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. 
 



Massey University Library 
Thesis Copyright Form 

· 1 /. ·s 7~o,,..., /1/o-/cJ ;, ;:x/1-, Title of thesisr?f,c ,K:rr,O/C..J ,-ore /k ~ e..i . • 

o.f:7 s-er vc. :rS' o,,-,cv <f :::x/1,<,,,-✓-6 v-"4 o,-,cv ;><ho 1.:t '-' /s I n ~o/?,s 1 

(1) (aV I give permission for my thesis to be made available to G-.3 1/·..r~ 

V readers in Massey University Library under conditions 
determined by the Librarian. 

(b) I do not wish my thesis to be made available to readers 
without my written consent for ... months. ' 

(2) y I agree that my thesis, or a copy, may be sent to 
another institution under conditions determined by the 
Librarian. 

(b) I do not wish my thesis, or a copy, to be sent to 
another institution without my written consent for 
months. 

(3) J.ai I agree that my thesis may be copied for Library use. 

(b) I do not wish my thesis to be copied for Library use for 
... months. 

Signed~~ v..;/<(' 

Date :2 6 · .,e. 9..i' 

. The copyright of this thesis belongs to the author. Readers must 
sign their name in the space below to show that they recognise 
this. They are asked to add their permanent address. 

NAME AND ADDRESS ·DATE 

15 



ARE FEMALES MORE HELPLESS THAN MALES? 

AN OBSERVATIONAL AND ATTRIBUTIONAL ANALYSIS 

IN MATHS AND ENGLISH 

A thesis presented in partial fulfilment of 

the requirements for the degree of 

Master of Arts in Psychology 

at Massey University 

Gillian Naomi Hawke 

1993 



ii 

ABSTRACT 

The present study investigated the under-achievement and participation of females in 

mathematical areas within the context of the attribution theory. Gender differences 

for causal attributions and achievement-related beliefs were investigated in Maths and 

English, employing a methodology which allowed for the subjective construction of 

the situation by the student. Subjects were 97 form five Maths and English students 

(50 males and 47 females). Overall, there were no consistent gender differences in 

attributions for success and failure in Maths and English. Although males perceived 

themselves as more competent in Maths, there were no gender differences in 

achievement-related beliefs. However, females displayed more mastery-oriented 

cognitions in English . 

Additionally, the relationship of gender and teacher-student interactions in Maths and 

English classrooms were investigated, in an attempt to conceptualise the role they 

have in sustaining gender related behaviours. It was hypothesized that males and 

females were being treated differently in Maths and English, which in some way 

affects their attributions for achievement outcomes, and subsequent achievement­

related beliefs. Four classrooms (two Maths and two English) were observed for five 

hours each. Contrary to predictions, there were few significant differences in the 

contingencies of evaluative feedback given to students, with respect to its frequency, 

its typical referents, and the specificity of its use. 

The results were discussed in terms of their relationship to other studies, and the 

implications for past and future methods of studying students' causal attributions in 

mathematical and verbal achievement situations. Alternative mechanisms by which 

females" self-derogating beliefs might inhibit their participation and achievement in 

maths-related areas were also considered. 



iii 

ACKNOWLEDGEMENTS 

I wish to thank my supervisors Dr. Ross Flett for his supportive assistance, and 

unique humour which inspired me to continue when things were particularly 

challenging, and Dr. Alison George for her guidance and expertise. I know it wasn't 

always easy fitting me into an already busy schedule. 

Thanks also to the students, teachers and principal of the college I attended for their 

co-operation and assistance, making this study possible. 

I also wish to express my deepest thanks to my family for their constant support, 

both emotionally and financially. I am especially grateful to my mother who so 

willingly proof-read the manuscript. 

Finally, very special thanks go to my flatmates and friends for their encouragement, 

and support, and for bearing with me at times when I know I wasn't easy to live 

with. In particular, I would like to thank Rachel, as writing this thesis would not 

have been the same without her friendship. 



TABLE OF CONTENTS 

CHAPTER page 

Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ii 

Acknowledgements ...................................... iii 

Table of Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iv 

List of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii 

1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 

Gender Differences in Educational Participation 

and Achievement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 

Approaches to Differential Gender Participation 

in Education and Employment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 

2 REVIEW OF LITERATURE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 

Attribution Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 

Causal Consequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 

Learned Helplessness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 

Learned Helplessness in the Classroom . . . . . . . . . . . . . . . . . . . . . . . 16 

Gender Differences in Learned Helplessness . . . . . . . . . . . . . . . . . . . . 17 

An Alternative Explanation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 

Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 

The Effect of Feedback on Attributions . . . . . . . . . . . . . . . . . . . . . . . 22 

Attributions and the Sex-Appropriateness of the Task . . . . . . . . . . . . . 27 

Attributions and Perceived Competence . . . . . . . . . . . . . . . . . . . . . . . 32 



3 THE PRESENT STUDY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 

Recapitulation and Hypotheses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 

Part One - Observational Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 

Gender Differences in Feedback . . . . . . . . . . . . . . . . . . . . . . . 36 

Gender Differences in Feedback in Maths 

and English . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 

Part Two - Attributional Questionnaire . . . . . . . . . . . . . . . . . . . . . . . . 39 

Gender Differences in Attributions . . . . . . . . . . . . . . . . . . . . . 40 

Gender Differences in Mastery and Learned 

Helplessness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 

Gender Differences in Perceived Competence . . . . . . . . . . . . . . 43 

Sex-Consistent versus Sex-Inconsistent Tasks . . . . . . . . . . . . . . 44 

4 METHOD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 

Overview and Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 

Subjects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 

Instruments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 

Development of the Attribution Questionnaire . . . . . . . . . . . . . . . 48 

Observation Schedule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 

Pilot Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 

Observer Training . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 

Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 

Introduction of Study to Students . . . . . . . . . . . . . . . . . . . . . . . 53 

Classroom Observation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 

Questionnaire Administration . . . . . . . . . . . . . . . . . . . . . . . . . . 54 

Summary of Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 



5 RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 

Section A: Observational Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 

Overall Feedback . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 

Feedback Differences in Maths and English . . . . . . . . . . . . . . . . 63 

Summary ........................................ 69 

Section B: Attributional Questionnaire . . . . . . . . . . . . . . . . . . . . . . . 70 

Gender Differences in Attributions . . . . . . . . . . . . . . . . . . . . . . 71 

Gender Differences in Mastery and Learned Helpless 

Attributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 

Gender Differences in Perceived Competence . . . . . . . . . . . . . . . 7 4 

Sex-Consistent versus Sex-Inconsistent Tasks . . . . . . . . . . . . . . . 75 

Gender Differences in Achievement-Related 

Beliefs in Maths and English . . . . . . . . . . . . . . . . . . . . . . . . . . 78 

Summary ........................................ 81 

6 DISCUSSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 

Observational Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 

Attributional Questionnaire . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 

Gender Differences in Attributions . . . . . . . . . . . . . . . . . . . . . . 87 

Relationships Between Attributions and 

Achievement-Related Beliefs . . . . . . . . . . . . . . . . . . . . . . . . . . 88 

Gender Differences in Attributions in Maths 

and English . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 

Gender Differences in Perceived Competence . . . . . . . . . . . . . . . 90 

Proposed Links Between Observational and 

Attributional Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 

Suggestions for Future Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 

Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 

REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 



APPENDICES ............................................. 106 

1 Questionnaire . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 

2 Observational Schedule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 

3 Information Letter to Parents and Caregivers ............... 112 

4 Verbal Instructions to Students ......................... 113 

5 Consent Form ..................................... 114 

6 Correlation Matrix for Combined Variables . . . . . . . . . . . . . . . . 115 



TABLE 

Table 1 

Table 2 

Table 3 

Table 4 

Table 5 

Table 6 

Table 7 

Table 8 

LIST OF TABLES 

page 

Subjects Taken by all Secondary Pupils at 

1 July, 1990 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 

University Degrees Completed by Level of Degree 

and Main Subject for the Year Ending with the 

Graduation Ceremony in 1990 . . . . . . . . . . . . . . . . . . . . . . . . . 5 

Occupation (Major Group) by Sex for Population 

Resident in New Zealand Aged 15 Years and Over 

Gainfully Employed in Full-time and Part-time 

Labour Force . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 

Percentages of Negative and Positive Feedback 

Given to Males and Females ........................... 57 

Percentages of Intellectual, Nonintellectual, 

and Conduct by Gender .............................. 58 

Percentages of Negative, Positive, and Absent 

Feedback Given for Conduct, Intellectual and 

Nonintellectual Activities by Gender . . . . . . . . . . . . . . . . . . . . . 60 

Percentages of Specific vs General Negative 

and Positive Feedback by Gender ....................... 62 

Percentages of Negative and Positive Feedback 

Received by Subject and Gender ........................ 64 



Table 9 

Table 10 

Table 11 

Table 12 

Table 13 

Table 14 

Table 15 

Table 16 

Table 17 

Table 18 

Percentage of Total Feedback Contingent Upon 

Intellectual and Nonintellectual Work Aspects, 

and Conduct by Subject and Gender . . . . . . . . . . . . . . . . . . . . . 65 

Percentages of Positive Feedback Contingent on 

Intellectual and Nonintellectual Work Aspects, 

and Conduct by Subject and Gender . . . . . . . . . . . . . . . . . . . . . 66 

Percentages of Negative Feedback Contingent on 

Intellectual and Nonintellectual Work Aspects, 

and Conduct by Subject and Gender . . . . . . . . . . . . . . . . . . . . 67 

Percentages of Specific vs General Feedback Given 

to Students by Subject and Gender ...................... 68 

Mean Scores for Attributional Dimensions 

by Gender ....................................... 72 

Students Defined as Learned Helpless or 

Mastery-oriented by Gender ........................... 73 

Pearson Correlations Between Helpless and Mastery 

Attributions, and Achievement-Related Beliefs .............. 74 

Pearson Correlations Between Perceived Competence 

and Achievement-Related Beliefs ....................... 75 

Mean scores for the Attributional Dimensions for 

Success (s) and Failure (f) in Maths and English ............. 78 

Mean Scores for Achievement-Related Beliefs 

by Gender and Subject . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 



CHAPTER 1 

INTRODUCTION 

GENDER DIFFERENCES IN EDUCATIONAL PARTICIPATION AND 

ACHIEVEMENT 

1 

For whatever reasons, few mathematicians are women. A plethora of research exists 

which documents discrepancies between males' and females' mathematical 

achievement. Although there are few gender differences in mathematical ability at 

the primary school level (Tittle, 1986), by secondary school, males are frequently 

outperforming females on maths achievement tests, especially those which involve 

problem solving (Gold, 1990; Hyde, Fennema, & Lammon, 1990; Fennema, Peterson, 

Carpenter, & Lubinski, 1990; Linn & Petersen, 1985). Moreover, research continues 

to show that females opt to take fewer advanced Maths courses than males (Elmore 

& Vasu, 1986), exhibit lower expectations for success and lower estimates of their 

competence in Maths (Fennema, 1985), have a more negative personal belief system 

pertaining to Maths, and a greater attrition for females than males in entry to 

mathematically-related careers (Fennema et al., 1990). 

