Massey u,,;VP.rsity lil--rary New Zealand & Pacific Collection DROP OUT FROM WANGANUI REGIONAL COMMUNITY POLYTECHNIC 1992-1993 A thesis presented in partial fulfilment of the requirements for the degree of Master of Educational Administration, Massey University Christine Elizabeth Coutts 1995 MASSEY UNIVERSITY LIBRARY THESIS COPYRIGHT FORM Title of thesis: Drop Out From Wanganui Regional Community Polytechnic 1992-1993. 1 . I give permission for my thesis to be made available to readers in Massey University Library under conditions determined by the Librarian. 2. I agree that my thesis, or a copy, may be sent to another institution under conditions determined by the Librarian. 3. I agree that my thesis may be copied for Library use. Signed The copyright of this thesis belongs to tlhe 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 II ABSTRACT In New Zealand recent changes In legislation have increased both the autonomy and accountability of the tertiary education sector. As a consequence, polytechnics have become painfully aware of the cost of student drop out from programmes both to their credibility as quality education providers and to their coffers. The case study of student drop out from Wanganui Regional Community Polytechnic (WRCP) was undertaken in response to administrators' concerns that, as a small, recently established polytechnic, costs associated with the loss of students might threaten its economic viability. WRCP administrators feared an increase in the rate of drop out as changes to the funding of tertiary study created increased financial pressures for students. The case study employed a range of techniques, including a survey of archived information and student questi0nnaires, augmented by the perceptions gained from informal d1scuss1ons with staff to ascertain the extent and nature of drop out. Analysis of data collected over a two year period (1992-1993) at WRCP revealed a pattern of drop out in terms of student characteristics. style and timing of withdrawal. Overall 11 showed that there was little to discriminate between the characteristics of leavers and those who persisted on a course. The two main theoretical conceptualisations or drop out. the 'integra11on' and 'investment' theories, were found to share the assumption that drop out 1s the consequence of 'cosVbeneht' analysis. However, despite increased fees and reduced allowances for many students. few cited financial reasons for withdrawal. An analysis of students' self-reported reasons for leaving indicated that finances are but one of many factors which affect a students determination of the costs and benefits of continued attendance. Many reasons given were outside the control of the Polytechnic. However, three Polytechnic academic schools were consistent in their reporting ol high drop out numbers, suggesting the need for further research into programme related reasons for withdrawal in these areas. These findings were consistent with recent models, which portray drop out as a complex process influenced by a multitude of factors, including student background and characteristics on entry, environmental changes and institullonal factors. From these findings some strategies to enhance the 'fit' of students and hence improve their retention are suggested. Continued monitoring and further research of a more phenomenological nature are recommended in order to gain a greater understanding of student drop out. LIST OF TABLES LIST OF FIGURES LIST OF APPENDICES GLOSSARY ACKNOWLEDGEMENTS INTRODUCTION TABLE OF CONTENTS CHAPTER 1 INTRODUCTION 1. 1 The New Zealand Context 1 .2 Theoretica l Interpretations of Drop Out 1.3 Case Study of Drop Out from Wanganui Regional Community Polytechnic 1 .4 Overview REVIEW OF THE LITERATURE CHAPTER 2 DROP OUT THEORIES Ill vi vii viii ix x , 2 3 4 2. 1 Drop Out: Relevant Research 7 2.2 Methodological Approaches to the Study of Drop Out 8 2.3 Problems in Defining the Population of Study: Terms Associated 9 with the Leaving Behaviour of Students 2.4 Drop Out Rates and Patterns: The Institutional Perspective 13 2.4.1 Drop Out and Graduation Rates 13 2.4.2 The Characteristics of Institutions as a Factor in Drop Out 14 2.4 3 Drop Out Within Institutions 17 2.4.4 Summary 22 2.5 Drop Out: The Student Perspective 22 2 5 1 Characteristics of Drop Out 22 2.5.2 Summary 24 2.6 There are Few Differences Between Leaving Students and Those 24 Who Stay: The Results of Comparative Studies 2.6. 1 Demographic and Socio-Economic Characteristics 24 2.6.2 Academic Factors 26 2.6.3 Psychological and Situational Variables in Drop Out 29 2.6.4 Wl1y Students Leave or Fall to Return to Study - A Search 32 for theCauseof Drop Out Through the Perceptions of Students and Significant Others 2.7 Theories to Account for the Complex Process of Drop Out 37 2.7.1 linto's (1975) 'Integration' Model 38 2.7.2 Commitment, Satisfaction and Reward in Student 42 Retention and Drop Out 2.7.3 A Multi-Variate Framework 48 2. 7 .4 Summary 50 2.8 Conclusion 53 CHAPTER 3 POLYTECHNIC EDUCATION IN NEW ZEALAND 3.1 Background 3.2 The Current Situation 3.2.1 Vocational Education 3.2.2 Community Education 3.3 Financial Support for Students 3.3. 1 Recent Changes to Fees, Allowances, and Student Financial Support Policies 3.4 Conclusion 59 64 64 67 69 70 73 IV METHODOLOGY CHAPTER4 METHODOLOGY 4.1 Research Aims and Questions 4.2 Method 4.2 1 Overall Approach 4.2.2 Selection of the Case and Some Associated Issues 4.2.3 The Boundaries of the Case 4.2.4 Types of Data and Data Gathering Techniques 4.2.5 Ethical Considerations 4.2.6 Issues of Reliability and Validity 4.3 Summary RESULTS 79 80 80 83 85 90 102 104 106 CHAPTER 5 THE INSTITUTION AND ITS ENVIRONMENT 5.1 The Institution: A Current Pic ture 109 5 1 1 Educational Provision at WRCP 110 5 1 2 Organisational Structure 111 51.3 Policies 112 5.2 Student Population 115 5 2 1 Part Time and Full Time Students 115 5 2.2 Pattern of Growth 116 5.2.3 Student Partic1pahon by Type of Programme 117 5 2 4 Students' Origins 118 5 2.5 Student Accommodation and Transpon 119 5 2 6 Demographic Charactenst1cs of Part1c1pants 119 5.3 Historical and Political Context 124 5.3 l Background 124 5.3.2 The Administrative Structure Over the Period of Study 125 5 3 3 Institutional Fundrng and Student Financial Support 125 5.4 Conclusion 126 CHAPTER 6 THE EXTENT AND NATURE OF DROP OUT AT WANGANUI REGIONAL COMMUNITY POLYTECHNIC 6.1 Extent of Drop Out 129 6 1 .1 Institutional Wrthdrawal Rates 130 6. 1 2 'Hidden Drop Out', Student Attrition and Revised 130 Drop Out Rates 6.1 3 Summary 131 6.2 Nature of Drop Out: Full Time and Part Time Withdrawal 131 6 2 1 Withdrawal by Nature of Attendance 131 6 2.2 Summary 132 6.3 Differences Between the Drop Out Rates of Different Types 133 of Programmes 6 3 1 Drop Out from 'Formal' Programmes 133 6 3 2 Programme Withdrawal Rates 1 34 6 3 3 Summary 134 6.4 Polytechnic Schools Have Different Drop Out Rates 134 6 4 1 Total Student Withdrawals by School 134 6.4 2 Full Time Withdrawals by School 136 6 4.3 Part Time Student Withdrawal by School 137 6.4 4 School Attrition Rates 137 6.4 .5 Summary 140 6.5 The Demographic Characteristics of Drop Outs 141 6.5.1 The Age Profile of Those Who Withdraw from WRCP 141 Programmes Overall and 'Formal' Programmes 6.5.2 Ethnic Differences Amongst Drop Outs from WRCP 143 Programmes and 'Formal' Programmes DISCUSSION v 6.5.3 Gender Patterns in Drop Out from WRCP Programmes 144 Overall and 'Formal' Programmes 6.5.4 Disability and Drop Out 145 6.5.5 Summary 145 6.6 Timing of Drop Out 146 6.6.1 The Annual Pattern of Student Withdrawal 146 6.6.2 Variance Over Time by Nature of Enrolment 146 6.6.3 The Perceptions of Service Sector Staff on 147 Withdrawal Patterns 6.6.4 School VarianceOver Time - Heads of Schools' Perspectives 148 6 .6.5 Summary 149 6.7 Reasons for Leaving 149 6.7.1 Student Withdrawal from New Zealand Polytechnics 149 6.7.2 Student Withdrawal from WRCP 151 6 .7.2.1 Reasons for Leaving the Programme 152 6. 7 .2.2 Destinations 159 6 .8 Summary 163 CHAPTER 7 DISCUSSION CONCLUSION 7.1 The Institution 165 7 1 1 The Extent of Drop Out and the S1gn1hcance of the 165 Institution's Drop Out Rate 7.1.1 .1 Charactenshcs of the Institution as Factors in 167 the Drop Out Rate 7.1 2 Drop Out Patterns W1th1n the lnstatullon 169 7 1 2.1 Oifferenllal Drop Out Rates Bet\'.een Programmes 169 7 1.2.2 Polytechnic Schools Have Different Drop Out Rates 171 7 .1.2.3 Timing of Withdrawal 173 7 1.2 .4 The Style of Dropping Out 175 7.2 The Student 177 7 2. 1 Student Demographic Characteristics 177 7.2.2 Drop Out by Nature of Enrolment Pan ume vs Full Time 182 7 2 3 Reasons for Leaving the Programme 185 7.3 Summary 194 CHAPTER 8 CONCLUSIONS, IMPLICATIONS ANO RECOMMENDATIONS 8.1 The Conclusion 8.2 Implication for WRCP 8.3 Recommendations for Future Research 8 .4 Summary REFERENCES APPENDICES 197 203 207 208 209 219 vi LIST OF TABLES TABLES Page Table 2.1 A Convergence or Ideas: Similarities Between Drop Out Models 51 Table 3.1 'Formal' Education Programmes: Number of Enrolments 66 Table 3.2 Number or Enrolments in Community Education 67 Table 5.1 Participation by Nature of Enrolment (full and part time) and Type of 115 Programmes at WRCP Table 5.2 A Comparison of MOE EFTS Increases {1991-1994) and Associated 117 Funding Grants Table 5.3 Number of Enrolments by Type of Programme at WRCP 118 Table 5.4 Comparison of the National, Wanganui, and the WRCP Student 121 Population Profiles by Age. Ethnicity and Gender. Table 6.1 Withdrawal by Nature of Enrolment and Type of Programme at 130 Wanganu1 Regional Community Polytechnic Table 6.2 'Formal' Programme Drop Out Comparison with Official 133 Wrthdrawal Rates Table 6.3 Student Withdrawal by School Size in MOE EFTS 136 Table 6.4 Drop Out by School · A Combination of Otf1cial and TOP Withdrawal 139 and Attrition Table 6.5 Comparison of Wanganui Regional Community Polytechnic Student 142 Population with the Population Profile of Formal Enrolees Table 6.6 Student Withdrawal from New Zealand Polytechnics: Factors 151 Identified as •very importanr in the Decision to Leave the Polytechnic Table 6.7 Planned Destinations of Students Completing Formal Programmes 162 of Study, as a Percentage of the Total Number of Students Enrolled, by Category VII LIST OF FIGURES FIGURES Page Figure 2.1 Kember's (1989) Adaptation of Tinto's (1975) 'Integration' 39 Model to Account for Drop Out. Figure 2.2 'CosVbenefit' Analysis (Roberts et al., 1991) 40 Figure 2.3 The Investment Model: Schell & Thornton {1985) 43 Figure 2.4 Organisational Determinants, Satisfaction, Institutional 44 Commitment and Drop out: Bean (1980). Figure 2.5 Selection and Socialisation. Bean (1985). 45 Figure 2.6 The Effects of Environmental Values. Bean and Metzner (1985) 47 Figure 2.7 A Mult1-vanate Framework. Powell, Conway and Ross (1990). 