0.84 5 .94* 
( 1 . 1 5) ( 1 .88) ( 1 ,84) ( I .  73) ( 1 .33) ( 1 ,93) 
Social Competition -0.80 - 1 .8 1  9.23** 0.76 -0.36 23 .5 1 *** 
Maintaining Status �1 .291 �1 .3 1} �I , 84) �1 .3 1 )  {0.96} �1 , 95} 
• p < .05, .. p< .0 1 ,  . . .  P < .001 
Note: (i) standard deviations and degrees of freedom are in brackets. 
(ii) n 's  may vary due to missing values. 
Combined SamEle 
Female Male F 
1 08 75 
1 .36 0.4 1 1 4.40***  
( 1 .68) ( 1 .63) ( I ,  1 79) 
- 1 .25  -0.62 4.92* 
( 1 .50) (2. 1 0) ( I ,  1 62) 
- 1 .63 -0.89 1 1 .60***  
( 1 .4 1 )  ( 1 .5 1 ) ( I ,  1 78) 
-0. 1 5  -0.80 9.43** 
{1 . 5 12 ( 1 .262 {t , 1 8 1 )  
OccuEations (dominated b�} 
Female Male 
86 97 
0.69 1 .2 1  
( 1 .98) ( 1 .42) 
- 1 .04 - 1 .0 1  
( 1 .84) ( 1 .7 1 )  
- 1 .47 - 1 . 1 9  
( 1 .4 1  ) ( 1 .56) 
- 1 .07 0. 16  
( 1 .362 �1 .26) 
0\ 00 
F 
4.3 1 * 
( l , 1 79) 
.0 1 
( 1 , 1 62) 
1 . 55  
( I ,  1 79) 
40.38***  
{I , 1 8 1 )  
1 69 
outgroup social competition, than the high status group (M = -0.26). Social competition 
beliefs (maintaining the status) were also different between the status groups, F (2, 
1 79) = 26.24, P < .0 1 .  The low status group (M = 0.92) had higher social competition 
(maintaining status) beliefs than the equal group (M = -0.8S) and the high status group 
(M = -0.60). 
Social Change Beliefs and Majority and Minority Groups 
As shown in Table S I  there was one significant difference between the minority 
group and the majority group. The majority group had lower social competition 
(maintaining the status) beliefs than the minority group, F ( 1 ,  1 8 1 )  = S .09, p < .0 1 .  
Social Change Beliefs and Gender Identification 
Total Sample 
As shown in Table S2 (p. 1 74) a positive correlation was found between gender 
identification as a full scale and ingroup social competition (maintaining status), r = .22, 
P < .0 I .  High gender identification was related to high social competition for 
maintaining status. A negative correlation was found between gender identification and 
outgroup social creativity, r = -.29, P <. 0 I .  High gender identification was related to 
low support of social creativity beliefs in the outgroup. The Cognitive/Emotional 
gender identification subscale was positively correlated with outgroup social mobility, r 
= .23, P < .OS.  High identification on the emotional and cognitive aspects of group 
membership was related to high support of social mobility beliefs in the outgroup. 
Hypothesis 36 that gender identification would be negatively related to outgroup social 
competition was not supported, r = .02, p >.OS. 
Table 50 
Mean Out grOUP Social Change Beliefs for each Status Group 
Outgroup Social Mobility 
Outgroup Social Competition 
Outgroup Social Creativity 
Social Competition Maintaining Status 
· · p< .O I , ··· p < .OOI 
Low Equal High F 
37 97 48 
1 .50 \ .08 0.40 4.69**  
( 1 .07) ( \ .93) ( 1 .49) (2, 1 77) 
- \ , 1 4  - 1 .40 -0.26 6 .79***  
( 1 .63) ( \ . 55) ( \ .98) (2, 1 60) 
- 1 .30 - \ .50 -0.95 2 . 1 9  
( \ .72) ( 1 .52) ( 1 .2 1 )  (2, 1 77) 
0.92 -0.85 - .60 26.24***  
( 1 . 1 21 _____ (1 .36) (\ ,22) (2, 1 79) 
Note: (i) standard deviations and degrees of freedom are in brackets. 
(ii) n's may vary due to missing values. 
L>H 
E<H 
L> E  
L>H 
(iii) For equal variances the Bonferroni Post Hoc test was used, for unequal variances the Dunnett T3 Post Hoc test was used. 
-.l 
o 
Table 5 1  
Mean Out grOUP Social Change Beliefs for the Minority and Majority Group 
Outgroup Social Change Beliefs Minority Group Majority Group 
Female Male F Female Male F Minority Majority F 
n =  45 23 63 52 68 1 1 5 
Outgroup Social Mobility 1 .67 -0.56 6 1 .06*** 1 . 1 4  0.83 0. 8 1  0.90 1 .00 0. 1 6  
( 1 .0 I )  ( 1 .27) ( 1 , 64) (2.00) ( 1 .60) ( I ,  1 1 3) ( 1 .54) ( 1 .83) ( 1 , 1 79) 
Outgroup Social Competition - 1 .2 1  0.6 1 9.88** - 1 .27 -0.84 1 .64 -0.89 - 1 .08 0.4 1 
( 1 .34) (2.48) ( 1 ,50) ( 1 .62) ( 1 .97) ( I ,  1 1 0) ( 1 .7 1 )  ( 1 .79) ( 1 , 1 62) 
Outgroup Social Creativity - 1 .6 1  -0.98 1 .87 - 1 .65 -0.84 1 2.0 1 * *  - 1 .40 - 1 .28 0.2 1 
( 1 .73) ( 1 .88) ( 1 ,64) ( 1 . 1 5) ( 1 .33) ( 1 ,  1 1 3) ( 1 . 80) ( 1 .29) ( I ,  1 79) 
Social Competition (Maintaining Status) 0.76 - 1 .8 1  58.5 1 ***  -0.80 -0.36 4.25* -0. 1 1 -0.60 5 .09* 
( 1 .3 1 )  ( 1 .3 1 )  ( 1 , 66) ( 1 .29) (0.96) ( 1 , 1 1 3) ( 1 .79) ( 1 . 1 7) ( 1 ,  1 8 1 )  
• p < .05, • •  p <  .0 1 ,  ••• P < .001 
Note: (i) standard deviations and degrees of freedom are in brackets 
( i i) n's may vary due to missing values 
..... 
--..) ..... 
1 72 
Table 52 
Correlations between Gender Identification Scales and Out group Social Change Beliefs 
N =  207). 
Variable 2 3 
I .  Gender Identification full scale 
2. Gender Identification Cognitive/Emotional .92*** 
3 .  Gender Identification Individual/Group Opposition .36*** -.03 
4. Outgroup Social Mobility . 1 5  .23* -. 1 2  
5 .  Outgroup Social Creativity -.29** -. 1 9* - .34* *  
6 .  Outgroup Social Competition .02 .05 -.07 
7 .  I ngroup Social Competition (maintaining status) .22** .33** -. 1 8* 
* p < .05, ** P < .0 1 ,  *** P < .00 1 ,  
Note: n's may vary due to missing values. 
Gender Groups 
The relationships between gender identification and social change beliefs were 
also explored separately for each gender group; the correlation results are shown in 
Table 53 .  For men in the sample gender identification was positively relation to 
outgroup social mobil ity r = .45, p < .00 1 ,  and ingroup social competition (maintaining 
status) r = .46, p < .00 1 .  High gender identification was associated with high support of 
social mobility strategies in the outgroup, and high ingroup social competition beliefs. 
Gender identification was negatively related to outgroup social competition, r = .26, p < 
.05. High gender identi fication was associated with low support of outgroup social 
competition strategies. For the women in the sample gender identification was related 
to ingroup social competition (maintaining status), r = .40, p < .00 1 .  
Table 53 
Correlations between Gender Identification Scales and OutgrouQ Social Change Beliefs for men and women. 
Variable Male Female 
n =  75  1 05 
2 3 2 3 
I .  Gender Identification full scale 
2 .  Gender Identification Cognitive/Emotional .93* *  .9 1  *** 
3 .  Gender Identification Individual/Group Opposition .48***  . 1 4  .29*** - . 1 3  
4. Outgroup Social Mobility .45 *** .52***  .00 . 1 4 . I S  -.05 
5. Outgroup Social Competition -.26* -.24 - . 1 6  .0 1  .04 -.06 
6. Outgroup Social Creativity -.20 -. I 3 -.38*** -. 1 6  -.09 -. 1 6  
7. In8!0uE Social ComEetition {maintaining status) .46***  .57***  -.04 .40*** .53 ***  -.30** 
* p < .05, ** p < .0 1 ,  * **  P < .00 1 ,  
1 74 
Status 
The relationship between gender identification and outgroup social change 
beliefs was also examined for the three status groups. The results are shown in Table 
54. As predicted (Hypothesis 35) for the Iow status group, gender identification was 
negatively related to outgroup social creativity, r = -.40, p < .05. High gender 
identification is associated with Iow support of social creativity beliefs in the outgroup. 
Contrary to hypothesis 3 1  for Iow status respondents gender identification was not 
related to outgroup social mobility beliefs, r = - .20, P > .05. 
For respondents of equal status gender identification was positively related to 
outgroup social mobility beliefs, r = .29, P < .0 1 ,  which was in the opposite direction to 
that predicted in hypothesis 3 1 .  High gender identification was associated with high 
social mobility beliefs. Gender identification was also related to ingroup social 
competition (maintaining status), r = .46, P < .00 1 .  High gender identification was 
associated with high ingroup social competition beliefs. Individual/group opposition 
was negatively related to outgroup social creativity, r = - .24, P < .05, and outgroup 
social competition, r = -.29, p < .0 1 .  Low opposition between individual and group 
needs was associated with low support of social creativity and social competition beliefs 
in the outgroup. 
For the high status group gender identification was not related to outgroup social 
change beliefs. Hypothesis 30 that gender identification would positively related to 
outgroup social mobility beliefs was not supported, r = -. 1 2 , p  >.05. Hypothesis 34 that 
gender identification would be positively related to out group social creativity was not 
supported, r = .08, p >.05 . 
Majority and Minority Groups 
The relationship between gender identification and social change beliefs was 
analyzed separately for the majority and minority groups. The results are shown in 
table 55 .  For the minority group gender identification was positively correlated with 
outgroup social mobility, r = .38, p < .0 1 ,  and ingroup social competition (maintaining 
1 75 
status), r = .42, P < .00 1 .  High gender identification was associated with high support 
of social mobility beliefs in the outgroup and high social competition beliefs. Gender 
identification was also negatively correlated with out group social creativity, r = - .37, p 
< .0 1 .  High gender identification was associated with low support of social creativity 
beliefs in the outgroup. 
For the majority group gender identification was positively related to social 
competition (maintaining status), r = .3 1 ,  p = < .0 1 .  High gender identification was 
associated with high social competition beliefs. The individual/group opposition 
subscale was negatively correlated with outgroup social creativity, r = -.25, p < .0 1 .  
High opposition between individual and group needs was associated with high social 
creativity beliefs. 
Social Change Beliefs, Legitimacy and Stability 
Social change beliefs were also compared for those holding different legitimacy 
and stability beliefs. As predicted (Hypothesis 39) respondents who saw the intergroup 
situation as legitimate showed less support for social competition beliefs in the outgroup 
than those who saw the situation as il legitimate, F ( 1 , 1 60) = 6.96, p < .0 1 .  In the 
opposite direction to that predicted (Hypothesis 37) respondents who saw the situation 
as legitimate had lower beliefs in social competition (maintaining the status) than those 
who saw the situation as i llegitimate, F ( 1 , 1 78) = 1 0.97, p < .0 1 .  Hypothesis 32 that 
respondents who see the situation as legitimate will have higher support for outgroup 
social creativity beliefs was not supported, F ( 1 , 1 60) = 2.40, p >.05 . The results are 
shown in Table 56. 
Hypothesis 38 that respondents who see the situation as unstable will have 
higher social competition (maintaining status) beliefs than respondents who see the 
intergroup situation as stable, was not supported, F ( I ,  1 75) = 2.64, p > .05. Hypothesis 
40 that respondents who see the intergroup situation as stable would have lower support 
for outgroup social competition than those who see it as stable was not supported, F ( 1 ,  
1 57) = 1 .23 , p >.05. 
Table 54 
Correlations between Gender Identification Scales and OutgrOUP Social Change Beliefs for Each Status Group 
Variable 
n =  
Gender Identification Full scale 
Gender Identification CognitivelEmotional 
Gender Identification Individual/Group Opposition 
Outgroup Social Mobility 
Outgroup Social Competition 
Outgroup Social Creativity 
IngrouE Social ComEetition {maintaining status} 
* P < .05, * *  p< .0 1 ,  *** P < .00 1 
Note: n 's  may vary due to missing values 
Low Status 
37 
2 
.92** 
.39** .00 
-.20 -. 1 0  
. 1 4  . 1 2  
-.40* -.43 ** 
-.25 -. 1 4  
Equal Status 
97 
3 2 3 
.93** 
.45***  . 1 0  
-.24 .29** .29**  .08 
. 1 0  -.07 .02 -.29** 
-. 1 8  - . 1 1 -.05 -.24* 
-.28 .46***  .50***  .05 
High Status 
48 
2 3 
.93*** 
.4 1 **  . 1 1 
-. 1 2  -.06 -. 1 8  
-.0 1 -. 1 0  . 1 1 
.05 .08 -.23 
.00 .09 -. 1 9  
-
-..l 0\ 
Table 55 
Correlations between Gender Identification Scales and OutgrouQ Social Change Beliefs for Majority and Minority GrouQs. 
Minority Majority 
67 1 1 3 
Variable 2 3 2 3 
I .  Gender Identification Full Scale 
2 .  Gender Identification CognitivelEmotional .89***  .94***  
3 .  Gender Identification Individual/Group Opposition .34**  -. 1 3  .44***  . 1 2  
3 .  Outgroup Social Mobility .38** .5 1  *** -. 1 6  . 1 1 .09 .02 
4 .  Outgroup Social Competition -. 1 4  -. 1 5  .0 1 - .01  .02 -. 1 3  
5 .  Outgroup Social Creativity -.37** -.32** -.22 .08 . 1 7  -.25 ** 
6 .  I ngroup Social Competition (maintairung status) .42***  .64*** -.32* *  .28**  .3 1 * * *  -.04 
* *  p< .0 1 ,  ***  p < .00 1 
1 78 
Hypothesis 33  that respondents who see the intergroup situation as stable would have 
higher support for outgroup social creativity beliefs than respondents who see the 
situation as unstable was not supported, F ( 1 , 2.48) = 2.48, p >.05. 
Social Change Beliefs and Acceptance 
Table 57 shows the correlations between acceptance of the ingroup and 
acceptance towards the outgroup and outgroup social change beliefs, for the total 
sample. Acceptance of the ingroup was negatively related to ingroup social competition 
(maintaining status), r =  -.32, p < .00 1 .  High acceptance of the ingroup is associated 
with low ingroup social competition beliefs. Acceptance towards the outgroup was 
negatively related to outgroup social creativity, r = -. l 9, p  < .05. High acceptance of the 
outgroup is associated with not supporting social creativity beliefs in the outgroup. 
Hypothesis 29 that acceptance would be related to out group social mobility beliefs was 
not supported, r = .02, p >.05. 
Table 57 
Correlations between Acceptance and Out grOUP Social Change Beliefs eN =207) 
Variable 
I .  Acceptance of the ingroup 
2. Acceptance towards the outgroup 
3. Outgroup Social Mobility 
4. Outgroup Social Competition 
5. Outgroup Social Creativity 
6. Ingroup Social Competition (maintaining status) 
* p < .05, *** P < .00 1 .  
Note: n's may vary due to missing values 
2 
. 1 3  
.02 .09 
. 1 1 - . 1 5  
.08 -. 1 9* 
-.32*** .01 
Table 56 
Mean Social Change Beliefs for Legitimacy and Stability Beliefs 
Scale Legitimacy Beliefs 
Legitimate I llegitimate F 
n= 97 83 
Outgroup Social Mobility 1 . 1 4 0.79 1 .89 
( 1 . 88) ( 1 .47) ( 1 , 1 76) 
Outgroup Social Competition - 1 .35 -0.64 6 .96** 
( 1 .53) ( 1 .9 1 )  ( 1 , 1 60) 
Outgroup Social Creativity - 1 .48 - 1 . 1 4 2 .40 
( 1 .53) ( 1 .45) ( 1 , 1 76) 
Social Competition -0.75 -0.05 1 0.97* *  
(Maintaining Status) ( 1 .4 1) ( \ .42) ( \ ,  \ 78) 
** p < .O I , 
Note: (i) standard deviations and degrees of freedom are in brackets 
(ii) n ' s  may vary due to missing values 
Stability Beliefs 
Stable Unstable 
1 3 1  46 
0.95 0.97 
( 1 .76) ( 1 .63) 
- 1 . 1 1 -0.76 
( 1 .85) ( 1 . 5 5) 
- 1 .43 - 1 .02 
( 1 .49) ( 1 . 55) 
-0.55 -0. 1 4  
( 1 .46) (1.43) 
F 
.00 
( \ ,  1 73) 
1 .23 
( 1 , 1 57) 
2 .48 
( 1 , 1 73) 
2.64 
( 1 , 1 75) 
...... 
-...l \0 
1 80 
Table 58 shows the correlations between acceptance of the ingroup and 
acceptance towards the outgroup and outgroup social change beliefs for each gender 
group. For the men in the sample acceptance towards the out group was negatively 
related to outgroup social creativity, r = - .38, p < .0 1 .  High acceptance of the outgroup 
(women) was related to low support of social creativity beliefs among women. For the 
women in the sample acceptance of the ingroup was negatively related to social 
competition (maintaining status), r = -.3 7, p < .00 1 .  High feelings of acceptance were 
associated with low social competition beliefs. 
Table 58 
Correlations between Acceptance and Out grOUP Social Change Beliefs (men and women) 
n =  
Variable 
I .  Acceptance of the ingroup 
2. Acceptance towards the outgroup 
3. Outgroup Social Mobility 
4 .  Outgroup Social Competition 
4. Outgroup Social Creativity 
5. Ingroup Social Competition (maintaining status) 
** p < .0 1 ,  *** P < .00 1 . 
Note: n's may vary due to missing values 
. 2 1  
. 1 2  
. \ 0  
.04 
-. 1 6  
Male 
66 
Female 
2 2 
.09 
.02 .04 .02 
-. 1 2  .08 -.07 
-.38** .07 .0 1  
-.25 -.37*** .06 
Table 59 shows the correlations between acceptance of the ingroup and towards 
the outgroup and outgroup social change beliefs for each status group. For the low 
status group acceptance was not related to out group social change beliefs. For the equal 
status group acceptance was not related to outgroup social change beliefs. For the high 
status group acceptance towards the outgroup was negatively related to outgroup social 
creativity, r = -.40, p <. 05. High acceptance of the outgroup was associated with low 
support of social creativity beliefs in the outgroup. 
Table 59 
Correlations between Acceptance and Out group Social Change Beliefs for Each Status Group 
Variable Low Status Equal Status High Status 
n= 34 87 40 
2 2 2 
I .  Acceptance of the ingroup 
2 .  Acceptance towards the outgroup - . 1 9  .56*** . 1 0  
3. Outgroup Social Mobility .25 -. 1 4  . 1 3  . 0 1  -.02 .2 1 
4 .  Outgroup Social Creativity .23 -. 1 4  -.0 1 -. 1 2  . 1 9  -AO* 
5. Outgroup Social Competition . 3 1 -.29 -.05 - .06 .22 - . 1 2  
6 .  Ingroup Social Competition (maintaining status) -. 1 8  - . 1 0  -. 1 2  -.0 1 -.29 -.0 1 
* P < .05, *** p < .00 1 
Note: n 's  may vary due to missing values 
00 ..... 
1 82 
Table 60 shows the correlations between acceptance of the ingroup and 
acceptance towards the outgroup for the minority and majority groups. For the minority 
group acceptance of the ingroup is negatively related to ingroup social competition, r = ­
.43 , p  < .00 1 .  High acceptance of the ingroup is associated with low social competition 
beliefs for the minority group. For the majority group acceptance towards the outgroup 
is negatively related to outgroup social competition, r = -.22, p < .05, and outgroup 
social creativity, r = -.26, p < .05. For the majority group high acceptance of the 
outgroup is associated with low support of social creativity and social competition 
beliefs in the outgroup. 
Table 60 
Correlations between Acceptance and Out grOUP Social Change Beliefs for Majority and 
Minority Groups. 
n=  
Variable 
I .  Acceptance of the ingroup 
2. Acceptance towards the outgroup 
3. Outgroup Social Mobility 
4. Outgroup Social Competition 
4. Outgroup Social Creativity 
5. Ingroup Social Competition (maintaining status) 
* p < .05 , * *  p< .0 1 ,  ***  P < .00 1 
Note: n's may vary due to missing values 
. 1 5 
-.09 
.2 1 
.09 
Minority 
67 
2 
- .04 
-.08 
-. 1 5  
- .43*** -.08 
Majority 
96 
2 
.39*** 
. 1 1 . 1 7 
.09 -.22* 
.07 -.26* 
-.06 -.05 
In order to examine the relative importance of the variables detennining each 
social change belief a stepwise regression analysis was perfonned, the results are shown 
in Table 6 1 .  For the total sample the results indicated that the best predictor of 
supporting social mobility belief in the outgroup was gender, � = -.34, p < .00 1 .  Males 
were lower in their support of social mobility bel iefs in the outgroup than females. The 
second best predictor of outgroup social mobility beliefs was gender identification � = 
.26, P < .0 1 .  High gender identification was associated with outgroup social mobility 
1 83 
beliefs. The best predictor of outgroup social creativity was gender identification, p = -
.42, P < .00 1 . The second best predictor of out group social creativity beliefs was 
occupation, p = - .34, p < .0 1 .  High identification with the group was associated with 
low support of social creativity strategies in the outgroup. The best predictor of 
outgroup social competition beliefs was legitimacy, p = .35,  p < .0 1 .  The second best 
predictor of outgroup social competition beliefs was acceptance towards the outgroup p 
= -.2 1 ,  p < .05 . The best predictor of ingroup social competition beliefs was gender p = -
.34, p < .00 1 . The second best predictor was occupation p = - .38, p < .00 1 .  