Many assert that gender differences in cognitive abilities are accountable for this 

disparity, as it is argued that females perform better than males in verbal areas, and 

males have a superior mathematical ability (Dweck & Licht, 1980). However recent 

analyses show that previous cognitive gender differences in verbal ability, spatial 

visualisation, and mathematical computation and concepts have declined, and no 

longer exist (Jacklin, 1989; Linn & Hyde, 1989). Thus, it appears that any under 

representation of women in mathematical areas is much larger than any found in 

other cognitive skills. 

Initially, the majority of psychological research which sought to explain the 

educational, occupational, and social status of women concentrated mainly on the 



study of individual differences, and the search for biological explanations of such 

differences (Tittle, 1986). Recently however, the emphasis has changed, with the 

research focus being not on the different biological features of the two sexes, but 

rather, the social interactions during which individuals construct gender related 

perceptions and responses (Jacklin, 1989). 

Of particular relevance to the present study is the quantity of educational research 

afforded to exploring the ways in which the school experience may facilitate the 

realisation of gender related differences in achievement, and further contribute to the 

sexual division of labour in adulthood. Sex differences have primarily been 

attributed to gender related differences in experiences, including education, and 

recent research contributes to a better understanding of the characteristics of these 

experiences (Tittle, 1986). 

In order to conceptualise and illustrate the issues fundamental to the present study, 

recent New Zealand statistics in this area will be presented. Table 1 displays the 

subjects taken by males and females at secondary school. It is clear to see that 

females prefer to study language, humanity, domestic and arts subjects, whereas 

males tend to concentrate more in the "hard science" and technical areas. 

Interestingly, the numbers of males and females studying English differs only 

slightly. However, more males take both Maths, and Maths with statistics. This 

pattern of subject differentiation greatly limits females' choices for further education, 

and many career opportunities (Loveridge, 1986). 

Table 2 indicates that the subject specialisation observed in university, mirrors the 

distribution of males and females in the same areas at secondary level. Females are 

consistently obtaining more degrees in humanities and languages than males. Not 

surprisingly, university mathematics and computer graduates are predominantly male, 

with the numbers of Doctorates and Masters awarded to males far exceeding those 

obtained by females. 

2 



Table 3 presents the major occupations of New Zealanders, over 25 years old. It is 

evident from this table that females are still entering jobs which have been 

stereotypically defined as appropriate for females (clerical, sales and service 

workers), and males dominate the field in legislation and administration, trade 

workers, machine operators and assemblers, and agriculture and fishery workers. It 

is interesting to observe that the number of professional females exceeds the number 

of professional males. This is probably a function of this category including Health 

Professionals (e.g., nurses, midwives, dentists, doctors etc.) and Teaching 

Professionals (e.g., early childhood educators; primary, secondary, and tertiary 

teachers; special needs teachers). 

Data presented in these tables demonstrates that there are clearly discrepancies 

between males' and females' educational participation and achievement in the New 

Zealand contemporary education system. Although the data suggests gender 

differences in many subject areas, the focus of the present study is on the biases 

observed in Maths (traditionally a masculine domain) and English (traditionally a 

feminine domain), as literature documents that unequal participation and achievement 

is greatest in areas which are typically perceived to be sex-typed. 

3 



4 

Table 1 

Subjects Taken by All Secondary Pupils At 1 July 1990. 

Subject Males Females Total 

English 112 510 112 865 225 375 
Language 8 874 10 596 19 470 
French 9 854 19 110 28 964 
German 2 856 6 152 9 008 
Japanese 5 017 7 425 12 442 
Chinese 0 2 2 
Spanish 65 203 268 
Russian 91 86 177 
Indonesian 68 82 150 
Pacific Languages 87 105 192 
Latin 1 856 1 372 3 228 
Classical Studies 1 486 3 102 4 588 
Language Studies 487 328 815 
Maori Studies 4 048 3 430 7 478 
Cultural Studies 19 6 25 
Social Studies 53 162 51135 104 297 
History 10 813 13 580 24 393 
Geography 20 790 19 433 40 223 
Mathematics 104 067 97 325 201392 
Statistics 6 783 5 329 12 112 
Science 74 802 70 440 145 242 
Earth Science 33 5 38 
Biological Science 10130 15 569 25 699 
Environmental Studies 82 124 206 
Human Biology 944 2 058 3 002 
Chemistry 9 010 6 983 15 993 
Physics 13 889 4 900 18 789 
Physical Science 994 438 1 432 
Art (Core) 22 977 21 667 44 644 
Art (Special) 11764 13 467 25 231 
Art History 1 176 2 955 4131 
Art Practical 3 609 4 507 8 116 
Music (Core) 25 199 24 980 50 179 
Music (Special) 4 311 5 441 9 752 
Drama 4 348 5 815 10 163 
Physical Education 87 440 82 527 169 967 
Technical Drawing 38 674 11 330 50 004 
Workshop Craft 30 528 7 311 37 839 
Engineering Shopwork 7 045 678 7 723 

Source: 1991 Educational Statistics of New Zealand. 



5 

Table 2 

University Degrees Completed by Level of Degree and Main Subject for the Year 
Ending with the Graduation Ceremony in 1990. 

Subject Doctorate Masters Bachelors Totals 
M F M F M F M F 

Humanities 
Arts 1 11 22 12 22 
Ancient History 3 3 3 3 
Asian Languages 4 16 4 16 
Chinese 1 2 4 2 5 
Classics 1 5 19 5 20 
English 3 5 12 27 118 346 133 378 
French 1 1 4 10 43 11 48 
German 2 1 11 35 13 36 
Greek 1 1 3 1 4 
History (Economic) 1 19 8 20 8 
History 6 1 11 11 153 173 170 185 
Indonesian 1 2 1 2 
Italian 2 1 7 1 9 
Japanese 12 19 12 19 
Latin 1 1 1 1 2 
Linguistics 1 1 7 10 8 11 
Maori 5 8 5 8 
Philosophy 1 8 2 40 15 49 17 
Russian 1 4 9 5 9 
Spanish 3 1 3 1 6 

Sub-total 11 8 39 54 407 746 457 808 
Mathematics & Computer Science 
Applied Maths 19 6 19 5 
Computer Science 3 12 2 172 25 187 27 
Management Science 3 1 3 1 
Computing Tech 2 5 2 7 2 
Information Systems 27 15 27 15 
Management Science 20 6 20 6 
Mathematics 2 1 6 1 100 57 108 59 
Operations Research 6 2 22 12 28 14 
Pure Mathematics 27 19 27 19 
Quantitative Studies 1 1 0 
Statistics 2 1 6 7 8 8 

Sub-total 5 1 28 6 402 149 435 156 

Source: 1991 Educational Statistics of New Zealand 



Table 3 

Occupation (Major Group) by Sex for Population Resident in New Zealand Aged 
15 Years and Over Gainfully Employed in Full-time and Part-time Labour Force. 

Occupation Males Females 

Legislators, Administrators 
and Managers 109 893 52 395 
Professionals 77 805 89 760 
Technicians 88 326 62 382 
Clerks 36 075 164 736 
Service and 
Sales Workers (1) 66 690 111 744 
Agriculture and 
Fishery Workers 97 191 40 170 
Trades Workers 140 637 8 790 
Machine Operators 
and Assemblers 101 262 30 444 
Elementary 
Occupations(2) 65 373 34 401 
Not Adequately 
Defined 11 820 10 509 

Source: 1991 Census, National Summary. 

6 



APPROACHES TO DIFFERENTIAL GENDER PARTICIPATION IN 

EDUCATION AND EMPLOYMENT 

Many researchers speculate possible and probable explanations for these inequities. 

Three principal theoretical approaches have emerged to explain the under­

representation of women in formal education, and professional and executive 

positions (Fiorentine, 1988). The first is the structural barriers approach which 

contends that inequities in males' and females' achievement in these positions is a 

result of sex-discrimination that limits females opportunity. That is, because the 

female gender may be considered a "discrepant status" women may encounter 

7 

barriers to their mobility. Available research suggests that regardless of profession, 

women are not expected to perform as well as men (Frieze, Fisher, Hanusa, McHugh, 

& Valee, 1978). Thus these prejudicial attitudes can affect the initial hiring and 

training of women, as well as undermining trust and certainty in their competence 

(Kanter, 1977). 

The second approach is the normative barriers approach. This approach exemplifies 

the attitude that females fear success, and are anxious in achievement situations 

because of their anticipation of negative consequences in the form of social rejection 

or loss of femininity (Homer, 1972). Correspondingly, Linn and Hyde (1989) 

contend that gender differences in career access are a function of specific cultural 

and situational contexts. 