49 Figure 3.1 Tertiary Education Enrolment Trends 65 Figure 3.2 Equivalent Full nme Students (EFTS) at Tertiary lnst11u11ons 65 from 1982-1992. Figure 3.3 Tertiary Enrolment Trends (Numbers} from 1979 to 1991 74 Figure 4.1 March Tnal - Report to Management Team 7 April 1992 101 Figure 5.1 Types of Quahflca11ons Offered 1986-1995 110 Figure 5.2 Percentage Comparisons of MOE Funding and Student Growth 126 Patterns 1991-1994 Figure 6.1 Withdrawing Students as a Percentage of Total Enrolments 132 1992-1993 Figure 6 .2 School Withdrawals and School Size (1992-1993) 135 Figure 6 .3 Student Withdrawals by Attrition by School as a Percentage o1 138 Total Leaving Students {1992-1993) Figure 6.4 Withdrawals by Month as a Percentage of Total Withdrawals 146 January to December {1992-1 993) Figure 6 .5 Reasons for Withdrawal (1992-1993) 152 Figure 6.6 Destination of Withdrawing Students as Percentage of Respondents 160 1992-1993 viii LIST OF APPENDICES Student drop out rates: summary of selected Overseas and New Zealand studies 2 Students receiving student allowances by institution type al 31 July 1990 3 Enrolment Data: Problems, Solutions and Limitations 4 Programme Evaluation Report Form (a) 1992 (b) 1993 5 Chief Executive's Consent for Study (letter) 6 Confidentiality through informed consent (letter) 7 Aotearoa Polytechnics' Student Union Questionnaire (form) 8 Telephone Survey (report to Management Team) 9 WRCP Withdrawal Form and Questionnaire 10 Reasons for Leaving - Developing Categories for Encoding in the WRCP Questionnaire 11 Programmes Offered al WRCP (a) Certificated (b) Community 12 Wanganui Polytechnic's Students Enroled by Programme Type - 30 July 1992/1993 13 Characteristics of Polytechnic Student Population and Officially Withdrawing Students (a) 1992 (b) 1993 14 Demographic Characteristics of Participants and Withdrawers from 'Formal' Programmes 15 Refund Data by School and Month (a) 1992 (b) 1993 16 Programmes Exhibiting Drop Out Rates of 15% or More 17 Reasons Given by Withdrawing Students over the Year (a) 1992 (b) 1993 18 Wanganui Temperatures: Monthly Mean Temperatures 19 Reasons for Leaving the Programme by Gender and Nature of Enrolment 20 Reasons for Leaving the Programme by Ethnicity and Nature of Enrolment (a) 1992 (b) 1993 21 Reasons for Leaving the Programme by Age and Nature of Enrolment 22 Destination of Withdrawing Students by Gender and Nature of Enrolment 23 Reasons Cited by Drop Outs Who Gave Their Planned Destinations 24 Polytechnic Students Facing Hardship (Press clipping) GLOSSARY OF TERMS The following definitions of various types of student leaving behaviour have evolved a common currency in the literature and will be used in this study to mean: drop out refers to any student who leaves the programme of study for which they were enrolled in a given year. This term includes both those who formally indicate their intention to leave as well as those lost by way of attrition. attrition describes the process whereby an enrolled student ceases to attend classes and fails to respond to follow up procedures initiated by the institution. Such students may simply stop coming after quite regular attendance, or more often they 'peter out'. missing lectures and getting behind with assignments before finally disappearing altogether: very rarely do students announce that they are withdrawing and explain why (after Smith, 1987). withdrawal describes the process by which a student (the withdrawer) who had enrolled subsequently informs the rnstitution that s/he no longer wishes to continue studying failure describes the process by which the student who had enrolled 1s prevented from continurng studies by the rnsutution because or failure to sallsfy regulations, such as not meetrng the standards necessary for minimum rate of progression. (Also referred to as push out or exclusion - after Abbott-Chapman et al, 1992.) wastage refers to students who finally enrolled but who did not gain a course credit, ie who did not complete the course successfully, either through w1thdrawrng before takrng the assessments or by failing the assessment. Implicit 1n the use of this term is the frnancial focus of the rnslltut1on's admrnistrat1on. discontinuation is the process whereby students do not return to progress from one year to another in a multi-level programme stop outs are students who are taking a break from study (stopping out), but intending to return - this term differentiates between permanent decisions to withdraw trom a programme and temporary breaks in study (after Abbott-Chapman et al, 1992). persisters are students who attend the programme they have enrolled in over the entire year or for the entire duration in the case of shorter courses. The persistence or retention rate compares the number of students completing the year with the initial number enrolled. wavering persisters is a term used to describe persisters who indicate they had thought of withdrawing or changing programmes. though for some reason they do not make the 'break' (after West et al, 1987). transfer - movement to another institution to continue studies drop down is said to occur when students lessen their study load, reducing the total number of units they are enrolled for, or dropping from a full-time study commitment to part-time. ix ACKNOWLEDGEMENTS I wish to acknowledge the support and guidance of my supervisors. Dr Janet Burns and Or Wayne Edwards in the completion of this study. x I am indebted to the Wanganui Regional Community Polytechnic for the opportunity to undertake this research as part of my professional development. Support has been both by way of practical assistance and encouragement. A number of staff have given freely of their time and expertise to participate in the study and to provide assistance and advice in the development of this work. In particular, I would like to acknowledge the assistance so willingly given by Hinemoa Ransom, in production aspects of this work and for her personal support towards its completion. Beverley Alexander, Carolyn Cameron. Judith Fraser and last. but not least, Margaret Gillespie. have all given a hand with the typing of the document at various times. Euan Mcleod has assisted with technical advice and in the development of figures and tables. Thanks are due to the Management Team for their cooperation in this study and to Lynley Cv1tanovich, Averill Lockhart and Stephen Town in particular, !or their timely advice and support. I would also like to acknowledge the value of the critique provided by Grace Sylvester (Manukau Polytechnic) and Dr John Harre (Fiji lnstitule of Technology) . And lastly, I wish to thank my mother, Betty Carding who spent hours proof reading, and my husband, Graeme, !or his constant encouragement for me to reach lrus goal. CHAPTER 1: INTRODUCTION Student 'drop out' is familiar to most higher education institutions but it is only since the introduction of legislation which changed the autonomy and accountability of tertiary institut ions in New Zealand that the polytechnic sector has become aware of the cost of leaving students both to the twenty-five institutions and the individuals concerned. The term 'drop out'1 in this study refers both to those students who officially withdraw and to those who do not complete the units of learning for which they enroled within a given year, many of whom 'quietly fade away' (student loss by attrition) without informing the polytechnic. 1 .1 The New Zealand Context Students who drop out may represent a loss of revenue as well as a waste of scarce human. physical, administrative and teaching resources because, since the rntroduct1on of 'bulk funding', polytechnics have been subsidised by the State on the basis of the number of Equivalent Full Time Students (EFTS) enroled in particular cost categories. 'Bulk funding' of polytechnics (Ministry of Education, 1991, b, 11) was implemented in 1991, as a direct consequence of educational reforms introduced by the Education Act 1989 (EA) and the Education Amendment Act 1990 (EAA). The intention of the EAA was to ensure more effiicient resource use and to make the tertiary sector more accountable for public funding, providing a mechanism for tl1e control of state expenditure in this area (Chief Executives of British Institutions, 1993, 4). Under 'bulk funding', any shortfall in institutional operating costs has to be met from sources other than the State. Subsequent reductions in cost category funding levels, which effectively reduced institutional income, were compensated for by the introduction of higher course fees in many polytechnics. When considered in conjunction with changes to government financial support for students,2 introduced at about the same time. and the potential loss of earnings while studying, higher course fees created a situation in Which the considerable financial investment required by students may have influenced drop out rates from polytechnics. Thus, it seems, there are costs associated with student drop out, both to the individual and to the polytechnic, in terms of loss of income and wasted financial investments. 2 1.2 Theoretical Interpretations of Drop Out Although drop out is not a new phenomenon in higher education, little polytechnic specific research was evident in the literature. One reason for this is the relative 'youthfulness' of polytechnics in New Zealand, most having been established within the last forty years. Another important reason is that drop out has only recently been regarded as a 'problem' in the polytechnic sector, as economic pressures have forced the optimisation of finite resources. However, two major conceptualisations of drop out were noted in the extensive international body of literature surrounding drop out from other higher educational institutions. Despite their different origins, the 'integration' and 'investment' theories3 share the assumption that drop out is the consequence of a 'cost/benefit' analysis. According to 'integration' theories, such as that put forward by Tinto (1975), a lack of 'integration' into the academic and social systems of the institution leads to a tack of commitment and eventually, when the costs of studying are found to outweigh the benefits, to drop out. A similar evaluation in which the student weighs up the value of their investment, in terms of rewards and costs, is proposed by the 'investment' theories, such as that developed by Bean (1980, 1985), Bean & Metzner (1985) and Schell & Thornton (1985) Whtie both the 'investment' and 'integration' theories highlighted the importance of 'costs' in relation to drop out, several writers, notably Johnes (1990), have highlighted the fact that finances are but one of many factors which affect a student's determination of the costs and benefits of continued attendance. Many models developed to describe and test the relat1onsh1p of factors thought to influence drop out from other higher education inst1tut1ons were found to have a number of important sim1larit1es: drop out was commonly portrayed as a complex process influenced by a multitude of factors, including student background and characteristics on entry, environmental changes and institutional factors. However, these theories were developed predominantly from research based on University drop out data. Whilst the effects which institutional characteristics have on drop out are not known, the different funding systems, entry criteria, length and level of programmes ottered and varying delivery styles alone suggest reservations in applying theories developed for other educational settings, in other countries. to the study of drop out from New Zealand polytechnics. This study, employs the findings from previous research only as a platform from which to launch an exploratory study of a particular instance, aiming to develop a greater understanding of polytechnic drop out from the study of a single polytechnic, Wanganui Regional Community Polytechnic. Subsequently, it may be shown that findings from this case study have general application across the polytechnic sector. 