< Previous page ^ Table of Contents Next page >

Indicators of Academic Performance: Notice

The Joint Committee on Academic Performance was established in July 1996 by the Council and the General Board to act as a central focus for discussion of matters relating to the performance of candidates in Tripos examinations, and to make recommendations to the central bodies and the Senior Tutors' Committee as appropriate on such matters. The Joint Committee recognized that whilst much valuable research had been undertaken by particular Faculties and Departments, a University-wide research project on factors affecting the academic performance of undergraduates would be valuable. Funding was subsequently secured for a project, entitled 'Indicators of Academic Performance' and overseen by the Joint Committee, which was completed in November 2001 on the submission of a substantial report by the principal investigator, Dr Christine Mann.

The Joint Committee has commissioned a summary report of the findings of the research which the General Board and Council have agreed to publish. During the current academical year the General Board will, through their Education Committee, be considering how to take forward the findings of the report, following consultation with those Faculties and Departments on which the report focused.

SUMMARY REPORT OF FINDINGS OF THE PROJECT ON INDICATORS OF ACADEMIC PERFORMANCE

3 January 2003

The enclosed paper is a summary report of innovative research undertaken in the University of Cambridge entitled 'Indicators of Academic Performance', which investigated the academic performance of undergraduate students. The project was funded by the General Board of the University from April 1997 to October 2001. Generous contributions to the cost of the project were also received from three Colleges: Trinity, Girton, and New Hall. From 1997 to 1999 the Project was conducted jointly by Research Associates, Dr Chris Mann and Dr Patrick Leman; from 1999 to 2001 Dr Mann (now Senior Research Associate) was the sole principal investigator. The qualitative research was undertaken by Dr Mann. The quantitative research was done by Dr Leman, and further statistical work was carried out by Ms Alice Kneen, Mrs Sally Roberts, and Dr Margaret Ely under the direction of Dr Mann. Dr Mann wrote the final, very substantial, report of the project, which was submitted in November 2001. This summary of the report has been produced by the Joint Committee on Academic Performance, a committee of the University Council and General Board, which has overseen the project. Dr Joan M. Whitehead, University Lecturer in the Faculty of Education, prepared this summary and undertook the analysis of the degree results for 2001-02 in the Appendix.

The full findings of the project have been received by the Education Committee of the General Board and that body has been charged with considering the University's response to the research including, as a first step, consultation with the Faculties and Departments on which the project focused. It is hoped that the findings will be useful nationally.

1. Background to the Research Project

Academic excellence is frequently defined in terms of examination performance. The purpose of the project was to devise and implement a research programme into the factors affecting the performance of the University's undergraduates in Tripos examinations. Although the project was initially set up in response to concern about the under-representation of women amongst those awarded first class marks in University examinations, particularly in Part II, it also had the wider brief of investigating factors that might affect student performance in general. The project had, therefore, two aims:

(i) to examine the extent of difference in Tripos performance between different groups at the University of Cambridge; and
(ii) to explore the reasons behind these differences.

1.1 Research Approach and Methods

The project has used both quantitative and qualitative methods drawn from research in the social sciences. To fulfil the first of the above aims quantitative statistical analysis was used to look at the student characteristics that might influence examination performance, and to look at achievement patterns over time. The second aim was achieved by the mapping of student experiences through a variety of qualitative methods including sequential e-mail interviews with a sample of a single cohort which began in October 1997 (the 'Graduates of the Millennium' project) and an 'exit-poll' survey of the retrospective perceptions of all students graduating in targeted subjects in 2000 or 2001. Further methodology details are given when each 'strand' of the research is discussed in more detail below.

1.2 Colleges

It is important to note at this point that the project concentrated on the differences in the performance of students within and between Faculties and Departments, and their perceptions of those Faculties and Departments, rather than differences in student performance between Colleges. Although it had been intended to include College membership in the analysis, the difficulties of doing this soon became apparent. For example if students are divided by Tripos taken at Part II, gender, class of degree, and College, then the number of students in each of the, very large, number of cells in the matrix were, in a number of cases, too small for any meaningful statistical analysis. It was agreed, therefore, that a systematic analysis involving College membership was not possible. The very important role that Colleges clearly play in the education and lives of the students, however, is apparent from the qualitative data, but again it was not possible to carry out any systematic analysis of College membership, owing to the small numbers involved within each College.