The third approach, the cognitive differences approach, is of most relevance to this 

thesis. This approach proposes that gender differences in educational and 

occupational participation are a consequence of gender differences in achievement­

related beliefs. That is, females have lower perceptions of competence and lower 

performance expectations than males (Eccles, Adler, & Meece, 1984; Parsons, 

Meece, Adler, & Kaczala, 1982), which results in a lack of confidence in their ability 

to perform successfully in a variety of achievement situations (Fiorentine, 1988). 



Closely allied with this approach is attribution theory, which refers to perceptions 

and inferences about the causes of one's own behaviour, and that of others. 

According to this approach, females lack prerequisite confidence in their abilities 

Consequently, successes are attributed to "external" or "unstable causes such as luck, 

and failures to "internal" or "stable" causes such as lack of ability (Weiner, Frieze, 

Kukla, Reed, Rest, & Rosenbaum, 1971). This attributional pattern, which discounts 

success while affirming failures, results in females choosing not to enter into and 

persist with, or perform well in, a wide range of achievement tasks (Fiorentine, 

1988). 

The purpose of the present study is to examine the nature of gender differences in 

achievement within the context of the cognitive approach, drawing significantly on 

the contribution of attribution theory. Specifically, this research will focus on the 

type of task (i.e., the perceived sex appropriateness of a task) as a determinant of 

gender differences in attribution and expectation. In search of an explanation for 

these differences, an observational study was also conducted. 

This thesis begins with a comprehensive account of the attribution theory and 

Weiner's three dimensional taxonomy for explaining success and failure. The 

concept of learned-helplessness will also be addressed in this section. The 

behavioural consequences of students' attributions will be detailed, and it will be 

illustrated how disparate responses to success and failure are associated with very 

different constellations of achievement cognitions. How these cognitions mediate 

student's expectations for future outcomes of behaviour, and subsequent achievement 

strivings will also be discussed. The next section will examine the available 

literature on cognitive factors that inhibit achievement in females, focusing in 

particular on the gender differences in causal attributions, and future expectancies. 

The subsequent section will then examine one explanation of why success and failure 

may have a more negative implication for females than males. Teacher-student 

interactions, and the role they play in sustaining gender related behaviours will be 

briefly examined. Predominantly, literature which details how feedback given to a 

8 



student by their teacher can acquire different meanings depending on the student's 

sex, and consequently result in different attributions and expectations, will be 

explored extensively. 

9 

The influence that the perceived sex appropriateness of a task has on attributions and 

subsequent cognitions will be outlined in the next section; and literature, which 

suggests that females' self-derogating attributional biases are more prominent in 

subject areas in which males are believed to be more competent than females, 

reviewed. This will also include some speculation on how differential patterns of 

feedback received by students in Maths and English may affect causal reasoning. 

Finally, the relationship between perceived ability, expectancies for future 

performance, and attributions will be addressed. 



CHAPTER2 

REVIEW OF LITERATURE 

ATTRIBUTION THEORY 

The general format of a cognitive model of motivation is stimulus--> cognition--> 

response. That is, an incoming stimulus, viewed as a source of information is 

encoded into a belief system that gives it "meaning". The subsequent response is 

then guided by the intervening structure of thought (Baldwin, 1969, cited in Weiner 

et al., 1971). 

In social psychology there exist many formulations of attribution (e.g., Jones & 

Davis, 1965; Kelly 1967; Nisbett & Ross, 1980; Reeder & Brewer, 1979), all of 

which emphasize "that people interpret behaviour in terms of its causes and that 

these interpretations play an important role in determining reactions to that 

behaviour" (Kelly & Michela, 1980, p. 458). However, it is the contribution of 

attribution theory in the educational setting which is of primary relevance to the 

present study. Thus, this thesis will encompass predominately those 

conceptualisations of theorists who consider attribution in the achievement context, 

specifically concentrating on the formulations of Weiner and his colleagues. 

10 

The attribution model of achievement motivation presented in this thesis is guided by 

the general cognitive approach outlined above. The model is based on the assumption 

that people generate explanations for their achievement outcomes, and that these 

beliefs about the causes of success and failure mediate between antecedent stimulus 

organism transactions and ensuing achievement behaviour. This approach emphasizes 

individuals' interpretation of events, not the events themselves, and is greatly 

influenced by the early propositions of Heider (1958). 



11 

Of particular interest to this thesis are the formulations proposed by Weiner (1984), 

who has provided much of the impetus for the development and application of the 

attribution theory to education. Based on the work of Heider (and others), Weiner 

(1984) proposed a cognitive dimension to earlier achievement motivation theories, 

suggesting that people's cognitive reactions to success and failure are important in 

understanding achievement-oriented behaviour. Weiner's attribution theory is based 

on the assumption that affective and cognitive reactions to, and thus behavioural 

consequences of, achievement outcomes depend on the perceived reasons for these 

outcomes. Implicit is the assumption that individuals naturally search for 

understanding about why events occur, especially when the outcome is important or 

unexpected (Stipek, 1988). These perceptions of the causes of achievement outcomes 

are referred to as causal attributions. 

The early postulations of Weiner and his colleagues (1971) included only two causal 

dimensions: locus of control and stability; subsequently in 1979 Weiner proposed a 

third dimension to the model, calling it controllability, and later renaming it as 

responsibility, which encompasses both controllability and intentionality. More 

recently, Weiner has relabelled the dimension of stability as constancy, although the 

majority of attributional research still prefers to address these dimensions 

controllability and stability. The model of primary reference in this thesis will 

encompass Weiner's most recent propositions. 

Weiner ( 1984) subsumes the causes of success and failure within a three dimensional 

taxonomy. The first of these is the internality dimension, and is associated with the 

field of locus of control (Rotter, 1976, cited in Lefcourt, 1976). From the point of 

view of the individual making the attribution, any causal factor can be described as 

residing within the person (an internal factor), or outside the person (an external 

factor). Ability, effort, mood and patience for example, are properties internal to an 

individual, whereas task difficulty, luck, and teacher bias are external or environ­

mental causes. It should be remembered that it is always the attributor's judgement 

that determines the location of a particular cause on the dimension. 



12 

A second dimension of causality, labelled stability, characterises cause on a stable 

(invariant) versus instable (variant) continuum. A stable cause is one that is per­

ceived as being relatively unchanging over time. For example, ability, difficulty of a 

task and patience; whereas luck, effort, and mood are more unstable. Luck implies 

random variability, effort may be augmented or diminished from one episode to the 

next, and mood is usually conceived as a temporary state (Frieze & Snyder, 1980). 

The third dimension discussed by Weiner (1984) is controllability. Some causes are 

seen as being within an individual's control, while others are not. For example, 

some causal factors such as effort, are likely to be perceived as controllable, whereas 

ability, mood, or task difficulty are uncontrollable causes. 

A fourth dimension, incorporated more recently into Weiner's theory, is globality 

(Weiner, 1986). A global or general cause (compared with a specific one) is seen to 

have wide, varying implications, whereas a specific cause is relevant to only a 

limited number of actions (Abramson, Seligman, & Teasdale, 1978). Thus, it is 

argued that global attributions imply that when individuals confront new situations, 

their responses will again have little relevance to the outcome. Hence, a student may 

attribute his or her failure to a specific lack of ability in a particular school subject 

(specific), or to a more general lack of ability (global). The ramifications of the 

latter attribution being that deficits in performance are likely to follow in other 

situations, whereas a specific attribution to lack of ability (eg., in Maths), implies 

that debilitation is likely to be restricted to Maths situations. 

Weiner et al. (1971) originally postulated that in attempting to explain a prior 

outcome of an achievement-related event, individuals utilize four main causal 

elements - ability, effort, task difficulty, and luck. That is, the individual assesses his 

or her ability level, the amount of effort that was expended, the difficulty of the task, 

and the magnitude and direction of experienced luck. More recently, a wider range 

of causal factors has been established, including fatigue, personality, mood, interest, 

illness and bias, as well as causes that are unique to specific situations (Elig & 

Frieze, 1979; Ross & Fletcher, 1985; Weiner, 1984). 



13 

However, the specific causal attributions are less important as determinants of 

behaviour than the underlying dimensions of the attributions, particularly when trying 

to predict future achievement-related behaviour (Thomas, 1989). Important to this 

discussion are students' conceptualisations of ability. Dweck (1986) maintains that 

there are two distinct notions of what constitutes intelligence. One is the opinion 

that intelligence is a fixed entity, that ability is a stable trait, unaffected by personal 

effort. If one lacks the prerequisite ability, no amount of effort will lead to success. 

The other view is that intelligence is an instrumental-incremental concept, which 

develops over time through instrumental behaviour, such as practice and study 

(Stipek, 1988; Thomas, 1989). 

The assumption that ability is a stable trait, and that attributing failure to ability will 

seriously and deleteriously effect expectations for future success and willingness to 

approach new achievement situations has traditionally been most popular. However, 

recent research suggests that ability is not always perceived as a stable capacity 

(Stipek, 1991). Therefore, depending upon students' notions of what constitutes 

intelligence or ability, different and contrasting behavioural consequences will be 

implicated. That is, if ability is perceived to be a variant trait, failure attributed to 

ability may not result in lower expectations or a desire to avoid achievement situ­

ations in future. 

Similarly, each of the other basic causes could be interpreted in terms opposite to 

Weiner's conceptualisations (Ross & Fletcher, 1985). For example, effort could be 

considered a stable personality trait (a lazy or industrious person), and task difficulty 

could be viewed as either a stable or unstable cause, as at different times tasks may 

be thought to be easier or harder (due to fatigue, hunger, temperature, etc). Due to 

this equivocation in interpretation of attributions, the emphasis in the present study is 

not on the attribution itself, but the underlying dimensions of the attribution. 

However, for illustration purposes, the traditional interpretation of the four basic 

causal dimensions will be employed in subsequent discussion. 