1.3 Case Study of Drop Out from Wanganui Regional Community Polytechnic 3 The investigation of student drop out from Wanganui Regional Community Polytechnic (WRCP) programmes was initiated because the institution's administrators were concerned that, as a small, recently established polytechnic, costs associated with the loss of students might threaten its economic viability. In particular, administrators feared an increase in the rate of drop out as planned changes to government financial support for students and higher course fees, resulting from reduced levels of funding to institutions, created increased financial pressures for students. WRCP offers a wide range of subjects and levels of study for students from throughout New Zealand. Since its establishment in 1984, the Polytechnic had grown rapidly in size. so that by 1991 it boasted 843 Equivalent Full Time Students (EFTS). Of these, 353 were known to be actual full time students enroled in 'Formal'' programmes of study leading to nationally recognised , largely vocational, qualifications. An unknown number of students, attending part time or participating in short seminars and workshops, contributed to the remainder of the EFTS figure. In 1991. the first year of bulk funding, records provided by WRCP's central admtn1strat1on indicated that 148 students officially withdrew. However, the Polytechnic administrators were not certain of the significance of this figure in relation to WRCP's total student population. Did this f 1gure indicate a high drop out rate? What proportion of drop outs left because of financial hardship? It was recognised that some level of drop out was probably inevitable as many students were known to withdraw for personally positive reasons such as beginning a new job or the opportunity for alternative further training. However, in light of increased resources invested in the area of student support, and the implementation of quality management processes to gain feedback and improve education delivery, the numbers of students leaving programmes at WRCP appeared, to staff working on these initiatives, to be disappointingly high. Further, staff handling complaints recorded, for the first time, students demanding 'value for money'. Higher course fees apparently brought with them raised expectations of the Polytechnic's education services. Drop out costs arise not only because leaving students may represent a toss of revenue and a waste of institutional resources but also because of the possible sense of personal failure and dissatisfaction for the individual. The WRCP Student Loans Officer, for example, indicated that 4 withdrawing students often expressed such feelings, whatever other reasons they gave for withdrawing. WRCP administrators feared that, if the students' reasons for leaving were associated with dissatisfaction with the programme or service delivery, then the reputation of the Polytechnic as a quality education provider may be at risk, giving rise to longer term marketing problems and possible loss of competitive edge. Gedye (1992) confirms this possibility, noting that dissatisfied customers are known to talk to between 20 and 30 people about their experiences, whereas satisfied customers tend to talk only to one or two others. These concerns lead to a consideration of the factors influencing student drop out, in particular, finances and institution related factors for leaving the polytechnic. 1.4 Overview Adopting a case study approach, this research investigates the nature and reasons for student drop out from Wanganui Regional Community Polytechnic by using a variety of methods. From understandings gained about both the drop out process and the extent of the 'problem', this study also aimed to develop strategies to reduce drop out, (should this intent prove to be appropriate). A major function of the study was to establish a baseline of data against which ongoing monitoring could ascertain changes and identify possible future trends. The study seeks firstly (in Chapter 2) to identify factors thought to affect drop out through an analysis and synthesis of previous higher education drop out literature. By categorising the findings from studies based on different research designs and methodological approaches, a greater understanding of apparently contradictory findings develops. This process enables issues and theories relevant to the study of drop out from WRCP are identified. Whilst the drop out literature establishes the importance of a student's personality and previous experience as factors in drop out, external factors relating to the institution and its environment are also identified as having a major influence. The significance of external factors in drop out was of particular interest because around the time of the study a number of far reaching changes occurred as a result of government economic and educational policies. Thus, the second section of the literature review (Chapter 3) addresses the broad social, economic and political issues surrounding polytechnic education in New Zealand. T his chapter indicates both internal and external factors which, through their impact on the administrative structure and resourcing of individual polytechnics, affect the very culture of these.organisations including the delivery of programmes, the provision of student support services and possibly an individual student's propensity to drop out. 5 Establishment of the context for this research and the development of a platform of current knowledge about drop out allowed clarification and refinement of the research questions specific to the case study at WRCP. In Chapter 4, the case study approach is described and the subjects and data sources (both qualitative and quantitative) are identified. The results are reported in two chapters. The first, Chapter 5, overviews educational provision at WRCP, its history, administrative structure, programmes and student population, and provides a specific context for the interpretation of the drop out results reported in Chapter 6. The pattern of drop out at WRCP, over the two year period from 1992 to 1993, highlights several critical issues which are discussed in Chapter 7. This section reflects on the significance of differential and 'hidden' drop out rates and, through an analysis of who drops out and why, seeks to assess the importance of finances and institution related factors amongst the many reasons given for students leaving prior to programme completion. The concluding section (Chapter 8) addresses key implications for WRCP and identifies areas for future research. 1 A full glossary of terms associated with drop out and other forms of leaving behaviour is provided on page 1x aoullined in Chapter 3 •The 'integration" and 'investmenr theories are descnbed more fully as part of the rnerature review in Chapter 2. • 'Formal' programmes ate a Ministry of Education defined category of particular significanc~ in drop out and have been identified in this way to avoid confusion between the common usage of the term 'formal'. 6 7 CHAPTER 2: LITERATURE REVIEW: DROP OUT THEORIES Using selected studies to illustrate major shifts in methodologica l approach, this first section of the literature review traces the search for the cause of drop out. Initially the search focus was on the characteristics of leavers. but later research suggests that drop out is a complex phenomenon influenced by a multitude of interdependent variables and the heterogeneity of students. Recent developments have seen two major conceptualisations proposed which endeavour to move beyond description to explain the processes that lead to drop out. 2.1 Drop Out: Relevant Research There has been extensive research into student drop out from University courses both in New Zealand (Hooper, 1988 and 1989; Jones, 1978; Tremaine, 1979) and overseas (de Rome and Wieneke. 1982; Johnes, 1990: Knight, 1991 ; Metzner. 1989; Phythian & Clements. 1982; Powell, Conway & Ross, 1990; Roberts, Boyton, Buete & Dawson, 1991; Smith, 1987; Sweet, 1986; Tight. 1992; Woodley & Parlett, 1983, to name but a few of the more recent studies). However, although drop out is not a new phenomenon in higher education, it 1s only recently that it has been regarded as a problem by polytechnics m New Zealand. This is reflected in the dearth of sector specific research on this topic, limited to Ostman et al. (1988), Polyview Teaching and Learning Task Force (1984), and Zajkowski (1991). Another factor contributing to the limited hterature on polytechnic drop out 1s, of course, the recent development of polytechnic education in New Zealand. Since the introduction of the Education Amendment Act 1990, the term 'polytechnic' has been officially defined in New Zealand, to include 'technical institute' and 'community college'. However, overseas these terms encompass a wide variety of different educational deliveries, which may or may not equate to the New Zealand polytechnic experience. Hence, whenever possible, local polytechnic studies have been selected to illustrate major research trends. A blurring of the traditionally recognised boundaries within the tertiary sector has also occurred as a result of legislation, allowing polytechnics to offer degree programmes. Overseas and University based studies are included in the literature review on the grounds that they contribute to the researcher's 'pre understanding' (after Gummesson, 1988) and to the platform of knowledge from which this study of polytechnic drop out has evolved. 8 Building on Pantages & Creedon's {1978) comprehensive (largely American) evaluation of drop out research conducted over the period 1950 to 1975 and West, Hore, Bennie, Browne & Kermond's (1987) review of recent Australian studies, this chapter provides an overview of drop out in higher education, through the addition of New Zealand findings (these were limited) and the results of more recent overseas studies, where these gave new perspectives. Munro (Cited in West et al., 1987) encapsulates some of the frustrations experienced in reviewing the drop out literature which she describes as "a rather murky body of knowledge". ft would be convenient if drop out research fell into distinct categories such as that associated with student leaving styles or with factors under the control of the institution or with broad social issues but, while the literature includes these perceptions, they are often blended together, Munro (Cited in West et al., 1987) explains. A further complication is that different methodologies and operational definitions employed in drop out research restrict cross study comparisons. Thus, through an analysis of previous research categorised by research design, an understanding of factors which are thought to affect drop out from higher education institutions is developed. 2.2 Methodological Approaches to the Study of Drop Out Three major research designs have been identified by West (1985, a), calling on Terenzini's (cited in Willett & Singer, 1991) analysis of drop out studies: (i) the autopsy, or post hoc design, where students who withdraw are identified and a sample is surveyed or interviewed (also referred to as 'retrospective' studies, according to Willett & Singer, 1991 ); (ii) the cross-section design in which both a sample of withdrawers and a (sometimes matched) sample of non-withdrawers is included; (iii) the longitudinal design in which a cohort sample identified at enrolment is followed over time, with information collected at the beginning of the programme and at various points over the lifetime of the study, until an expected date of graduation and sometimes longer. In an international overview of factors influencing drop out and reasons given tor leaving, Thomas {1985) discovered that the bulk of the literature prior to that time comprised demographic studies of a descriptive nature. Gekowski & Schwartz (cited in Pantages & Creedon, 1978) warn that the findings from this type of research are suspect, because there is a focus on either the characteristics of drop outs or of persisting students, with no comparisons made. They note that early drop out studies typically focused on single factors when, they 9 argue, multiple factors operate concurrently to produce drop out. Many subsequent studies support this view - notably that conducted in a New Zealand Polytechnic by Ostman, Wagner & Barrowclough (1988). Investigations adopting post hoc designs commonly select a group of students who have already dropped out and then attempt to discover the reasons tor dropping out. West (1985, a) explains that the reliability of this approach is questionable because of the possible influence which subsequent events have on the students' recollections. West also doubts the honesty of the reasons given for withdrawing because of the negative connotations associated with withdrawal. This view is shared by Knight (1991) who maintains that many of the reasons given by drop outs are merely socially acceptable attributions and not the real cause of withdrawal. Although West suggests that indepth follow up discussions with students and the gaining of others perspectives can be used to overcome the inherent weakness of post hoc studies, Jex & Merrill (cited in Pantages & Creedon, 1978) maintain that this approach should be abandoned In favour of longitudinal studies. The longitudinal approach permits an exploration of factors that affect dropping out at the very time they are exerting their effects and the approach provides a "clearer view of the complex interaction of factors that influence a student to withdraw ... ," Pantages & Creedon (1978, 50) argue. This approach also enables the researcher to distinguish more accurately between students who drop out and return to the same institution and those who transfer to another place. Whilst many drop out researchers agree that the longitudinal design is superior, costs and time factors limit its general applicability in favour of cross-sectional designs. Sunmrnn; This section described the major research designs which are employed to study drop out, highlighting limitations which need to be taken into account when considering the findings across a variety of studies. 