1.3 Characteristics of Cambridge Students - The 1997 Cohort

The Project Data contain detailed information on the cohort of students who entered the University as undergraduates in 1997 in terms of gender, social class, ethnicity, type of school attended, and A level results.1 This particular cohort was chosen as they were the students who would graduate in 2000 or 2001 and some of them, therefore, would be participating in the 'Graduates of the Millennium' project. What follows is a brief discussion of the main characteristics of this cohort, giving information about the composition and background of the students. This part of the project looked at a large sample of students and used quantitative statistical analysis to identify characteristics across the population as a whole.

In terms of gender, 52.6% of the cohort are male and 47.4% are female. The cohort as a whole is predominately middle class and white.2 In terms of class, 83% of the cohort are middle class compared with 17% who are lower middle/working class. The distribution is very similar within gender: 84.5% of women and 81.5% of men are middle class compared with 15.2% of women and 18.5% of men who are lower middle/upper working class. In terms of ethnicity, 88.7% of the sample is white, 90.2% of women and 87.4% of men. The non-white students came from a wide range of ethnic backgrounds.

In terms of school background, 45.8% come from independent schools (48.4% of men and 42.8% of women); 22.3% from comprehensive schools (22% of men and 22.6% of women); 13.6% from grammar schools (12.8% of men and 14.6% of women); and 18.9% from 'other' state schools (16.7% of men and 20% of women);3 39.8% of women and 36% of men came from single sex schools, mainly independent and grammar schools.

1.4 The 'Dominant Group' Hypothesis

One of the arguments frequently put forward to explain why some students do not do as well, academically, at Cambridge is because they do not 'fit in' with what is perceived as the dominant group, this group being most frequently characterized as men from independent schools, who are assumed to provide a particular 'ethos' which works to their advantage and disadvantages others. As the following discussion demonstrates, it is actually quite difficult to identify any group of students, from the 1997 intake, as being a dominant group. Clearly men per se are not a dominant group as the number of men in the sample is only slightly larger than the number of women, nor are independent school pupils a dominant group as just over half of the cohort come from the state sector. What is clear, however, from the above statistics is that, in terms of class and ethnicity, working class students and those who are from non-white ethnic groups can be identified as minority groups within the student population.

If, however, we look at type of school in addition to social class and ethnicity, the picture changes. It shows that although students may come from a different social class or ethnic background, they will share other aspects of their experience with students in the majority groups.

Social class and type of school

Below are the percentages of the middle and lower middle/working class students who went to the different types of school arranged according to size of group.
Middle class, independent school boys 61221.3%
Middle class, independent school girls 51117.9%
Middle class, comprehensive school boys 2729.5%
Middle class, comprehensive school girls 2669.3%
Middle class, other girls2027.6%
Middle class, other boys1966.8%
Middle class, grammar school girls1645.7%
Middle class, grammar school boys1515.3%
Working class, independent school boys1144.0%
Working class, comprehensive school boys712.5%
Working class, independent school girls632.5%
Working class, other girls572.0%
Working class, other boys572.0%
Working class, comprehensive school girls441.5%
Working class, grammar school boys441.5%
Working class, grammar school girls371.3%

Although middle class independent school pupils are the largest group (some 39.2% of the total population), the males who come into this category are only slightly more numerous than females. It is difficult, therefore, to argue that they form a 'dominant group' in terms of a prevailing ethos, constituting as they do only one fifth of the population.

Looking at class and type of school within each gender group, and adding in single sex or co-education school experience, demonstrates even more clearly the difficulty of identifying a dominant group in terms of home background and previous school experience within the University student population.

Women

School Middle
class No.
Middle
class %
Working
class No.
Working
class %
Comprehensive26619.8%443.3%
Grammar16412.2%372.1%
Independent51138%624.6%
Other20215%574.2%

The largest group of women is middle class women from independent schools, which at 38% is twice the size of the next largest group (middle class comprehensive school women). The same pattern can be seen within the working class sample.