14 

CAUSAL CONSEQUENCES 

It is the behavioural consequences of students' attributions which make them relevant 

to the discussion of motivation in the classroom. Although Weiner and his col­

leagues also examine the effect of different attributions on students' emotional 

reactions in achievement situations, to avoid complexity, the affective consequences 

of attributions were not addressed in the present study. 

Much research shows that the types of attributions a person makes, and the way they 

will behave in achievement situations are closely related to each other. The original 

theory proposed that causal attribution patterns are related in systematic ways to 

expectations for future performance, to subsequent achievement strivings, and to the 

affect associated with achievement outcomes (Dweck & Licht, 1980; Forsyth & 

McMillian, 1981; Howard, 1984; see Harvey & Weary, 1984; Weiner, 1986). 

Logically, within this attributional paradigm certain implications for various causal 

attributions are apparent. When success is perceived as being the result of possess­

ing ability (an internal, stable factor), maximum security can be derived. Effort 

attributions for success offer less security than since continued effort must be exerted 

to maintain positive outcomes, (Frieze & Snyder, 1980). However this attribution is 

still favourable, since effort is controllable and changeable. Attributing success to 

external factors, e.g., good luck or an easy task, offers even less security that success 

will recur. 

An opposite pattern of consequences follows failure attributions. Specifically, 

attributing failure to a lack of ability implies that there is no way the failure could 

have been avoided, and that it can't be changed to a success in future (except for 

occasional instances of good luck). In failure situations however, luck attributions 

are favourable, as future success can be anticipated as luck changes. Similarly, effort 

attributions for failure are desirable as, unlike ability, this cause is changeable and 

controllable. 



15 

Thus, it is not hard to see why effort attributions are generally considered the most 

productive for learning. If a student attributes failure to low effort, it would still 

leave hope for success in the future (assuming he or she was prepared to try harder). 

When failure is attributed to lack of ability, it is a lot less likely that the student will 

exert more effort on future tasks, as without the prerequisite ability, no amount of 

effort would lead to success in the future (Stipek, 1988). 

Similarly, attributing success to effort is desirable. This perception implies that 

although the student realises he or she has the required ability, they would not have 

succeeded without exerting some effort. However, it has been suggested that 

attributing success solely to ability is not favourable. Students who do this perceive 

effort as being unnecessary for success, and because they do not try their best, often 

they will perform at a level below their capacity. Hence, it appears that attributing 

success to both ability and effort are most beneficial. Indeed, Nicholls (1976) 

showed that students would prefer to be seen to both have ability, and to have tried 

hard when successful, concluding that students want to be perceived as able, but 

virtuous too. 

LEARNED HELPLESSNESS 

When failure is attributed to stable factors that the individual is unable to control 

(eg., a lack of ability), maladaptive behaviour coined "learned helplessness" can 

develop. Learned helplessness refers to a particular phenomenon by which an 

individual learns, over a series of trials, that she/he has no control over the outcome 

of events. One sees no relation between effort and changes in surroundings, or the 

attainment of a goal (Thomas, 1979). Failure is perceived as insurmountable, and is 

followed by a deterioration in performance. 

The term learned helplessness was originally advanced by Seligman and Maier in 

1967 (Abramson, et al., 1978), to describe the debilitated escape-avoidance response 

shown by dogs exposed to uncontrollable shocks in the laboratory. Seligman found 



16 

that dogs who first were subjected to shocks from which they cannot escape, behaved 

quite differently when exposed to avoidable shock, compared to dogs who had not 

previously experienced inescapable shock. They appear to lose the ability and 

motivation to respond to aversive stimulation and simply "give up", acquiring what 

might be referred to as a "sense of helplessness" . 

Just as organisms can learn contingencies, so, too, can they learn about the absence 

of contingencies (Abramson, et al., 1978). It is proposed also that this pattern of 

maladaptive responding can be generalised to similar situations, thus significantly 

reducing the probability that the same animal will attempt to make effective 

responses in the future, and therefore recognize the presence of a contingency when 

one does actually exist (Abramson et al., 1978). 

These early findings with dogs have subsequently been extended to a variety of other 

species. More importantly, analogous experiments with human beings have yielded 

similar results. In human helplessness studies, people who have been subjected to 

inescapable noise, or inescapable shock, or who have been confronted with unsolv­

able problems, fail later to escape noise and shock which is escapable, or solve 

simple problems (e.g., Hiroto, 1974;· Thornton & Jacobs, 1971). 

LEARNED HELPLESSNESS IN THE CLASSROOM 

Helplessness Versus Mastery-Orientation: The parallels between this general 

description of learned helplessness, and the behaviour of students who give up when 

they confront failure in achievement situations are remarkable. Dweck and her col­

leagues (e.g., Diener & Dweck, 1978; Dweck, 1986; Dweck & Reppucci, 1973) have 

studied extensively the achievement behaviour of "learned helpless" and "mastery­

oriented" students. Essentially, those students who attribute failures to internal, 

stable, and uncontrollable causes, such as ability, rather than effort, are considered to 

be helpless. In contrast, those students who attribute both success and failures to 



effort are referred to as mastery-oriented (Hokoda, Fincham & Diener, 1990; 

Thomas, 1989). 

17 

The studies of Dweck and her associates, evidence that learned-helpless students tend 

not to try very hard in any achievement-related tasks, show low task persistence, low 

goal-setting, low expectations for future performance, and low overall academic 

achievement. These students believe that there is nothing they can do to escape 

failure, and that irrespective of any response, failure is inevitable. This perception of 

failure as insurmountable is associated with attributions of failure to stable, but 

uncontrollable factors, predominately, a lack of ability (Dweck, 1986; Thomas, 

1989). 

In contrast, mastery-oriented students show increased persistence or improved 

performance after failure, and tend to attribute their failure to variable, controllable, 

internal factors such as lack of effort, or variable situational factors such as task 

difficulty. The cognitions of these mastery-oriented students imply that their 

successes are replicable, and their mistakes rectifiable (Dweck, 1986; Eccles et al., 

1984; Hokoda et al., 1990;). 

GENDER DIFFERENCES IN LEARNED HELPLESSNESS 

Much educational research has been devoted to examining possible gender differ­

ences in causal attributions, with the results frequently denoting that females appear 

to be more helpless in failure situations than males. (Bar-Tal & Frieze, 1977; Deaux 

& Farris, 1977; Diener & Dweck, 1978; Dweck & Bush, 1978; Dweck & Repucci, 

1973; Frieze & Snyder, 1980; Howard, 1984; Licht & Dweck, 1984; Simon & 

Feather, 1973; Sweeney, Moreland, & Gruber, 1982). 

That is, females, compared with males make causal attributions which could inhibit 

learning (Deaux, 1984; Dweck & Licht 1980; Fox & Ferri, 1992; Frieze, Whitely, 

Hanusa, & McHugh, 1982; Howard, 1984; Nicholls, 1980). They are more likely 



18 

than males to attribute success to external, unstable, factors such as luck and task 

ease, and are less likely to attribute success to their own abilities (Fox & Ferri, 

1992). Thus, females do not experience the enhanced self-esteem, and security of 

future success commonly afforded by males in similar achievement situations (Stipek, 

1984). 

Moreover, females tend to take responsibility for their failures, attributing them to 

internal, stable, factors such as a lack of ability (Bar-tal & Frieze, 1977; Fox & Ferri, 

1992) neglecting to use effort attributions after failure (Deaux & Farris, 1977). 

These attributions are accompanied by differential expectancies for success in males 

and females, with females tending to have lower initial expectancies for success than 

males (Deaux, 1984; Frieze & Snyder, 1980). Deaux (1984) proposed that it may 

not be the process of making attributions that distinguishes males and females, but 

rather their initial expectancies for success that cause the differences. 

According to her model, females initially have low achievement expectations. 

Hence, they attribute an unexpected, successful outcome to an external or unstable 

cause such as luck, since a success is not consistent with their low expectations. 

Since luck is ever changing, there is no reason to expect future success, self-esteem 

is not increased, and future expectancies for success remain low. Alternatively, 

failure, which is consistent with expectancies, serves only to confirm females' low 

expectations, consequently increasing expectancies for subsequent failure. The 

implication of this low expectation cycle is that females may be condemning them­

selves to failure through a self-fulfilling prophecy (Erkut, 1983). 

New Zealand data on students' responses to success and failure are limited, but 

generally confirm the pattern of causal attributions, and subsequent achievement­

related behaviour observed in other studies. For example, Nicholls (1980) reported 

that among students of high socio-economic status (SES), males were more likely to 

choose a difficult task rather than an easier one. Additionally, females were more 

inclined to ascribe their performance on these tasks to luck than were males. 



19 

Another of Nicholl's studies (1984) involved lower SES students. Results indicated 

that although both males and females performed at a similar level, females were less 

likely to employ ability attributions for success, and more likely to attribute this 

success to luck. Males were more likely to attribute failure to a lack of effort or an 

uninteresting task. 

Nicholls (1978) also performed two studies focusing only on reading. The first 

indicated that despite higher perceived and actual achievement, females still 

attributed failure to a lack of ability. The second study however did not confirm 

these results, finding instead that females were more inclined to attribute success to 

ability, and less likely than males to blame a lack of ability for failure, but attribute 

it instead to bad luck. Males were still more likely than females to ascribe their 

failure in reading to a lack of effort. 

Lawes (1983) studied gender differences in New Zealand students' attributional 

constellations. Specifically, Lawes investigated causal attributions for performance in 

the School Certificate English examination. Her results indicated that males and 

females did not differ in the explanations they gave for success, but only in failure 

attributions. Males were more likely to attribute failure to unstable, external causes, 

whereas females utilised stable, internal causes. 