2.3 Problems in Defining the Population of Study: Terms Associated with the Leaving Behaviour of Students The wide range and lack of clarity of operational definitions which are employed to study drop out, and the diversity of data presentation, also pose problems in attempting to compare the findings of various studies. Many different terms' are used in educational literature to describe the leaving behaviour of students, including discontinuation, wastage, failure, drop out, attrition and withdrawal. However, the various terms which are associated with drop out are "used very loosely and mean different things to different people" (Woodley & Parlett, 1983, 2). 10 Tinto (1975, 90) was one of the first writers to draw attention to the limitations imposed on theory development by the inadequate attention which is given to questions of definition, leading researchers to Mlump together ... forms of leaving behaviour that are very different in character". For example, it is common for researchers not to distinguish between institution initiated drop out resulting from academic failure and voluntary withdrawal. Similarly, permanent drop outs are often categorised with transferring students and temporary withdrawers who plan to return and this lack of consideration of the numbers of students flowing back into the education system can lead to over-estimation of drop out (Willet & Singer, 1991). Tinto captu res the essence of the problem well by explaining that the failure to make such distinctions has resulted in "findings contradictory in character and/or misleading in implication" (1975, 90). By way of illustration, he notes that inadequate distinctions between academic failure and voluntary withdrawal have resulted in findings which indicate scholastic ability to be inversely related to drop out. unrelated to drop out and directly related to drop out. In contrast, Pantages & Creedon (1978) claim that the d1stinct1on between ·voluntary' and 'non­ voluntary· withdrawal ought to be abandoned on the basis that both forms of withdrawal are determined by social forces: why else do students qualified for admission get poor grades, they question. Bean (1985) 1s of the same view. He argues that students who leave an insutullon involuntarily, due to v1olat1on of social or academic standards, represent failures of either soc1allsat1on or selection. However. selection criteria are variable between institutions and, over lime. w1th1n an ins\llution. Dilferenl types of institutions have graduation rates which reflect their selection policies. For example, the number of drop outs can be expected to be low 1n an lnsutut1on with a highly selective system of admissions. Conversely, the implementation of 'open door' and 'affirmative action' policies allows greater numbers of 'at risk' students to have the opportunity to participate 1n higher education with the consequence that drop out may be high from institutions which adopt such policies, as they must still maintain the academic standards of their qualifications. Such examples of the diversity of interpretation of leaving behaviour highlight the need for drop out studies to clearly specify the context of institutional policies and practices, in order to develop a better understanding of the factors which affect drop out, although few attempt to do this. How the term 'drop out' was defined operationally was another major difficulty in comparing findings. What constituted 'drop out' varied enormously according to the period over which various studies were conducted, and the point at which data collection commenced. Depending on the research design, data were found to be concerned with drop out in a given year, with 11 continuation from one academic year to the nex1, or with graduation rates. Willett & Singer (1991) maintain that each year more and more students are taking longer to complete2, so that at the time chosen as the end point of a study such students may be neither graduates nor drop outs. There were also differences as to when an applicant was formally recognised as a student and, therefore, eligible to be termed a 'drop out' if studies were discontinued. Data collection began variously: - at the point of a student's first contact with the institution, - after an application form was submitted, - once selected onto a course, - once enrolment procedures were completed, - after the full fee had been paid - after a certain specified period of time, prior to which a full refund was given if the student withdrew. In most New Zealand polytechnics normal practice is to consider a student enroled on completion of an enrolment form and payment of enrolment fees (or arrangement of a student loan). From this point, the student's progress becomes the concern of the polytechnic and the student becomes a potential drop out. In contrast, a review3 of the literature indicated that a system of provisional enrolment operates in many universities. Provisional enrolment gives prospective students the chance to try the new educational experience without penalty as withdrawal before final registration (several weeks after provisional registration) is not considered as 'dropping out' and is not part of an individual's academic record. Neither is it considered in the final drop out figures published by the institution, and hence provisional enrolment would seem to result in artificially low drop out rates for those institutions which allow this practice. Once again, the importance of an understanding of an institution's policies is highlighted as an important factor in interpreting and understanding drop out findings. The purposes for which data were collected appear to determine the way in which the information is presented, once again posing problems in comparing the findings from various studies. Measures of performance, or rates, which compared leavers with the total number of participants in the population of study, were found to be largely institution, departmental or course based, reflecting the financial impact which drop out has on activities at each of these levels within an educational organisation. By far the largest proportion of institution-focused drop out studies were found to emanate from distance providers, because they were concerned that they were experiencing a much higher drop out rate than their conventional counterparts, although rarely was this assumption checked. Similarly course drop out tended to become the focus of study only when drop out from a programme was considered as being unacceptably 12 high. Business/accounting and management programmes were a frequent target. It was uncommon to find such studies placed in the context of drop out from other programmes in the same institution, or referenced to drop out from similar programmes offered by other providers. There was no way of identifying the degree to which these cases were unique/common to other situations, a problem exacerbated by the lack of consistency and clarity in the operational definitions applied to drop out research in higher education. Whilst a few of the earlier researchers considered drop out from the students' perspective, more recent literature indicates a shift from an institutional towards a student-centred approach to drop out, evidenced by a trend to classify drop outs and persisters into a greater number of categories as suggested by Tinto (1975) and later by Pantages & Creedon: "finer discriminations between the categories will yield better results and clea;er interpretation ... and will improve our understanding ... ," (1978, 55). Various criteria have been employed to categorise students ranging from demographic and personality characteristics to reasons for, and timing of withdrawal and these are discussed in greater depth in Sections 2.5 and 2.6 of this chapter. According to Tinto (1975), the failure of previous research to resolve questions of differentiation of the target population are of concern because they significantly impact on policy and planning in higher education. The identification of target populations which require specific assistance is important to polytechnic administrators because they need to consider strategies, such as flexible admission and tutoring systems, which can respond to the heterogeneity of the student population caused by learners who transfer or return after a sometimes lengthy period away from study. Administrators also need this information to make appropriate financial provision to take account of both the numbers and special needs of these students. However, West, Hore, Bennie, Browne & Kermond (1987} point out that gross statistics, such as the percentage of students failing to complete their course, do not take into account qualitative aspects such as whether the decision to withdraw was 'good' or 'bad' from the student's perspective. For example, transferring students are clearly losses from the institution's point of view, although this process may result in a positive outcome for the students concerned. Similarly the institution might regard students as 'unsuccessful' if they do not gain full credits in a given year for the courses in which they have enroled. However, a student who withdrew from three courses but passed one might still consider themselves as successful. This consideration is an important one, particularly for polytechnics, as New Zealand Qualifications Authority (NZQA}4 requirements for standards based assessment and a 'units of learning' approach to programme curricula have lead to a questionning of the value of traditional measures of drop out, such as the proportion of students completing a programme. 13 Sum man; Investigations in this field have been characterised by inconsistency 1n defining target groups, both theoretically and operationally and the importance of considering drop out findings in the context of institutional policies and procedures has been signalled. Indeed, Panos & Astin (1968, cited in Pantages & Creedon, 1978) maintain that the results of many studies are not comparable because they deal with different phenomena. Munro (1987, cited in West et al.. 1987) is in agreement with this view, noting that "findings are rarely generalisable outside the situations in which the research is conducted". Failure to resolve questions of definition. and an inability to move beyond description towards an explanation of the processes lhat lead to withdrawal, are factors which have hampered our understanding of the nature of the drop out process from higher education, according to Tinto (1975), one of the early drop out theorists. Unfortunately httle progress appears to have been made on this issue in the intervenmg decades. Whilst the student perspective is highlighted as an important cons1derat1on. most studies were found to reflect the institution's porn! of view, because drop out has a heavy 1mpac1 on finance and operations: 1t is "a waste of private and social investment rn higher education". according to Johnes (1990, 87). Section 2.4 examines drop out rates and patterns. focusing on the institullonal perspective. 2.4 Drop Out Rates and Patterns: The Institutional Perspective In terms of sheer numbers, drop out deserves the attention of those interested rn. and affiliated with , institutions of higher education, Pantages & Creedon (1978, 49) maintain. Certainly drop out rates reported by previous researchers would suggest that drop out is a problem of some magnitude for many tertiary institutions. 