If, however, we add to these figures the type of school in terms of single sex or co-educational, a slightly different pattern emerges. The figures show that roughly two thirds (63.8%) of middle class independent school women went to a single sex school (a similar pattern emerges for middle class grammar school women), while just over 90% of middle class comprehensive school women went to co-educational schools. By adding in this variable the balance of the groups alters slightly. Thus among middle class women there is one group of women who went to single sex independent schools which forms approximately 25% of the population and a group of women who went to co-educational comprehensive schools which forms approximately 18% of the total sample. In between these two largest groups are middle class women from grammar schools and working class women from a number of different schools who may share one or more elements of their previous experience with one or other of these larger groups. There are working class women from single sex independent schools and from co-educational comprehensive schools. Thus there is no clear class/type of school divide within the sample, that is working class women from co-educational comprehensive schools versus middle class women from single sex independent schools.

Men

School Middle
class No.
Middle
class %
Working
class No.
Working
class %
Comprehensive27218%715%
Grammar15110%44 3%
Independent61240%1147.5%
Other19613%573.7%

A very similar pattern emerges for males.

The largest group of men is middle class men from independent schools, more than twice the size of any other group at 40%, followed by middle class comprehensive school men. A similar pattern emerges among working class men.

Adding in single sex or co-educational schools, once again, changes the balance. Approximately half of middle class independent school boys went to single sex schools, 20% of the whole sample, while 86% of middle class comprehensive school boys went to co-educational schools forming approximately 16% of the whole sample. As with girls, the other groups share one or more elements of their previous experience with one or other, or both, of these larger groups.

The inter-relationships between the variables of gender, social class, and type of school show complex patterns of overlapping experiences, with the result that no really 'dominant' group emerges, in terms of these variables, that could provide an overwhelming 'ethos' within the University.

Ethnicity and type of school

The majority of non-white students in the sample went to independent schools (60.4%) with 12.5% having been at comprehensive schools, 8.9% at grammar schools, and 18.1% at other schools.

Conclusion

From the above discussion, therefore, we can see that it is very difficult to argue that there is a dominant student group within the University that is providing an all pervading ethos. Although it is clear that students from working class backgrounds and from non-white ethnic groups are, in terms of class or ethnicity, in a minority, in terms of their school experience, however, they will have experiences in common with larger groups within the student body. The question still remains, however, as to whether any of these 'background' factors (alone or in combination) are more likely to be associated with success in University examinations than others.

2. Student Characteristics and Academic Success

Student characteristics and academic success were investigated by looking at the variation in class of examination awarded to students who graduated in 1997 and 1998,4 again using quantitative analysis to identify trends within the sample. The number in the sample was 4,866 which is 73.6% of the total number of those graduating in those two years, the remaining 26.4% (1,755) having to be excluded from the analysis because the data on these students were incomplete.

The variables included in the analysis were subject of study, gender, social class, ethnicity, and type of school. Excluded from the analysis were the students' A level scores at the time of entry to Cambridge. They were excluded because there is so little variation among Cambridge students on this variable. 90% of all students who come to the University have achieved at least three A grades at A level with only 4.3% achieving less than AAB; 98.5% of students enter the University within two years of leaving school. Students who did not complete were also excluded, in this case because the numbers are so small; 97.5% of Cambridge students successfully complete their degree. Although these factors may be important in other universities, the homogeneity of the Cambridge population means that these variables do not relate to variation in degree results. This homogeneity, however, has an advantage because, as Dr Leman, one of the Research Associates on the project says, it 'creates particular conditions in which the impact of social factors upon undergraduates' academic performance can be clearly examined' (1999, p. 237).

Using loglinear analysis, because the data are ordinal (categorical) in nature, two way tests of association were considered for each of the variables given above with class of degree. Three way interaction analyses were performed using different combinations of variables, for example gender, subject of study, and class of degree.

The two way tests of association showed that three variables were significantly related to class of degree, subject studied, gender, and ethnicity; and that two variables, social class and type of school attended, were not significantly related to class of degree.