Loveridge (1986) also sought to explore gender differences for causal attributions in 

New Zealand students. Diverging away from the methodology common to most 

attributional research, Loveridge measured students' attributions and expectations 

through an extensive interview, which allowed for the subjective construction of a 

situation by the student. This method also allowed students to locate causes on an 

interval scale for each proposed dimension, rather than have the researcher assign the 

causes to the dimensions, which is common in most other attributional research. 

The results of this investigation were only partially conclusive, with males attributing 

success in science to "involvement and interest" more often than females, but not to 

"ability and skill" as predicted. The hypothesis that females would be more likely to 



20 

attribute success in reading to "involvement and interest", and "ability and skill" was 

unsupported. 

AN ALTERNATIVE EXPLANATION 

Although gender differences in achievement-related attributions have been observed 

by many researchers, results are far from conclusive. Despite the abundance of 

published studies that report that a subject's sex is a determinant of behaviour and 

attitudes, including attributions, many researchers remain sceptical, challenging 

established findings. 

For example, Sweeny, et al. (1982) found that in failure situations males tended to 

employ internal attributions, while females blamed external causes. Dweck & 

Repucci (1973), and Parsons et al. (1982), also failed to find gender differences in 

ability attributions for failure. Further, it is unknown how many of the unpublished 

studies have found no significant differences in causal attributions (Megarry, 1984). 

Frieze et al. (1982) performed a meta-analysis of 21 studies which examined gender 

differences in attributions for success and failure. They concluded that in actuality, 

there was not substantial support for any of the theoretical models proposed to 

account for such differences. The only consistent differences their analyses yielded, 

was the tendency of females, compared with males to make greater luck attributions 

for both success and failure. 

Similarly, Sohn (1982) conducted an effect-size analysis on studies pertaining to 

students' achievement attributions. Again, it was concluded that the empirical 

evidence supports only the proposition that females utilise luck attributions more 

frequently than males when explaining a success. No consequential relationships 

were found between sex and the other three main types of achievement self-attribu­

tions. Thus, it may be that proposed gender differences in cognitions following 

success and failure are all in the mind. 



21 

METHODOLOGY 

The procedure used in attributional research has frequently been criticised, with many 

investigators asserting that the way in which students' causal attributions are obtained 

may be directly responsible for the equivocal findings pertaining to gender differ­

ences in attributions (McHugh, Frieze, & Hanusa, 1982; Parsons et al., 1982), 

concealing true differences. 

The plurality of studies investigating the nature of gender differences in causal 

reasoning in educational settings have employed structured questionnaires that were 

baseq on rating scales (Elig & Frieze, 1979; Fox & Ferri, 1992). As structured 

questionnaires have had much widespread appeal, there has been little exploration 

into other possible techniques which may also be potentially useful in obtaining attri­

butions. However, the structured questionnaire has been criticised due to little 

attention being given to the function and meaning of the behaviour for the people 

concerned (Maehr & Nicholls, 1980). Frieze et al. (1982), emphasized that students' 

own judgements of an achievement situation need to be integrated into the achieve­

ment process. In other words, "it is, fina,lly, the subjective construction of the 

situation by the individual tha.t is of critical importance" (Maehr, 1983, p.90). 

Rarely, both structured and open-ended questionnaires have been used (e.g., Parsons 

et al., 1982). 

Hence, Forsyth and MacMillan (1981) recommend that instead of the researcher 

imposing causal dimensions upon students' attributions as they interpret them, it may 

be more appropriate for students to locate causes on an interval scale for each pro­

posed dimension themselves. Further, McHugh et al. (1982) advocate that if attri­

butional research is to advance, various situational determinants of sex differences in 

attributions need to be considered. The majority of studies of sex differences 

measure attributions about outcomes in experimentally contrived tasks or by requiring 

participants to make causal attributions for a series of hypothetical outcomes (Stipek, 

1984). An incredible 87% of the studies reported in the review by Sohn (1982), 

employed such experimentally-contrived achievement tasks. 



22 

Although it is believed that this may be the most effective way to measure the 

generalized attributional tendencies or attributional styles of the student (McHugh et 

al., 1982), this approach is frequently criticised as these tasks can have only a limited 

relevance to the subjects, and possibly do not correspond to the attributions they 

would give for a more salient event (Frieze, Sales & Smith, 1991). 

THE EFFECT OF FEEDBACK ON ATTRIBUTIONS 

Another factor which has been extensively explored in an attempt to conceptualize 

the nature of gender differences in attributions, is the effect of different types of 

feedback on casual reasoning. It is hypothesized that females learn to blame them­

selves for their failures because both teachers and other adults see them as lacking 

competence, and somehow communicate this to them. Yet, research does not support 

this assumption. It seems that females are on average more successful in achieve­

ment situations than males, and consistently get higher grades. Females are also 

rated more favourably on behavioural and personality characteristics, and female 

teachers see the behaviour of females as being closer to their conception of the ideal 

student than males' (Goebes & Shore, 1975). 

Further, females receive far less disapproval, both "official" (e.g., detention) and 

informal from both male and female teachers, even though females and males did not 

differ on any observed behaviours (Lee, 1973; Meyer & Thompson, 1956). Thus, in 

general, it seems that females are treated and regarded more favourably in many 

areas. Why is it then that despite females consistently achieving higher than males, 

and receiving less disapproval from teacher, that they frequently display learned help­

lessness? 

For a long time it was assumed that it is males' and females' general socialisation 

histories that determine their response to failure feedback (Dweck, Davidson, Nelson 

& Enna, 1978). That is, it was believed that males, due to greater independence 



23 

training develop their own standards of excellence, so external evaluation means little 

to them. 

Alternatively, females are thought not to develop these independent standards in the 

way males do, but remain dependent on the feedback of others to gain information 

on the adequacy of their performance (Dweck et al., 1978). Females may be more 

oriented to teacher approval, and self-reinforce for effort, neatness, and good 

behaviour (Nicholls, 1980). Hence, when a male receives negative feedback, he can 

accept it or reject it depending on whether it corresponds with his own standards. 

The same feedback can have different meaning to a female, who will probably infer 

that this is a true indication of her performance and abilities (Dweck & Goetz, 1978). 

Recently, it has been suggested that it is not generalised response tendencies that 

determine differential responses by females and males to failure feedback, but instead 

specific histories with particular agents, namely adults, as they are the major 

evaluators in all academic environments (Dweck & Goetz, 1978). Implicit in this 

assumption is that failure feedback must somehow acquire different meanings to 

males and females. Dweck and her colleagues (1978) sought to conceptualize the 

nature of gender differences in responses to failure feedback by analysing the pattern 

of feedback given to students by their teachers. The focus of their study was on 

teachers' use of negative feedback, its frequency relative to positive feedback, its 

typical referents, and the specificity of its use. 

Results of this study have helped resolve the puzzle. Although females seemed to 

receive more positive and less negative evaluation than males overall, this was only 

with respect to nonintellectual aspects of their work (neatness, instruction following, 

speaking clearly). In terms of absolute amounts of feedback for the intellectual 

quality of work (competence or correctness) there appeared to be no overall tendency 

for one sex to receive a greater amount of either positive or negative feedback. 

What was relevant however, was the way in which negative feedback was used, and 

the type of attributions which followed it. Overall, males received a greater amount 



24 

of negative feedback. Moreover, negative feedback was more ambiguous in its 

referents for males, with nearly half relating to form and presentation aspects, rather 

than content. This compared with about 90 percent of the of the positive feedback 

relating directly to their intellectual competence. 

In contrast, less than 80 percent of the positive feedback received by females 

addressed the intellectual quality of their work, with about 20 percent relating to 

presentation aspects. Conversely, the majority of negative feedback they received 

was directly concerned with intellectual aspects of their work. Hence, it is obvious 

that there are distinct gender differences in the degree to which positive and negative 

feedback are valid indicants of their intellectual ability as demonstrated by their 

academic performance. Warren and Cairns (1972), similarly advocate that when 

evaluative feedback is used indiscriminately, and for a wide range of nonintellectual 

behaviour, it loses its meaning as an evaluation of the quality of the student's 

performance, and criticism can easily be attributed to some characteristic or attitude 

of the agent, rather than to the adequacy of performance. 

Another striking difference was evident when the explicit attributions made by 

teachers for students' successes and failures were examined. When males failed, 

teachers generally attributed it to lack of effort, a failure to follow instructions, or the 

messiness of the work (i.e., motivational factors). In fact, teachers attributed males' 

failures to lack of motivation eight times more often than they did females'. 

Females' failures, on the other hand, were attributed to errors in the work itself. 

This criticism was often also accompanied by praise for motivation and effort. 

In_ conclusion, Dweck's research reveals that the differential pattern of praise and 

criticism given to males and females may lead them to interpret evaluation in 

contrasting ways. As a result of these feedback patterns, when a male receives fail­

ure feedback, he was able to conclude that the evaluator simply didn't like him, or 

that it was nonintellectual aspects of their work which were at fault. Even when 

males did conclude that their work was intellectually inadequate, they were able to 

blame motivational factors. 



25 

In contrast, the negative feedback given to females was highly specific with regard to 

intellectual aspects of their work. Moreover, little of the negative feedback given to 

females was for conduct or nonintellectual aspects of their work, although males 

were often criticised for these reasons. Females are accustomed to receiving praise, 

and know that teachers generally view them as highly conscientious and hard 

working in their academic work. Thus, they are less likely to interpret criticism as a 

result of a teacher who simply doesn't like them, and because insufficient motivation 

is a less salient factor in determining females' failures, they have little choice but to 

accept that intellectual failures can only reflect an objective evaluation, and a lack of 

true ability. 

This deduction lends itself to the belief that they are incapable of future academic 

success, thus leaving them with a sense of helplessness. Yet males who experience 

this same failure are accustomed to criticism in achievement-related contexts, can 

more easily disregard failure feedback, and can continue to perceive themselves as 

capable (Dweck & Licht, 1980). 