2.4.1 Drop Out and Graduation Rates Rates of drop out from higher education have remained strikingly constant over the past 100 years despite marked growth in participation rates and radical alterations to the tertiary system, Tinto (1982) claims. Analysis of major research reviews conducted between 1913 and 1975 lead Pantages & Creedon (1978) to a similar conclusion. They maintain that drop out rates had not changed significantly over this period, averaging at just under the 40% level. In contrast, an analysis3 of selected drop out studies conducted in New Zealand and overseas revealed a wide variation in drop out rates reported by individual institutions, ranging from 12% to 82%. Whilst previous discussion (Section 2.3) would suggest that this variation is partially attributable to the myriad of operational definitions employed by researchers, changes in the 14 political, economic and social environment experienced over the last 20 years may also have contributed to these differences in drop out rates. More detailed analysis of drop out rates suggests that, on average, less than half the entering students graduated within the normal programme term, although many completed their studies at a later date. Indeed. unbroken participation appears to be the exception rather than the rule. For example Thomas (1985) found that graduation in the minimum time was never the norm for 40% of students. However, of the students who dropped out for any length of time, Garrison (1987) found 70% of them re-enroled at either the same institution at a later date or at another one. Similar findings relating to drop out and graduation rates were corroborated by reviews of recent Australian studies (Clark, 1989; West et al., 1987). Whilst West et al. noted that 25% of drop outs did so without failure, a number of New Zealand researchers (Boshier, 1969; Hooper, 1988; and Ostman et al., 1988) found that students who evidenced past failures, or a history of dropping out, were more likely to do so again. Most of the studies were found to be retrospective studies of cross sectional design and subject to the criticisms outlined previously (Section 2.2). especially the lack of distinction made between temporary and permanent withdrawal. The findings from Eckland's (1964) longitudinal study (cited in Pantages & Creedon, 1978) suggest that previous investigations have over­ estimated drop out rates by about 10% once returning and transferring students are taken into account. Panos & Astin (cited in Pantages & Creedon, 1978) however, question the desirability of distinguishing between temporary and permanent withdrawal, noting that researchers wishing to make such a distinction would have to wait until all the students in the sample either re-enroled or died. Summary Although many studies demonstrated some consistency of drop out and graduation patterns over time, the huge variation of rates evidenced across the whole spectrum of institutions studied and the variety of operational definitions employed gives rise to criticism regarding the validity of the findings. As West et al. (1987, MF,3) conclude "these crude figures hide more than they expose". The following sections, therefore, focus on a finer analysis of both the institution and groups of students in order to gain a better understanding of the drop out phenomenon. 2.4.2 The Characteristics of Institutions as a Factor in Drop Out Early researchers assumed that the institution is a constant for all students. Therefore, the institution received little attention as a factor in drop out until lffert's (1957) survey (cited in Pantages & Creedon, 1978) revealed its importance. Hence, it is only recently that research has focused on the effects of the institutional environment on student drop out and retention. 15 "Dropping out is an interaction between an individual and an institution ... ," Cope & Hannah (1975, 9) explain. The institution and its environment influence the formation of goals, produce changes in students who attend, and also act as a selective device prior to enrolment. Different types of institutions have different public images and, therefore, appeal to different types of students. Where the educational goals of the student are compatible with the educational philosophy and orientation of the institution then the chances of the student persisting are enhanced. This notion is described in the drop out literature as 'fit' (Astin, 1975; Cope & Hannah, 1975; Marinaccio, 1985; Pantages & Creedon. 1978). The 'fit' between student and institution is thought to account for most of the transferring, stopping out, and dropping out of students from institutions. Prospective students are interested in choosing an institution to maximise the benefits they expect from higher education, whereas tertiary administrators hope to recruit and select the applicants best qualified to capitalise on the particular education programmes offered. Astin (1975), therefore, deduced that selection and recruitment procedures were also about 'fit': if the student attended an institution at which the socia l backgrounds of other students resembled their own background. especially regarding town size, religion, and 'race', then the likelihood of persistence was enhanced, he argued. The notion of 'fit'. described by Pantages & Creedon (1978) as one of the best theoretical focuses ror understanding drop out. has important ramifications for the implementation of equal educational opportunity (EEd0)5 policies which are common In polytechnics. EEdO policies aim to detect and take steps to remove barriers to education for all potential learners, but particularly those exhibiting low participation and success rates in polytechnics. Application of the 'fit' theory may be able to explain why minority groups do not participate/persist in higher education and may also prove useful in developing strategies to overcome such barriers. Type According to overseas studies reviewed by Pantages & Creedon (1978) higher drop out rates were indicated at state supported institutions than at private ones, and at technical institutes compared to liberal arts colleges and universities. Many studies indicated higher drop out rates from distance education providers than from conventional institutions, although the "openness and accessibility," (Powell et al., 1990) and part time nature of study (Tremaine, 1979) were major factors in distance drop out. As few New Zealand polytechnics evidenced similar administrative structures or subject specialisations to those overseas institutions surveyed, this particular line of research was not found to be particularly useful, except to suggest that polytechnics may evidence differential drop out rates according to their origin: that is whether they were originally established as a technical institute or community college.6 16 Size Some earlier studies identified a relationship between size and persistence, while others concluded there was none, according to research reviews undertaken by Cope & Hannah (1975). Astin's (1975) comprehensive, longitudinal study found that small institutions (fewer than 500 students) had higher drop out rates than expected although otherwise there was no apparent correlation between institutional size and persistence. Astin observed that larger institutions were seen by students as offering greater curriculum diversity and national prominence, whereas smaller ones were able to give personal attention and provide a collegial environment. Thus, it would appear that size alone was not the critical factor in drop out. What appeared to be important was the match or 'fit' between what the student expected and what the institution could offer because of it's size. Tinto (1975) was critical of many of the earlier studies which focused on institutional size and drop out: many simply categorised institutions as above or below a given size without controls for the type or quality of the institution. Calling on the work of Rock, Centra & Linn (1970), which took into account levels of institutional income per student, a factor considered to indicate high quality institutions, Tinto suggested that "very good" smaller colleges might be as effective in promoting students to graduate as larger "high quality" institutions, but in different ways. The smaller institution may be able lo enhance persistence through increased student-staff interaction because of its normally lower student-faculty ratio. This should lead to greater academic 'integration' as evidenced by good grades and intellectual development, he theorised. Larger institutions, normally more heterogeneous in student composition, may enhance persistence through their ability to provide a wider variety of student sub- cultures, leading to greater social integration. So although the theoretical foundation offered by Tinto, based on lack of integration, differs somewhat from Astin's notion of 'fit' both of these earlier researchers suggest that the size of an institution alone cannot account for differences in drop out rates between institutions. Bradley, Mclachlan & Sparks (1990) recent study on the nature and extent of stress experienced by Australian students attending two different sized tertiary institutions (one with 3,000 students and the other with less than 900) also failed to add support to the notion of a direct relationship between institutional size and drop out. This may have been because of the very distinctive characteristics of the institutions selected for study, as these were extremely diverse on a range of factors other than size. Problems in controlling the number of factors involved when conducting research across widely different systems and settings have been commented on by other researchers (notably Taylor et al., 1986) and, as a result, little other cross-institutional work on sizes as a factor in drop out appears to have developed these ideas further. 17 Mode of DelivenJ As most polytechnics in New Zealand are now moving to include more 'open learning'7 modes of delivery in their programmes, the traditional differentiation between 'distance' and 'conventional' teaching/learning processes is becoming somewhat blurred. Distance teaching institutions appear to be associated with significantly lower rates of successful completion of courses and programmes of study than campus-based institutions according to Powell et al. (1990). And an 'horrendous' non-completion rate of over 50% has been observed in systems where the institution permits students to work at their own pace without absolute deadlines, Daniel & Marquis (1979) say (in Ost man et al.. 1988). In their study of drop out from a New Zealand distance polytechnic, Ostman et al. ( 1988) acknowledge that drop out statistics validate the criticism of distance learning as not being cost effective and confirm that distance education students are more likely to drop out than are classroom students (with reference to Feasley, 1982). However, Tremaine's (1979) New Zealand University study led her to conclude that problems which resulted in the students leaving their extramural programme were related to being part-time rather than to the distance mode of delivery itself. Thus. it appears that both the mode of dehvery and the nature of enrolment (full time or part time) are factors which need to be considered within the mix which determines the unique character of that institution. Summary Apart from the high drop out rate noted for distance education institutions, there was little clarity in the relationship between institutional type and drop out. However, the 'fit' between the institution and the student has been identified by a number of researchers as an important concept in understanding drop out behaviour. The characteristics of individual institutions differ according to many factors including the variety of students enroled, the range and level of subjects taught, the styles of teaching, modes of delivery, and types of assessment procedures used. These aspects are picked up in the next section which focuses on drop out patterns observed within an institution. 