The fact that social class and type of school are not directly related to class of degree is both significant and interesting. It shows that students from state schools, whether comprehensive or selective, perform just as well in examinations as those from independent schools. Thus attendance at an independent school does not advantage students as is often assumed. It is also clear that the minority of students from working class backgrounds perform just as well as the majority of students who come from middle class backgrounds.

In terms of gender, men as a group get more firsts than women and more thirds and lower seconds. In terms of the relationship between ethnicity and class of degree, the significant relationship is produced primarily by the performance of black students. White, Indian, Chinese, and other Asian students have a very similar distributions across all classes of degree. The distribution for black students, however, is different and skewed towards the lower end of the class distribution, as the table below shows.


Ethnicity Number and
% of those
gaining a
first or a 2i
Number and
% of those
gaining a
2ii or a third
Total
 
White2,80680%70220%3,508100%
Indian10678.5%2921.5%135100%
Chinese4275%1425%56100%
Other Asian8972.4%3427.6%123100%
Black2062.5%1237.5%32100%

The differences between black students and other ethnic groups is most marked at the extremes of the distribution as the table below shows.


Ethnicity Number and % of
those gaining a first
Number and % of
those gaining a third
 
White73821%852.4%
Indian3223.7%64.4%
Chinese1017.9%11.8%
Other Asian1512.2%10.8%
Black13.1%515.6%

Thus we can see that of all the ethnic minority groups it is black students who most differ from the majority white students with only 3.1% of black students gaining a first compared with 21% of white students while 15.6% of them get a third compared with 2.4% of white students. Although the numbers in non-white ethnic groups are small, and particularly so for black students, the difference between black students and other students is marked enough to give cause for concern. A significant minority of these students clearly do not thrive in the Cambridge context. Further research looking specifically at this group of students is therefore needed.

Although both gender and ethnicity have a significant effect on the class of degree obtained, by far the most important variable in this respect is the subject studied, with students standing a far higher chance of gaining a first in some subjects than others. The next stage of the analysis, the three way interactions with Tripos class, shows, however, that different variables are associated with success in different subject areas.

The analysis of gender, subject area, and class of degree showed that the results for men and women vary across subjects. In the two years under study, for example, women and men taking Law tend to perform equally well, achieving similar levels of representation in each class while in English men significantly outperform women, gaining twice as many firsts as they do, while women are awarded twice as many lower seconds than men.

The performance of different ethnic groups also varies according to subject. For the three way analysis, because of the small numbers of students from non-white backgrounds, the sample was divided into only two groups, white and all other ethnic minorities. Dr Leman goes on to say: 'Finer category distinctions inevitably get lost in this transformation. Yet in order to demonstrate variation along ethnic lines and subject area all that is required is the minimal case, i.e. differences between one ethnic group and the others. So if ethnicity interacts with subject and degree class the binary distinction between ethnic groups will pick it out' (p. 242). The analysis showed that performance in the different subject areas did differ according to ethnicity. For example, ethnic minority students performed less well than white students in Engineering and History but better than them in Mathematics.

Although type of school showed no significant interaction with class of degree in the two way analysis, it did show a marginally significant effect in the three way analysis when combined with gender and ethnicity. In terms of gender, there was no significant interaction for men between type of school, subject area, and class of degree but there was for women. State school women obtained slightly more firsts than other women in Law and Physics, while independent school women achieved slightly more firsts in History. In terms of ethnicity, an analysis of the interaction between class of degree, type of school, and the two largest ethnic groups (white and Indian) showed that although there was no significant interaction for white students, there was for Indian students, with students from independent schools gaining many more firsts and fewer thirds (no Indian student from an independent school was awarded a third) than Indian students from other schools.

Conclusion

The summary of findings from Dr Leman's analysis show that the factors associated with academic success at Cambridge are very complex. The strongest associations are between class of degree and subject area, followed by gender and to a lesser extent ethnicity. No association is found between class of degree and either social class or type of school attended. However, the three way interactions show that, in different subject areas, the students' gender, ethnicity, and to a lesser extent the type of school they attended, interact in different ways to produce variations in academic outcomes depending on the subject area. Given the extent of the variation it would appear that each subject area has its own 'profile' in terms of which combination of variables is likely to produce success. For example, the 'profile' of a successful student in History is very different from that of a successful student in Law.