Interested in the implications of this research, Dweck and her colleagues continued 

their investigations, this time seeking to determine whether these observed patterns of 

feedback serve as causes of sex differences in the interpretation of failure feedback. 

They sought to understand how the use of negative feedback determined its meaning 

for males and females. In order to do this, they applied the different contingencies 

of work-related feedback observed in the previous study, under experimental condi­

tions. From this, they were able to evaluate the impact of this criticism on students' 

subsequent performance in an achievement situation (where all subjects experienced 

failure), and their attributions for this failure. 

The results of this study confirmed that any student (male or female) exposed to the 

contingencies of feedback that females commonly receive ("teacher-girl" condition), 

where all negative feedback referred to the correctness of their answers, will gen­

erally interpret failure feedback as indicating a lack of ability. Regardless of the 

students's sex, only 25% of students in the "teacher-girl" condition used effort as an 



explanation for their failure. In comparison, more than 75% of the students in the 

"teacher-boy" condition (those who received criticism common to males in the 

classroom, referring sometimes to correctness, but other times to neatness), did not 

ascribe their failures to a lack of ability, opting instead to use effort attributions for 

their failure. 

26 

In other words, those students who received a mixture of solution-relevant and 

solution-irrelevant feedback, were much more likely to see failure as a reflection of 

their effort or of the negative attitude of the evaluator, than of a lack of ability. 

Hence, it is evident that the pattern of criticism students receive is very influential on 

their interpretation of their failures. 

A further interesting aspect of the differential influences of failure feedback on males 

and females emerged from another study by Dweck and Bush (1976). Results of this 

investigation indicated that failure feedback can acquire different meanings and 

impact for the two sexes, depending upon the sex and age of the evaluator. That is, 

when feedback was given to females by an adult evaluator, in particular a female 

adult, they evidenced greater helplessness than did males. They were more likely to 

blame their lack of ability for their failures, and their subsequent performance was 

typically impaired. With a male adult evaluator however, there was a greater 

tendency to attribute failures to insufficient effort, which led to a greater improve­

ment in performance. 

In comparison, males tended to attribute their failures to a lack of effort with an 

adult evaluator of either sex, and continue to persevere in the face of failure. These 

results provide further evidence that it is not the differing socialisation histories of 

males and females that influence their responses to failure, but rather their differing 

evaluative interactions with particular agents, suggesting again that what should be 

analyzed is the pattern of evaluative feedback that males and females receive from 

different evaluators. 



27 

Neale (1978) performed an observational study very similar to that of Dweck et al. 

(1978) in New Zealand classrooms. In particular, Neale was interested in studying 

the ways male and female teachers gave positive and negative feedback to male and 

female students. 

She observed that both male and female teachers were involved in significantly more 

evaluative interactions with males, regardless of whether the feedback was work­

related, or to do with the student's conduct She found little difference in the way in 

which teachers used positive feedback with students of either sex, with the majority 

of it referring to intellectual competence. However, there were differences in the 

way aegative feedback was used. While criticism generally referred to the intellec­

tual qualities of females' work, males on the other hand received their negative 

feedback for both intellectual and nonintellectual aspects of their work, and for 

conduct. It appears therefore that the findings from this study greatly mirror those of 

Dweck and her colleagues, and that the same implications for students' subsequent 

failure attributions are relevant. 

Neale found also that although female teachers gave more feedback to males, they 

still treated males and females in a relatively similar way. What differed was the 

way in which male teachers gave feedback. They tended to give significantly less 

evaluative feedback to females, and have overall a lot fewer interactions with them. 

When they did criticise females, it usually related directly to their intellectual 

competence. 

ATTRIBUTIONS AND THE SEX-APPROPRIATENESS OF THE TASK: 

Thus, although females' attributional biases have been depicted in many studies, 

there is some inconsistency in the evidence. This has led to many researchers now 

modifying their research question to "under what conditions do females evidence 

debilitating attributional biases?" (Stipek, 1984). The focus of much attributional 

research now considers that differences in attributions about performance outcomes, 



28 

and expectancies for future success and failure may vary not only between males and 

females, but also from task to task, most specifically, across school subjects (Maehr, 

1980). 

Licht and Dweck (1984) hypothesized that certain academic areas are more likely 

than others to pose difficulties at the start of new units. Thus, it follows that the 

necessity of surmounting difficulties will probably be easier for students of certain 

achievement orientations, namely mastery-oriented students. They proposed that 

Maths is a subject area where failures are common, as one is constantly confronted 

with new concepts and operations. In comparison, failures are less likely in verbal 

areas (such as English), as once the basic concepts are mastered, acquisition of new 

material is more piecemeal, with the difficulty level increasing gradually. 

Consistent with their supposition, they found that Maths is more likely than verbal 

areas to involve failure and confusion when new material is introduced, and that 

"mastery-oriented", rather than "helpless" students learned more effectively in these 

situations. 

As females are more likely to display a helpless achievement orientation than males, 

this seems a plausible explanation for why males often out-perform females in 

Maths. Additionally, a further implication of this work is that females will show less 

confidence in their Maths abilities than males, while no such differences will exist in 

their verbal abilities. 

Masculine versus Feminine Tasks 

It may also be that school subjects differ in their perceived sex-appropriateness, re­

sulting in differences in perceived ability. For the purpose of this thesis, a "mascu­

line" task is defined as one on which males perform better on average than females, 

with the opposite being true on a perceived "feminine" task. Research has often 

shown that females have superior verbal skills in comparison to males (Maccoby & 

Jacklin, 1974; Ryckman & Peckman, 1987), and that males perform better in Maths 

achievement tasks (Fennema & Sherman, 1977), and that these actual differences in 



29 

performance are related to students' beliefs· that various cognitive tasks favour one 

sex or the other (Mc Mahan, 197 6). It has been hypothesized that this has lead to 

sex-role stereotyping of achievement domains, with Maths typically being seen as a 

masculine domain and English a more feminine domain (Dweck & Licht, 1980). 

Weiner et al. (1971) began investigations into attributions following performance on 

sex-consistent and inconsistent tasks several years ago. Their initial studies showed 

that an expected outcome is readily attributed to the performer's ability (by an 

observer), while an outcome which is unexpected is more likely to be ascribed to 

chance or luck. Deaux and Emswiller (1974) concluded from this research that 

possibly an analogous argument can be made for the evaluation of performance by 

males and females in which expectations are sex-linked. That is, males and females 

will have low or high expectancies depending upon whether the task is considered 

predominately masculine or feminine. They hypothesized that when a task was sex­

consistent, performance would be attributed to internal factors, such as ability. In 

comparison, chance would be a more common explanation for performance on a sex­

inconsistent task. 

Results of this study demonstrated that this was indeed the case. When the per­

former was male, and the task "masculine", performance was attributed to skill. 

Whereas, on the same task if the performer was now female, luck attributions were 

more frequently employed. However, the reverse was not true on feminine tasks. 

Contrary to what was predicted, under these conditions, both males and females were 

rated very similarly. Overall, independent of the task, males were considered more 

skilful, while females were seen as being luckier. Many other more recent studies 

have reached the same conclusion - that males make more egotistical attributions 

(i.e., internal attributions for success, and external attributions for failure) than 

females. 

Stein, Pohly, and Mueller (1971) also carried out studies of this nature, concentrating 

on students' own expectations for success and failure. Results of their investigation 

indicated that when students were told that a task measured typically masculine or 



feminine traits, high expectancies of success were predominant on sex-appropriate 

tasks, and low expectancies prevailed on sex-inappropriate tasks. 

30 

Rosenfield's (1976) studies yielded corresponding results. He argued that the 

different degrees of ego-involvement that males and females may have in the tasks 

could account for gender differences in males and females attributional patterns. If 

a task appears to be characteristically "masculine", it would be expected that males 

would be more ego-involved than females. Consequently, males would be more 

likely to take responsibility for their successes but blame their failures on elements 

out of their control. However, on a "feminine" task (i.e., ones which involve 

feminine traits) the opposite would be expected, with males attributing success less 

internally and failure less externally than females. 

Results of this study found that again, males did make more egotistical attributions 

than females when the task was described as masculine, and on the feminine task 

females made more egotistical attributions than males. Thus, it is shown that females 

do not always attribute their successes less internally, or their failures more external­

ly than males. Additionally, it appears that males and females do not vary in how 

egotistical they are, with attributions being dependent on a student's ego-involve­

ment. 

Parsons et al. (1982) also investigated the hypothesis that females were more likely 

exhibit self-derogating attributions in Maths than males. This study revealed no 

overall gender differences in students' self-concepts of Maths ability. Additionally, 

no significant differences were evident in males' and females' attributional patterns 

for success or failure. The only significant difference was that females, compared to 

males rated their expectations for future success slightly lower. Hence according to 

this investigation it would appear that females are not more learned helpless on 

Maths than males. 

These trends have not been replicated across recent studies however. For example, 

Stipek (1984), found that females were more likely to exhibit helpless attributions on 



31 

tasks in which males are believed to be more competent (Maths). However, on tasks 

in which females are thought to be more competent (spelling), males did not evi­

dence these same attributional biases, despite males actually achieving at a lower 

level than females. 

Similarly, Gitelson, Petersen and Tobin-Richards (1982) examined students' expect­

ancies of success, evaluations of performance, and attributions for success and failure 

for verbal and spatial tasks. Results showed that irrespective of the task, females 

expected to perform less well than males, although no sex differences in performance 

were found. For the spatial task, females attributed less ability to themselves than 

did males, and rated the task as more difficult. Females also evaluated their perform­

ance more negatively on this task. No such differences were found on the verbal 

task. Again, these results demonstrate that helpless achievement expectancies, 

evaluations and attributions on sex-inappropriate tasks exist for females but not for 

males. 

A more recent study by Stipek and Gralinski (1991), which measured females' and 

males' achievement related beliefs in Maths, found that females rated their ability 

lower than did males, and also expected to perform less well on a Maths exam than 

did males. When females failed they were more likely to attribute it to a lack of 

ability, and were less likely to ascribe their success to high ability than males. 