Table 6 1  
Standardized Regression Coefficients between Out group Social Change Beliefs and Predictor Variables (Total Sample N= 207) 
Predictor Variables 
Social Change Beliefs: Identification Legitimacy Stability Acceptance Acceptance Discriminatio Gender Occupation RZ F 
I ngroup OutgrouE n 
Out group Social .26**  -. 1 2  .06 . 1 1  -.0 1 .07 -.34***  . 1 7  1 7.42**  
Mobility (.07) ( .0 1 )  (.00) ( .0 1 )  (.00) (.00) (. I l ) * 
( 1 , 1 40) 
Outgroup Social -.42***  .05 .08 . 1 6  -. 1 6  .0 1  . 1 4  -.34* *  . 1 7  6.98** 
Creativity (.09) (.00) (.00) ( .02) (.03) (.00) (.02) (.05) ( 1 , 1 39) 
Outgroup Social -. 1 0  .35* *  -.04 .2 1 * -.2 1 * . 1 5  6.47* 
Competition ( .0 1 )  (. 1 0) (.00) (.04) ( .05) ( 1 , 1 3 1 )  
Ingroup Social .08 .02 .03 -.08 -.03 .24***  -.34***  -.38***  .43 1 3 .45* *  
Competition (.00) (.00) (.00) (.0 I )  (.00) (.08) ( . 1 5) (.09) * 
{maintaining Status} {I , 1 4 1} 
* P < .05 , *** f < .00 I 
Note: partial r are shown in brackets. 
-
00 � 
1 85 
Table 62 shows the free answer responses to the question that asked what are the 
types of strategies you notice the outgroup using to improve their situation. For the 
female dominated occupation sample, men most often noticed women using social 
competition strategies, which included campaigning for childcare and derogatory 
comments about men, such as "the only reason he got that job is because he is a man". 
The majority of women did not work with male nurses, so did not notice any strategies. 
Female nurses mentioned men using social competition, social creativity and social 
mobility strategies in equal proportions. 
For the respondents in male dominated occupations women noticed men using 
social competition strategies, which included various forms of networking to the 
exclusion of women such as ' the old boy network ' 'playing golf' and fishing trips. The 
next largest group of women noticed no strategies used by men. The men most 
frequently noticed women using networking and social support to improve their 
position. Men also noticed women using social mobility strategies, which included 
extra study, and presenting papers to improve their visibility, and demphasizing their 
gender. 
Table 62 
Social Change Strategies Used by the Out group 
Strategy Noticed Female Dominated Male Dominated 
Occupations Occupations 
Male Female Male Female 
n= 1 1  2 1  28 24 
Social Competition 45.5% 1 9% 14 .3% 58.3% 
Social Support 27.3% 4.8% 4 1 .7% 0% 
Social Creativity 1 8. 1 8% 1 9% 0% 0% 
Social Mobility 9. 1%  1 9% 32. 1 %  0% 
None 1 8.2% 38% 2 1 .4% 20.8% 
Other 0% 0% 1 6.7% 
1 86 
CHAPTER SEVEN DISCUSSION 
The present study examined men and women in three gender dominated 
occupations. The research focused on the fol lowing specific areas of S IT gender 
identification, identity salience, beliefs about the intergroup situation, and social change 
beliefs for achieving and maintaining status. Each is discussed in turn in this chapter. 
7.1  Gender Identity 
Group identification is a key aspect of S IT, which states that individuals strive to 
achieve positive social identities by achieving positive distinctiveness for their groups. 
The measure of gender identification used in the current research measured general 
attitudes towards belonging to the gender group, which included cognitive/emotional 
factors and opposition between individual and group needs. 
Gender Identification and Gender Groups 
Hypothesis 1 that men would have higher gender identification than women was 
supported. Men also had higher gender identification on the cognitive/emotional 
subscale than women. This supports SIT's (Tajfel & Turner, 1 979) claim that high 
status groups will have higher group identification because identification with a high 
status group (men) contributes more positively to the self-concept than identification 
with a low status group (women). Very few studies have studied the gender 
identification of men in comparison to the gender identification of women. A study by 
Swan and Wyer ( 1 997) found that when men became conscious of their gender they 
emphasised their ingroup attributes whereas women demphasised their ingroup 
attributes. Swan and Wyer explain that it is more beneficial for men to be seen as a 
member of their gender group because their status makes more of a positive 
contribution to their self-concept than does women's. For women the opposite is true: 
they are motivated to draw attention away from their gender group membership- or have 
low identification because their category membership is not seen positively. Other S IT 
researchers have also found the high status group had higher identification than low 
1 87 
status groups. For example Ellemers, Doosje, Van Knippenberg and Wilke, ( 1 992) 
manipulated status groups in an experimental setting, and found that the higher status 
group had higher group identification than the low status group. 
The present research offers some support for S IT's assumption that high status 
groups identify more strongly with their group than low status groups. Women's lack 
of identification with each other can decrease group action, which will decrease the 
chance of changing women's status position. 
Gender Identification and Occupations 
Hypothesis 2 that gender discrimination and gender identification would be 
related was not supported. Gender discrimination was only related to the 
individual/group opposition subscale. Experiences of gender discrimination were 
associated with increased opposition between individual and group needs. One of the 
items in this subscale is "I feel held back by my group", which reflects an element of 
discrimination. The absence of a relationship between discrimination and identification 
is contrary to previous findings that discrimination increases awareness of gender 
(Gornick 1 983;  Kanter, 1 977). S IT does not explicitly predict a relationship between 
discrimination and identification but discrimination is likely to be a factor that increases 
gender salience that should increase the likelihood of identifying with the group. A 
possible explanation is that awareness of category membership does not necessarily lead 
to identifying with the group. Discrimination could change attitudes towards the 
outgroup more than attitudes towards the ingroup. Only a small number of respondents 
had experienced discrimination with a low mean on this scale. 
Minority group members (female engineers and male nurses) had higher gender 
identification than majority group members. The different subscales of gender 
identification performed slightly differently. The minority group was higher in gender 
identification on the cognitive/emotional subscale than the majority group, while the 
majority group was higher on the individual/group opposition subscale than the 
minority group. Therefore, although the minority group identified more strongly with 
their group, they experienced more conflict between individual and group needs than 
1 88 
did the majority group. SIT (Tajfel, 1 978) predicts that low status group members wil l  
identify less with the ingroup because of the negative contribution this makes to their 
self-concept. However SIT does not necessarily equate being a numerical minority with 
being of low status. 
Low status groups can identify highly with their group under some conditions. 
For example Ellemers, Van Kruppenberg, and Wilke ( 1 990) found that if respondents 
saw the intergroup situation as unstable, the low status group strongly identified with 
their group. In another experiment on group conditions Ellemers, Wilke & Van 
Knippenberg, ( 1 993) found that when a low status group is i l legitimately of low status 
their identification with the group increases. A high status minority group can also be 
seen very favourably and make a positive contribution to the self-concept (Ellemers et 
al . ,  1 992). 
There are other reasons that may cause minority or low status groups to identify 
more strongly with their group. Brown and Smith ( 1 989) and Lee ( 1 993) found that the 
minority groups in their studies saw their groups as more homogenous than the majority 
groups. This may make it easier for the minority group to identify with their group as an 
abstract idea of a group, as they see the group as more homogeneous, and actually have 
less within group contact than the majority group. The majority group in Brown and 
Smith's study also saw the minority group as more homogeneous than the majority. 
The effect of the expectations of the majority group may also play a part in the 
identification of the minority group. The majority may expect them to identify with 
each other, because of their shared category membership. 
Minority status in itself may create a common bond between the group 
members. This is reflected in comments by one of the male nurse respondents who said 
" there's a bit of male camaraderie among other nurses in the wards, like you'd see 
someone else and you'd go hey man and you feel a bit of a bond". The perceptions of 
homogeneity between the majority and minority groups may also explain why women 
in the majority group saw women as more different to each other than the minority 
women did. This was reflected in the number of comments about the difference between 
women in the gender identification free responses. 
1 89 
Gender identification was higher for those respondents in male dominated 
occupations than female dominated occupations. Gender identification was not higher 
for men or women within either occupation. When analysing identification by subscale, 
men in male dominated occupations experienced less individual/group opposition than 
women. 
The high individual/group opposition that the female engineers experience and 
the feeling that they do not fit in with other members of their group, may be related to 
the fact that they do not have opportunities to work or socialise with other women in 
their profession. They may also have difficulties fitting in with the support staff who 
are usually women. For example one woman engineer in the study said, "in engineering 
I feel that sometimes other women staff can resent the fact that you 're the engineer and 
they are admin." Another female engineer stated that an important part of trying to fit 
in was not socialising with other women "Trying to spend more time in staff 
rooms/lunch time with engineers (predominately male) than administration staff (always 
female), to be seen as one of that group". Metcalfe ( 1 987) also found that women in 
male dominated occupations experienced pressure to separate from their gender 
identity, which was so they could take on the masculine characteristics seen to be 
needed for their work. 
Women in male dominated occupations may have more masculine self-concepts 
than other women, which may lead to their dislike of working with other women 
(McLean & Kalin, 1 994). The female engineers in the current study expressed dislike 
of working with other women in the free answer section. Many women reported that 
they enjoyed working with men. These factors may also increase their feelings of not 
fitting in with other women. 
Gender Identification and Status 
Hypothesis 1 0, that gender identification would be higher for those respondents 
who saw themselves of high status than respondents who saw themselves of low status 
was not supported on the full gender identification scale. The low status group had 
1 90 
higher gender identification than the equal group on the cognitive/emotional subscale. 
There was no difference between the high and the low status groups, or between the 
equal and the high status group. SIT claims that the high status group will have higher 
identification than the low status group because their group membership makes a more 
positive contribution to the self-concept than does membership in a low status group. 
SIT does not consider the position of equal status groups, and focuses on objective 
status differences rather than the sUbjective self-ratings of status that were used in this 
study. 
Finding no difference in identification between the high and low status groups is 
inconsistent with previous research within the S IT framework. For example, Ellemers 
et al . ( 1 990) found the high status group was seen as more attractive and the low status 
group members were more dissatisfied with their group membership than the high status 
group. One of the few studies to examine equal groups found that members of equal 
groups felt more positive about their group membership than low groups, although this 
study focussed more on power differentials than status (Bourhis, 1 992). 
The first possible explanation for the low identification of the equal status group 
is that the equal status group do not have the same motivation or need to think 
collectively as a low status group does. For the low and high status groups it is in their 
interest to think collectively, to improve or maintain their position. Group membership 
may not be as important for those who have equal status, as there may be no need for 
group action. Alternatively respondents who identify less with their group may have 
less cause or opportunities to make comparisons between their group and the outgroup 
so do not perceive their group to be of low status. 
The low status group had higher conflict between individual and group needs 
than the high status and equal status group�. The difference between the low status 
group and the equal and high status groups on individual/group opposition, reflect the 
nature of the questions on this subscale. The items that make up this subscale cover how 
well  the respondent fits in and how they feel with other members of their gender group 
and whether they feel held back by their gender group. The perception that the ingroup 
is of low status may be related to the feeling of being held back by group membership. 
1 9 1  
These results also show that low status group members have trouble fitting in with their 
group. 
The results do not support the SIT assumption that high status groups will have 
higher gender identification than low status groups. They do provide information about 
the identification of groups of equal status. The feeling of being held back by the group 
does not diminish identification with the group on cognitive emotional aspects. 
Gender Identification and Majority and Minority Groups 
C.Williams ( 1 989) states that male minorities have more reason to identify with 
each other as their gender group is seen more positively by the majority group, whereas 
women in male dominated occupations try to reduce the attention to their category 
membership. Comparing the minority groups shows that minority women had higher 
gender identification than the minority men, which is in contrast to C. William's 
findings. 
The difference between the subscales is important in explaining these findings. 
Women identified less highly on the subscale that measured how well they get on with 
other women, and whether they feel held back by their group, than male nurses. These 
items measure more specific behaviours than the cognitive/emotional subscale items, 
which tend to be more abstract. Therefore, women engineers can identify highly with 
other women in the abstract, but when it comes to fitting in with other women, they 
have more difficulty. 
Gender Identification Summary 
The gender identification results give mixed support to S IT's prediction that 
members of high status groups have higher gender identification than members of low 
status groups. Men have a higher status in society than women and as predicted, they 
had higher gender identification than women. On the other hand, the minority group in 
the workplace had higher gender identification than the majority group, and minority 
1 92 
women had higher gender identification than minority men. Among the status groups, 
there was no difference in identification between the high and low status groups, the 
equal group had the lowest gender identification. The differences among the groups 
may reflect the specificity of gender identification that was measured. The measure of 
status used in this study, was a self-report measure of respondents perceived status, in 
the workplace. However, the gender identification measure focussed on identification 
with the gender group as a whole not specifically within the occupations, which may 
explain why the SIT prediction was upheld in the gender groups. Future research could 
usefully explore the differences between gender identification with the entire gender 
group and gender identification with subgroups. Researchers could for example study 
identification with the group women engineers or male nurses. 
7.2 Identity Salience 
The discussion of identity salience is organised into three sections identity 
salience and gender groups, identity salience in gender dominated occupations and 
identity salience and SIT. Two aspects of gender salience were studied, gender 
salience in the self-concept and gender salience as a basis for categorising others. 
Identity Salience and Gender Groups 
Hypothesis 5 that women would have higher self-concept gender salience than 
men was not supported. No differences were found between the gender groups in 
gender salience in the self-concept, occupational salience in the self-concept or the 
Gender Salience Scale. These results are consistent with McGuire, McGuire and 
Winton ( 1 979) who found no difference in gender salience between the males and 
females in their sample. 
One of the problems mentioned by McGuire and Padawer-Singer with the 
Spontaneous Self-Concept measure, that could have affected the salience rating in the 
1 93 
present study, is the small numbers of respondents who actually mention the attributes 
of interest, such as gender and occupation in the present study. There are also problems 
with measuring gender salience as if it is a stable construct when SIT states that 
different identities wil l  be salient at different times, according to relevant 
categorizations in the situation. 
Identity Salience and Occupations 
There were no differences in gender salience between the occupations or 
between men and women within occupations. These findings are inconsistent with 
previous research, which suggests that the emphasis on gender that occurs within gender 
dominated occupations wil l  increase the salience of gender (Kanter, 1 977). The 
findings are also in contrast to some of the comments made by respondents. The 
respondents highlight some of the characteristics of the job that emphasise gender, and 
seem likely to make gender salient. For example there are practical differences in the 
jobs that male and female nurses are able to do male nurses have "to protect themselves, 
- bathing little girls, for example male staff nurses tend to not do that sort of work". The 
work that nurses do seems to include many situations that emphasise the physical 
differences between the genders. On the other hand for female engineers it seems to be 
the workplace culture that emphasises their different gender, for example the men spent 
time, "Talking about cars, engines, and mountain bike parts, so that females are 
excluded from conversations, and kept on the outer ring". The attitudes of managers 
and peers to the female engineers may also highlight their gender. For example one 
woman reported "being referred to in front of a client by her manager as ' little Sarah"'. 
She felt that ber manger would never have referred to a male employee as ' little' or 
'young' . 
1 94 
Identity Salience and Majority and Minority Groups 
Hypothesis 6 that the minority group would have higher gender salience than the 
majority was not supported by the results. This is inconsistent with previous research; 
McGuire and Padawer-Singer ( 1 976) found that the numerical minority or majority 
status of gender affected gender salience, creating higher gender salience for the 
minority gender 
Hypothesis 8 that the minority group would have an inverse relationship 
between occupational and gender salience was not supported, and that the majority 
would have a positive relationship between occupational and gender salience was not 
supported. Hypothesis 9 that the minority group would have higher gender salience 
than occupational salience was not supported. In fact, the minority and the majority 
groups both had higher occupational salience than gender salience. These results are 
inconsistent with other research that has demonstrated that gender is more salient for 
gender minorities than occupational membership (Macke, 1 98 1 ). The findings of the 
current study suggest that being a gender minority may not always diminish 
occupational salience. 
The majority group did have higher scores on the Gender Salience Scale. The 
majority group members have more traditional attitudes towards using gender for 
categorising others, than the minority group members. This has implications for the 
minorities working in gender dominated occupations as they will find that the majority 
groups use gender as a basis for attitudes and decisions about others. This is consistent 
with Kanter's ( 1 977) research, she found that the dominant group emphasised category 
membership and used it to make assumptions about the minority group members. 
For the majority group high occupational salience was related to low numbers of 
the outgroup in the workplace. This is an unexpected finding. It means that the 
presence of the outgroup in the workplace affects the salience of occupation as part of 
the self-concept. Majority group members' ideas about their occupation may be 
affected by their gender and the gender of their workmates. Martin ( 1 994) states that 
the entry of women into male dominated occupations can highlight the feminine and 
1 95 
service elements of an occupation, which men prefer to hide from themselves and 
others. The emphasis on the characteristics of the occupation associated with opposite 
gender may have decreased the salience of occupation in the self-concept of the 
respondents. 
Identity Salience and SIT 
Hypothesis 3 from SIT that individuals who have a member of the opposite 
gender in their immediate workgroup would have higher self-concept gender salience 
than those who do not work with a member of the opposite gender, was not supported. 
According to S IT the presence of an outgroup increases salience for both the ingroup 
and outgroup members. This prediction from SIT has been supported by an experiment 
on gender groups by Abrams, Thomas and Hogg ( 1 990). They found that gender was 
more salient for those with outgroup members in their group than individuals in a 
homogenous group. 
It was also hypothesised (Hypothesis 4) that the number of same gender 
colleagues or the distinctiveness of category membership would not be related to gender 
salience. This hypothesis was supported by the data; the number of in group members 
was not related to the gender salience of the respondents. The presence of groups in the 
workplace did not increase self-concept gender salience and the distinctiveness of 
gender group membership did not increase gender salience. This is inconsistent with 
previous research that shows that distinctiveness is related to identity salience (McGuire 
& McGuire, 1 98 1 ). Many factors other than distinctiveness can influence the salience 
of gender in the self-concept. Oakes and Turner ( 1 986) found that novel category 
memberships were not automatically more salient; the task on which the individuals in 
their experiment worked had more of an effect than novelty. Sex Role Orientation may 
also influence gender salience (Abrams, Thomas & Hogg, 1 990). 
Hypothesis 2 that gender discrimination and gender salience would be correlated 
was not supported. This is inconsistent with previous research that has highlighted how 
1 96 
gender discrimination increases awareness of gender (Kanter, 1 977). A possible 
explanation for this finding is once again the lack of respondents actually mentioning 
gender as part of the self-concept and the small numbers of respondents who actually  
experienced discrimination. Skevington and Dawkes ( 1 988) also found very low 
numbers of nurses of either gender reporting gender discrimination. 
A correlation was hypothesised (hypothesis 6) between self-concept salience and 
the GSS. This was not found for the majority or minority group. Self-concept gender 
salience did not make respondents more likely to categorise others by gender. Rather the 
GSS was related to majority group and age with older people having higher scores. 
Older people categorise individuals by gender more than younger people. 
Identity Salience Summary 
The predictions made relating to identity salience were not supported, with no 
differences between men and women, status or occupational groups. Gender as a basis 
for categorising others was higher in the majority group reflecting more traditional 
attitudes in the majority group. No support was found for either distinctiveness theory 
or S IT. The lack of support for any of the salience hypotheses suggests that the measure 
of salience may have been inadequate. The nature of the spontaneous self-concept scale 
means that its reliability and validity cannot be assessed, and there are few alternative 
measures with which to compare it. To add to knowledge of the causes of salience 
future research should explore the factors within contexts such as gender dominated 
occupations, which increase the gender salience of individuals. 
7.3 Beliefs about the Intergroup Situation 
Status 
Status, Gender Groups and Occupations 
1 97 
Status beliefs were not independent of gender or occupation. Minority or 
majority group membership also affected the distribution of beliefs about status. The 
largest percentage of the minority group saw their situation as low status. The largest 
percentage of the majority group saw their situation as equal . The majority of women 
saw their status as equal, very few women saw their status as high and very few men 
saw their status as low. Research findings suggest that some women, especially 
younger women, now consider that women have achieved equality with men in society 
(Cockbum, 1 99 1 ,  Condor, 1 989). The assumption that many researchers make that 
women constitute a group of low status should be used with caution considering the 
number of women who see themselves as of equal status, although the perceptions of 
the respondents may not necessarily reflect reality. 
Hypothesis 1 1  that the majority of female engineers would see themselves of 
low status and that the majority of male engineers will see themselves of high status was 
supported. The history of women in male dominated occupations is that their societal 
status and low numerical representation within an occupation interact increasing the 
disadvantage that they experience. Zimrner ( 1 988) emphasises the role societal sexism 
has on tokenism processes, and that minority status should not be considered alone 
when examining the position of tokens in the workplace. The history of the genders in 
the occupation is also an important factor. Women in engineering did see themselves as 
disadvantaged, giving examples such as the lack of female role models, the lack of 
women in management and senior positions, and the attitudes of men to their presence. 
Hypothesis 1 2  that the majority of male nurses would see themselves of low 
status and that the majority of female nurses would see themselves as high status, was 
not supported. Hypothesis 1 3  that the majority of female nurses would see themselves 
of low status and that the majority of male nurses would see themselves of high status 
was not supported. No difference was found in the perceptions of status between the 
1 98 
men and women in female dominated occupations. The combination of status in 
society and being a minority group may interact to make a situation of equality between 
the gender groups in nursing. Ellemers et al. ( 1 992) found that being a minority does 
not necessarily make the group low status, and a high status minority was seen in their 
study the most positively of all .  Men in nursing could be described as a high status 
minority as their societal status and status within the organisation (in terms of 
representation at higher levels) is high, and yet they are a numerical minority. 
Skevington and Dawkes ( 1 988) in their study of nurses found that men in nursing did 
not see themselves as disadvantaged by being a numerical minority. In contrast to 
Kanter ( 1 977) who thought that numbers had the most influence on creating a token 
situation these results highlight the importance of considering more than just numbers 
when studying minorities in the workplace. Zimmer ( 1 988) states that societal status 
and social attributes may also contribute to tokenism processes, which results in 
different experiences for female and male tokens. 
Legitimacy and Stability 
Hypothesis 1 7  that legitimacy and stability would not be independent was 
supported. This is consistent with research by EIIemers et al . ( 1 990) who found that the 
intergroup situation was fairer when the situation was also seen as stable. Hypothesis 
1 8  that more high status respondents will perceive the situation to be stable than low 
status respondents was not supported there was no difference between the high and low 
status group in perceptions of stability. The comments on the free response section 
demonstrate that even the low status respondents felt that the intergroup situation would 
be very slow to change, a female engineer commented, "I 'd l ike to think that males and 
females will have equal advantages in engineering in the future but I doubt that it will 
be in my l ifetime". 