The main conclusion, therefore, that can be drawn from Dr Leman's analysis is that differences in degree results cannot be ascribed to any one single variable. Instead, we have to look at complex patterns of interacting variables in order to understand success in the different subject areas. The two largest sources of variation in class of degree, however, are the subject studied and gender. The fact that working class students and students from non-white ethnic groups are in a minority within the student body does not affect their academic performance, with the single exception of black students.

Questions for Further Study

Three questions, therefore, remained unanswered following this analysis:

1. Why should there be such variation in academic success between subjects studied?

2. Why should being male or female be so important in determining academic success?

3. What combination of factors are most likely to lead to success in the different Triposes?

It is reasonably clear from the above analysis that the answers to these questions are most likely to be found in the day-to-day experiences that students have, both academically and socially, in Faculties, Departments, and Colleges. The next phase of the research, therefore, was aimed at 'tapping into' students' experiences and perceptions of life as a student at Cambridge and how these relate to academic success.

3. Selection of the Sample for Further Study - Departments and Faculties

An in-depth study of student experiences clearly could not be carried out on all students in the University. The first process of selection, therefore, was by Faculty and Department to cover as representative a cross-section of the University as possible. Their selection was guided by the following considerations:

1. The Faculties/Departments selected should include some examples of where:

(i) women achieved a higher proportion of firsts than men in Part II of the Tripos,
(ii) men achieved a higher proportion of firsts than women,
(ii) men and women tended to perform equally well.

2. The Faculties/Departments selected should include some which could be described as arts, some as science and some as 'intermediate' (such as Law and Geography) in orientation.

3. The Faculties/Departments selected should have sufficient numbers of students taking Part II to ensure meaningful sample sizes for any statistical analysis.

4. The Faculties/Departments selected should include some:

(i) awarding a relatively high number (approximately 20% or more) of firsts in Part II,
(ii) awarding an average number (approximately 15%) of firsts,
(iii) awarding a relatively low number (approximately 10%) of firsts.

Using these criteria, based on examination results from 1990 to 1996, the Faculties or Departments shown in the table below were selected.

The study of the Natural Sciences Tripos subsequently focused on the Departments of Physics and Chemistry, in both of which subjects the number of men and women gaining firsts was fairly equal but slightly favoured men. It should also be noted that by the end of the study the results for the Department of Engineering were dramatically different with women doing as well as, and in some courses better than, men (a point which will be discussed later).



Faculty or DepartmentGender Differences (% differences)Subject AreaNumber of Firsts Awarded
Physics and Chemistry (Natural Sciences Tripos)Men slightly better = 4%ScienceHigh
ClassicsWomen slightly better = 1.4%ArtsAverage
EngineeringMen much better = 13.2%ScienceHigh
EnglishMen much better = 12%ArtsAverage
GeographyMen slightly better = 3.3%IntermediateLow
LawMen slightly better = 3.4%IntermediateLow
MathematicsMen much better = 20.7%ScienceHigh
Modern and Medieval LanguagesMen much better = 13.2%ArtsAverage

3.1 Selection of Students and Staff within Faculties and Departments

Two separate groups of students and some academic staff were used for the second phase of the research.

1. For a pilot study, some staff in the selected Faculties and Departments and some of the students who obtained a first in the Part II examinations of 1997 were interviewed.

2. For the main study, students who began their university career in October 1997 were followed throughout their time at Cambridge - the Cohort Study.

The purposes of the pilot study interviews with students and staff were to:

(i) allow comparisons to be made between the perspectives of students and staff within individual disciplines;
(ii) identify the terms in which high achieving students make sense of their academic careers in each subject area; and
(iii) identify similarities and differences between students' experiences.

The interviews with staff, combined with analysis of prospectuses and other material describing course structure and requirements, helped to build a picture of forms of knowledge and teaching practices in the target Faculties and Departments. The interviews with students were exploratory, to try to identify areas for further investigation and to provide information on the best way to proceed with that further investigation in terms of the most appropriate methodology to use.