Moreover, they employed more luck attributions when explaining their failures. 

Consequently, females also expressed a greater desire to avoid future Maths achieve­

ment situations, and held lower expectations for future performance in Maths. When 

asked about whether they believed success was achievable by hard work, females, 

less frequently than males, claimed that anyone could do well if they tried hard 

enough. 



32 

ATTRIBUTIONS AND PERCEIVED COMPETENCE 

One assumption of the attribution theory is that students' achievement behaviours are 

mediated by perceptions of ability. Ability perceptions have been shown as an 

important factor in determining achievement behaviour (e.g., Chapman, 1988; Eccles 

et al., 1984; Thomas, 1989). Students who believe that they are competent will be 

more likely to choose more difficult tasks, to choose achievement-related tasks, and 

to evidence more effort and persistence when faced with a difficult task, or lack of 

success. Alternatively, students with low perceptions of their abilities will expect 

failure and attribute this to low ability. They will choose tasks at which failure is 

highly likely, and performance is likely to be impaired (Stipek, 1988; Eccles et al., 

1984). The literature indicates that people who believe they are competent are more 

confident in approaching achievement situations, have higher achievement-related 

expectancies, and consequently perform better on these tasks (Nicholls, 1980; 

Thomas, 1989). 

That is, students who perceive themselves to be competent at a task are more likely 

to attribute success to their abilities, and failure to some other cause. Attributing 

success to internal factors, such as ability, is likely to result in favourable affective 

reactions, which serve to reinforce the person's perception of him/herself as compet­

ent, and also increase expectations of future success. The competent person who 

fails is likely to ascribe his/her failure to external factors such as luck or task ease, 

since it is an unexpected outcome and has a low expectancy for recurrence (Nicholls, 

1980; Stipek, 1988; Weiner, 1979). 

In contrast, students who do not believe they are competent will probably attribute 

failure to this lack of ability, and will explain success using other external factors. 

The student who has a low self-concept of ability tends to expect failure, so can 

readily attribute it to an internal factor. When success occurs unexpectedly, it will be 

attributed to external causes such as good luck and teacher assistance (Covington & 

Omelich, 1979; Erkut, 1983). Moreover, implicit in ascribing success to external 

factors, is low expectancies for future success. Failure, however, confirms the 



student's belief that he/she is incompetent and lacks the prerequisite ability to 

achieve. Consequently continued failure is expected and achievement-related 

situations are avoided in future (Nicholls, 1980). 

33 

That males and females have different perceptions of their competence has been well 

documented already, with males generally perceiving themselves as more able than 

females, and having higher estimates of their performance and expectations for future 

success in achievement situations, despite there being little, if any, difference in 

actual performance (eg Deaux, 1976; Dweck & Goetz, 1978; Frieze et al., 1978; 

Parsons et al., 1982). 

Research performed with New Zealand students has yielded a similar constellation of 

attributions and subsequent achievement-related behaviours. For example, in a study 

with middle-class students, Nicholls (1976, cited in Nicholls, 1980) found that again, 

despite females performing better than males on a reading task, and perceiving 

themselves as more competent, females attributed failure to lack of ability more fre­

quently than did males. This led to the conclusion that even when females perform 

as well or better than males, and perceive themselves as doing so, they are still less 

likely to employ attributions that makes for continuing achievement behaviour. 

Another New Zealand investigation (Hunt, 1989) also found that females have lower 

perceived ability than males in Maths, but they had a slightly higher mean score on 

English perceived ability. 



34 

CHAPTER3 

THE PRESENT STUDY 

RECAPITULATION AND HYPOTHESES 

Gender differences in cognitive performance have been documented extensively, with 

males consistently performing at a higher level on measures of spatial ability, and 

females on measures of verbal ability. There has been much research into potential 

mechanisms which could account for these differences. Of particular irJerest in the 

present study are several of the social factors which have been hypothesized to 

mediate gender related differences in intellectual performance, including expectancies 

of success, perceived competence, the sex-appropriateness of the task and attributions 

made about success and failure. 

Much available literature reports that females show greater evidence of learned 

helplessness than males in achievement situations. The present investigation seeks to 

explain gender differences in attributions and expectations as a function of the way in 

which teachers interact with their students. Focusing on the nature of evaluative 

feedback given by teachers to their students after academic responses, the links 

between teacher behaviour and students' cognitive responses to success and failure 

will be explored. It is hypothesized that males and females are being treated 

differently in the classroom, which in some way affects their cognitions in achieve­

ment situations, and subsequent achievement behaviour. 

The present investigation will consist of two major parts - the first being an observa­

tional study (loosely following the format of Dweck et al., 1978), and the second 

being in the form of a questionnaire. 



35 

PART ONE - OBSERVATIONAL STUDY 

Although this type of study has been carried out already in New 2.ealand, only the 

sex of the evaluator has been manipulated, and not the subject being studied. 

Considering the evidence from studies quoted on differential effects of male and 

female teachers, it would appear logical to include sex of the evaluator as a predictor 

of students' attributions. However, while the present study acknowledges the import­

ance of this variable, practical constraints meant that all teachers observed in this 

study, regardless of the subject being taught, were male. Male teachers, rather than 

female teachers, were chosen as they predominated in the teaching of Maths and 

English classes. By keeping this variable constant, it cannot be argued that any 

significant results are a function of the evaluator's sex. 

Dweck and Licht (1980) suggest that subject areas may differ in the degree to which 

persistence after failure is necessary for success. Given that females respond in a 

more debilitating way following failure, it is likely that they would perform most 

poorly in those subject areas where they are left with little choice but to view failure 

feedback as a condemnation of their ability, such as in subjects which emphasize 

intellectual competence and correctness, and where little evaluation is given for the 

nonintellectual quality of their work. 

In contrast, this same pattern of feedback might acquire a different meaning for 

males, given their feedback history with teachers. To them, increased criticism 

would not necessarily imply a lack of ability, but instead be viewed as reflecting the 

teachers' negative attitude towards them. Therefore subsequent performance should 

not be adversely affected in the same ways that of females is. It is more likely that 

males would evidence debilitated performance in subjects where nonintellectual 

aspects of work are emphasized (Dweck & Licht, 1980). 

In Maths, in contrast to English, there is a stronger emphasis on intellectual work 

aspects, errors are likely to be more frequent and highly salient, and the criteria for 

success in Maths is more objective than in verbal areas (Dweck & Licht, 1980). 



36 

Thus the majority of evaluation given to students would pertain to intellectually 

relevant aspects of their work, suggesting that criticism given to females is even 

more likely than usual to be interpreted as a lack of competence. In English, while 

intellectual qualities are still important, a student is able to compensate for perceived 

intellectual inadequacies by improving some other quality of their work (e.g., 

spelling, grammar, imaginative ideas), as the criteria for success is more ambiguous. 

Given that females are shown to be more debilitated in Maths than English, it is the 

prediction of the present study that evaluative feedback in these subjects may differ 

in several qualitative and quantitative ways, which may subsequently seriously and 

adversely inhibit females' learning in a traditionally "masculine" domain. 

GENDER DIFFERENCES IN FEEDBACK 

QUESTION 

Are there gender differences in the nature and frequency of feedback received by 

students? 

HYPOTHESES 

1.1 That males and females will not differ in the amount of overall feedback they 

receive. 

1.2 That males, compared with females, will receive more negative feedback. 

1.3 That females, compared with males, will receive more positive feedback. 

1.4 That males, compared with females, will receive a greater amount of feedback 

for nonintellectual aspects of work, and conduct. 



1.5 That males and females will not differ in the total amount of feedback they 

receive for intellectual aspects of their work. 

37 

1.6 That males, compared with females, will receive more positive feedback for the 

intellectual qualities of their performance (i.e., females will receive more positive 

feedback for intellectually irrelevant work aspects). 

1. 7 That negative feedback given to females will be primarily for intellectual 

inadequacies in their performance. 

1.8 That negative feedback given to males will refer little to the intellectual quality 

of their work, but instead refer to conduct, and the nonintellectual aspects of their 

work as well. 

1.9 That positive and negative feedback given to males is more diffuse and 

ambiguous in its referents than with females, who will receive more specific 

feedback. 

GENDER DIFFERENCES IN FEEDBACK IN MATHS AND ENGLISH 

QUESTION 

Are there gender differences in the nature and type of feedback received by students 

in Maths and English? 

HYPOTHESES 

1.11 That males and females will not differ in the amount of overall feedback they 

receive in Maths and English. 



1.12 That males, compared with females, will receive more negative feedback in 

Maths. 

1.13 That females, compared with males, will receive more positive feedback in 

Maths. 

1.14 That males and females will not differ in the total amount of feedback they 

receive for intellectual aspects of their work in Maths and English. 

38 

1.15 That females, compared with males, will receive less feedback for nonintellec­

tual aspects of their work, and conduct in Maths. 

1.16 That females, compared with males, will receive less positive feedback for the 

intellectual qualities of their performance, in Maths than English (i.e., females will 

receive more positive feedback for intellectually irrelevant work aspects). 

1.17 That negative feedback given to females will refer predominantly to the 

intellectual quality of their work, and little to conduct or the nonintellectual aspects 

of their work in Maths. 

1.18 That negative feedback given to females, compared with males, is more 

specific in its referents, in Maths than English. 

1.19 That positive feedback given to females, compared with males, is more diffuse 

and ambiguous in its referents, in Maths than English. 

The first part of this study investigated the patterns of evaluative feedback given to 

males and females in the Maths and English classrooms. The purpose of the second 

part of the study is to attempt to explain gender differences in achievement by 

examining gender differences in students' attributions and expectations in these 

subjects. Subsequent discussion will then include speculation into potential links 



between teachers' appraisals following academic responses, and students' cognitive 

and behavioural responses to success and failure. 