Hypothesis 1 8  that more high status group members would see the situation as 
legitimate than low status respondents was not supported. There was no difference 
between the high and low status groups on perceptions of legitimacy. A large number 
1 99 
of the high status group acknowledged the illegitimacy of the status relations. It is 
difficult to tell how honest the high status group was when answering the questionnaire, 
there may have been a strong influence to be 'political ly correct' .  The majority of the 
equal group saw the situation as legitimate and stable. More members of the minority 
group saw the situation as il legitimate than did members of the majority. 
Gender, Occupations, Legitimacy and Stability 
There were no differences in legitimacy beliefs between men and women in the 
sample. The majority of respondents in female dominated occupations saw the 
intergroup situation as legitimate, and the majority of respondents in male dominated 
occupations saw the situation as i llegitimate. 
There were no differences in legitimacy beliefs between the men and women in 
male dominated occupations. The legitimacy of the situation in gender dominated 
occupations has not been previously studied. The acknowledgment of the i llegitimacy 
of the intergroup situation by male engineers may have been influenced by the method 
of recruiting engineers. Male engineers interested in gender relations and with views 
that are more liberal may have been more likely to fill out questionnaires. The majority 
of men in female dominated occupations saw the situation as illegitimate and the 
majority of women saw the situation as legitimate. 
The majority of respondents saw the situation as stable. No differences in 
stability beliefs were found between male and female dominated occupations and 
between males and females in each occupation. Perceptions of the stability of status 
relations were unaffected by majority or minority status 
Acceptance 
Perceived acceptance of the ingroup and acceptance towards the outgroup were 
taken as a measure of the permeability of group boundaries. 
200 
Acceptance, Gender and Occupations 
Acceptance was not influenced by gender alone, with no differences in 
acceptance of the ingroup or acceptance towards the outgroup between the men and 
women in the total sample. Respondents in female dominated occupations felt that the 
ingroup was more accepted than did respondents in male dominated occupations. There 
was no difference between male and female dominated occupations in acceptance 
towards the outgroup. The minority group felt less accepted than the majority group, 
and they were more accepting of the out group than the majority group. 
The acceptance result may reflect the nature of the work of nurses and engineers 
nurses must work in teams whereas many of the engineers work by themselves or in a 
much more competitive environment than nursing. Oaker and Brown ( 1 986) claim that 
there is "a strong ethos of co-operation" in nursing (p. 767). 
In male dominated occupations females felt less accepted than males. In male 
dominated occupations males were less accepting of the outgroup than were females. 
This finding supports previous research into the situation of women in male dominated 
occupations and into engineering in particular. For example, McIlwee and Robinson 
( 1 99 1 )  found that women often struggled to fit into the tool obsessed male culture that 
exists in engineering. It is encouraging that the males in the male dominated sample 
acknowledge that they are not accepting towards the outgroup, as that is a first step in 
changing the atmosphere. Although like legitimacy beliefs acceptance could have been 
influenced by the fact that male engineers with liberal views may have been more likely 
to fill in the questionnaires. 
In female dominated occupations acceptance of the ingroup was the same for 
men and women, and males were more accepting to the outgroup than females. 
Hypothesis 1 6  that women would be more accepting of men in female dominated 
occupations, than men in male dominated occupations are of women was supported. 
Skevington and Dawkes ( 1 988) found that men had very positive attitudes towards 
women in nursing. Previous studies have shown that women in female dominated 
occupations view men positively and men are also accepted with little gender 
20 1 
discrimination (C. Williams, 1 992). Benokraitis and Feagin ( 1 986) also claim that 
women have more difficulty being accepted into traditional male occupations than 
males do into female occupations. This indicates that it may be possible for males to 
successfully assimilate into female dominated occupations. 
Societal status affects positions within occupations resulting in men being 
accepted into female dominated occupations more easily than women are into male 
dominated occupations. Fairhurst and Snavely ( 1 983a) found that male tokens were not 
as socially isolated as female tokens. Comments from female nurses in the present 
study demonstrated that women felt that they made an effort to include the male nurses 
for example, one of the female nurses said "I think they're often included quite readily 
so that they're not made to feel you know out on a limb on their own". 
Perceived acceptance of the ingroup was higher when there were more ingroup 
members in the workplace. Acceptance towards the outgroup was also higher when 
there was more out group members in the workplace. These findings contrast with those 
of Wharton and Baron ( 1 987) who studied the effect of gender segregation on men's 
well-being. They found that the men in the mixed gender setting experienced lower job 
satisfaction and self-esteem than the men in the male dominated or the female 
dominated work setting. They concluded that token males might actually interact better 
with females in the female dominated setting than in the mixed setting. The relationship 
of acceptance to numbers also challenges SIT's prediction that larger numbers of the 
outgroup are seen as a threat to the ingroup, and offers support for researchers such as 
Kanter ( 1 977) who suggest that tokenism processes will decrease as the numbers of the 
minority in the occupation increase. 
Acceptance and Status 
Hypothesis 1 4  that low status respondents would have perceived lower 
acceptance of the ingroup than high status respondents was supported. Lalonde and 
Silverman ( 1 994) found that there was a relationship between legitimacy and 
permeabi lity. Individuals who saw the situation as highly permeable saw the situation 
202 
as more legitimate. Respondents in the present study who believed the situation to be 
i llegitimate had lower mean scores on acceptance of the ingroup. SIT states that 
accepting some members of the minority legitimises the position of the high status 
group. Status was associated with acceptance beliefs and acceptance of the outgroup 
legitimises the intergroup situation, so that group action is less likely to be taken. The 
hypothesis that high status respondents will express lower acceptance of the outgroup 
than the low status respondents was supported. 
Acceptance and Gender Identification 
Hypothesis 1 5  that high gender identification would be related to lower 
acceptance towards the outgroup was only supported by the individual/group opposition 
subscale of gender identification. The less opposition between group and individual 
needs the less accepting towards the outgroup the ingroup was seen to be. One of the 
aspects of individual/group opposition is fitting in well with the ingroup. Fitting in wel l  
with the ingroup may decrease accepting the outgroup. Lower perceived acceptance of 
the ingroup was related to higher gender identification on the cognitive/emotional 
subscale. The more individuals in gender dominated occupations identified with their 
group the less they felt accepted by the outgroup. This is reflected in comments made 
by female engineers about the ways they minimise stereotypical aspects of femininity in 
order to be accepted at work, which includes not wearing "pink butterfly clips" and 
being "as professional as possible and never cry !" 
The relationship between identification and acceptance is consistent with SIT 
that predicts that high identification, in this case low opposition between the individual 
and the group, will be associated with differentiation between the groups. The results of 
previous studies looking at differentiation and identification have had mixed results. 
Hinkle, Taylor, Fox-Cardamone and Crook ( 1 989) claim that subscales of identification 
may have differing effects on differentiation. 
The results of the current study show that the different sub scales of gender 
identification have different effects on acceptance towards the outgroup. This adds to 
the knowledge of which aspects of identification are important influences on other 
203 
intergroup factors. For respondents working in gender dominated occupations the better 
individuals fit in with their ingroup the less l ikely they may be to make efforts to accept 
the outgroup. 
Beliefs about the Intergroup Situation Summary 
To summarise there were differences in perceived status between gender groups 
and occupations. Legitimacy and stabil ity beliefs were not independent of each other or 
of status beliefs. In line with previous research, perceived acceptance of the ingroup 
was higher in female dominated occupations than in male dominated occupations and 
male nurses felt more accepted than females in male dominated occupations. More 
research is needed into the perceptions and beliefs of equal groups. 
7.4 Social Change Belief Systems -
Achieving Positive Distinctiveness 
Social Change Belief Systems, Gender and Occupations 
There were differences between men and women on social change beliefs. 
For the total sample men scored higher on social mobility than did women. Women 
scored higher on social competition and social creativity than men. 
There were differences between the occupations. Members of male dominated 
occupations scored higher on social mobility than respondents in female dominated 
occupations, members of female dominated occupations scored higher on ingroup social 
creativity than members of male dominated occupations. 
Within each occupation, there were differences between men and women. In 
both female and male dominated occupations females scored higher on ingroup social 
competition and ingroup social creativity than men. In the female dominated 
occupation there was no difference between men and women on ingroup social mobility 
beliefs. In male dominated occupations females scored lower on ingroup social 
204 
mobility than men. There are very few studies on gender differences in social change 
beliefs. Although Skevington and Dawkes ( 1 988) found that male nurses had higher 
individual ambition (equivalent to social mobility) than did female nurses. This was not 
supported by the current study in which there were no differences between male and 
female nurses on social mobility beliefs. 
Men scored higher on individual action, and women scored higher on collective 
action measures. This shows that men may not be as competitive against women as 
women are against men, as the women were higher in social action than men. Research 
into gender dominated occupations supports these findings. O 'Farrel l  and Hanlan 
( 1 982) found well over half the men in the male dominated occupation they studied 
approved of women's presence. Bourhis ( 1 992) conducted an experiment manipulating 
power between groups and gender. He found that males with 70% power discriminated 
less against female outgroup members than males with 1 00% power. Arnancio ( 1 989) 
also found differences in differentiation between men and women. He found higher 
differentiation from men than women. J. Williams ( 1 984) suggested that females have 
different types of beliefs than men. She states that women emphasise affiliation 
between groups and group members, which she labelled communal. Men focus on 
differentiation and competition J. Will iams labelled these beliefs agency. In contrast to 
Arnancio and Williams, the results of the current study show that women can be just as 
competitive as men if not more so on gender issues. 
Social Change Belief Systems and Status 
The low status group had lower ingroup social mobility beliefs than the equal 
and high status group. Low social mobility may be related to the low status group's 
acceptance beliefs as they also felt less accepted than the other groups which rules out 
social mobility as a strategy for improving individual status. The low status group had 
the highest ingroup social competition scores followed by the equal group then the high 
status group. That the low status group would want to improve its position the most is 
consistent with the theory and previous research. However, SIT also predicts the high 
status group will be trying to maintain its status. Bourhis (1992) found increasing 
differentiation between the low, equal and high power conditions. 
205 
The low status group had higher social creativity beliefs than the equal and high 
status groups. Social creativity beliefs are expected to be used more by the low status 
group, as a high status group has more power to choose the more efficient social 
competition strategy to maintain its position. 
The minority group had lower social mobility beliefs than the majority group 
and higher social competition beliefs than the majority. The majority group supports 
individual action and the minority favours collective action. 
Social Change Belief Systems and Identification 
Total Sample 
Gender identification was related to high social mobility and social creativity 
beliefs. Low individual group/opposition was related to low beliefs in social 
competition. Previous studies have not always found the predicted relationship 
between identification and social change beliefs in natural groups. For example Kelly 
(1988) found identification was related to differentiation (a form of social competition). 
Brown et al. (1986) found that identification was not related to differentiation. Turner 
(1999) criticises these previous studies for only focussing on one social change belief 
(differentiation). In addition, other factors may play a part in the relationship such as 
the perceived legitimacy of the status relations. Oaker and Brown (1986) found that the 
amount of contact between the groups of nurses in their study affected the amount of 
differentiation between the groups. SIT research has yet to adequately explore the 
relationship between identification and social change beliefs. 
206 
Gender Groups 
For men gender identification was positively related to ingroup social mobility, 
ingroup social creativity and negatively to ingroup social competition. The relationship 
between identification and social mobility is in the opposite direction predicted by SIT. 
It may be possible that gender is associated with strategies in a manner not predicted by 
S IT. Social mobility seems to be associated with a masculine image of individual 
ambition and assertiveness, which may result in increased identification with men 
leading to increased social mobility beliefs. 
For women gender identification was related to ingroup social creativity, which 
is a collective strategy. J. Williams ( 1 984) states that women may emphasise 
communality, or a focus on affil iation and attachments to others. Men on the other hand 
men emphasise agency or differentiation from others. J. Williams assumes that this 
means that men focus on competitive behaviour and women focus on affiliative 
behaviour. The findings in the current study suggest that perhaps the agency that men 
exhibit is more focused on individual mobility or achievement. The attachment or 
communality that women exhibit does not rule out competitive behaviour, but it may 
mean that women are more l ikely to undertake collective action than social mobility 
strategies. 
Status 
Low Status Group: 
High gender identification was associated with high social creativity beliefs. 
Hypotheses 20 and 28 that high gender identification would be related to low ingroup 
social mobility beliefs and high ingroup social competition beliefs, were not supported 
on the full gender identification scale. However, analysing gender identification by 
subscale shows that low individual/group opposition was related to low ingroup social 
207 
competition beliefs. High gender identification was associated with high ingroup social 
creativity beliefs. High individual group/opposition was related to high ingroup social 
competition beliefs. Feeling held back by the group and not fitting in with the group 
may motivate group action to improve the group's status. 
Equal Group: 
High identification was related to high ingroup social creativity. Respondents 
high in identification were expected not to support social mobility beliefs, as according 
to S IT social mobility is associated with low identification. However high identifiers 
among the equal group did support social mobility for the in group and as a belief for the 
outgroup. This may be due to the unique position of the equal group that has seldom 
been studied. The equal group may not experience the same conflicts between 
individual ambition and group action as for the other status groups. Therefore, 
identification with the group does not rule out individual ambitions. Tajfel ( 1 978) did 
propose that fear of social sanctions from the group can prevent strategies for individual 
mobility. The opinions of the ingroup may influence the choice social change beliefs. 
For example, Barreto and Ellemers (2000) found that accountability to the group 
influenced choice of group or individual action, although Barreto and Ellemers only 
studied groups of low status. For a group of equal status group identity may not be as 
important to the group, as reflected in the lower identification in this group. Therefore, 
the group may not sanction individual strategies. Individual ambition and taking of 
opportunities may even be emphasised by those who believe in the equality of 
opportunities for both groups. 
High identifiers among the equals also supported ingroup social creativity 
strategies and in group social competition for maintaining status. S IT predicts high 
identification to be related to social competition and social creativity. An explanation 
for why high identifiers among the equal group support individual and group strategies 
may be that they must maintain their equality of opportunities. Alternatively support of 
208 
these strategies may have assisted or contributed to the groups achievement of equal 
status. Individuals who believe the status is equal may stil l  support action to maintain 
and achieve status while emphasising individual action. 
High Status Group 
The hypotheses (20 and 28) that predicted relationships between identification 
and social change beliefs for the high status groups were not supported. The only 
relationship was that low individual/group opposition was associated with high social 
mobility beliefs. The identification of the high status group has seldom been studied. 
Oaker and Brown ( 1 986) did study high and low status groups of nurses. Contrary to 
their hypothesis they found a negative relationship between identification and 
differentiation for their high status group. The lack of a direct relationship between 
identification and social change beliefs for the high status group suggests that factors 
other than identification are influential in determining the beliefs of the high status group. 
One of these factors is legitimacy. For example, Finchilescu and de lay Rey ( 1 99 1 )  
found social competition beliefs of the high status group were affected by their 
legitimacy beliefs. Another factor that can affect social change beliefs other than 
identification is favourable treatment to the outgroup (Commins & Lockwood, 1979). 
The social change beliefs may also be affected by different factors depending on whether 
they are attitudes towards the out group or ingroup beliefs. Brown ( 1 995) suggests that 
different motives may be responsible for outgroup attitudes and ingroup bias. 
It is also possible that the measure of social change beliefs for the high status 
group was inadequate. The items were adapted from items designed for low status 
groups and may have not adequately identified the relevant dimensions on which the 
groups make efforts to maintain their status in their workplace. The measure of 
identification may have been inadequate although it has been used successfully in many 
previous studies. Turner and Reynolds (in press) suggest that measuring gender identity, 
as an individual differences variable as was done in this study, may not adequately 
reflect the group nature of identification. 
Social Change Belief Systems and Acceptance 
209 
Acceptance was also hypothesised to relate to social change beliefs. Separate 
hypotheses were formed for the low equal and high status groups. Gender groups were 
also considered separately. 
Total Sample 
For the total sample high acceptance of the in group was associated with high 
social mobility beliefs and low social creativity beliefs as predicted by SIT. Low 
acceptance of the out group was associated with high belief in social mobility, and high 
belief in social competition. 
Gender Groups 
For the men in the sample acceptance of the ingroup was not associated with 
social change beliefs. High acceptance towards the outgroup was associated with low 
social mobility beliefs, which is not as predicted by SIT. For women high acceptance of 
the ingroup was associated with higher beliefs in social mobility. This relationship is 
predicted by SIT, permeabi lity of the group boundaries, which was measured by 
acceptance in the current study, will affect the choice of group or individual action. 
2 1 0  
Status 
Low Status 
Hypothesis 1 9  that high acceptance of the ingroup would be related to high 
ingroup social mobility beliefs was supported. For social mobility to be possible, the 
group must feel that it can be accepted into the high status group. SIT predicts this 
relationship and it has been supported by research findings in a variety of experimental 
groups (Ellemers et aI . ,  1 990; Lalonde & Silverman, 1 994). Acceptance can decrease 
group action, which can be beneficial for the high status group. 
Hypothesis 23 that high acceptance would be related to low ingroup social 
creativity beliefs was supported. If the group cannot join the outgroup then it will think 
of other ways of improving its status which can include using social creativity methods. 
Experiments by lackson et a1. ( 1 996) found that social creativity strategies were 
employed more when the intergroup situation was impermeable. The current study 
offers support from natural low status groups to the relationships predicted by S IT.  
Equal  Status 
The hypotheses that acceptance of the ingroup will be positively related to 
ingroup social mobility and negatively related to all other social change beliefs was not 
supported. Acceptance towards the outgroup, or acceptance of the ingroup was not 
related to social change beliefs for the equal group. There is very l ittle research into the 
social change beliefs of the equal group. If the group perceives the intergroup situation 
as equal, acceptance is not an important factor in determining their social change 
beliefs. There may be other factors that affect the equal group's social change beliefs. 
2 1 1 
High Status 
The hypotheses that high ingroup acceptance will be related to high outgroup 
social mobility, low outgroup social creativity, low outgroup social competition, high 
ingroup social competition and social creativity and low ingroup social mobility were 
not supported. 
Acceptance towards the outgroup was also not related to the social change 
beliefs of the high status group. These results are not consistent with Ellemers et al. 
( 1 990) who found that when the intergroup situation is permeable individuals would 
strive for individual position regardless of their group position. Their research suggests 
that the high status group should be high on social mobility beliefs even when the 
boundaries are open. 
Social Change Belief Systems, Legitimacy and Stability 
Social change beliefs were compared between those who see the situation as 
legitimate and stable. As predicted (Hypothesis 2 1 )  respondents who saw the intergroup 
situation as legitimate had higher social mobility beliefs than those who saw the 
situation as i l legitimate. This may be because acceptance of the group or open 
boundaries between groups legitimises the groups' position, and increases social 
mobility beliefs. Mogghaddam and Perreault ( 1 992) also found that that belief in the 
legitimacy of the intergroup situation was associated with individual action. Hypothesis 
26 that respondents who see the intergroup situation as legitimate would have higher 
social competition beliefs than respondents who saw the situation as i l legitimate was 
not supported. The hypotheses (22, 24 and 27) that social change beliefs would be 
different for respondents who saw the situation as stable or unstable were not supported. 
2 1 2  
Social Change Belief Systems and Predictor Variables 
Social change beliefs are predicted by S IT to be influenced by identification 
with the ingroup, acceptance or permeability of the group boundaries, and legitimacy 
and stability beliefs. These predictions by SIT have had mixed support from previous 
research and the current research. Different factors seem to be important in determining 
each social change belief for different groups. Regression analysis gave more 
information about the relative importance of the variables to each belief. 
Social mobility beliefs are determined by gender, socialisation may encourage 
men to be more individually ambitious than women. Social mobility beliefs are also 
determined by experiences of discrimination. Discrimination discourages social 
mobility, which suggests that discrimination may make individuals more aware that 
individual ambition is not enough for their success. 
Social creativity was best predicted by gender identification. This was predicted 
by SIT and has been found in previous research. Gender is the second best predictor of 
social creativity beliefs, with women favouring social creativity more than do men. 
Social competition beliefs (achieving status) are predicted best by gender and 
experiences of discrimination. Women again favour group action, more than the men 
do, and experiencing discrimination increases group action. The two variables that best 
predict each social change belief are summarised in Table 63 . 
Table 63 
Variables Predicting Ingroup Social Change Beliefs 
Social Change Belief First Variable Second Variable 
Ingroup Social Mobility Gender Discrimination 
Ingroup Social Creativity Identification Gender 
Ingroup Social Competition Gender Discrimination 
(achieving) 
2 1 3  
Social Change Belief Systems - Achieving Positive Distinctiveness 
Summary 
In the choice of social change beliefs to improve group status men and 
respondents who saw themselves of high status favoured individual beliefs -social 
mobility. Women and respondents who saw themselves of Iow status favoured group 
actions. For women and respondents who saw themselves of low status high gender 
identification was related to high social creativity beliefs. For men high gender 
identification was associated with high social mobility and social creativity beliefs. For 
the high status group identification was not related to social change beliefs. Legitimacy 
stability and acceptance also affect choice of social change beliefs. The regression 
analysis shows that different variables predict each social change belief. A better 
approach for future research is to study social change beliefs separately, considering the 
different factors that predict each social change belief. 
7 . 5  Social Change Belief Systems -
Maintain ing Positive Distinctiveness 
Social Change Beliefs, Gender and Occupations 
Respondents in female dominated occupations showed lower support for social 
mobility beliefs in the outgroup than respondents in male dominated occupations. 
Similarly, the ingroup social mobility beliefs were also lower in the male dominated 
occupations. Respondents in female dominated occupations had lower social 
competition (maintaining) beliefs than those in male dominated occupations. This may 
be because female dominated occupations are more accepting so there is less need for 
social competition. 
From the overall sample women showed higher support for social mobility 
beliefs in the outgroup than men, and higher social competition (maintaining status) 
beliefs than men. Women's support of social mobility beliefs in the outgroup is 
surprising given their low scores on ingroup social mobility, social mobility beliefs 
2 14  
appear to be seen as a masculine characteristic. Women showed lower support for 
social competition beliefs in the outgroup than men, and lower support for social 
creativity beliefs in the outgroup than men. This demonstrates that men support action 
by women to improve their status, however women do not support men's attempts to 
maintain their status. This may make it difficult for men in gender dominated 
occupations to improve their position. 
In both the female dominated occupations and the male dominated occupations 
women were higher in support for outgroup social mobility and social competition 
(maintaining) than men, and lower in their support for outgroup social creativity than 
men. In female dominated occupations women were lower on supporting outgroup 
social competition than men. Men in the female dominated occupation supported 
women's attempts at improving their status. These findings are consistent with research 
conducted by Skevington and Dawkes ( 1 988) who found that men in nursing had liberal 
sex role orientations that were very similar to the female nurses. 