2.4.3 Drop Out Within Institutions Drop Out Patterns Associated with the Level of Study Most withdrawing students are new to tertiary study, with a greater drop out rate noted in first year (Bean, 1985; and Clark, 1989) or first level (foundation) programmes (Garrison, 1985; and Smith, 1987). Bean (1985) attributed the higher first year drop out rate to the effect of grades, 18 which were thought to take their toll early, as students left because they expected to fail. Alternatively, Boshier's (1969) research into New Zealand University extension drop outs suggested that lack of familiarity with the requirements and expectations of tertiary study may be a key factor. West et al's (1987) multi-institutional Australian study would appear to support this notion because early withdrawers predominantly cited reasons associated with academic preparedness. D1ff1cult1es experienced in making the transition from school to less structured study and the ·unusual expectations some students have been given of tertiary education," (West. 1987,48) were the most significant cluster of problems identified and these were reflected in dissatisfaction with either the course or the institution, whereas finance and job related reasons became more important for withdrawers beyond the first year. They found that 'first timers' were more likely to drop down to part-time study than withdraw altogether. Bean explains that if ins11tut1onal 'fit' results from socialisation and not selection. one would expect 'fit' to increase over time, so that students could be expected to become more committed to finishing, the closer they get to completing their qualtfication. In contrast. Woodley & Parlett's (1983) analysts of Open University (United Kingdom) programmes showed that the highest drop out rates were found in third and fourth level courses. Although the iump from second to third level courses was considered by respondents to be a significant factor most were mature students studying part-lime extramurally and ·as people get older and spend longer in the system, they are less inclined to have the time and energy~ (Woodley & Parlett. 1983, 42). Differences 1n drop out rates noted between first and higher levels of study at the Open University may, thus, be a function of the mode of delivery, (distance learning) or a function of part time study. However, further analysis of the drop out rates recorded for each level at the Open University revealed that if the numbers of students fail ing to complete final registration had also been Included in the calculations, drop out among new students would be about as great as second level, highlighting, once again. the problems of definition in drop out research and the dangers inherent in cross-institutional comparisons. Although little research appears to have been done in either New Zealand or Australia on the possible relahonsh1p between drop out and the level of study. especially drop out amongst first time tertiary students. it was clear from overseas research that this was an area worth exploring further in relation to the polytechnic sector. Differences Between Subjects, Courses and Departments Differences in drop out rates between subjects, courses and departments were recorded in many studies. Science and maths drop out rates were consistently higher than arts and social sciences in distance education offered by the Open University (Woodley & Parlett, 1983; Phythian & Clements 1982). de Rome & Wieneke's (1982) longitudinal study at the University of New South Wales found engineering students were the most likely group to discontinue because of failure to meet academic requirements. They noted that early leavers were most likely to be arts students although this may have been attributable to the nature of their enrolment as most were part- timers. 19 Smith's (1987) study of participants on foundation programmes at the University of Newcastle, Australia, revealed that some subjects placed greater strain on students than others, but the results did not indicate the arts/science dichotomy apparent in some of the studies referred to previously. A similar view was proffered by Hooper's (1989} New Zealand study of Massey University extra mural accounting drop outs. He noted that drop out rates varied widely amongst courses with no conformity among related subjects. Hooper claimed that drop out rates were random across all faculties although he provided little evidence to support this.8 Differential drop out rates across subject areas were also noted in New Zealand studies conducted in the polytechnic sector by the Polyview Teaching & Learning Taskforce (1984}. Whilst this group observed variation both within the programmes and amongst departments, they reported that the Arts and Community Studies9 department at Christchurch Polytechnic exhibited the lowest overall drop out rate of any department. as well as consistently low drop out rates in individual classes. However, this was found to be a function of the length of programmes in this area (most were short courses) rather than of the nature of subjects offered. It would appear that subject effects were also influenced by the course context, particularly by whether the course was career oriented. A cross-disciplinary study conducted by Jones (1978) at Auckland University revealed that those who enroled in the 'professional' faculties were more likely to return after the first year than those enroled in general arts or science courses, a trend attributable to the perceived career benefit of professional courses. Hooper's (1988) study of accountancy drop outs indicated that vocational 'award' courses made greater demands on students than non-vocational ('recreational') courses. Hooper also confirmed Hibbett's (1986} earlier findings that drop outs had expected their course to be interesting, although this expectation was not found to be so significant among successful students. These somewhat inconsistent findings suggest that departmental course ownership is a rather crude category, masking many course characteristics which Influence drop out, including the overall course design and delivery and intrinsic difficulties of subject matter. 20 Timing of Drop Out Within any year, there was a tendency for drop out to occur early in the programme. Hooper's (1989} Massey University study indicated that most drop outs. half the final number, occurred before the due date or the first assignment. This early leaving group commonly reported being in the wrong course, or taking too many papers as reasons for dropping out. Hooper found a further 9% of students left before the second assignment, citing work pressure and course demands as their main reasons. The remaining students who left later in the academic year, dropped out because they found the course too hard to cope with. Research conducted by Barger & Hall (1965}, de Rome & Wieneke (1982), Tremaine (1979), West et al. ( 1987), and Woodley & Parlett (1983) supported the notion that particular factors were associated with the trme or withdrawal. Although finances. dissatisfaction with the institutional environment, lack of motivation and the gaining of full time employment were reasons for leaving given by those who withdrew early, many found the course different from what they had expected. The early withdrawers were more likely not to be in the programme or institution of their first choice and to have lacked a real commitment to further study, sheltenng in tertiary education while they sought employment. Later drop outs were more likely to give academic reasons, these studies found. Thus. it seems that it is the student who is 'new' to higher education, the entering students, the first year students, who are more 'at risk' of dropping out, a notion in keeping with Anderson & Darkenwald's (1979) study which revealed that participants on adult basic education courses (the equivalent of foundation courses, or second chance education in New Zealand) were four times more likely to drop out than other adult education participants, a view supported by Boshier's earlier (1969) New Zealand study. However, Anderson & Darkenwald did not confirm the pattern of early drop out, noting that basic education participants were Mno more or less likely to drop out at the beginning or end of a course", (1979, 27). These contradictory findings may be attnbutable to differences between the student populations of the various studies. Indeed, West et al. (1987} maintain that such findings confirm the need for researchers to disaggregate withdrawing students into different types. Part ol the problem in obtaining unequivocal answers about factors affecting withdrawal, has been the failure of previous studies to separate different types of students, according to Tinto (1982). Whilst the forces that lead to drop out in the early stages of the studenfs academic career are quite d1ff erent from those that influence drop out later, it appears that these may also differ for different types of students. 21 Drop Out by Nature of Enrolment: Part-time versus Full-time The apparent contradiction in findings regarding drop out rates at different levels and in different subjects may be attributable to other factors, including whether the students are studying full­ time or part-time. Thomas (1985, 43) shifted the focus from the programme to the participant, arguing that "Dropping out is a highly individual matter". He noted that part-time students were more likely to be mature age'0 and tended to have higher drop out rates than full time students coming directly from school. Smith's (1987) study revealed that as many as half the part-time mature age students dropped out, mostly through attrition. O'Shea & Corrigan's {1979) indepth study of working class participants in adult education appeared to confirm the view that many folk commenced tertiary study in a tentative way. Often the students' first taste of tertiary study is to attend a hobby or non-assessed introductory course. probably as a part-time student. As confidence grows. the student may feel capable of enrolling in a more academic programme or extending studies to a higher level, and may also make the commitment to attend full-time. It was. then. not surprising to find that several studies reported a higher drop out rate for pan-timers than for lull-t1me students {Jones. 1978; Katz & Barrett, 1972: Knight. 1991 ; Tremaine. 1979). Part time drop outs may be first timers expenencmg d1fftcult1es in adiustrng to the culture and expectations of ter1iary study (Boshier, 1969) or alternatively, they may be leaving because they are ltnd1ng the conflicts between study and family or work commitments unresolvable (Hackman & Dysinger, 1970). S 11111111nry A number of common patterns of leaving behaviour were able to be identified, despite concerns about the misleading nature of drop out rates (Willett & Singer, 1991 ). Those new to tertiary study were generally found to be more at risk of withdrawing than those enroled in subsequent years of a programme or on more advanced types of programmes. A differential drop out rate was noted between subjects and departments in a number of studies, although there was little consistency in findings, suggesting that this factor was only one of many affecting drop out. Part-time, mature age enrollees and people on longer courses were found to be more likely to drop out than younger, full -time participants and those on shorter courses, features apparently related to time management and the pull of external responsibilities. Early drop outs (before the first assessment) were found to form the largest proportion of the total number of those who dropped out in any one year, and the reasons given for leaving suggest implications for career counselling prior to enrolment However. students rarely announced their intention to withdraw. They just stopped attending or gradually dropped out after a period of irregular attendance, posing difficulties for institutions in providing drop out support. 