PART TWO - ATTRIBUTION QUESTIONNAIRE 

39 

Much educational literature documents the dramatic effects failure can have on 

performance. It has been extensively illustrated that there are marked differences 

among students in the way they respond to failure in achievement situations. For 

some students, these effects are positive and lead to increased effort and intensified 

concentration. Persistence is increased, and problem-solving strategies become more 

sophisticated, resulting in enhanced overall performance. For other students, however 

failure has a different meaning, and many of the above consequences are reversed. 

For these students, failure leads to a reduction in effort, the deterioration of 

strategies, and performance is severely disrupted. Moreover, these students often are 

unable to complete tasks they had previously been able to accomplish with ease 

(Dweck & Licht, 1980). 

Although the behaviour of these two groups preceding failure is very similar, once 

failure occurs, vastly different consequences become apparent. Much research shows 

that these two groups start out with virtually identical performance - equivalent 

speed, accuracy, and sophistication of problem solving strategies, as well as similar 

intelligence levels. It is the proposition of this thesis that what distinguishes these 

two groups, are their cognitions following success and failure. 

There is considerable evidence which indicates that students' perceptions of the cause 

of achievement-related events have important implications for their subsequent 

affective responses, expectations for success and failure, self-esteem, achievement­

related behaviour, and essentially, their performance in ensuing achievement 

situations (Dweck & Goetz, 1978; Weiner, 1979). 

Much research reports gender differences in achievement related attributions and 



40 

behaviours. Females have often been observed to be more helpless than males in 

achievement situations, even on tasks at which they have clearly demonstrated their 

ability, or in areas in which they have clearly out performed males. This often 

results in females predicting lower grades for themselves, or poorer performance on 

novel tasks, and avoid tasks that pose a challenge. 

It is evident that New Zealand data on gender differences in attributions for success 

and failure is sparse. Moreover, the studies that are available yield equivocal, and 

often contradictory results. It has been suggested that inconsistent findings across 

studies may be a function of the instrumentation used to collect attributional data. 

Most studies have employed structured questionnaires, based on rating scales, and 

measured attributions by presenting subjects with experimenter-defined achievement 

tasks. 

The design of the investigation reported in this thesis attempted to take these factors 

into consideration. Attributions were measured in a natural achievement setting, the 

classroom. The questionnaire used to obtain students' causal reasons for success and 

failure required students to locate causes on an interval scale for each of the three 

dimensions (stability, controllability, and locus) themselves, rather than have the 

researcher assign the causes to the dimensions. Also, to avoid forcing students to 

attribute outcomes to factors which may not be perceived as the most salient, the 

attribution question required subjects to generate their own reasons for being success­

ful or failing. 

It has been shown also that students' causal attributions for achievement outcomes 

are reliable predictors of their responses to success and failure and subsequent 

achievement behaviour. However, it is apparent that this relationship is mediated by 

several factors. It has been suggested that the perceived sex-linkage of the task is a 

crucial mediator of students' expectations and attributions, with both males and 

females expectancies for success being greatest on sex-appropriate tasks. 

Results of sex-typed studies demonstrate that females perceive Maths as an "un­

feminine" choice, and at a time when they are self-conscious about making gender­

inappropriate choices, they decide not to continue with Maths (Megarry, 1984 ). 



41 

Following from these studies, the present investigation assumed that both males and 

females would stereotype Maths as predominantly a "masculine" subject, and view 

English as more appropriate for females to study. It appears that males and females 

have differential academic attitudes and values, resulting from different intellectual 

demands, instructional activities and socialisation experiences in Maths compared to 

verbal areas (Jacklin, 1989). 

It was hypothesized that there would be gender differences in students' achievement 

behaviour and cognitions, following success and failure in Maths and English, not 

only because these subjects have typically been sex-typed, but also due to 

characteristics of the subjects which differentially appeal to males and females. Licht 

and Dweck (1984) suggest that these subject areas differ in the degree to which one 

must persist after failure to achieve success, with mathematical areas possessing 

qualities least suited to females, and verbal areas having qualities which render them 

least attractive to males. 

Gender differences in perceived ability, as a moderating variable, have also received 

considerable attention. Research suggests that regardless of actual performance, 

males, in comparison to females, believe they are more able. Thus, the present 

research advances the view that considering the general belief in greater male 

competence, if a task is perceived as masculine, there will be significant gender 

differences in expectations and attributions. However, on a feminine task, males will 

have lower perceptions of competence, and expectancies of success, than on a 

masculine task. These lowered expectancies, when combined with the more general 

belief in male competence, are predicted to produce little or no overall gender 

differences in expectations and attributions on feminine tasks. 

Another factor which is hypothesized to contribute to the gender differences in Maths 

is the role of effort in success. There is some evidence in learned helplessness 

literature which suggests that males, in comparison to females, place more 



42 

importance on the role of effort, and other motivational factors, for success (Stipek, 

1991). As females generally believe themselves to be harder working than males 

(Dweck, Goetz & Strauss, 1980), they may believe that success in Maths is less 

obtainable through effort than males do. Thus it is probable that females will try to 

avoid Maths achievement situations more than males, and they will exert less effort 

on Maths related activities, as increased effort is not expected to result in success. 

Little attributional research has included these factors when discussing gender 

differences in causal reasoning (Stipek, 1991), hence, the present study will assess 

males and females beliefs that success can be achieved through effort, and their 

desire to avoid achievement situations, in both Maths and English. 

GENDER DIFFERENCES IN ATTRIBUTIONS 

QUESTIONS 

Do males and females use different causal attributions for success and failure? 

HYPOTHESES 

2.1 Females are predicted to attribute success to external, unstable, uncontrollable 

causes, and failure to internal, stable, uncontrollable causes. Males will evidence the 

opposite attributional pattern, attributing success to internal, stable, and controllable 

causes, and failure to external, unstable and controllable causes. 



GENDER DIFFERENCES IN MASTERY AND LEARNED HELPLESSNESS 

QUESTIONS 

Do females develop learned helplessness more readily than males? 

Are learned helpless and mastery attributional patterns related to perceptions of 

competence, avoidance behaviours, and expectations for future performance? 

HYPOTHESES 

2.2 Females, in comparison with males, are predicted to make learned helpless 

attributions for success and failure. 

43 

2.3 Mastery-oriented attributions will be related to high perceptions of competence, 

high expectations for future success, and a low desire to avoid future achievement­

related situations. Learned helpless attributions will be related to low perceptions of 

competence, low expectations for future success, and a high desire to avoid future 

achievement-related situations. 

GENDER DIFFERENCES IN PERCEIVED COMPETENCE 

QUESTIONS 

Do males perceive themselves to be more competent than females? 

Are perceptions of competence related to expectations and attributions? 

HYPOTHESES 

2.4 Males, in comparison to females, will perceive themselves as more competent. 



44 

2.5 Regardless of a student's sex, high perceptions of competence will be related to 

high expectations for future performance, the perception that the task is relatively 

easy, low avoidance desires, expending a lot of effort on the task, and the belief that 

success is obtainable through effort Low perceptions of competence will yield the 

opposite pattern of achievement-related beliefs. 

SEX-CONSISTENT VERSUS SEX-INCONSISTENT TASKS 

QUESTIONS 

Do students perceive Maths as a "masculine" subject and English as a "feminine" 

subject? 

Do students' attributions for success and failure depend on the perceived sex-appro­

priateness of the task? 

Are there gender differences in perceived competence on a sex-consistent task versus 

a sex-consistent task? 

Are there gender differences in expectations for future performance on a sex­

consistent versus a sex-inconsistent task? 

Are there gender differences in desire to avoid a sex-inconsistent versus a sex­

inconsistent task? 

Are there gender differences in how hard students try on a sex-consistent versus a 

sex-inconsistent task? 

Are there gender differences in perceived task difficulty on a sex-consistent versus a 

sex-inconsistent task? 



45 

Are there gender differences in students' beliefs that success can be achieved through 

effort on a sex-consistent versus a sex-inconsistent task? 

HYPOTHESES 

2.6 Both males and females will perceive Maths as predominantly a "masculine" 

subject and English as a "feminine" subject 

2.7 Females, in comparison with males will make more helpless attributions in 

Maths. No such differences are expected in English. 

2.8 Females, in comparison with males are predicted to have lower perceptions of 

competence in Maths, and lower expectations for success. No such differences are 

expected in English. 

2.9 Females, in comparison with males are predicted to expect to do less well in 

future in Maths. No such differences are expected in English. 

2.10 Females, in comparison with males are predicted to express a stronger desire to 

avoid Maths performance situations in future. No such differences are expected in 

English. 

2.11 Females, in comparison with males are predicted to not try as hard in Maths. 

No such differences were expected in English. 

2.12 Females, in comparison with males are predicted to rate Maths as more 

difficult than English. 

2.13 Females, in comparison with males are predicted to believe that success can not 

readily be achieved through effort in Maths. No such differences are expected in 

English. 



46 

CHAPTER4 

METHOD 

OVERVIEW AND DESIGN 

The present investigation comprised two parts. The first assessed the nature and 

frequency of evaluative feedback received by males and females in the classroom, 

following a format similar to that used by Dweck et al. (1978). Four classrooms -

two English, and two Maths, all taught by male teachers, were observed for a total of 

five hours each. 

The second part of the study used a questionnaire to measure students' attributions 

for success and failure, their perceptions of competence, evaluations of the difficulty 

of the task, expectations for future performance, avoidance behaviour, and the 

perceived sex-appropriateness of the subject (Maths or English) in question. 

SUBJECTS 

Subjects comprised 97 Form Five students who took at least either Maths or English. 

Of this sample 50 were male and 47 female. Their ages ranged from 15 to 17 years 

old. 80.4% considered themselves to be European, 11.3% Maori or Polynesian, and 

3.1 % Other. 5 students did not indicate ethnicity. 12.3% of 1991 New Zeal