Information on the attitudes that the groups have about the outgroup's social 
change beliefs expands on initial predictions by SIT that have been neglected by 
previous research. The findings demonstrate that men do support women's efforts to 
change the intergroup situation. 
Social Change Belief Systems and Status 
The low status group showed higher support of social mobility beliefs in the 
outgroup, than the high status group. The equal status group showed lower support for 
outgroup social competition than the high status group. This may be because the equal 
group does not want to compete, and the outgroup beginning to take action may threaten 
theu equal status. 
The low status group had higher social competition (maintaining status) beliefs 
than the equal group and the high status group. In this situation the high status groups 
are not concerned with maintaining their status. This could be due to the high numbers 
of high status respondents who saw the intergroup situation as illegitimate. 
2 1 5  
Alternatively, it could be because the group's social position is so strong that 
competition to maintain status is unnecessary. The majority group also had had lower 
social competition (maintaining the status) beliefs than the minority group. 
Social Change Belief Systems and Identification 
Total Sample 
High gender identification was related to high ingroup social competition 
(maintaining status). High gender identification was related to low support of social 
creativity bel iefs in the outgroup. 
Gender Groups 
For men high gender identification was related to high support of social mobility 
beliefs in the outgroup, and high ingroup social competition (maintaining status) . High 
gender identification was also related to low support of social competition beliefs in the 
outgroup. 
Status 
Low Status 
Hypotheses 3 1  and 36 that high identification would be related to low out group 
social mobil ity and low support of outgroup social competition were not supported. 
Hypothesis 35 that high identification would be related to low support of outgroup 
social creativity beliefs was supported. High identifiers in the low status group do not 
support attempts by the outgroup to improve the outgroup's status. S IT states that 
attempts to maintain status by the out group wil l  be seen as threats to the status of the 
ingroup (Abrams & Hogg, 1 990b). This may be especially true for members of the low 
status group who identify strongly with their group. The support for the hypothesis that 
high identifiers in the low status group do not support social creativity beliefs in the 
2 1 6  
outgroup elaborates on the predictions made by S IT about how groups react to each 
other. Perceptions of the beliefs of the out group are important in trying to change 
groups attitudes towards each other. High identification with the group for low status 
group members results in resistance to social change for the outgroup. 
Equal Status 
Hypothesis 3 1  that identification would be related to low outgroup social 
mobility was not supported. In contrast, to the hypothesis high identification was 
related to high support of outgroup social mobility beliefs. Other hypotheses were not 
formed for the equal group, as SIT does not directly predict the beliefs of equal groups. 
High Status: 
For the high status group identification was not related to social change beliefs 
for maintaining positive distinctiveness. Hypotheses 34 and 30 were not supported. The 
current research provides an insight into how identification contributes to the different 
belief systems proposed by SIT (Tajfel and Turner, 1 979), particularly it highlights the 
difference between beliefs about the ingroup and the outgroup. The different social 
change beliefs are affected to differing degrees by different factors, although this was 
not explicitly predicted by Tajfel and Turner ( 1 970). Identification in the present study 
was related to beliefs about social creativity for the in and out groups. Identification 
therefore does seem to be an important factor in beliefs about the use of social creativity 
to improve the intergroup situation, but not as an important factor in the other social 
change beliefs 
2 1 7  
Social Change Belief Systems and Acceptance 
Acceptance was also hypothesised to relate to social change beliefs for 
maintaining status. Separate hypotheses were formed for the low equal and high status 
groups. Gender groups were also considered separately. 
Total Sample 
For the total sample high acceptance of the ingroup was related to in group social 
competition (maintaining status) High acceptance towards the outgroup was associated 
with not supporting social creativity beliefs in the outgroup. 
Gender Groups 
For the men in the sample high acceptance of the outgroup (women) was related 
to low support of social creativity beliefs among women. If a group is accepting 
towards the outgroup they may feel that the outgroup is not entitled, or does not need 
group action. For the women in the sample high feelings of acceptance were associated 
with low social competition (maintaining beliefs). When the group boundaries are 
permeable (or a group feels accepted) there is less need for them to maintain their status 
through social competition strategies. 
Status 
Low and Equal Status 
Hypothesis 29 that acceptance would be related to outgroup social mobility 
beliefs was not supported. For the low and equal status groups' acceptance was not 
related to outgroup social change beliefs. 
2 1 8  
H igh Status: 
For the high status group acceptance towards the outgroup was negatively 
related to outgroup social competition. Similarly to the men in the sample if the high 
status group is very accepting they do not see a need for the out group to take group 
strategies to improve their status. However if the high status group is accepting towards 
the outgroup it is expected that they would support social mobility strategies in the 
outgroup (hypothesis 30) this was not supported. 
Social Change Belief Systems, Stability and Legitimacy 
Hypothesis 39 that respondents who saw the inter group situation as legitimate 
will show less support for social competition beliefs in the outgroup than those who saw 
the situation as illegitimate was supported. This is similar to research conducted on 
natural groups by Finchilescu and de la Rey ( 1 99 1 )  who found that among the white 
high status group in their study those who saw the situation as illegitimate showed less 
discrimination towards the outgroup. 
Hypothesis 37 was not supported, in contrast respondents who saw the situation 
as legitimate had lower beliefs in social competition (maintaining status) than those who 
saw the situation as illegitimate. This result is inconsistent with SIT that predicts more 
discrimination when the situation is perceived as legitimate. However there may be less 
need to maintain the position for those who are securely or legitimately in their position. 
Hypothesis 32 that respondents who see the intergroup situation as legitimate will have 
higher support for out group social creativity beliefs was not supported. S IT predicts 
that social creativity can be encouraged in the outgroup as a way of maintaining status 
as it does not change the intergroup situation. However, it is also possible that social 
creativity threatens the position of the ingroup so may not be related to legitimacy. 
Contrary to Hypotheses 33,  38 and 40 and similarly to the ingroup measures, stability 
beliefs were not related to social change beliefs. 
2 1 9  
Social Change Belief Systems and Predictor Variables 
Rather than a simple and direct relationship between all social change beliefs 
and identification, acceptance or legitimacy a complicated relationship exists. Different 
factors are more important for each social change belief. For some acceptance may be 
the key factor determining choice of strategy for others it may be identification or 
legitimacy. Other social change beliefs seem to be most affected by gender. The 
stepwise regression analysis gives more information about the relative importance of the 
variable to each social change belief. The predictors of each social change belief are 
summarised in Table 64. Gender and gender identification are the best predictors of 
outgroup social mobility beliefs. Males were lower in their support of social mobility 
beliefs in the outgroup than females. Ambition and striving for individual position, 
which are characteristics of social mobility strategies, may stil l  be seen as undesirable 
qualities in women, and associated with the masculine role. More research would need 
to be conducted to further explore this suggestion. High gender identification was 
associated with outgroup social mobility beliefs. Respondents who identify highly with 
their gender group may support social mobility in the outgroup as a method of 
controlling the entry of the outgroup and protecting their own group's status. 
Outgroup social creativity is best predicted by gender identification and 
occupation. High identification with the group was associated with low support of 
social creativity strategies in the outgroup; once again this reflects a strategy of 
protecting the group's status. Individuals who identify highly with the group do not 
want the outgroup using strategies that may improve their position. 
Outgroup social competition was best predicted by legitimacy and acceptance. 
This is consistent with the research by Finchilescu and de la Rey ( 1 99 1 )  who found that 
the high status group was less discriminatory when the intergroup situation was 
perceived as i l legitimate. These findings have implications for changing the status of 
groups in gender dominated occupations. To change the attitudes of the high status 
group, it may be necessary to first demonstrate the i llegitimacy of the status relations. 
Similarly, to beliefs about achieving status, women were higher in support for social 
competition beliefs (for maintaining status) than men were. 
220 
Table 64 
Variables Predicting Out grOUP Social Change Beliefs 
Social Change Belief First Variable Second Variable 
Outgroup Social Mobility Gender Identification 
Outgroup Social Creativity Identification Occupation 
Outgroup Social Competition Legitimacy Acceptance 
l ngroup Social Competition 
(maintaining) 
Gender Occupation 
Social Change Belief Systems -Maintaining Positive Distinctiveness 
Summary 
Measuring the attitudes of the group to the outgroups' social change beliefs adds 
to knowledge about S IT. Social change beliefs are affected by gender. Female 
respondents show higher support for social mobility beliefs in the outgroup than men 
and higher social competition (maintaining status) than men. Women also showed 
lower support for social competition and social creativity beliefs in the out group than 
men. Status also affects support of social change beliefs in the outgroup, the low status 
group showed higher support of social mobility beliefs in the outgroup than the high 
status group, and higher social competition beliefs than the equal and high status 
groups. The equal group had lower support for outgroup social competition than the 
high status group. In the low status group identification was related to outgroup social 
creativity, for the high status group identification was not related to social change 
beliefs for maintaining positive distinctiveness. Acceptance was not related to social 
change beliefs for the low and equal group and was negatively related to outgroup social 
competition in the high status group. In a similar manner to the social change beliefs 
for achieving status, different beliefs are best predicted by different variables. Jngroup 
and outgroup beliefs were predicted by the same variables, gender best predicted 
ingroup and outgroup social mobility. Gender was also the best predictor of social 
22 1 
competition for achieving and maintaining status. Gender identification was the best 
predictor of both ingroup and outgroup social creativity. 
7.6 Conclusion 
In conclusion, the present study contributes to the research within the SIT 
paradigm by investigating social change beliefs of groups of low, equal and high status. 
The study contributed to a gap in research by studying each groups' beliefs about the 
outgroups' strategies. The information gathered can be useful in improving the welfare 
of minority members of gender dominated occupations. It also highlights the 
importance of not treating men working in female dominated occupations in the same 
manner as women in male dominated occupations, because they do have different 
situations and experiences. 
Female engineers felt conflict between their group membership and their 
individual needs, but they identified strongly with their gender group. Female engineers 
saw themselves of low status and disadvantaged by their gender in the workplace. The 
intergroup situation in both gender dominated occupations was seen as slow to change. 
Female engineers felt less accepted than did male nurses, and overall nursing was more 
accepting than engineering. Individuals in both occupations who identified highly with 
their gender groups did not feel as accepted as individuals who did not identify with 
their gender group. Males entering nursing find it much easier to assimilate into the 
female dominated profession than do women entering engineering, who must minimise 
their gender group membership to feel accepted. A large number of both male and 
female nurses saw their positions as equal . Although more male nurses saw their 
position as i l legitimate than did female nurses. Acceptance of the group legitimises 
the position of the groups and is associated with individual beliefs for improving status. 
When a group sees itself as accepting towards the outgroup it does not support social 
competition beliefs in the outgroup. 
There were larger differences between men and women in their social change 
beliefs for addressing the status position of the groups than within occupations. 
222 
Although engineers focussed more on individual strategies for improvement than did 
nurses. Men were higher in individual mobility beliefs or ambition than group action. 
However they realise that individual mobility may not be possible for women, as they 
do not support this as a strategy for women. Women are higher in collective beliefs for 
improving their status than men. Men support women's attempts to improve their 
status. For men high gender identification was related to individual strategies and for 
women high gender identification was related to collective strategies, this might be 
reflecting the reality of the different status positions of the genders. For women 
individual strategies may not be as effective as group action. Experiencing 
discrimination seems to highlight the futility of individual action and discourage 
individual mobility beliefs in both men and women. 
There were l imitations with some of the measures used in the present study. 
Improvements in the methodology used within the S IT framework is important, for 
example standard measures would increase the comparability between studies. The 
measure of social change beliefs may not have adequately found items of relevance to 
the high status group, and some of these scales had low internal consistency. In 
addition, the measure may not have been specific enough to each occupation. There 
were also problems with the measures of salience in the current study. S IT does not 
claim that salience is stable, which makes it difficult to measure adequately. 
The study confirmed many of the predictions of SIT, and began investigation 
into the predictors of each social change belief. Research should continue studying the 
different variables that predict social change beliefs rather than focussing narrowly on 
whether identification affects differentiation, as has been done in the past. More 
research into the beliefs of high status groups and equal groups will also add to 
knowledge about intergroup relations. 
REF ERENCES 
Abrams, D. ( 1988). Comments on the motivational status of self-esteem in social 
identity. European Journal of Social Psychology, 18, 3 1 7-334. 
223 
Abrams, D. ( 1989). Differential association: Social developments in gender identity 
and intergroup relations during adolescence. In S. Skevington & D. Baker 
(Eds.), The social identity of women (pp. 59-83). London: Sage. 
Abrams, D. ( 1990). How do group members regulate their behaviour? An integration 
of social identity and self-awareness theories. In D. Abrams, & M. Hogg 
(Eds.), Social Identity Theory: Constructive and critical advances (pp. 89-
1 1 2). New York: Springer-Verlag. 
Abrams, D. ( 1 992). Processes of social identification. In G. Breakwell (Bd.). Social 
psychology of identity and the self-concept (pp. 57-1 00). London: Surrey 
University Press in Association with Academic Press. 
Abrams, D., & Hogg, M. A. ( 1 988). Comments on the motivational status of self­
esteem in social identity and intergroup discrimination. European Journal of 
Psychology, 18, 3 1 7-344. 
Abrams, D. , & Hogg, M. A. ( 1 990a). Social identification, self-categorization and 
social influence. European Review of Social Psychology, /, 1 95-228. 
Abrams, D., & Hogg, M. A. (Eds.). ( 1990b). Social Identity Theory: Constructive and 
critical advances. New York: Springer-Verlag. 
Abrams, D., Sparkes, K., & Hogg, M. A. ( 1 985). Gender salience and social identity: 
The impact of sex of siblings on educational and occupational aspirations. 
British Journal of Educational Psychology, 55, 224-232. 
Abrams, D., Thomas, J., & Hogg, M. A. ( 1 990). Numerical distinctiveness, social 
identity and gender salience. British Journal of Social Psychology, 29, 87-92. 
Acker, J .  ( 1 990). Hierarchies, jobs and bodies: A theory of gendered organizations. 
Gender and Society, 4, 1 39- 1 58. 
Amancio, L. ( 1 989). Social differentiation between 'dominant' and 'dominated' 
groups: Toward and integration of social stereotypes and social identity. 
European Journal of Social Psychology, 19, 1 - 1 0. 
Ashmore, R. D. & del Boca, F. K. (Eds.). ( 1 986). The social psychology of m ale­
female relations: A critical analysis of central concepts. Orlando, FL: 
Academic Press. 
Bailyn, L. ( 1 987). Experiencing technical work: A comparison of male and female 
engineers. Human Relations, 40, 299-3 1 2. 
224 
Baker-Miller, J .  ( 1 976). Towards a new psychology of women. Boston: Beacon Press. 
Barreto, M. ,  & Ellemers, N. (2000). You can't always do what you want: Social 
identity and self-presentational determinants of the choice to work for a low­
status group. Personality and Social Psychology Bulletin, 26, 891 -906. 
Beauvais, C. & Spence, J. T. ( 1 987) Gender, prejudice and categorization. Sex Roles, 
16, 89- 1 00. 
Bell, D. J .  ( 1 982). Policewomen: Myths and reality. Journal of Po lice Science 
Administration, 1 0, 1 1 2- 1 20. 
Bern, S.  ( 1 98 1 ). Gender Schema Theory: A cognitive account of sex typing. 
Psychological Review, 88, 354-364. 
Benokraitis, N.  V. ,  & Feagin, J. R. ( 1 986). Modern sexism: Blatant subtle and covert 
discrimination. Englewood Cliffs, NJ: Prentice Hall . 
Biernat, M. ,  & Vescio, T. K. ( 1 994). Stil l another look at the effects of fit and novelty 
on the salience of social categories. Journal of Experimental Social 
Psychology, 30, 399-406. 
Bill ig, M. ( 1 976). Social identity and social categorisation. In Billig, M. Social 
psychology and intergroup relations (pp. 322-369). London: Academic Press. 
Billig, M. ( 1 985). Prejudice, categorization and particularization: From a perceptual to 
a rhetorical approach. European Journal of So cia I Psychology, 15, 79- 1 03 .  
Billig, M .  & Tajfel, H. ( 1 973) Social categorization and similarity i n  intergroup 
behaviour. European Journal of So cia I Psychology, 3, 27-52. 
225 
Blascovich, J., Wyer, N. A., Swart, L. A., & Kibler, J .  L. ( 1 997). Racism and racial 
categorization. Journal oJ Personality and Social Psychology, 72, 1 364- 1 372. 
Bond, T. & Kemp, S.  ( 1 99 1 ). Are women perceived as underpaid? _New Zealand 
Journal of Psychology, 20, 49-55. 
Bourhis, R. Y .  ( 1 992). Power gender, and intergroup discrimination: Some minimal 
group experiments. In M. Zanna & 1. Olson (Eds.), The Psychology of 
prejudice: The Ontario symposium on personality and social psychology (pp. 
1 7 1 -208). Hillsdale, N J: Erlbaum Associates. 
Bourhis, R. Y, Sachdev, 1 . ,  & Gagnon, A. ( 1 992). Intergroup research with the Tajfel 
matrices: Methodological notes. In M .  Zanna & J. Olson (Eds.), The 
Psychology oJ prejudice: The Ontario symposium on personality and social 
psychology (pp. 209-232). Hil lsdale, NJ: Erlbaum Associates. 
Breakwell ,  G. M .  ( 1 979). Women: Group and identity. Women 's Studies 
International, 72, 1 364- 1 372. 
Breinlinger, S.  & Kelly. C. ( 1 994). Women's responses to status inequality: A test of 
social identity theory. Psychology oJ Women Quarterly, 18, 1 - 1 6  
Brenner, M .  ( 1 985). Survey Interviewing. In M. Brenner, J .  Brown & D. Canter 
(Eds). The research interview: Uses and approaches (pp. 9-36). London: 
Academic Press. 
Brewer, M. B. ( 1 99 1 ). The social self: On being the same and different at the same 
time. Personality and Social Psychology Bulletin, 1 7, 475-482. 
Brewer, M. B. ( 1 993). The role of distinctiveness in social identity and group 
behaviour. In M .  Hogg, & D. Abrarns, (Eds.). Group motivation: Social 
psychological perspectives (pp. 1 - 1 6). New York: Harvester Wheatsheaf. 
Brown, R. J. ( 1 978). Divided we fall :  An analysis of relations between sections of a 
factory workforce. In H .  Tajfel (Ed.), Differentiation between social groups: 
Studies in the social psychology oJ intergroup relations (pp. 395-429). 
London: Academic Press. 
Brown, R. J. ( 1 995). Prejudice: Its social psychology. Oxford, United Kingdom: 
Blackwell. 
226 
Brown, R 1. , Condor, S., Mathews, A., Wade, G., & Williams, 1 . ( 1 986). Explaining 
intergroup differentiation in an industrial organisation. Journal of 
Occupational Psychology, 59, 273-286. 
Brown, R, & Smith, A. ( 1 989). Perceptions of and by minority groups: The case of 
women in academia. European Journal of Social Psychology, 1 9, 6 1 -75 .  
Brown, R, & Williams, 1 . ( 1 984). Group identification: The same thing to all people? 
Human Relations, 37, 547-564. 
Buchan, l .  ( 1 995). Male nurses: Losing their job advantage? Nursing Standard, 9, 30. 
Cockbum, C. ( 1 99 1 ). In the way of women: Men IS resistance to sex equality in 
organizations. Basingstoke: MacMillan. 
Commins, B. ,  & Lockwood, 1 . ( 1 979). The effects of status differences, favoured 
treatment and equity on intergroup comparisons. European Journal of Social 
Psychology, 9, 28 1 -289. 
Condor, S.  ( 1 989). 'Biting into the future ' :  Social change and the social identity of 
women. In S.  Skevington & D. Baker (Eds.). The social identity of women (pp. 
1 5-39). London: Sage. 
Condor, S. ( 1 990). Social stereotypes and social identity. In D. Abrams & M. Hogg 
(Eds.). Social Identity Theory: Constructive and critical advances (pp. 230-
249). New York: Springer-Verlag. 
Cota, A. A., & Dion, K. L. ( 1 986). Salience of gender and sex composition of ad hoc 
groups: An experimental test of distinctiveness theory. Journal of Personality 
and Social Psychology, 50, 770-776. 
Crocker, 1 . ,  & McGraw, K. M. ( 1 984). What's good for the goose is not good for the 
gander: Solo status as an obstacle to occupational achievement for males and 
females. American Behavioral Scientist, 2 7, 357-369. 
Crouch, B. M. ( 1 985). Pandora's  box: Women guards in men's prisons. Journal of 
Criminal Justice, 13, 535-548. 
Deaux, K. ( 1 992). Personalizing identity and socializing self. In G. Breakwell (Ed.). 
Social psychology of identity and the self-concept. London: Surrey University 
Press in association with Academic Press. 
227 
Del Boca, F.  K., & Ashmore, R. D. ( 1 986) Male-female relations: A summing up. In 
R. Ashmore, & F. del Boca (Eds.), The social psychology of male-female 
relations: A critical analysis of central concepts (pp. 3 1 1 -332). Orlando, FL: 
Academic Press. 
Duveen, G. & Lloyd, B. ( 1 986) The significance of social identities. British Journal of 
Social Psychology, 25, 2 1 9-230. 
Ethier, K. A., & Deaux, K. ( 1 994). Negotiating social identity when contexts change: 
Maintaining identification and responding to threat. Journal of Personality and 
Social Psychology, 67, (2) 243-25 1 .  
Ellemers, N. ,  Doosje, B. ,  van Knippenberg, A., & Wilke, H. ( 1 992). Status protection 
in high status minority groups. European Journal of Social Psychology, 22, 
1 23- 1 40. 
Ellemers, N. ,  van Knippenberg, A. ,  & Wi1ke, H .  ( 1 990). The influence of permeability 
of group boundaries and stability of group status on strategies of individual 
mobility and social change. British Journal of Social Psychology, 29, 233-246. 
Ellemers, N., Wilke, H., & van Knippenberg, A. ( 1 993). Effects of the legitimacy of 
low group or individual status on individual and collective status-enhancement 
strategies. Journal of Personality and Social Psychology, 64, 766-778 .  
Emler, N. ,  & Hopkins, N. ( 1 990). Reputation, social identity and self. In  D. Abrams, 
& M .  Hogg, Social Identity Theory: Constructive and critical advances (pp. 
1 1 3- 1 30). New York: Springer-Verlag. 
Evetts, J. ( 1 993). Women and management in engineering: "The glass ceiling" for 
women's careers. Women in Management Review, 8, 1 9-25 .  
Faihurst, G .  T . ,  & Snavely, B .  K .  ( 1 983a). A test of  the social isolation of  male tokens. 