22 2.4.4 Summary Although a number of considerations 'militate' against comparing drop out data derived from different institutions (Pantages & Creedon, 1978, 56), several useful patterns have been revealed . While Kennedy & Powell do not dispute the value of such basic research, they point out that "the important question of why students drop out together with the related issues of why they enrol ... in the first place, are left unresolved." (1976, 61}. The next section builds on these early findings to focus on the characteristics of drop outs in an attempt to develop the type of insight for which Kennedy & Powell argue. 2.5 Drop Out: The Student Perspective Because of the possible effect which drop out can have on the institution's finances, most studies were found to reflect the institutions point of view. However. Johnes (1990) argues, financial costs are only one part of the total costs: drop out also incurs non-pecuniary costs which can only be guessed at. For this reason, Pantages & Creedon (1978) highlight the importance of cons1denng the student's point of view, while recognising that this aspect is difficult to gauge. This section traces development towards an understanding of drop out, initially focusing on the identification of leaving students. 2.5.1 Characteristics of Drop Outs Many earlier studies focused on the characteristics of leaving students in an attempt to predict which students might be at risk of dropping out. Hore (1985, 46) noted a tendency of this early research "to place the 'blame' for drop out on the shoulders of the student", questioning their background and ability, their performance, motivation, and commitment. Two New Zealand studies of this type, that of Hooper (1988) at Massey University, and Jones ( 1978} at the University of Auckland, typify the findings of many overseas researchers, evidencing inconclusive and contradictory findings in regard to the relationship between student characteristics (such as gender, age and entry qualifications} and drop out. Gender Hooper's study of those who failed and those who withdrew within their first year of accountancy found that while women were more likely to withdraw early they were also more likely to 23 complete the course. These findings confirmed earlier studies, such as that conducted by Demos (1968), which reported that men dropped out at significantly higher rates than women. In contrast, Astin's (1964) longitudinal, multi-institutional study of personal and environmental factors associated with drop out among high aptitude students revealed women had a significantly higher drop out rate than men, and less of them (84% compared to 94% men) stated that they planned to return to college at a later date to complete their studies. However, Jones' inter-disciplinary study of non-returning first year enrollees indicated that gender was not a factor as men and women stayed away in equal proportions. Similarly Cope & Hannah (1975) found that men and women students discontinue, stop out and transfer in about equal proportions, but for different reasons. They noted that men tended to give reasons related to competency, adequacy or identity searching, whereas women more often gave intellectual, aesthetic and social dimensions. In their extensive review of earlier research, Pantages & Creedon (1978) also concluded that there were no significant differences in the overall drop out rates of men and women, maintaining that the apparent discrepancies of findings could be accounted for by the influence of other factors such as previous education experience or type of institution. Age Astin (1975), Anderson & Darkenwald (1979), Hayes {1976), Jones (1978) and Thomas (1985) found mature participants were more likely to drop out or fail to return in a subsequent year of a programme. although Hooper (1988), noted that mature students were more highly motivated and hence more likely to persist. Hooper (1988) suggested there was a correlation between increasing age and increased likelihood of course completion, or decreased tendency to drop out. Students on vocational courses (such as accountancy) were found to be younger than average and less likely to drop out if they were full time. Hooper found those under 20 years had the lowest drop out rate, although the small number of students in this categor'f suggests some caution as to the generalisability of these findings. Indeed, Pantages & Creedon's review of previous research found drop out rates were similar for students who were either older or younger, concluding that age was not a primary factor in drop out. Entry Qualifications Students who achieved well at school generally performed well at University, according to Jones (1978). However, art students were found to be an exception - their chances of successful course completion could not be predicted from qualifications on entry. 24 While Hooper's (1988) correlation of enrolment data with student University performance did not suggest any significant relationships, students whose records showed a high proportion of past failures were not likely to complete the accounting course either. He found that the number of prior papers completed and the qualifications held on entry were not useful in predicting whether a student would be successful or drop out from their accounting course, findings consistent with Boshier's (1969) earlier New Zealand comparative study (discussed further in Section 2.6.1 ). 2.5.2 Summary Research into the characteristics of 'drop outs' provided few consistent correlations. The contradictory nature of these findings is partially attributable to the problematic nature of the definition of terms. A second methodological flaw apparent in many of the earlier studies was the absence of a control group (the persisters) in their research design. The next section summarises key findings developed from comparative studies. 2.6 There are Few Differences Between Leaving Students and Those who Stay: The Results of Comparative Studies Studies of drop out emphasise differences between drop outs and persisters Hayes (1976) explains, on the assumption that the more marked the ditterences, the greater the probability of successfully predicting drop out. Comparative studies appear to have been marked by the influence of sociological11 thinking in their approach to the problem of student drop out. 2.6.1 Demographic and Socio-economic Characteristics Drop Outs and Persisters Share the Same Demographic Characteristics Apart from educational background 12 , most of the comparative studies (most recently Knight, 1991) found little to differentiate the two groups in terms of demographic characteristics. Socio-Economic Status Although New Zealand is commonly referred to as a classless society (Beatson, 1990), a system of social practices exists, shaping the form of people's daily lives and maintaining a social hierarchy (or class structure) against the vision of equality which New Zealanders have of themselves (Wilkes, 1994). Whilst Beatson (op. cit.) warns of the dangers of equating socio­ economic scales (SES), with class, their use can sometimes indicate relationships between various factors and particular groups. 25 In a review of earlier drop out studies, Summerskill (1962, cited by Pantages & Creedon, 1978) indicated that although socio-economic factors (frequently cited are father's occupation, parental education, family income, and ethnicity) are commonly believed to influence drop out, research has provided equivocal results. One of the earliest comparative studies was Boshier's (1968, 1969) New Zealand longitudinal study of 252 first year University extension students from a range of programmes. Boshier was particularly interested to ascertain whether the findings of previous overseas studies highlighting the role of socio-economic status were relevant to drop out from higher education in New Zealand, but he concluded that there was little to differentiate drop outs from course participants on these variables, a finding supported by Birch (1975), but at variance with the findings of Anderson & Darkenwald (1979), Hayes (1976), and Astin (1964). Whilst Hayes' (1976) research at the University of New South Wales concluded that the two groups of students appeared to be indistinguishable when they entered University, a greater percentage of persister's mothers were found to have achieved some tert iary education. Astin found that students who dropped out were more likely to come from lower socio-economic backgrounds. 'Racial' status had little effect on participation in tertiary educatron although the category they described as 'blacks' was found by Asttn to be under-represented because of their low educational status, and 'black' adults were more likely to drop out according to Anderson & Darkenwald. Spoonley, however, is critical of the concept of 'race', noting that the interchange of genes makes "racial boundaries nonsense," (1994, 82) and. he argues. the assumption that physical differences provide an accurate marker of other differences is invalid. Spoonley explains that the notion of ethnicity is more acceptable, classifying people according to their self claimed group affiliations and shared cultural traditions, beliefs and values. A study conducted at the New Zealand Open Polytechnic (Ostman et al., 1988), suggested a relationship between ethnicity and drop out, because more Maori and Pacific Islanders discontinued their studies. These groups were not well represented in the student sample studied, however, raising questions as to the generalisability of these findings. Few other New Zealand studies appear to have looked at socio-economic variables in relation to drop out from higher education, possibly because of the sensitivity of tertiary students in revealing their own income, or that or parents were applicable, because of current regulations which 'means test' the allocation of student allowances13• Whilst Lauder et al's {1985) study or education and social inequity is one which has been criticised for the "regrettable imprecision,• (in Beatson, 1990, 90) with which the SES levels of the Elley-lrving scale have been merged and referred to as classes, this work did show that socio-economic status of students' families played a significant role in the level of 26 educational success achieved by Christchurch secondary school leavers. Because education had already been identified as a key differentiating factor between drop outs and persisters, this finding may add some support to Astin's suggestion that there is a relationship between SES and drop out through the level of education gained prior to entry. Overall, however, no clear relationship between drop out and socio-economic variables is apparent in the literature. Eckland (cited in Pantages & Creedon, 1978), explains that this lack of correlation exists because the composition of the student population at each institution tends to be relatively uniform with respect to SES, and most studies have been based on a single institution . These contradictory findings were not unexpected, given the diverse cultural settings in which drop out research has been conducted, and the variety of socio-economic scales and concepts applied. Su111111an; Few significant differences in demographic characteristics or SES factors were reported between those who left and those who stayed. However, persisters and drop outs were differentiated by their educational experiences, and this notion 1s further explored together with other academic factors in the following section. 