The complexity of the personal and interpersonal issues raised by these interviews with students, combined with the consistent message from them that an individual's experience of Cambridge changed over time, suggested that a longitudinal approach to student experience would be a valuable next step. Although all the students had been academically successful, many of them described periods or events that had nearly taken them 'off-course' - as well as key periods or events which they saw, retrospectively, as being beneficial. These 'moments of epiphany' (which could be related to College or Faculty or Department life) often seemed to crop up in a particular year of study and, depending on how they were experienced and dealt with, could affect the course of study in subsequent academic terms or for longer.

3.2 The Cohort Study (Graduates of the Millennium Project)

The aim of the cohort study was to track the progress of students from the start of their time at Cambridge (October 1997) until the end of their undergraduate careers. The study was called the 'Graduates of the Millennium' project as many of the students graduated in the year 2000.

Students who were intending to study in the selected Departments and Faculties from October 1997 were sent a questionnaire before they arrived. Of the 500 questionnaires sent out, 355 were returned. It is these 355 students who constitute the cohort study database and who were used for the statistical analysis. The questionnaire asked students if they were prepared to be re-contacted. About 200 students agreed to be re-contacted and these students became the ones who generated the qualitative data in the cohort study, mapping their day to day experiences and perceptions of life as a student. All of the students were again contacted at the end of their course in either 2000 or 2001 and asked to take part in a semi-structured questionnaire about their time at Cambridge. This survey had several core questions across all subjects and some subject-specific questions. Survey data provided retrospective impressions of Cambridge experiences. The survey also mapped the immediate career intentions of those about to graduate.

3.3 Approach to Data Collection

As the main purpose of the cohort study was to investigate the perceptions and experiences of students in terms of their 'on-going' life as a student, a Grounded Theory approach was adopted (Glaser and Strauss 1967).5 The development of Grounded Theory arose from Glaser and Strauss's work on the generation of hypotheses. They argued that in areas where there was little knowledge about the phenomena under study the generation of meaningful hypotheses for empirical testing was, at best, difficult and, at worst, impossible. Because of the paucity of knowledge researchers were, in effect, simply generating hypotheses on the basis of ill informed guesses which might result in the exclusion of significant information about the phenomena under investigation, and hence produce misleading results and conclusions. As an alternative, Glaser and Strauss proposed that hypotheses should be generated as a result of 'in-depth' study of the phenomena using the 'unstructured' approach of participant observation. They argued that the observers should aim to enter a research situation with no prior theoretical preconceptions and then create, refine, and revise theory in the light of further data collected. The resulting 'Grounded' hypotheses generated through actual observation would, therefore, be more accurate than those produced by speculation, whether theoretically underpinned or not, however inspired. The Grounded Theory approach has been developed by others since the early work of Glaser and Strauss. These developments have made a major contribution in introducing methods of data collection, other than participant observation, and methods of analysis collectively known as qualitative methods. The basic principle, however, remains the same and is well summarized by Patton (1980):6 'the cardinal principle of qualitative analysis is that causal relationships and theoretical statements be clearly emergent from and grounded in the phenomena studied. The theory emerges from the data; it is not imposed upon it'.

A qualitative approach, therefore, was the most appropriate way to investigate the way student 'lifestyle' influenced academic achievement because it fulfilled the conditions in which Grounded Theory is most valuable - when little is known about the phenomena under investigation.

The method chosen for actually collecting the data was an innovation on the traditional diary approach often used in qualitative research. Rather than keeping a written diary of their thoughts and activities over a specific period of time for subsequent analysis, participants communicated with the researcher, Dr Chris Mann, by e-mail. Data were collected in two main ways. Direct questions were e-mailed to them by Dr Mann to which participants were asked to respond; in addition they were encouraged to e-mail any thoughts or observations that occurred to them which they felt were relevant to the study.

The use of e-mail (involving responses to direct questions, combined with this more spontaneous 'diary' approach) provided a flexible research tool which adapted well to the unpredictable lifestyle of many students.7 The regular contact mapped changing perceptions of College and Department/Faculty experience. The data identified how students learn, and the kinds of support and pressure they meet in different learning environments. On-going analysis also tracked and analysed students' references to confidence, resilience, motivation, social adjustment, organizational skills, and making choices (both personal and course based).