Academy of Management Journal, 26, 353-36 1 .  
Fairhurst, G .  T. ,  & Snave1y, B .  K. ( 1 983b). Majority and minority group relationships. 
Academy of Management Review, 8, 292-300. 
Finchilescu, G., & de la Rey, C. ( 1 99 1 ). Understanding intra-group variations in 
prejudice - The role of perceived legitimacy and stability. South African 
Journal of Psych 0 logy, 21, 225-232 
Floge, L., & Merill, D. M. ( 1 986). Tokenism reconsidered: Male nurses and female 
physicians in a hospital setting. Social Forces, 64, 925-947. 
Glidden-Tracey, C.E. ,  and Wagner, L. ( 1 995). Gender salient attribute x treatment 
interaction effects on ratings of two analogue counselors. Journal of 
Counseling Psychology, 42, 223-23 1 .  
Gornick, V.  ( 1 983). Women in science: 1 00 journeys into the territory. New York: 
Touchstone. 
228 
Gray, J .  ( 1 993). Women arefrom Venus and Men are from Mars. London: Thorsons. 
Greenwood, A. K., & Glidden-Tracey, C. E. ( 1 996). Development and psychometric 
evaluation of the Gender Salience Scale. Paper presented at the Annual 
meeting of the APA. Toronto. 
Griffin, C. ( 1 989). ' I 'm not a women's libber but. . . ' :  Feminism, consciousness and 
identity. In S .  Skevington & D.  Baker (Eds.), The social identity of women (pp. 
1 73- 1 93). London: Sage. 
Gurin, P. ,  & Markus, H. ( 1 989). Cognitive consequences of gender identity. In S .  
Skevington & D.  Baker (Eds.), The social identity of women (pp. 1 52- 1 72). 
London: Sage. 
Gurin, P. ,  & Townsend, A. ( 1 986). Properties of gender identity and their implications 
for gender consciousness. British Journal of So cia I Psychology, 25, 1 39- 1 48 .  
Haslam, S .  A. (2000). Social identity, deindividuation and organizational behaviour: 
The upside to groupthink, information over-sampling and soldiering. In T. 
Postmes, R. Spears, M. Lea, S. Reisher (Eds.). SIDE issues centre stage: 
Recent developments in studies of deindivduation in groups. Amsterdam: 
Royal Netherlands Academy of Arts and Sciences. 
Haslam, S. A., Powell, C. ,  & Turner, J. C. (2000). Social identity, self-categorization, 
and work motivation: Rethinking the contribution of the group to positive 
sustainable organizational outcomes. Applied Psychology: An International 
Review, 49, 3 1 9-339. 
Heikes, E.1 . ( 1 99 1 ) .  When men are the minority: The case of men in nursing. 
Sociological Quarterly, 32, 389-40 1 .  
229 
Heskin, K., & Power, V. ( 1 994). The determinant of Australians' attitudes toward the 
gulf war. The Journal of So cia I Psychology, 134, 3 1 7-330. 
Hinkle, S . ,  Taylor, L. A., Fox-Cardamone, D. L., & Crook, K. F. ( 1 989). Intragroup 
identification and intergroup differentiation: A multi component approach. 
British Journal of So cia I Psychology, 28, 305-3 1 7 . 
Hinkle, S . ,  & Brown, R.l. ( 1 990). Intergroup comparisons and social identity: Some 
links and lacunae. In D. Abrams, & M .  Hogg (Eds.), Social Identity Theory: 
Constructive and critical advances (pp. 48-70). New York: Springer-Verlag. 
Hoelter, 1. W. ( 1 983). The effects of role evaluation and commitment on identity 
salience. Social Psychology Quarterly, 46, 1 40- 1 47. 
Hogg, M. A. ( 1 992). The social psychology of group cohesiveness: From attraction to 
identity. New York: New York University Press. 
Hogg, M. A., & Abrams, D. ( 1 988). Social identifications: A social psychology of 
intergroup relations and group processes. London: Routledge. 
Hogg, M. A. ( 1 985). Masculine and feminine speech in dyads and groups: A study of 
speech style and gender salience. Journal of Language and Social Psychology, 
4, 99- 1 1 1 . 
Hogg, M. A., & Sunderland, 1. ( 1 99 1 ). Self-esteem and intergroup discrimination in 
the minimal group paradigm. British Journal of Social Psychology, 30, 5 1 -62. 
Hogg, M. A., & Turner, 1.  C. ( 1 987). Intergroup behaviour, self-stereotyping and the 
salience of social categories. British Journal of Social Psychology, 26, 325-
340. 
Hurtig, M. & Pichevin, M. ( 1 990). Salience of the sex category system in person 
perception. Sex Roles, 22, 369-395. 
Institute of Professional Engineers New Zealand ( 1 999, November 26). The gender 
makeup - How does engineering compare? Retrieved April 0 1 ,  200 1 ,  from 
the World Wide Web: 
http://www.ipenz.org.nzldocuments/research/ipenz%5Fresearch7.htm 
230 
Jackson, L., Sullivan, L., Hamish, R,  & Hodge, C. ( 1 996). Achieving positive social 
identity: Social mobility, social creativity, and permeability of group 
boundaries. Journal 0/ Personality and Social Psychology, 70, 24 1 -254. 
Jagacinski, C. M., LeBold, W. K., & Linden, K. W. ( 1 987). The relative career 
advancement of men and women engineers in the United States. Work and 
Stress, 1, 235-247. 
James, K., & Greenberg, J. ( 1 989). In-group salience, intergroup comparison, and 
individual performance and self-esteem. Personality and Social Psychology 
Bulletin, 15, 604-6 1 6. 
Jetten, J. , Spears, R., & Manstead, A. ( 1 997). Distinctiveness threat and prototypicality: 
Combined effects on intergroup discrimination and collective self-esteem. 
European Journal o/Social Psychology, 27, 635-657. 
Jurik, N. C. ( 1 985). An officer and a lady: Organizational barriers to women working 
as correctional officers in men's prisons. Social Problems, 32, 375-388. 
Kanter, R M. ( 1 976). The impact of hierarchical structures on the work behaviour of 
women and men. Social Problems, 23, 4 1 5-430. 
Kanter, R. M. ( 1 977). Some effects of proportions on group life: Skewed sex ratios and 
responses to token women. American Journal of Sociology, 82, 965-990. 
Katz, P. A. ( 1 986). Gender identity: Development and consequences. In R. Ashrnore, 
& F. del Boca, (Eds.), The social psychology ofmale-female relations: A 
critical analysis of central concepts (pp. 2 1 .68). Orlando, FL: Academic Press. 
Kelly, C. ( 1 988). Intergroup differentiation in a political context. British Journal of 
Social Psychology, 27, 3 1 9-332. 
KiwiCareers ( 1 998). Hospital nurse: Job market. Retrieved January 1 1 , 200 1 ,  from the 
World Wide Web: http://www.careers.co.nzIjobs/3a_nur/j35 1 1 I e.htrn 
Lalonde, R N., & Silverman, R A. ( 1 994). Behavioural preferences in response to 
social injustice: The effects of group permeability and social identity salience. 
Journal of Personality and Social Psychology, 66, 78-85. 
Lee, Y. T. ( 1 993). Ingroup preference and homogeneity among African American and 
Chinese American students. Journal o/Social Psychology, 133, 225-235 .  
23 1 
Lemaine, G. ( 1 974). Social differentiation and social originality. European Journal of 
Social Psychology, 4, 1 7-52. 
Liff, S. ,  & Aitkenhead, M. ( 1 992). Equal employment opportunities: An attempt to 
restructure employment relations. In J. Hartley & G. Stephenson (Eds.), 
Employment relations: The psychology of influence and control at work (pp. 
27 1 -289). Oxford: Blackwell. 
Lindeman, M., & Sundvik, L. ( 1 995). Evaluative bias and self-enhancement among 
gender groups. European Journal of Social Psychology, 25, 269-280. 
Lips, H. M. ,  & Colwil l ,  N. L. ( 1 988). Psychology addresses women and work: 
Canadian research in 1 980s. Canadian Psychology, 29, 57-68. 
Locke, E., McClear, K., & Knight, D. ( 1 996). Self-esteem and work. In C. Cooper, & 
I. Robertson (Eds.), International review of industrial and organisational 
psychology, Vol. 2 (pp. 6-32). Chicester, United Kingdom: John Wiley and 
Sons. 
Luhtanen, R., & Crocker. J. ( 1 992). A collective self-esteem scale: Self evaluation of 
one's social identity. Personality and Social Psychology, 18, 302-3 1 8 . 
Lorenzi-Cioldi , F. & Doise, W. ( 1 990). Levels of analysis and social identity. In D.  
Abrams, & M.  Hogg (Eds.), Social Identity Theory: Constructive and critical 
advances (pp. 7 1 -88). New York: Springer-Vedag. 
Macke, A. S. ( 1 98 1 ). Token men and women: A note on the salience of sex and 
occupation among professionals and semi-professionals. Sociology of Work 
and Occupations, 8, 25-38. 
Major, B. ( 1 994). From social inequality to personal entitlement: The role of social 
comparisons, legitimacy appraisals, and group membership. Advances in 
Experimental Social Psychology, 26, 293-355. 
Marshall, H. ( 1 99 1 ). The social identities of women lawyers. Operant Subjectivity, 
13, 1 06- 1 2 1 .  
Marshall ,  H. ,  & Wetherell, M .  ( 1 989). Talking about career and gender identities: A 
discourse analysis perspective. In S. Skevington & D.  Baker (Eds.). The social 
identity of women (pp. 1 06- 1 29). London: Sage. 
232 
Martin, S.E. ( 1 979). Policewomen and policewomen : Occupational role dilemmas and 
choices of female officers. Journal of Police Science and Administration, 7, 
3 1 4-323. 
Martin, S.E. ( 1 994). "Outsider within" the station house: The impact ofrace and 
gender on black women police. Social Problems, 41, 383-400. 
McCann, C.D. ,  Ostrom, T.M., Tyner, L. K. , & Mitchell, M.L. ( 1 985). Person 
perception in heterogeneous groups. Journal of Personality and Social 
Psychology, 49, 1 449- 1 459. 
McCauley, C. & Thangavelu, K. ( 1 99 1 ). Individual differences in sex stereotyping of 
occupations and personality traits. Social Psychology Quarterly, 54, 267-279. 
McClean, H. M. ,  & Kalin, R. ( 1 994). Congruence between self-image and 
occupational stereotypes in students entering gender dominated occupations. 
Canadian Journal of Behavioural Science, 26, 1 42- 1 62. 
McGuire, W. J . ,  & McGuire, C. V. ( 1 98 1 ). The spontaneous self-concept as affected 
by personal distinctiveness. In M. Lynch, A. Norem-Hebeison, & K. Gergen 
(Eds.), Self-Concept: Advances in theory and research (pp. 1 47- 1 72). 
Cambridge, MA: BalIinger. 
McGuire, W. J . ,  McGuire, C. V., Child, P. , & Fujioka, T. ( 1 978). Salience of ethnicity 
in the spontaneous self concept as a function of one 's  ethnic distinctiveness in 
the social environment. Journal of Personality and Social Psychology, 36, 
5 1 1 -520. 
McGuire, W. J . ,  McGuire, C .  V., & Winton, W. ( 1 979) . Effects of household sex 
composition on the salience of one 's gender in the spontaneous self-concept. 
Journal of Experimental Social Psychology, 1 5, 77-90. 
McGuire, W. J . ,  & Padawer-Singer, A. ( 1 976). Trait salience in the spontaneous self­
concept. Journal of Personality and Social Psychology, 33, 743-754. 
McIlwee, J., & Robinson, J .  G. ( 1 992). Women in engineering: Gender power and 
workplace culture. Albany, NY: State University of New York Press. 
Metcalfe, B. A. ( 1 987). Male and female managers: An analysis of biographical and 
self-concept data. Work and Stress, 1, 207-2 1 9. 
Miles, M. ,  & Huberman, M .  ( 1 994). Qualitative data analysis: An expanded 
sourcebook (2nd ed.). Thousand Oaks, CA: Sage. 
Millward, L. 1. ( 1 995). Contextualizing social identity in considerations of what it 
means to be a nurse. European Journal of Social Psychology, 25, 303-324. 
233 
Moghaddam, F. M., & Perreault, S. ( 1 992) . Individual and collective mobility 
strategies among minority group members. The Journal of Social Psychology, 
132, 343-357. 
Morgan, S. ,  Schor, S.  M. ,  & Martin, L. R. ( 1 993). Gender differences in career paths in 
banking. Career Development Quarterly, 41, 375-382. 
Morrison, A. M., White, R. P., & Van Velsor, E. ( 1 987). Breaking the glass ceiling: 
Can women reach the top of America 's largest corporations? Reading, MA: 
Addison-W esley. 
Mummendey, A., & Simon, B. ( 1 989). Better or different III: The impact of 
importance of comparison dimension and relative ingroup size upon intergroup 
discrimination. British Journal of Social Psychology, 28, 1 - 1 6. 
New Zealand Department of Statistics. ( 1 993). Social trends: Work. Wellington, New 
Zealand: Author. 
New Zealand Ministry of Women's Affairs. ( 1 992). Status of New Zealand Women 
1 992 . Wellington, New Zealand: Author. 
Nicholson, L. ( 1 994). Interpreting gender. Signs, 20, 77- 1 05 .  
Oaker, G . ,  & Brown, R. J .  ( 1 986). Intergroup relations in  a hospital setting: A further 
test of Social Identity Theory. Human Relations, 39, 767-778. 
Oakes, P .  ( 1 987). The salience of social categories. In J .  C.  Turner, (with M. Hogg, P. 
Oakes, S.  Reicher, & M. Wetherell). Rediscovering the social group: A Self­
Categorization Theory (pp. 1 1 7- 1 4 1 ). Oxford: Basil Blackwell. 
Oakes, P., & Turner, J .  C. ( 1 986). Distinctiveness and the salience of social category 
memberships: Is there an automatic perceptual bias towards novelty? 
European Journal of Social Psychology, 1 6, 325-344. 
234 
O'Farrell, B. & Harlan, S.L. ( 1 982). Craftworkers and clerks: The effect of male co­
workers hostility on women's  satisfaction with non-traditional jobs. Social 
Problems, 29, 252-265. 
Ott, E. ( 1 989). Effects of the male-female ratio at work: Policewomen and male 
nurses. Psychology of Women Quarterly, 13, 41 -57. 
Pease, B., & Pease, A ( 1 998). Why men don 't listen and women can 't read maps. 
Mona Vale, NSW: Pease Training International . 
Qualifications play second fiddle. (2000, October 1 1 ). The New Zealand Herald 
Reiser, P. ( 1 995). Couplehood. Sydney, Australia: Bantam Books. 
Robinson, G. J ,  & McIlwee, J ( 1 989). Women in engineering: A promise unfulfilled. 
Social Problems, 36, 455-472. 
Robinson, G. J. , & McIlwee, J ( 1 99 1 ). Men women and the culture of engineering. 
The SOCiological Quarterly, 32, 403-42 1 .  
Ryan, R. ( 1 99 1 ). Dangerous opportunities? Women and the process of workplace 
reform. In J. Sayers and M. Tremaine (&is.), The vision and the reality: Equal 
employment opportunities in the New Zealand workplace (pp. 234-246). 
Palmerston North, New Zealand: Dunmore Press. 
Schneer, J. A, & Reitman, F. ( 1 994). The importance of gender in mid-career: A 
longitudinal study ofMBA's. Journal of Organizational Behaviour, 32, 307-
320. 
Serbin, L. A ,  & Sprafkin, C. ( 1 986). The salience of gender and the process of sex 
typing in 3 to 7 year old children. Child Development, 57, 1188-1199. 
Sherif, C. W. ( 1 982). Needed concepts in the study of gender identity. Psychology of 
Women Quarterly, 6, 375-398. 
Simeone, A ( 1 987). Academic women: Working towards equality. Boston, MA: 
Bergin and Garvey. 
Skevington, S. M. ( 1 98 1 ). Intergroup relations and nursing. European Journal of 
Social Psychology, 1 1, 43-59. 
235 
Skevington, S .  M. ( 1 989). A place for emotion in social identity theory. In S. 
Skevington, & D. Baker (Eds.), The social identity oJwomen (pp. 40-58). 
London: Sage. 
Skevington, S. M. ,  & Baker, D. (Eds.) .  ( 1 989). The social identity oJwomen. London: 
Sage. 
Skevington, S. M . ,  & Dawkes, D. A. ( 1 988). Minorities at work: Men in a woman's 
world. In D. Canter, J . Jesuino, L. Soczka, & G. Stephenson (Eds.), 
Environmental Social Psychology (pp. 272-280). London: Kluwer. 
Snavely, B .  K., & Fairhurst, G. T. ( 1 984). The male nursing student as a token. 
Research in Nursing and Health, 7, 287-294. 
Smith, B. ( 1 988). Woman: The one and the many. In E. Spelman (Ed.), Inessential 
woman: Problems oJ exclusion in feminist thought (pp. 1 33 - 1 59). Boston: 
Beacon Press. 
Spelman, E. V. (Ed.). ( 1 988). Inessential woman: Problems of exclusion infeminist 
thought. Boston: Beacon Press. 
Statistics New Zealand. ( 1 997) 1 996 census average income by sex for persons aged 1 5  
years over gainfully employed in the full-time labour force. Wellington, New 
Zealand: Author. 
Statistics New Zealand, & New Zealand Ministry of Women's Affairs. ( 1 999). New 
Zealand now: Women. Wellington, New Zealand: Authors. 
Struch, N. ,  & Schwartz, S. H .  ( 1 989). Intergroup aggression: Its predictors and 
distinctness from ingroup bias. Journal of Personality and Social Psychology, 
56, 364-373. 
Swan, S . ,  & Wyer, R. S .  ( 1 997). Gender stereotypes and social identity: How being in 
the minority affects jUdgments of self and others. Personality and Social 
Psychology Bulletin, 23, 1 265- 1 276. 
Swim, J. K., Aikin, K. J., Hal l ,  W. S., & Hunter, B. A. ( 1 995). Sexism and racism: 
Old-fashioned and modem prejudices. Journal of Personality and Social 
Psychology, 68, 1 99-2 1 4. 
Tajfel, H .  ( 1 974). Social identity and intergroup behaviour. Social Science 
Information, J 3, 65-93 . 
236 
Tajfel, H. ( 1 975). The exit of social mobility and the voice of social change: Notes on 
the social psychology of intergroup relations. Social Science Information, 1 4, 
1 0 1 - 1 1 8 . 
Tajfel, H. (Ed.). ( 1 978). Differentiation between social groups: Studies in the social 
psychology of intergroup relations. London: Academic Press. 
Tajfel, H .  ( 1 98 1 ). Human Groups and Social Categories. Cambridge, United 
Kingdom: Cambridge University Press. 
Tajfel, H. ( 1 982). Social psychology of intergroup relations. Annual Review of 
Psychology, 33, 1 -39. 
Tajfel, H., Billig, M., Bundy, R., & Flament, C. ( 1 97 1 ). Social categorization and 
intergroup behaviour. European Journal of Social Psychology, 1, 1 49- 1 78.  
Tajfel, H. ,  & Turner, J .  C. ( 1 979). An integrative theory of intergroup conflict. In W. 
Austin & S. Worchel (Eds.), The social psychology of intergroup relations. 
(pp. 33-48). Monterey, CA: Brooks/Cole. 
Tajfel, H., & Turner, J .  C. ( 1 986). The Social Identity Theory of inter group behaviour. 
In S. Worchel & W. Austin (Eds.), The psychology of inter group relations 
(2nd ed., pp. 7-24). Chicago, IL: Nelson-Hall . 
Tajfel, H. ,  & Wilkes, A. ( 1 963) .  Classification and quantitative judgement. British 
Journal of Psychology, 54, 1 0 1 - 1 1 4. 
Taylor, D., & Moghaddam, F .  M .  ( 1 994). Social Identity Theory. In D.  Taylor & F. 
Moghaddam. Theories of inter group relations: international social 
psychological perspectives (2nd ed., pp 6 1 - 1 94). Westport, RI: Praeger. 
Taylor, S . ,  Fiske, S . ,  Etcoff, N. ,  & Ruderman, A. ( 1 978). Categorical and contextual 
bases of person memory and stereotyping. Journal of Personality and Social 
Psychology, 36, 778-793. 
Taylor, S., & Thompson, S. ( 1 982). Stalking the elusive "vividness" effect. 
Psychological Review, 89, 1 55- 1 8 1 .  
Thoreson, R., Shaughnessy, P., Cook, S.,  & Moore, D. ( 1 993). Behavioral and 
attitudinal correlates of masculinity: A national survey of male counselors. 
Journal of Counselling and Development, 7 J, 337-342. 
Turner, 1.  C. ( 1 999). Some current issues in research on Social Identity and Self 
Categorisation Theories. In N. Ellemers, R. Spears, & B. Doosje (Eds.), 
Social identity: Context, commitment, and content (pp. 6-34). Oxford: 
Blackwell .  
237 
Turner, le., Brown, R. l,  & Tajfel, H. ( 1 979). Social comparison and group interest 
in ingroup favouritism. European Journal o/Social Psychology, 22, 227-239. 
Turner, 1. C., Hogg, M. A ,  Oakes, P. , Reicher, S.,  & Wetherell, M. ( 1 987). 
Rediscovering the social group: A Self-Categorization Theory. Oxford, 
United Kingdom: Basil BlackwelI .  
Turner, J.  e. ,  & Reynolds, K .  (in press). The social identity perspective in  intergroup 
relations: Theories, themes and controversies. In R. Brown, & Gaertner 
(Eds.), Blackwell handbook in social psychology: Vol. 4: Intergroup 
processes. Oxford: Blackwell .  
Turner, 1 .  e., Sachdev, 1., & Hogg, M. A ( 1 983). Social categorisation, interpersonal 
attraction and group formation. British Journal o/Social Psychology, 22, 227-
239. 
van Knippenberg, A, & Ellemers, N. ( 1 993). Strategies in intergroup relations. In M 
Hogg, & D. Abrams, (Eds.), Group motivation: Social psychological 
perspectives (pp. 1 7-32). New York: Harvester Wheatsheaf 
Vaughan, G., & Hogg, M. A. ( 1 995). Introduction to social psychology. Sydney, 
Australia: Prentice-Hall. 
Villeneuve, M. J. ( 1 994). Recruiting and retaining men in nursing: A review of the 
literature. Journal of ProfeSSional NurSing, 1 0, 2 1 7-228. 
Waddell ,  K. ( 1 995). Nursing a gender. Nursing Standard, 9 (30) 62. 