2.6.2 Academic Factors Previous educational experiences, study habits and level of academic performance on the programme are areas which researchers have explored in order to determine which of the many academic factors are significant in student drop out. Previous Ed11catio11nl Experience Anderson & Oarkenwald (1979) and Boshier (1969) found that students with fewer years of formal schooling were more likely to drop out. However, Astin (1975) warned that higher education was a long way from being able to identify students who would be unable to cope with the academic demands of tertiary study. For example, while some studies noted positive correlations between qualifications on entry (Jones, 1978) and successful completion of programmes, Hooper (1988) and Tremaine (1979) found that entrance qualifications and previous tertiary education experience had no influence on withdrawal. Carr & Ledwith's (1980) examination of factors influencing the success of disadvantaged adults beginning studies with the British Open University, Morrison's (1981) analysis of success and failure amongst first year Auckland University students, and Johnes' (1990) study of the effects of academic entry requirements on non-completion at Lancaster University provide more recent evidence in support of Astin's concern. These researchers found that mature applicants, who often lacked formal academic entry qualifications, were just as likely to succeed in higher education as their school leaver colleagues. Motivation was thought to play the major role in 27 the success of these students and thus, raising academic entry requirements was not supported as a strategy for reducing 'wastage'. Jones (1978) theorised about the experiences of school compared to University as a factor in student drop out. An analysis of the different requirements of arts compared to other subject areas led him to suggest that prior knowledge of a conceptual framework, ability to problem solve and the flexibility to adapt to the freedom of study (the loose structure of University) were factors likely to be associated with success and re-enrolment in successive years of a programme. West's (1985, b) longitudinal study of BA students at Monash University (Australia) demonstrated that school leavers were more likely to successfully complete their first year of higher education if they had received most of their secondary education at a government, rather than independent schooL These findings (controlled for social background and ethnicity) would seem to offer support to Jones' point regarding the importance of the school environment in preparing students to meet the demands of higher education. However. Baldwin & Howe (1982) found that previous experience with a subject did not necessarily confer any advantage. Comparison of 498 first year University accounting students. revealed little difference overall between those with a background in lhe subject and those without, rn either exam performance or drop ou1 rate. However, students who had taken accounting at high school tended to drop out late in the semester, suggesting they may have been over confident at the beginning and unprepared for the extra effort required as new and more rigorous topics are introduced. Level of Academic Pe1formance on the Programme "All or even most of the potential drop outs cannot be 1dent1fied and, therefore, eliminated before applicants enrol", Pantages & Creedon (1978, 64) explain and, thus, many researchers turned their attention to predicting drop out after enrolment. Summerskill (op. cit..) found a highly significant relation between drop out and first semester grades in all 35 studies he examined. This finding was supported by Pantages & Creedon's analysis of later research, leading them to conclude that good grades function as effective reinforcers in maintaining and strengthening a student's academic performance, and, thus, decreasing their chances of dropping out. Astin's (1975) comprehensive cross-institutional study yielded similar results, supporting the notion of a correlation between achievement and persistence, although grades did not always indicate academic potential and many with very high potential dropped out. High grades were thought 28 by Astin to correlate with persistence because they affected motivation. They appeared to influence persistence directly, independent of initial variations in ability and family background, financial aid, employment during study, residence and type of institution. Although Astm's ( 1975) study found only one quarter of the respondents attributed withdrawal to academic difficulties, other reasons may have been 'handy rationalisations' for drop out because of poor performance, he said. Counsellors interpretations of the reasons for drop out in Demos' (1968) study would support this view, although he maintained that poor grades per se did not cause the student to withdraw. Expressed variously as course "too advanced", or course ''too hard", poor academic performance was highly rated amongst the reasons given by respondents in many later drop out studies, notably those conducted in New Zealand polytechnics (Ostman et al., 1988; Polyview Teaching and Learning Taskforce, 1984). Schedvin's (1985) study found that more than one third of the first years who discontinued were students for whom academic difficulties gave rise to such an acute fear of failure, that they withdrew rather than face it. These findings raise questions regarding the level of learning support given to students. SI 11dy Hnl11/s Astin (1975), Ostman et al. (1988) and Pantages & Creedon (1978) reported that the most drop out prone students had poor study habits. Although Garrison (1985, 31) noted that" ... persisters worked a greater number of hours than drop outs", some studies found there was little difference in hours spent on study between stayers and leavers (Jones, 1978, and Tremaine, 1979). Jones found that a large number of students who did not re-enrol had performed poorly m their end of year examinations and although this may have been a reason for not continuing, the student may already have 'given away' the idea of continuing before the exams and hence performed poorly. Summary Whilst academic variables have been described as the strongest single-variable predictors presently available in the study of drop out, (Pantages & Creedon, 1978), the level of academic performance whilst actually on the programme, measured by assessment grades, was the only factor able to provide a consistent relationship with drop out. However, Pantages & Creedon (1978, 65) conclude, poor grades are not generally sufficient "in and of themselves to cause attrition, but must be coupled with other non-intellectual factors". Thus, it appears that, although academic factors are an important group of variables affecting success in tertiary study, they are 29 not the only variables that contribute to drop out. The influence of advances in psychology1• becomes apparent as researchers turned their attention to psychological and situational differences between those who leave and those who persist on a programme, in order to make progress towards understanding drop out. 2.6.3 Psychological and Situational Variables in Drop Out Iffert (1957) and Summerskill (1962), both cited by Pantages & Creedon (1978), documented that the most prominent reasons given by drop outs in their decision to leave were motivational. Students' own expectations (influenced by their level of aspiration, fear of failure and parental attitudes), their reasons for attending tertiary education (including their educational interests and career goals) and peer group influence were areas of motivation Pantages & Creedon (op. cit.) found previous studies had touched on. although they considered that earlier researchers had failed to establish relationships among levels of motivation, commitment to the institution, strength of educational goals and drop out. More recent research explores these, and other psychological aspects, further. Reasons for Studying Several writers have suggested that the reason for study (Astin, 1975; Boshier, 1973; Hibbelt, 1986; Hunt, 1982) 1s an important factor in persistence/drop out and in overcoming any difficulties encountered. Some of the wide variety of reasons students indicated for attending University were also found to be related to re-enrolment probab1hty. including the desire to qua hf y for a well paid or mterestmg JOb, interest 1n the subject and to have a 'good time'. Jones (1978) found career orientation to be a powerlul incentive for re-enrolling at University for all students. West et al. (1987) suggest that student drop out may not be a sign of failure but an indication of a re-evaluation of career goals or an assessment that tertiary education is not for them. Students who chose subjects at University because they found them interesting at school, or because they were good at them, were more likely to continue within a year (Hooper, 1988) or to re-enrol in a subsequent year (Jones, 1978). Interestingly, Jones found those who rated "having a good time" as an important reason for continuing were more likely to return to complete subsequent years of a programme especially in the non-arts areas. This finding contradicts the 'common-sense view' which predicts poor performance for students who are out to 'have a good time'. Whilst many studies which aimed to ascertain the reasons for leaving a programme also looked into reasons for joining the programme, the assumption that these aspects are related was 30 rarely challenged. Similarly, the factors involved in the initial decision to join a programme may not be the same as those involved in the decision to re-enrol fo r subsequent year on a longer programme. The reliability of the data in many studies is also questionable, as subsequent events may influence students recollections of reasons given for both enrolling and leaving. Course Choice and Commitment De Rome & Wieneke's (1982) longitudinal study of 1.375 first year students at the University of New South Wales made a significant differentiation between persisters and drop outs: ii appeared that drop outs were more likely to have enroled under less than propitious conditions, although few other differences were noted. The study revealed that the University of New South Wales was the first choice of persisting students - they had consulted relevant publications when making enrolment decisions, and they had a long-standing commitment to the course. Students who had left within two months of their 1n111al enrolment were less likely to have had that University as their first choice. They tended not to have consulted pubhcat1ons or career advisors when making enrolment decisions. and they were not particularly committed to their course. indicating they would leave if they found employment. Students who left because of failure to meet minimum requirements to continue, were also more l ikely not to have consulted prior to enrolment. Thus, the quality of information received prior to enrolment and the students' commitment to the programme are factors which may be associated with withdrawal. Satisfaction In a follow up survey, de Rome & Wieneke (1986) found the content of courses was not as expected by almost half the respondents: persisters were likely to refer to differences in emphasis or workload, and drop outs to poor teaching or the course being less interesting than they had expected (also noted by Hibbert, 1986, and Hooper, 1988). A proportion of both persisters and drop outs in this study were not satisfied with their course. Anderson & Darkenwald (1979) maintain that the most powerful predictor of persistence in adult education is satisfaction with the learning activity in terms of its helpfulness in meeting the students objectives. However. this simplistic measure of satisfaction did not allow for complex responses, for explanation, nor for any distinction between types of satisfaction. Satisfaction appeared to account for a small part of the drop out equation, with only 10% of drop outs reporting dissatisfaction as a reason for leaving the programme. 31 Motivation The association between motivation and drop out has been suggested by many writers. notably Birch (1975), Boshier (1973), Hayes (1976, 1977) and Jones (1978). Hayes' concluded that drop outs could "not be considered as a group with homogenous characteristics differentiating ttiem from persisting students," (1977, 148). She argues that the term 'drop out' must be seen as encompassing a wide range of personal characteristics, mot