Wharton, A. S., & Baron, J .  N. ( 1 987). So happy together? The impact of gender 
segregation on men at work. SOCiological Review, 52. 574-587. 
238 
Wharton, A. S.,  & Baron, 1. N. ( 1 99 1 ). Satisfaction? The psychological impact of 
gender segregation on women at work. Sociological Quarterly, 32, 365-397. 
Williams, C. L. ( 1 989). Gender differences at work: Women and men in 
nontraditional occupations. Berkeley, CA: University of California Press. 
Williams, C. L. ( 1 992). The glass escalator: Hidden advantages for men in the 
"female" professions. Social Problems, 39, 253-267 . 
Wil liams, J. A. ( 1 984). Gender and intergroup behaviour: Towards an integration. 
British Journal o/Social Psychology, 23, 3 1 1 -3 1 6. 
Williams, J .  A., & Giles, H. ( 1 978). The changing status of women in society: An 
intergroup perspective. In Tajfel, H. (Ed.). Differentiation between social 
groups: Studies in the social psychology of intergroup relations. London: 
Academic Press. 
Yoder, J .  ( 1 994). Looking beyond numbers: The effects of gender status, job prestige 
and occupational gender-typing on tokenism processes. Social Psychology 
Quarterly, 57, 1 50- 1 59. 
Young, I .  M. ( 1 994). Gender as seriality: Thinking about women as a social collective. 
Signs, 19, 7 1 3-738. 
Zimmer, L. ( 1 988). Tokenism and women in the workplace: The l imits of gender­
neutral theory. Social Problems, 35, 64-77. 
Zuppan, L. ( 1 986). Gender-related differences in correctional officers' perceptions and 
attitudes. Journal o/Criminal Justice, 14, 349-36 1 .  
239 
APPENDICES 
Appendix A: 
Questionnaire One: Female Nurses 
SECTION A:  
Please tell m e  about yourself: 
Write down in the space provided the first five important pieces of information about yourself 
that you think of: 
1 )  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  . 
2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  . 
3 )  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  . 
4) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  . 
5 )  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  . 
240 
SECTION B 
Please circle the number on the right that most accurately reflects your opinion of the 
following statements: 
Strongly Neutral Strongly 
Disagree Agree 
1 .  I identify with women as a group. -4 -3 -2 - I  0 1 2 3 4 
2. 1 am glad to be a woman. -4 -3 -2 - I  0 I 2 3 4 
3 .  I feel held back because 1 am a woman. -4 -3 -2 - I  0 I 2 3 4 
4. I think that women work well together. -4 -3 -2 - I 0 I 2 3 4 
5 .  1 see myself as  an important part of  the group 
of working women. -4 -3 -2 - I  0 1 2 3 4 
6. 1 do not fit in well with other women. -4 -3 -2 - 1  0 1 2 3 4 
7 .  I consider my gender to be  important. -4 -3 -2 - I  0 I 2 3 4 
8 .  I feel uneasy with other women. -4 -3 -2 - I  0 1 2 3 4 
9. I feel strong ties to other women. -4 -3 -2 - I  0 I 2 3 4 
Optional : Please use this space if you wish to comment on your choices: 
24 1 
SECTION C 
INSTRUCTIONS: Each of the following groups of statements describes various possible 
opinions about a particular situation or context. For each group of statements, please circle 
the letter next to the one statement which best represents your own attitude within that 
context. Please read all of the statements in a group before choosing the one that most 
clearly reflects your opinion. Choose only one statement within each of the groups of 
statements. 
GROUP 1 
A. Men and women share equal responsibil ities for all kinds of household tasks, whether 
indoor or outdoor. 
B .  Although gender may have some influence over the types of household tasks individuals 
choose, there are other factors such as interest and ability that play a bigger role in 
determining household responsibi l ities. 
C .  There are often differences between men's and women's typical household 
responsibilities, but the differences are not all related to gender. 
D .  Gender i s  certainly one of the big factors that influence which domestic tasks members 
of a household will be responsible for. 
E .  The particular household tasks that an individual takes responsibility for are virtually 
always determined by gender differences in interest and ability. 
GROUP 2 
A. In a family with children, the parent' s  child care responsibilities depend on factors other 
than the gender of either parent. 
B. The gender of each parent in a family is the factor which typically determines who wil l  
be primarily responsible for child care. 
C. Although gender plays a small role in determining family child care responsibilities, 
other factors are more important. 
D.  In a family with children, the differences in parental responsibilities for child care are 
due almost exclusively to the differences between men and women. 
E. The gender differences between men and women at least partially influence a family's 
decision about which parent wil l  have primary responsibility for child care. 
GROUP 3 
A. The choice of a career that will be personally satisfying depends on interests or 
personality factors that are completely independent of a person's gender. 
B .  Gender plays only a minor role in  determining how appropriate and satisfying an 
individual 's career choice will be. 
242 
C. Choosing a career that is typically associated with a person's own gender may partly 
influence career satisfaction, but other factors contribute to career satisfaction as well .  
D.  Gender differences in career interests and opportunities are among the primary factors 
that determine people's satisfaction with their career choices. 
E. A career choice that is consistent with an individual 's masculinity or femininity will be 
the most satisfying. 
GROUP 4 
A.  I notice that my interactions with different friends are largely influenced by their 
gender. 
B. Sometimes I have a preference for spending time with a friend of a particular gender, 
while other times the gender of my friends seems irrelevant. 
C. I behave much differently with my female friends than I do with my male friends. 
D. I find myself acting the same way with different friends, regardless of their gender. 
E .  The way I act with different friends are only minimally influenced by their gender. 
GROUP 5 
A. It's  not obvious why members of one gender should change their names after they are 
married while members of the other do not. 
B. Gender provides a clear basis for determining the names by which people are addressed 
once they are married. 
C. How people are addressed after they get married should be based on factors entirely 
independent of an individual 's  gender. 
243 
D .  There are reasons why gender i s  a convenient basis for determining married names, but 
couples should take many other considerations into account in deciding their married 
names. 
E. Although a few other considerations may be relevant, it makes the most sense to use 
gender as a rule to determine the name each spouse will be addressed by after marriage. 
GROUP 6 
A.  An individual 's gender is among the major factors that influence the way he or she 
makes financial decisions. 
B .  An individual's gender i s  totally irrelevant to the way a person makes financial 
decisions. 
c .  Gender has little to do with an individual's style of making financial decisions. 
D .  Although gender partially determines the way an individual goes about making financial 
decisions, other individual variables are also equally important influences. 
E. Men make very different financial decisions than women make. 
GROUP 7 
A. Men and women very typically have equal access to higher education, without any 
discrimination on the basis of gender. 
B. Because of basic gender differences, men and women have very different options for 
gaining access to higher education. 
c. One of the criteria for determining who will be admitted to a programme in higher 
education is the gender of the applicant. 
D .  One of  the major factors that determine whether or  not an individual will be accepted as 
a candidate for higher education is the individual's gender. 
E .  The differences between men and women in general have little to do with their ability 
to gain access to higher education. 
GROUP 8 
A. Differences in how individuals deal with someone in authority over them might be 
related to gender differences of the individuals involved, but are more related to a lot 
of other important personality or situational factors. 
B. Gender doesn't have any influence on how individuals deal with persons in authority 
over them. 
244 
C. Men's and women's approaches to dealing with authority figures are influenced partly 
by their genders and basic differences between men and women. 
D. The interaction between an individual 's gender and the gender of an authority figure 
totally determines the ways the individual wil l  behave in authority situations. 
E. The way an individual deals with authority figures is mostly determined by the 
individual's gender and the gender of the person in authority. 
GROUP 9 
A. The quality or character of a male same-sex friendship is likely to be slightly different 
than a female same-sex friendship because of gender-differences. 
B. Men's same-sex friendships are very different from women's same-sex friendships 
because of significant differences between the genders. 
c. Gender is a major factor that determines what a same-sex friendship wil l  be like. 
D. There are no distinguishable differences between male friendships and female 
friendships. 
E. Men's friendships and women's friendships are different partly because of gender 
differences. 
GROUP 1 0  
A. Men and women in positions of authority act in ways that are completely unrelated to 
their gender. 
B. Because of basic gender differences, male authority figures operate in a way that is 
completely different from the way female authority figures behave. 
C. Men and women in authority positions almost always behave differently because of 
differences between their genders. 
D. Gender distinctions sometimes come into consideration when looking at the ways 
individuals deal with being in positions of authority. 
245 
E. A person's gender isn't a very important factor when considering the way she or he acts 
when she or he is the person in authority. 
GROUP 1 1  
A. Gender differences between men and women have a reasonably big impact on what 
mixed-sex friendships are normally like. 
B. The character of a friendship between a man and a woman is completely independent of 
their gender differences. 
C .  Gender differences might play a small role in the nature of a friendship between a man 
and a woman, but other factors are more critical. 
D.  Gender differences between men and women make mixed-sex friendships very different 
from same-sex friendships. 
E. The nature of a mixed-sex friendship is partly determined by the differences between 
the genders. 
GROUP 1 2  
A.  It makes a lot of sense to me to use gender as the main rule for deciding the name I or 
my spouse would use following marriage. 
B .  Ideally, I feel that gender should be irrelevant to  deciding what name I or  my spouse 
would be called by after marriage. 
C .  In the ideal situation, considerations other than gender would carry the most weight in 
deciding my or my spouses name after marriage. 
D .  In  my ideal marriage, gender would definitely be the basis for determining my or my 
spouse's  name after I get married. 
E. Ideally, it makes sense to me to use gender as part of the basis for deciding my or my 
spouse's name following marriage. 
246 
GROUP 1 3  
A. In my ideal family, gender would be a convenient way to make many of the decisions 
about specific household responsibi lities. 
B. I think ideally, gender should have only a minor influence on the types of household 
chores that members of a family perform. 
C. In my ideal family, gender would have nothing to do with who takes responsibility for 
which specific household tasks. 
D.  Ideally, gender distinctions would be the appropriate way to determine the specific 
household tasks for which my family members take responsibil ity. 
E. An ideal division of household tasks would not be completely dictated by each family 
member's gender, but gender may play some role in making those decisions. 
247 
SECTION D 
1. Which oJ these groups do you think has the most advantages in your workplace, at the 
present time. 
Female Nurses have the most advantages. 
If you ticked this box continue completing this section (0).  then complete Section E on page 9, 
followed by Section H on page 1 9. 
Male Nurses have the most advantages. 
If you ticked this box continue completing this section (0).  skip Section E and complete 
Section F on page 12 .  followed by Section H on page 1 9. 
Female and Male Nurses have equal advantages. 
If you ticked this box continue completing this section (O) .  skip Sections E and F and 
complete Sections G and H beginning on page 1 5. 
2. Which oJ these groups do you think should have the most advantages in your 
workplace. 
Female Nurses should have the most advantages 
Male Nurses should have the most advantages 
Female and Male Nurses should have equal advantages. 
3 . Which oJ these groups do you think will have the most advantages in your workplace, in 
the Juture. 
Female Nurses will have the most advantages. 
Male Nurses will have the most advantages. 
Female and Male Nurses will have equal advantages. 
4a. Rate the level oJ difficulty/ease that you think male nurses experience in integrating 
with and being accepted by Jemale nurses: 
Difficult 
-4 -3 -2 
Neither Easy or Difficult 
- 1  0 1 
Easy 
2 3 4 
4b. Rate the level oJ difficulty/ease that you thinkJemale nurses experience in integrating 
with and being accepted by male nurses: 
Difficult 
-4 -3 -2 
Neither easy or difficult 
- 1  0 1 
Easy 
2 3 4 
SECTION E 
Complete this section if you ticked that female nurses have the most advantages in question 1 of Section D. 
Please circle the number on the right that most accurately reflects the extent to which you agree that male nurses should feel this way: 
Strongly Neutral Strongly 
Disagree Agree 
1 .  If men in nursing had as many advantages as women 
then we would see more men in senior positions. -4 -3 -2 - 1  0 1 2 3 4 
2. Men are tougher and stronger than women. -4 -3 -2 - 1  0 1 2 3 4 
3 .  Drive and ambition are still important in  our workplaces. -4 -3 -2 - 1  0 1 2 3 4 
4. When male nurses see the kinds of jobs other men are doing 
they should be satisfied with their own achievements. -4 -3 -2 - 1  0 1 2 3 4 
5 .  In  general men make better work colleagues than women. -4 -3 -2 - 1  0 1 2 3 4 
6. If you want something badly enough you can achieve it, if you work hard enough. -4 -3 -2 - 1  0 1 2 3 4 
7. It is more important to be logical than it is to be intuitive. -4 -3 -2 - 1  0 1 2 3 4 
8 .  As a (numerical) minority men have a lower status than women within nursing. -4 -3 -2 - 1  0 1 2 3 4 
9. Most people just don't have enough drive and ambition to get on in their careers in nursing. -4 -3 -2 - 1  0 1 2 3 4 
1 0. The world would be a better place if women were more like men in their values. -4 -3 -2 - 1  0 1 2 3 4 
Strongly Neutral Strongly 
Disagree Agree 
1 1 . Women have an unequal share of status and types of job within nursing. -4 -3 -2 - 1  0 1 2 3 4 
12 .  Men are better at  handling situations that involve the potential for violence than women. -4 -3 -2 - 1  0 1 2 3 4 
1 3 . Men should rarely consider how much female colleagues get paid. -4 -3 -2 - 1  0 1 2 3 4 
1 4. Men should be angry about their low status in nursing. -4 -3 -2 - 1  0 1 2 3 4 
1 5 .  H is important that people make the most of  the opportunities they get i n  their careers. -4 -3 -2 - 1  0 1 2 3 4 
1 6. The role of breadwinner is the most important in the family. -4 -3 -2 - 1  0 1 2 3 4 
1 7 . Men must keep on fighting until they have equality with women in all occupations. -4 -3 -2 - 1  0 1 2 3 4 
1 8 . Men and women have equal opportunities in nursing, 
it's just a matter of having the right personal characteristics for the job. -4 -3 -2 - 1  0 1 2 3 4 
1 9. When male nurses hear about men who stay in the home 
they should feel good about their careers in nursing. -4 -3 -2 - 1  0 1 2 3 4 
20. If you're good at your job, then you will be promoted; 
it's got nothing to do with what gender you are. -4 -3 -2 - 1  0 1 2 3 4 
2 1 .  Women want to keep large numbers of men from becoming nurses 
so that they can keep a hold of the power and authority they've always had. -4 -3 -2 - 1  0 1 2 3 4 
22. There are so many men in senior positions, 
there is no excuse for other men not succeeding in nursing. -4 -3 -2 - 1  0 1 2 3 4 
Strongly 
Disagree 
23. Men's general values and outlook are superior to women's. -4 -3 -2 
24. When assessing their career progress men should try to compare themselves with other men 
rather than comparing themselves with women. -4 -3 -2 
25. It's up to each individual person to make the most of the opportunities they have in their workplace. -4 -3 -2 
26. There is still a lot of prejudice and discrimination within nursing 
which prevents men achieving their full potential. -4 -3 -2 
Please circle the number on the right that most accurately reflects your level of agreement with the following statements: 
Strongly 
Disagree 
1 .  Reverse discrimination is happening and men take advantage of it. -4 -3 -2 
2. Women are better suited emotionally and intellectually for senior positions in nursing than men. -4 -3 -2 
3 .  Women in nursing should be making efforts to  maintain and improve their status. -4 -3 -2 
4. Positive action for men has made women the disadvantaged 
gender in nursing. -4 -3 -2 
5 .  In nursing men are overrepresented in senior and administrative positions once held by women. -4 -3 -2 
Neutral 
- 1  0 1 
- 1  0 1 
- 1  0 1 
- 1  0 1 
Neutral 
- 1  0 1 
- 1  0 1 
- 1  0 1 
- 1  0 1 
- 1  0 1 
2 3 
2 3 
2 3 
2 3 
2 3 
2 3 
2 3 
2 3 
2 3 
Strongly 
Agree 
4 
4 
4 
4 
Strongly 
Agree 
4 
4 
4 
4 
4 
N VI o 
Optional : 
1 )  What are the types of strategies (if any) that you notice male nurses using to try and improve the general situation for men in your occupation? 
2) What are your reactions (if any) to the use of these strategies? 
Please use this space if you wish to comment on your choices: 
Now go stra i g h t  to S ection H (do n o t  com p lete sections F a n d  G ) .  
IV V'I ..... 
SECTION F 
Complete this section if you ticked that male nurses have the most advantages in question 1 of Section D. 
Please circle the number on the right that most accurately reflects your opinion of the following statements: 
Strongly 
Disagree 
1 .  I am proud of the female ability to maintain personal relationships. -4 -3 -2 - 1  
2 .  If companies and institutions provided child care facilities then we 
would see more women in senior positions in nursing. -4 -3 -2 - 1  
3 .  Drive and ambition are overrated i n  our workplaces. -4 -3 -2 - 1  
4 .  I f  women want to get ahead in their nursing careers there is little to stop them. -4 -3 -2 - 1  
5 .  When assessing my career progress I try to compare myself 
with other women rather than comparing myself with men. -4 -3 -2 - 1  
6. Women are better at doing many things at once than men. -4 -3 -2 - 1  
7. Women should be angry about their low status in nursing. -4 -3 -2 - 1  
8 .  I think it is  more important to be intuitive than i t  is  to be logical. -4 -3 -2 - 1  
9. It 's up to each individual woman to make the most of the opportunities she has in her workplace. -4 -3 -2 - 1  
1 0 .  When I see the kinds of jobs other women are doing I 'm satisfied with my own achievements. -4 -3 -2 - 1  
1 1 . Women's general values and outlook are superior to men's. -4 -3 -2 - 1  
Neutral 
0 1 2 
0 1 2 
0 I 2 
0 1 2 
0 1 2 
0 1 2 
0 1 2 
0 1 2 
0 1 2 
0 1 2 
0 1 2 
3 
3 
3 
3 
3 
3 
3 
3 
3 
3 
3 
Strongly 
Agree 
4 
4 
4 
4 
4 
4 
4 
4 
4 
4 
4 
N VI N 
Strongly Neutral Strongly 
Agree Disagree 
1 2 . Women must keep on fighting until they have equality with men in all aspects of life; 
at home and at work. -4 -3 -2 - 1  0 1 2 3 4 
1 3 . I regard the choice to stay at home to raise children as the most honourable and valuable contribution 
a person can make to society. -4 -3 -2 - 1  0 1 2 3 4 
1 4. When I see how l ittle some other women have I feel much better about myself. -4 -3 -2 - 1  0 1 2 3 4 
1 5 . In general women make better work colleagues than men. -4 -3 -2 - 1  0 1 2 3 4 
1 6. There is prejudice and discrimination within nursing which prevents 
women achieving their full potential . -4 -3 -2 - 1  0 1 2 3 4 
1 7. On the whole women are more versatile and flexible than men. -4 -3 -2 - 1  0 1 2 3 4 
1 8 . When I hear about women who stay in the home, I feel good about my career in nursing. -4 -3 -2 - 1  0 1 2 3 4 
1 9. As a woman I have a lower status in society than a man, even if he is in a similar career to me. -4 -3 -2 - 1  0 1 2 3 4 
20. The world would be a better place if men were more like women in their values. -4 -3 -2 - 1  0 1 2 3 4 
2 1 .  It is important now that women make the most of the freedoms they've got and drop 
this feminist stuff. -4 -3 -2 - 1  0 1 2 3 4 
22. Compared with Muslim women, Western women have no grounds for complaint, 
they should think themselves lucky. -4 -3 -2 - 1  0 1 2 3 4 
23. Men want women to stay in the home so that they can keep a hold on the power 
and authority that they've always had. -4 -3 -2 - 1  0 1 2 3 4 
24. Now that there are so many women in senior positions, there is no excuse for other women 
not succeeding in their nursing careers. -4 -3 -2 - 1  0 1 2 3 4 
25. Men have an unequal share of status and types of job within nursing. -4 -3 -2 - 1  0 1 2 3 4 
Strongly 
Disagree 
26. If you're good at your job, then you'l l  be promoted; it 's got nothing to do with what gender you are. -4 -3 -2 - 1  
27. Women are more sensitive and caring than men. -4 -3 -2 - 1  
28.  Men and women have equal opportunities within nursing, 
it 's  just a matter of having the right personal characteristics for the job. -4 -3 -2 - 1  
Optional : 
1 )  What are the types of strategies that you use (if any) to try to improve the general situation for women in your occupation? 
2) What sort of reactions are these strategies met with by other female nurses? 
3) What sort of reactions are these strategies met with by male nurses? . 
Please use this space if you wish to comment on your choices : 
Now go straight to Section H (do not complete Section G) .  
Neutral 
0 2 3 
0 1 2 3 
0 2 3 
Strongly 
Agree 
4 
4 
4 
IV 
Vl 
� 
SECTION G 
Complete this section if you ticked that male nurses and female nurses have equal advantages in question 1 of Section D. 
Please circle the number on the right that most accurately reflects the extent to which you agree that male nurses should feel this way: 
Strongly Neutral 
Disagree 
1 .  Men are tougher and stronger than women. -4 -3 -2 - 1  0 1 2 
2 .  When male nurses see the kinds of jobs other men are doing 
they should be satisfied with their own achievements. -4 -3 -2 - 1  0 1 2 
3 .  In general men make better work colleagues than women. -4 -3 -2 - 1  0 1 2 
4. If you want something badly enough you can achieve it, if you work hard enough. -4 -3 -2 - 1  0 1 2 
5 .  I t  i s  more important to  be logical than i t  i s  to  be  intuitive. -4 -3 -2 - 1  0 1 2 
6. Most people just don't have enough drive and ambition to get on in their careers in nursing. -4 -3 -2 - 1  0 1 2 
7. The world would be a better place if  women were more l ike men in their values. -4 -3 -2 - 1  0 1 2 
8 .  Women have an unequal share of status and types of job within nursing. -4 -3 -2 - 1  0 1 2 
9. Men are better at handling situations that involve the potential for violence than women. -4 -3 -2 - 1  0 1 2 
1 0. Men should be angry about their low status in nursing. -4 -3 -2 - 1  0 1 2 
1 1 . It is important that people make the most of the opportunities they get in their careers. -4 -3 -2 - 1  0 1 2 
1 2. The role of breadwinner is the most important in the family. -4 -3 -2 - 1  0 1 2 
3 
3 
3 
3 
3 
3 
3 
3 
3 
3 
3 
3 
Strongly 
Agree 
4 
4 
4 
4 
4 
4 
4 
4 
4 
4 
4 
4 
N VI VI 
Strongly Neutral Strongly 
Disagree Agree 
1 3 . Men must keep on fighting until they have equality with women in all occupations. -4 -3 -2 - 1  0 1 2 3 4 
1 4. When male nurses hear about men who stay in the home they should feel good about their careers. -4 -3 -2 - 1  0 1 2 3 4 
1 5 . Women want to keep large numbers of men from becoming nurses 
so that they can keep a hold of the power and authority they've always had. -4 -3 -2 - 1  0 1 2 3 4 
1 6. When assessing their career progress men should try to compare themselves with other 
men rather than comparing themselves with women. -4 -3 -2 - 1  0 1 2 3 4 
1 7. Men should be proud of the natural authority and leadership skil ls that they possess. -4 -3 -2 - 1  0 1 2 3 4 
1 8 . It 's up to each individual person to make the most of the opportunities they have in nursing. -4 -3 -2 - 1  0 1 2 3 4 
Please circle the number on the right that most accurately reflects your opinion of the following statements: 
Strongly Neutral Strongly 
Disagree Agree 
1 .  I am proud of the female ability to maintain personal relationships. -4 -3 -2 - 1  0 1 2 3 4 
2. If companies and institutions provided child care facilities then we 
would see more women in senior positions in nursing. -4 -3 -2 - 1  0 1 2 3 4 
3 .  Drive and ambition are overrated in our workplaces. -4 -3 -2 - 1  0 1 2 3 4 
4. Positive action for men has made women the disadvantaged gender in nursing. -4 -3 -2 - 1  0 1 2 3 4 
5 .  When assessing my career progress I try to  compare myself 
with other women rather than comparing myself with men. -4 -3 -2 - 1  0 1 2 3 4 
6. Women are better at doing many things at once than men. -4 -3 -2 - 1  0 1 2 3 4 
Strongly Neutral Strongly 
Disagree Agree 
7. Women should be angry about their low status in nursing. -4 -3 -2 - 1  0 1 2 3 4 
8 .  Reverse discrimination is  happening and men take advantage of it. -4 -3 -2 - 1  0 1 2 3 4 
9. If you're good at your job, then you' ll be promoted it's got nothing to do with what gender you are. -4 -3 -2 - 1  0 1 2 3 4 
1 0. Women's general values and outlook are superior to men's. -4 -3 -2 - 1  0 1 2 3 4 
1 1 . Women are better suited emotionally and intellectually for senior positions in nursing than men. -4 -3 -2 - 1  0 1 2 3 4 
1 2 . I regard the choice to stay at home to raise children 
as the most honourable and valuable contribution a person can make to society. -4 -3 -2 - 1  0 1 2 3 4 
1 3 . When I see how little some other women have I feel much better about myself. -4 -3 -2 - 1  0 1 2 3 4 
1 4. There is a lot of prejudice and discrimination within nursing which prevents 
women achieving their full potential. -4 -3 -2 - 1  0 1 2 3 4 
1 5 .  Now that there are so many women in senior positions, 
there is no excuse for other women not succeeding in their nursing careers. -4 -3 -2 - 1  0 1 2 3 4 
1 6. On the whole women are more versatile and flexible than men. -4 -3 -2 - 1  0 1 2 3 4 
1 7. Women in nursing should be making efforts to maintain and improve their status. -4 -3 -2 - 1  0 1 2 3 4 
1 8 . As a woman I have a lower status in society than a man, even if he is in a similar career to me. -4 -3 -2 - 1  0 1 2 3 4 
1 9. Compared with Muslim women, Western women have no grounds for complaint, 
they should think themselves lucky. -4 -3 -2 - 1  0 1 2 3 4 
20. Men are over represented in senior and administrative positions once held by women. -4 -3 -2 - 1  0 1 2 3 4 
Strongly 
Disagree 
2 1 .  Women are more sensitive and caring than men. -4 -3 -2 
22. Men and women have equal opportunities in nursing, it 's just a matter of having the right personal 
characteristics for the job. -4 -3 -2 
Optional : 
1 )  What are the types of strategies that you use (if any) to try to improve the general situation for women in this occupation? 
2) What sort of reactions are these strategies met with by other women? 
3) What sort of reactions are these strategies met with by men? 
Neutral 
- 1  0 1 2 
- 1  0 1 2 
3 
3 
Strongly 
Agree 
4 
4 
N 
V'I 
00 
4) What are the types of strategies (if any) that you notice male nurses using to try and improve the general situation for men in your occupation? 
P lease use this space if you wish to comment on your choices: 
P l ease Conti n u e  by Co m p l eti ng Section H 
260 
SECTION H 
In order for me to describe my sample can you please answer the following demographic 
questions: 
(Circle correct answers) 
1 .  Sex: M F 
2. Age: . . . . . . . . . . . . . . . .  (Years Only) 
3 .  What is your current position? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  . 
4. How long have you been in this current position? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  (Years Only) 
5 .  What is your current working situation? 
1 .  Full - Time Employment 
2. Part-time Employment 
(30 hours or more a week) 
( less than 30 hours a week) 
6. What is the highest level of formal education that you have achieved? 
1 .  School Certificate (or equivalent) 
2. University Entrance (or equivalent) 
3. Tertiary Qualification! Nursing Diploma: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  . 
4. Postgraduate diplomas or certificates: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  . .  
5 .  Postgraduate degree: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  . .  
7. How many female nurses (including yourself) are there in your immediate workgroup? 
8. How many male nurses are there in your immediate workgroup? . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  . 
9. How much discrimination or prejudice because of your gender have you experienced in 
your present position? 
No prejudice A medium amount 
or Discrimination of prejudice or Discrimination 
-4 -3 -2 - 1  o 
1 0. What is your current marital status? 
1 .  Single 
2. Married 
3 .  Divorced or separated 
4. Widowed 
5 .  Living with a partner 
I I .With what ethnic background do you primarily identify yourself? 
1 .  New Zealander of European descent 
2 3 
A lot 
Di! 
26 1 
2. Maori 
3 .  Pacific Peoples 
4. South East Asian 
5. North Asian 
6. Other . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  (please specify) 
262 
Questionnaire Two: Male Engineers 
SECTION A 
Please tell me about yourself: 
Write down in the space provided the first five important pieces of information about yourself 
that you think of 
1 )  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  . 
2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  . 
3 )  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  . 
4) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . .  . 
5 )  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  . 
SECTION B 
Please circle the number on the right that most accurately reflects your opinion of the 
following statements: 
263 
Strongly Neutral Strongly 
Disagree Agree 
l .  I identify with men as a group. -4 -3 -2 - 1  0 1 2 3 4 
2 .  I am glad to be a man. -4 -3 -2 - 1  0 1 2 3 4 
3 .  I feel held back because I am a man. -4 -3 -2 - 1  0 I 2 3 4 
4.  I think that men work well together. -4 -3 -2 - 1  0 1 2 3 4 
5 .  I see myselfas an important part of the 
group of working men. -4 -3 -2 - 1  0 1 2 3 4 
6. I do not fit in well with other men. -4 -3 -2 - 1  0 1 2 3 4 
7. I consider my gender to be important. -4 -3 -2 - 1  0 1 2 3 4 
8. I feel uneasy with other men. -4 -3 -2 - 1  0 1 2 3 4 
9. I feel strong ties to other men. -4 -3 -2 - 1  0 1 2 3 4 
Optional : Please use this space if you wish to comment on your choices: 
264 
SECTION C 
INSTRUCTIONS:  Each of the following groups of statements describe various possible 
opinions about a particular situation or context. For each group of statements, please circle 
the letter next to the one statement which best represents your own attitude within that 
context. Please read all of the statements in a group before choosing the one that most 
clearly reflects your opinion. Choose only one statement within each of the groups of 
statements. 
GROUP l 
A. Men and women share equal responsibilities for all kinds of household tasks, whether 
indoor or outdoor. 
B. Although gender may have some influence over the types of household tasks individuals 
choose, there are other factors such as interest and ability that play a bigger role In 
determining household responsibilities. 
C. There are often differences between men's and women's typical household 
responsibilities, but the differences are not all related to gender. 
D. Gender is certainly one of the big factors that influence which domestic tasks members 
of a household will be responsible for. 
E. The particular household tasks that an individual takes responsibility for are virtually 
always determined by gender differences in interest and ability. 
GROUP 2 
A. In a family with children, the parent's child care responsibilities depend on factors other 
than the gender of either parent. 
B .  The gender of each parent in  a family i s  the factor which typically determines who will 
be primarily responsible for child care. 
C. Although gender plays a small role in determining family child care responsibilities, 
other factors are more important. 
D. In a family with children, the differences in parental responsibilities for child care are 
due almost exclusively to the differences between men and women. 
E .  The gender differences between men and women at  least partially influence a family's 
decision about which parent will have primary responsibility for child care. 
GROUP 3 
A. The choice of a career that will be personally satisfying depends on interests or 
personality factors that are completely independent of a person 's gender. 
B .  Gender plays only a minor role in  determining how appropriate and satisfying an 
individual 's career choice will be. 
265 
C. Choosing a career that is typically associated with a person's own gender may partly 
influence career satisfaction, but other factors contribute to career satisfaction as well .  
D.  Gender differences in career interests and opportunities are among the primary factors 
that determine people's satisfaction with their career choices. 
E. A career choice that is consistent with an individual 's  masculinity or femininity will be 
the most satisfying. 
GROUP 4 
A. I notice that my interactions with different friends are largely influenced by their 
gender. 
B .  Sometimes I have a preference for spending time with a friend of a particular gender, 
while other times the gender of my friends seems irrelevant. 
C. I behave much differently with my female friends than I do with my male friends. 
D.  I find myself acting the same way with different friends, regardless of their gender. 
E. The ways I act with different friends are only minimally influenced by their gender. 
GROUP 5 
A.  I t 's  not obvious why members of one gender should change their names after they are 
married while members of the other do not. 
B .  Gender provides a clear basis for determining the names by which people are addressed 
once they are married. 
C. How people are addressed after they get married should be based on factors entirely 
independent of an individual 's gender. 
D .  There are reasons why gender i s  a convenient basis for determining married names, but 
couples should take many other considerations into account in deciding their married 
names. 
E. Although a few other considerations may be relevant, it makes the most sense to use 
gender as a rule to determine the name each spouse will be addressed by after marriage. 
GROUP 6 
A. An individual 's gender is among the major factors that influence the way he or she 
makes financial decisions. 
B. An individual 's gender is totally irrelevant to the way a person makes financial 
decisions. 
C. Gender has little to do with an individual's style of making financial decisions. 
266 
D. Although gender partially determines the way an individual goes about making financial 
decisions, other individual variables are also equally important influences. 
E .  Men make very different financial decisions than women make. 
GROUP 7 
A. Men and women very typically have equal access to higher education, without any 
discrimination on the basis of gender. 
B. Because of basic gender differences, men and women have very different options for 
gaining access to higher education. 
C. One of the criteria for determining who wil l  be admitted to a programme in higher 
education is the gender of the applicant. 
D. One of the major factors that determine whether or not an individual will be accepted as 
a candidate for higher education is the individual 's gender. 
E. The differences between men and women in general have little to do with their ability 
to gain access to higher education. 
GROUP 8 
A.  Differences in how individuals deal with someone in authority over them might be 
related to gender differences of the individuals involved, but are more related to other 
important personality or situational factors. 
B .  Gender doesn' t  have any influence on how individuals deal with persons i n  authority 
over them. 
C. Men's and women's approaches to dealing with authority figures are influenced partly 
by their genders and basic differences between men and women. 
D.  The interaction between an individual 's gender and the gender of an authority figure 
totally determines the ways in which the individual will behave in authority situations. 
E. The way an individual deals with authority figures is mostly determined by the 
individual 's gender and the gender of the person in authority. 
GROUP 9 
267 
A. The quality or character of a male same-sex friendship is likely to be slightly different 
than a female same-sex friendship because of gender-differences. 
B. Men's same-sex friendships are very different from women's same-sex friendships 
because of significant differences between the genders. 
c.  Gender is a major factor that determines what a same-sex friendship will be  like. 
D. There are no distinguishable differences between male friendships and female 
friendships. 
E. Men's friendships and women's friendships are different partly because of gender 
differences. 
GROUP 1 0  
A. Men and women in positions of authority act in ways that are completely unrelated to 
their gender. 
B. Because of basic gender differences, male authority figures operate in a way that is 
completely different from the way female authority figures behave. 
C. Men and women in authority positions almost always behave differently because of 
differences between their genders. 
D. Gender distinctions sometimes come into consideration when looking at the ways 
individuals deal with being in positions of authority. 
E. A person's gender isn't a very important factor when considering the way she or he acts 
when she or he is the person in authority. 
GROUP 1 1  
A. Gender differences between men and women have a reasonably big impact on what 
mixed-sex friendships are normally l ike. 
B. The character of a friendship between a man and a woman is completely independent of 
their gender differences. 
C. Gender differences might play a small role in the nature of a friendship between a man 
and a woman, but other factors are more critical . 
268 
D. Gender differences between men and women make mixed-sex friendships very different 
from same-sex friendships. 
E .  The nature of a mixed-sex friendship is partly determined by the differences between 
the genders. 
GROUP 1 2  
A. It makes a lot of sense to me to use my gender as the main rule for deciding the name I 
would use fol lowing marriage. 
B .  Ideally, I feel that my gender should be irrelevant to deciding what name I would be 
called by after marriage. 
C. In the ideal situation, considerations other than my gender would carry the most weight 
in deciding my name after marriage. 
D. In my ideal marriage, my gender would definitely be the basis for determining my name 
after I get married. 
E. Ideally, it makes sense to me to use my gender as part of the basis for deciding my name 
following marriage. 
GROUP 1 3  
A. In my ideal family, gender would be a convenient way to make many of the decision 
about specific household responsibilities. 
B. I think gender should have only a minor influence of the types of household chores that 
members of a family perform. 
c .  I n  m y  ideal family, gender would have nothing t o  do with who takes responsibility for 
which specific household tasks. 
D.  Ideally, gender distinctions would be the appropriate way to determine the specific 
household tasks for which my family members take responsibility. 
E. An ideal division of household tasks would not be completely dictated by each family 
member's gender, but gender may play some role in making those decisions. 
Optional :  Please use this space if you wish to comment on your choices: 
SECTION D 
1. Which of these groups do you think has the most advantages in your profession, at the present time. 
Women have the most advantages in my profession. 
Men have the most advantages in my profession. 
Women and men have equal advantages in my profession. 
2. Which of these groups do you think should have the most advantages in your profession. 
Women should have the most advantages in my profession. 
Men should have the most advantages in my profession. 
Women and men should have equal advantages in my profession. 
3. Which of these groups do you think will have the most advantages in your profession, in the future. 
Women will have the most advantages in my profession. 
Men will have the most advantages in my profession. 
Women and men will be equally advantaged in my profession. 
Optional :  Please use this space if you wish to comment on your choices: 
Please circle the number on the right that most accurately reflects your opinion: 
Low Status High Status 
4. I see women in my profession as having: -4 -3 -2 - 1  0 1 2 3 4 
5 .  I see men i n  m y  profession as having: -4 -3 -2 - 1  0 1 2 3 4 
Unfair Fair 
6.  I see women's status in my profession as: -4 -3 -2 - 1  0 1 2 3 4 
7. I see men's status in my profession as: -4 -3 -2 - 1  0 1 2 3 4 
Worse No Change Improved 
8. Compared to their current position, I think women's 
future status in my profession wil l  be: -4 -3 -2 - 1  0 1 2 3 4 
9. Compared to their current position, I think men's 
future status in my profession wil l  be: -4 -3 -2 - 1  0 1 2 3 4 
Difficult Easy 
1 0. Rate the level of ease or difficulty that you think women experience in integrating with and 
being accepted by men in your profession: -4 -3 -2 - 1  0 1 2 3 4 
1 1 . Rate the level of ease or difficulty that you think men experience in integrating with and 
being accepted by women in your profession: -4 -3 -2 - 1  0 1 2 3 4 
SECTION E 
Please circle the number on the right that most accurately reflects the extent to which you agree that female engineers should feel this way: 
Strongly Neutral 
Disagree 
1 .  Women should be proud of the female ability to maintain personal relationships. -4 -3 -2 - 1  0 1 
2 .  If  the more companies provided child care facilities 
then we would see more women in senior positions. -4 -3 -2 - 1  0 1 
3 .  Drive and ambition are overrated in our workplaces. -4 -3 -2 - 1  0 1 
4. If women want to get ahead in their careers as engineers there is little to stop them. -4 -3 -2 - 1  0 1 
5 .  When assessing their career progress women should try to compare themselves with other women 
rather than comparing themselves with men. -4 -3 -2 - 1  0 1 
6. Women are better at doing many things at once than men. -4 -3 -2 - 1  0 1 
7. Women should be angry about their low status in the engineering profession. -4 -3 -2 - 1  0 1 
8. It is more important to be intuitive than it is to be logical. -4 -3 -2 - 1  0 1 
9. It's up to each individual woman to make the most of the opportunities she has in her workp1ace. -4 -3 -2 - 1  0 1 
1 0 . When women engineers see the kinds of jobs other women are doing 
they should be satisfied with their own achievements. -4 -3 -2 - 1  0 1 
1 1 . Women's  general values and outlook are superior to men's. -4 -3 -2 - 1  0 1 
1 2 . Women must keep on fighting until they have equality with men in all aspects of life;  
at home and at work. -4 -3 -2 - 1  0 1 --
2 3 
2 3 
2 3 
2 3 
2 3 
2 3 
2 3 
2 3 
2 3 
2 3 
2 3 
2 3 
Strongly 
Agree 
4 
I 
4 I 
4 i 
4 
4 
4 
4 
4 
4 
4 
4 
4 IV -.) 
-
Please circle the number on the right that most accuratel� reflects the extent to which you agree that female engineers should feel this wa.r; 
Strongly Neutral Strongly 
Disagree Agree 
1 3 . The choice to stay at home to raise children is the most honourable and valuable contribution 
a person can make to society. -4 -3 -2 - 1  0 1 2 3 4 
1 4. When women see how l ittle some other women have they should feel much better about themselves. -4 -3 -2 - 1  0 1 2 3 4 
1 5 . In general women make better work colleagues than men. -4 -3 -2 - 1  0 1 2 3 4 
1 6. There is still a lot of prejudice and discrimination within the engineering profession which prevents 
women achieving their full potential. -4 -3 -2 - 1  0 1 2 3 4 
1 7. On the whole women are more versatile and flexible than men. -4 -3 -2 - 1  0 1 2 3 4 
1 8 . When women engineers hear about women who stay in the home, 
they should feel good about their careers. -4 -3 -2 - 1  0 1 2 3 4 
1 9. A woman has a lower status in society than a man, even if she is in a similar career to him. -4 -3 -2 - 1  0 1 2 3 4 
20. The world would be a better place if men were more l ike women in their values. -4 -3 -2 - 1  0 1 2 3 4 
2 1 .  It is important now that women make the most of the freedoms they've got and drop 
this feminist stuff. -4 -3 -2 - 1  0 1 2 3 4 
22. Compared with Muslim women, Western women have no grounds for complaint, 
they should think themselves lucky. -4 -3 -2 - 1  0 1 2 3 4 
23.  Men want women to stay in the home so that they can keep a hold on the power 
and authority that they've always had. -4 -3 -2 - 1  0 1 2 3 4 
24. Now that there are so many women in senior positions, 
there is no excuse for other women not succeeding in their careers as engineers. -4 -3 -2 - 1  0 1 2 3 4 
25. Men have an unequal share of status and types of job as engineers. -4 -3 -2 - 1  0 1 2 3 4 
26. If you're good at your job, then you' ll be promoted it's got nothing to do with what gender you are. -4 -3 -2 - 1  0 1 2 3 4 
Please circle the number on the right that most accurately reflects the extent to which you agree that fem ale engineers should feel this way: 
Strongly Neutral Strongly 
Disagree Agree 
27. Women are more sensitive and caring than men. -4 -3 -2 - 1  0 1 2 3 4 
28. Men and women have equal opportunities as engineers, 
it's just a matter of having the right personal characteristics for the job. -4 -3 -2 - 1  0 1 2 3 4 
Please circle the number on the right that most accurately reflects your level of agreement with the following statements: 
Strongly Neutral Strongly 
Disagree Agree 
l .  If companies and institutions gave men as many advantages as women 
then we would see more men in senior positions. -4 -3 -2 - 1  0 1 2 3 4 
2. Positive or reverse discrimination is happening and women take advantage of it. -4 -3 -2 - 1  0 1 2 3 4 
3 .  Men are better suited emotionally and intellectually for senior positions, 
as engineers than women. -4 -3 -2 - 1  0 1 2 3 4 
4. Men should be making efforts to maintain and improve their status as engineers. -4 -3 -2 - 1  0 1 2 3 4 
5 .  Positive discrimination towards women has now made men the disadvantaged gender. -4 -3 -2 - 1  0 1 2 3 4 
6. Women with children should stay at home and free up more jobs for men. -4 -3 -2 - 1  0 1 2 3 4 
Optional : 
1 )  What are the types of strategies (if any) that you notice female engineers using to try and improve the general situation for women in your profession? 
2) What are your reactions (if any) to the use of these strategies? 
Please use this space if you wish to comment on your choices: 
SECTION F 
Please circle the number on the right that most accurately reflects your opinion of the following statements: 
Strongly Neutral Strongly 
Disagree Agree 
1 .  If companies gave men as many advantages 
as women then we would see more men in senior positions. -4 -3 -2 - 1  0 1 2 3 4 
2. Men's general values and outlook are superior to women's. -4 -3 -2 - 1  0 1 2 3 4 
3 .  Drive and ambition and stil l important in  our workplaces. -4 -3 -2 - 1  0 1 2 3 4 
4. Most people just don't have enough drive and ambition, 
to get on in their careers as engineers. -4 -3 -2 - 1  0 1 2 3 4 
5. Men are better at handling situations that involve the potential for violence than women. -4 -3 -2 - 1  0 1 2 3 4 
6. Men should be angry about their low status as within the engineering profession. -4 -3 -2 - 1  0 1 2 3 4 
7. Men are tougher and stronger than women. -4 -3 -2 - 1  0 1 2 3 4 
8. If you want something badly enough you can achieve it , if you work hard enough. -4 -3 -2 - 1  0 1 2 3 4 
9. I think it is more important to be logical than it is to be intuitive. -4 -3 -2 - 1  0 1 2 3 4 
1 0. In general men make better work colleagues than women. -4 -3 -2 - 1  0 1 2 3 4 
1 1 . There is prejudice and discrimination which prevents men 
from achieving their full potential as engineers. -4 -3 -2 - 1  0 1 2 3 4 
1 2. I rarely consider how much my female colleagues get paid. -4 -3 -2 - 1  0 I 2 3 4 
1 3 . It is  important that people make the most of the opportunities they get in their careers. -4 -3 -2 - 1  0 1 2 3 4 
Please circle the number on the right that most accurately reflects your opinion of the following statements: 
Strongly Neutral Strongly 
Disagree Agree 
1 4. When I hear about men who stay in the home I feel good about my career as an engineer. -4 -3 -2 - 1  0 1 2 3 4 
1 5 . There are so many men in senior positions, there is no excuse for other men 
not succeeding as engineers. -4 -3 -2 - 1  0 1 2 3 4 
1 6. The role of breadwinner is the most important in the family. -4 -3 -2 - 1  0 1 2 3 4 
1 7 . Women have an unequal share of status and types of job as engineers. -4 -3 -2 - 1  0 1 2 3 4 
1 8 . Men and women have equal opportunities as engineers, 
it 's j ust a matter of having the right personal characteristics for the job. -4 -3 -2 - 1  0 1 2 3 4 
1 9. When I see the kinds of jobs other men are doing I 'm satisfied with my own achievements. -4 -3 -2 - 1  0 1 2 3 4 
20. Men must keep on fighting until they have equality with women in all occupations. -4 -3 -2 - 1  0 1 2 3 4 
2 1 .  The world would be a better place if women were more like men in their values. -4 -3 -2 - 1  0 1 2 3 4 
22. If you're good at your job, then you will be promoted 
it's got nothing to do with what gender you are. -4 -3 -2 - 1  0 1 2 3 4 
23. When assessing my career progress I try to compare myself with other men 
rather than comparing myself with women. -4 -3 -2 - 1  0 1 2 3 4 
24. It's  up to each individual person to make the most of the opportunities they have in their workplace. -4 -3 -2 - 1  0 1 2 3 4 
25. I think it is more important to be professional than it is to be caring. -4 -3 -2 - 1  0 1 2 3 4 
26. Compared with male nurses, male engineers have no grounds for complaint 
they should think themselves lucky. -4 -3 -2 - 1  0 1 2 3 4 
27. I am proud of the natural authority and leadership skills that men possess. -4 -3 -2 - 1  0 1 2 3 4 
28. Positive discrimination towards women has now made men the disadvantaged gender. -4 -3 -2 - 1  0 1 2 3 4 
Optional : 
1 )  What are the types of strategies that you use (if any) to try to improve the general situation for men in your profession? 
2) What sort of reactions are these strategies met with by other male engineers? 
3) What sort of reactions are these strategies met with by female engineers? 
Please use this space if you wish to comment on your choices: 
IV 
-...l 
-...l 
SECTION H 
In order for me to describe my sample can you please answer the following 
demographic questions: 
(Circle correct answers) 
1 .  Sex: M F 2. Age: . . . . . . . . . . . . . . . .  (Years Only) 
3 .  What i s  your working situation? . . . . . . . . . . . . . . . . . . .  (eg: self-employed, consultant,etc) 
4. What is your field of engineering? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  . 
S .  How long have you been in  this current position? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  (Years Only) 
6. What is your current working situation? 
1 .  Full- Time Employment (30 hours or more a week) 
2. Part-time Employment (less than 30 hours a week) 
7. What is the highest level of education that you have achieved? 
1 .  School Certificate (or equivalent) 
2. University Entrance (or equivalent) 
278 
3. Tertiary Qualification: :  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  (please specify) 
4. Postgraduate diplomas or certificates: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  (please specify) 
S .  Postgraduate degree: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (please specify) 
8. How many women are there in your immediate workgroup? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  . 
9. How many men (including yourself) are there in your immediate workgroup? . . . . . .  . 
1 O. How much discrimination or prejudice because of your gender have you 
experienced in your present position? 
No prejudice A medium amount A lot of prejudice 
or discrimination of prejudice or discrimination or discrimination 
-4 -3 -2 - 1  
1 1 . What is your current marital status? 
1 .  Single 
2. Married 
3 .  Divorced or separated 
4. Widowed 
S. Living with a partner 
o 1 2 
1 1 . With what ethnic background do you primarily identify yourself? 
1 .  New Zealander of European descent 
3 4 
279 
2. Maori 
3 .  Pacific Peoples 
4. South East Asian 
5. North Asian 
6. Other . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  (please specifY) 
Appendix B - Social Change Bel ief Systems Items 
Prepilot and Pi lot Study 
280 
Items also used in the main study are identified with an asterisk. Items changed 
for use in the main study are identified with a plus sign. The remaining items were not 
used in the main study. 
Female Questionnaire: 
Social Mobility 
1 Women have only themselves to blame for their position in society. 
1 2  Most women don't have enough drive and ambition to get on. 
I S  If you want something badly enough you can achieve it, if you work hard 
enough. 
1 8  I rarely think about women as a group. 
23* If women want to get ahead there is little to stop them. 
26 The only way women's position in society will change is if women like me 
continue to get better jobs. 
35* It's up to each individual woman to make the most of the opportunities she has. 
36* It is important now that women make the most of the freedoms they've got and 
drop this feminist stuff. 
38* Now that there are so many women in high status jobs, there is no excuse for 
other women not succeeding at work. 
39 I regard myself  as different to other women. 
42 I believe that if women were as ambitious as men they would be more 
successful .  
28 1 
44* If you're good at your job, then you 'l l  get promotion it's got nothing to do with 
what sex you are. 
45* Men and women have equal opportunities, it's just a matter of having the right 
personal characteristics for the job.  
Social Creativity 
New Dimension : 
4 I don't have to prove that I can do what a man can do to feel good about myself. 
1 3  * In general women make better work colleagues than men. 
1 9  Women are better at managing interpersonal relationships at work. 
20* In their general values and outlook, women are superior to men. 
2 1  + Women are good at doing lots of things at once. On the whole women are more 
versatile and flexible than men. 
43 * I don't envy the role of breadwinner, I think women get a much better deal out 
of l ife. 
46* Women are more sensitive and caring than men. 
Changing the Value: 
2+ I am proud of the female culture of emotion, intuition, love, and personal 
relationships and see them as the most essential human characteristics. 
5 *  Drive and ambition are overrated i n  our society. 
6* I regard the abi lity to nurture children as the most honourable and valuable 
contribution a person can make to society. 
1 6* I think it is as important to be intuitive as it is to be logical. 
1 7* The world would be a better place if men were more like women in their values. 
33  The ability to  care, sensitivity and selflessness are the most important qualities a 
person can have. 
Changing the Outgroup: 
8+ I try to compare myself with other women rather than comparing myself with 
men. 
282 
1 0* When I see the kinds of jobs other women are doing I 'm satisfied with my own 
achievements. 
22 My generation have had so much more opportunity than my mother's generation 
had and I am grateful for that. 
25* When I see how little some other women have I feel much better about myself. 
29* When I hear about women who stay in the home, I feel good about my job. 
37* Compared with Muslim women, western women have no grounds for 
complaint, they should think themselves lucky. 
Social Competition 
3* If companies and institutions treated women equally and provided child care 
facilities then we would see more women in powerful jobs. 
7 When you get right down to the basics, women are an oppressed group and men 
are the oppressors. 
9 I don't  think that the position of women at work will improve unless women act 
as a group rather than as individuals. 
1 1  + A lot of women are better suited emotionally and intellectually for positions of 
power than men. 
1 4  If there is a choice between a man and a woman who are equally qualified for a 
job then as a woman you should employ the woman. 
24 It make me angry that men and women can often earn different amounts for 
doing the same job. 
28 There is nothing intrinsic to women that automatically makes them better 
mothers and housekeepers. 
27* Women should be angry about their low status in society 
283 
30* Women must keep on fighting until they have equality with men in all aspects of 
life; at home and at work. 
3 1  A woman should always try and support other women at work. 
32* There is sti l l  a lot of prejudice and discrimination which prevents women 
achieving their ful l  potential. 
34 * As a woman I have a lower status in society than a man, even if he does a 
similar job to me. 
40* Men want to keep women in the home so that they can keep a hold on the power 
and authority that they've always had. 
4 1  + Men have an unequal share of status and wealth. 
Male Questionnaire 
Social Mobility 
Men have only themselves to thank for their position in society. 
1 2  * Most people just don 't have enough drive and ambition to get on. 
1 5  * If you want something badly enough you can achieve it, if you work hard 
enough. 
1 8  I rarely think about men as a group. 
23 If men want to get ahead there is little to stop them. 
26 The only way men will maintain their position in society is if men like me 
continue to get better jobs. 
35* I t 's  up to each individual person to make the most of the opportunities they have. 
36* It is important that people make the most of the opportunities they've got. 
38* There are so many men in high status jobs, there is no excuse for other men not 
succeeding at work. 
39 I regard myself as different from other men. 
42 I believe that ambitious people will be more successful. 
44* If you 're good at your job, then you will get promotion it's got nothing to do 
with what sex you are. 
284 
45* Men and women have equal opportunities, it's just a matter of having the right 
personal characteristics for the job. 
Social Creativity 
New Dimension:  
4 * Men are tougher and stronger than women. 
1 3  * In general men make better work colleagues than women. 
20* In their general values and outlook, men are superior to women. 
2 1  * Men are better at handling situations that involve the potential for violence than 
women. 
43* The role of breadwinner is the most important in the family. 
Outgroup New Dimension 
1 9  Women are better than men at managing interpersonal relations at work. 
46* Women are more sensitive and caring than men. 
Changing the Values 
5 *  Drive and ambition are stil l  important in our society. 
1 6* I think it is more important to be logical than it is to be intuitive. 
1 7* The world would be a better place if women were more like men in their values. 
Outgroup Changing the Va lues 
2+ I see emotion, intuition , love and the ability to maintain personal relationships 
as the most essential human characteristics. 
6+ I regard the ability to nurture children as the most honorable and valuable 
contribution a person can make to society. 
33  The ability to  care, show sensitivity and selflessness are the most important 
qualities a person can have. 
Changing the Outgroup 
8+ I try to compare myself with other men rather than comparing myself with 
women. 
1 0* When I see the kinds of jobs other men are doing I 'm satisfied with my own 
achievements. 
25* I rarely consider how much my female colleagues get paid. 
29* When I hear about men who stay in the home I feel good about my job. 
Outgroup Changing the Outgroup 
22 This generation of women have had so many more opportunities than their 
mother's generation has and they should be grateful for that. 
285 
37* Compared with Muslim women, western women have no grounds for complaint, 
they should think themselves lucky. 
Social Competition 
3 * If companies and institutions gave men as many advantages 
as women then we would see more men in powerful jobs. 
9 I don't think that men can hold their position in the workplace unless men act as 
a group rather than as individuals. 
1 4  If there is a choice between a man and a woman who are equally qualified for a 
job then as a man you should employ the man. 
27+ Men should be making efforts to maintain their status in society. 
28 Women intrinsically make better mothers and housekeepers than men. 
3 1  A man should always try and support other men at work. 
286 
32 Women are now equal with men in society and should drop this feminist stuff. 
34 As a man I have a higher status in society than a woman, even if she does a 
similar job to me. 
4 1  * Women have an unequal share of status and wealth. 
I tems retained in the main study for Social Competition (mainta in ing status). 
7 Positive or reverse discrimination is happening and women take advantage of it. 
1 1  Men are better suited emotionally and intellectually for positions of power than 
women. 
24+ Men with families to support deserve to be paid more than women. 
30 Positive action for women has now made men the disadvantaged sex. 
40 Women with children should stay at home and free up more jobs for men. 
287 
APPENDIX C - Social Change Bel ief Systems Items 
Main Study 
Male Ingroup Items 
Social Mobility 
9 Most people just don't have enough drive and ambition to get on in their careers 
. . m nursmg. 
6 If you want something badly enough you can achieve it, if you work hard enough. 
1 5  It is important that people make the most of the opportunities they get in their 
careers. 
1 8  Men and women have equal opportunities in nursing/engineering, it's just a 
matter of having the right personal characteristics for the job. 
20 If you're good at your job, then you will be promoted it's got nothing to do with 
what gender you are. 
22 There are so many men in senior positions, there is no excuse for other men not 
succeeding in their nursing/engineering careers. 
25 It's up to each individual person to make the most of the opportunities they have 
in their workplace. 
Social Creativity 
Changing the Values 
3 Drive and ambition are stil l  important in our workplaces. 
7 I think it is more important to be logical than it is to be intuitive. 
1 0  The world would be a better place if women were more like men in their values. 
27 I think it is more important to be clinical than it is to be emotional (nurses). 
I think it is more important to be professional than it is to be caring ( engineers). 
29 I am proud of the natural authority and leadership skills that men possess. 
New Dimension: 
2 Men are tougher and stronger than women. 
5 In general men make better work colleagues than women. 
288 
1 2  Men are better at handling situations that involve the potential for violence than 
women. 
1 6  The role of breadwinner is the most important in the family. 
23 Men's general values and outlook are superior to women's. 
Changing the Outgroup 
4 When I see the kinds of jobs other men are doing I 'm satisfied with my own 
achievements. 
1 3  I rarely consider how much my female colleagues get paid. 
1 9  When I hear about men who stay in the home I feel good about my career in 
nursing/ engineering. 
24 When assessing my career progress I try to compare myself with other men rather 
than comparing myself with women. 
28 Compared with male early childhood teachers, male nurses have no grounds for 
complaint, they should think themselves lucky (nurses). 
Compared with male nurses, male engineers have no grounds for complaint, they 
should think themselves lucky ( engineers). 
Social Competition 
1 If men in nursing/engineering had as many advantages. 
as women then we would see more men in senior positions. 
8 As a (numerical) minority I have a lower status than a woman within 
nursing/engineering. 
1 1  Women have an unequal share of status and types of job within nursing. 
1 4  Men should be angry about their low status in nursing/engineering. 
1 7  Men must keep on fighting until they have equality with women in all 
occupations. 
2 1  Women want to keep large numbers of men from becoming nurses so that they 
can keep a hold of the power and authority they've always had (nurses only). 
289 
26 There is stil l  a lot of prejudice and discrimination which prevents men achieving 
their full potential in nursing/ as engineers. 
28 Positive discrimination towards women has now made men the disadvantaged 
gender (engineers only). 
Male Outgroup Items 
Social Mobility 
4 If women want to get ahead in their careers in their nursing careers/as engineers 
there is little to stop them. 
9 Ifs up to each individual woman to make the most of the opportunities she has in 
her workplace. 
2 1  It is important now that women make the most of the freedoms they've got and 
drop this feminist stuff. 
24 Now that there are so many women in senior positions, there is no excuse for 
other women not succeeding in their careers in their nursing careers/as engineers. 
26 If you're good at your job, then you' l l  be promoted it's got nothing to do with 
what gender you are. 
28 Men and women have equal opportunities in nursing/ as engineers , it's just a 
matter of having the right personal characteristics for the job. 
290 
Social Creativity 
Changing The Values 
1 Women should be proud of the female ability to maintain personal relationships. 
3 Drive and ambition are overrated in our workplaces. 
8 It is more important to be intuitive than it is to be logical. 
1 3  The choice to stay at home to raise children is the most honourable and valuable 
contribution a person can make to society. 
20 The world would be a better place if men were more like women in their values. 
New Dimension 
6 Women are better at doing many things at once than men. 
1 1  Women's general values and outlook are superior to men's. 
15 In general women make better work colleagues than men. 
1 7  On the whole women are more versatile and flexible than men. 
27 Women are more sensitive and caring than men. 
Changing the Outgroup 
5 When assessing their career progress women should try to compare themselves 
with other women rather than comparing themselves with men. 
1 0  When women see the kinds of jobs other women are doing they should be 
satisfied with their own achievements. 
1 4  When women see how little some other women have they should feel much better 
about themselves. 
1 8  When female nurses/engineers hear about women who stay in the home, they 
should feel good about their careers in nursing/engineering. 
22 Compared with Muslim women, Western women have no grounds for complaint, 
they should think themselves lucky. 
Outgroup Social Competition 
2 If the police provided child care facilities then we would see more women in 
senior positions (in nursing). 
7 Women should be angry about their low status in nursing/ as engineers. 
291 
1 2  Women must keep on fighting until they have equality with men in all aspects of 
life, at home and at work. 
1 6  There is still a lot of prejudice and discrimination within nursing/engineering 
which prevents women achieving their ful l  potential . 
1 9  A woman has a lower status in society than a man, even i f  she is in a similar 
career to him. 
23 Men want women to stay in the home so that they can keep a hold on the power 
and authority that they've always had. 
25 Men have an unequal share of status and types of job within nursing/as engineers. 
Ingroup Social Competition (maintaining status) 
1 If companies and institutions gave men as many advantages as women then we 
would see more men in senior positions. 
2 Affirmative action is necessary for men to maintain and improve their status 
within nursing (nurses). 
Positive or reverse discrimination is happening and women take advantage of it 
( engineers). 
3 Men are better suited emotionally and intellectually for senior positions within 
nursing/engineering than women. 
4 Men should be making efforts to maintain their status within nursing/engineering. 
5 Positive action for women has now made men the disadvantaged gender 
(engineering only). 
6 Women with children should stay at home and free up more jobs for men. 
Female Ingroup Items 
Social Mobility 
292 
4 If women want to get ahead in their nursing/engineering careers there is l ittle to 
stop them. 
9 It 's up to each individual woman to make the most of the opportunities she has in 
her workplace. 
2 1  It is important now that women make the most of the freedoms they've got and 
drop this feminist stuff. 
24 Now that there are so many women in senior positions, there is no excuse for 
other women not succeeding in their nursing careers/ as engineers. 
26 If you 're good at your job, then you 'l l  be promoted it's got nothing to do with 
what gender you are. 
28 Men and women have equal opportunities in nursing/ as engineers , it's just a 
matter of having the right personal characteristics for the job. 
Social Creativity 
Changing The Values 
I am proud of the female ability to maintain personal relationships. 
3 Drive and ambition are overrated in our workplaces. 
8 I think it is more important to be intuitive than it is to be logical . 
1 3  I regard the choice to stay at home to raise children as the most honourable and 
valuable contribution a person can make to society. 
20 The world would be a better place if men were more like women in their values. 
New Dimension 
6 Women are better at doing many things at once than men. 
1 1  Women's general values and outlook are superior to men's. 
1 5  In general women make better work colleagues than men. 
1 7  On the whole women are more versatile and flexible than men. 
27 Women are more sensitive and caring than men. 
Changing the Outgroup 
293 
5 When assessing my career progress I try to compare myself with other women 
rather than comparing myself with men. 
1 0  When I see the kinds of jobs other women are doing I 'm satisfied with my own 
achievements. 
1 4  When I see how little some other women have I feel much better about myself. 
1 8  When I hear about women who stay in the home, I feel good about my career in 
nursing/as an engineer. 
22 Compared with Muslim women, Western women have no grounds for 
complaint, they should think themselves lucky. 
Social Competition 
2 If companies and institutions provided child care facilities then we would see 
more women in senior positions in nursing/engineering. 
7 Women should be angry about their low status in nursing/as engineers. 
1 2  Women must keep on fighting until they have equality with men in all aspects of 
l ife; at home and at work. 
1 6  There is still a lot of prejudice and discrimination within nursing/engineering 
which prevents women achieving their full potential. 
1 9  As a woman I have a lower status in society than a man, even if he is in a similar 
career to me. 
23 Men want women to stay in the home so that they can keep a hold on the power 
and authority that they've always had. 
25 Men have an unequal share of status and types of job within nursing/engineering. 
Female Outgroup Items 
Social Mobility 
294 
6 If you want something badly enough you can achieve it, if you work hard enough. 
9 Most people just don't have enough drive and ambition to get on in their careers 
in nursing/engineering. 
1 5  It is important that people make the most of the opportunities they get in their 
careers. 
1 8  Men and women have equal opportunities in nursing/engineering, it 's just a 
matter of having the right personal characteristics for the job. 
20 If you're good at your job, then you will be promoted it's got nothing to do with 
what gender you are. 
22 There are so many men in senior positions, there is no excuse for other men not 
succeeding in nursing/engineering. 
25 It's up to each individual person to make the most of the opportunities they have 
in their workplace. 
Social Creativity 
Changing the Values 
3 Drive and ambition are stil l  important in our workplaces. 
7 It is more important to be logical than it is to be intuitive. 
1 0  The world would be a better place if women were more like men in their values. 
27 It is more important to be clinical than it is to be emotional (nurses) 
It is more important to be professional than it is to be caring (engineers). 
29 Men should be proud of the natural authority and leadership skills that they 
possess. 
New Dimension 
2 Men are tougher and stronger than women. 
5 In general men make better work colleagues than women. 
295 
1 2  Men are better at handling situations that involve the potential for violence than 
women. 
1 6  The role of breadwinner is the most important in the family. 
23 Men's general values and outlook are superior to women's. 
Changing the Outgroup 
4 When male nurses/engineers see the kinds of jobs other men are doing the should 
be satisfied with their own achievements. 
1 3  Men should rarely consider how much female colleagues get paid. 
1 9  When male nurses/engineers hear about men who stay in the home they should 
feel good about their careers in nursing. 
24 When assessing their career progress men should try to compare myself with other 
men rather than comparing themselves with women. 
28 Compared with male early childhood teachers, male nurses have no grounds for 
complaint, they should think themselves lucky (nurses). 
Compared with male nurses, male engineers have no grounds for complaint they 
should think themselves lucky ( engineers). 
Outgroup Social Competition 
1 If men in nursing!engienering had as many advantages 
as women then we would see more men in senior positions. 
8 As a (numerical) minority men have a lower status than women within nursing 
(nurses). 
1 1  Women have an unequal share of status and types of job within nursing! 
engmeenng. 
1 4  Men should be angry about their low status in nursing/engineering. 
1 7  Men must keep on fighting until they have equality with women in all 
occupations. 
2 1  Women want to keep large numbers of men from becoming nurses so that they 
can keep a hold of the power and authority they've always had (nurses). 
Positive action for women has now made men the disadvantaged gender 
( engineers). 
296 
26 There is stil l  a lot of prejudice and discrimination in nursing/engineering which 
prevents men achieving their full potential. 
Ingroup Social Competition (maintaining status) 
Reverse discrimination is happening and men take advantage of it (nurses). 
2 Women are better suited emotionally and intellectually for senior positions in 
nursing/engineering than men. 
3 Women in nursing/engineering should be making efforts to maintain and improve 
their status. 
4 Positive action for men has made women the disadvantaged gender in 
nursing (nurses). 
Affirmative action is necessary for women to maintain and improve their status 
within engineering (engineers). 
5 Men are overrepresented in senior and administrative positions once held by 
women (nurses). 
Positive or reverse discrimination is happening and women should take 
advantage of it (engineers).