The great advantage of material generated in this way is that it provides a rich 'data base' of students' observations and perceptions about their life as students as well as offering many insights into gender, class, and ethnicity issues.

It does, however, like all methods of data collection in the social sciences, have its disadvantages. First of all, the students are themselves volunteers, a self-selecting group out of the 500 who were sent the questionnaire, which, may, therefore, not be representative of that cohort of students. There is little that can be done about this possible bias; clearly, one cannot force students to participate if they do not want to do so. It has also to be remembered that the 500 students are not a random sample of students from the University, but those who are studying in certain Faculties and Departments. Thus, any findings for the cohort may not necessarily be generalizable to students in other Faculties or Departments in the University.

The 'unstructured' method of data collection also has disadvantages. Although the students were all sent the same questions to respond to at periodic intervals, much of the data arose from spontaneous e-mail messages sent to Dr Mann. Thus, there is clearly an element of selection in what students chose to talk about, again raising questions about how generalizable the findings are. As stated above, however, the purpose of Grounded Theory is to generate questions or hypotheses and to raise issues that can then be subject to more systematic study, and the results of the cohort study certainly do that. The data generated provide insights into the students' experience of Cambridge which points to much good practice within the Faculties under scrutiny which in turn contributes to successful outcomes for both men and women. The data also highlight issues and areas of concern which may inhibit some students from thriving and reaching their potential, and this strongly indicates the need for further consideration and action. It is also interesting to note that, despite individual differences between students, there is in many areas a remarkable degree of consistence in the views expressed.

The main areas explored through the cohort study were:

(i) The impact of school-university transition on academic performance;
(ii) the implications for student performance of perceptions of both excellence and the means of achieving excellence in different subject areas;
(iii) the ways in which teaching, assessment procedures, and course curricula might lead to variations in students' academic performance;
(iv) the ways in which the cultural context of learning might impact on academic performance;
(v) the role of students' experiences (including social and personal) of Cambridge in general as an influence on how they perceived themselves;
(vi) the influence of students' perceptions of themselves and the University upon the choices they make about their lives academically, socially and personally.

4. Summary of findings from the Cohort Study: The Outcome

The detailed breakdown of the results for each Tripos is being discussed with each of the Faculties or Departments concerned; here we summarize general findings that can be drawn for the research as a whole. These are given below.

4.1 Characteristics of the Sample

The cohort comprised 355 students of whom 46.5% were men and 53.5% women (see Table 1 below). Women were therefore significantly over-represented in the cohort compared with the 1997 intake as a whole.

  No.% of total
GenderFemale19053.5%
 Male16546.5%
Total 355100%

The cohort included a higher proportion of students from independent schools (50.6%) than the whole 1997 entry (45.8%) and a lower proportion of students from other schools (13.4%) than the whole entry (18.3%). However, the proportions from comprehensive and grammar schools were similar and the overall distribution was not significantly different from that for the whole entry.

4.2 Part II Degree results for the 1997 Cohort - Gender Differences

Arts subjects: Although slightly more men than women were awarded firsts in Classics, the difference was not statistically significant. However, in English and Modern and Medieval Languages the difference was significant with men being awarded more firsts than women.

Natural Sciences: Although women were awarded a higher proportion of firsts both overall and in Physics, the differences were not statistically significant. In Chemistry the difference was significant with women being awarded more firsts than men.

Engineering: There was no significant difference in the number of firsts awarded, although women did slightly better in the Electrical and Information Sciences Tripos and the Manufacturing Engineering Tripos and men in Engineering and Chemical Engineering. This was a significant change in the trends prior to the start of the study.

Mathematics: a significantly higher percentage of men than women were awarded firsts.

Intermediate subjects: There was no significant difference in the number of firsts awarded in either Law or Geography.

There were, therefore, significant gender differences in the number of firsts awarded in only four subjects, English, Modern and Medieval Languages, and Mathematics, where men did better and Chemistry, where women did better.

There are a number of factors that can be identified that contribute to these gender differences in results, which are each discussed below.

1. Gender differences in the perception of the role of learning and examinations

Throughout the study, academic performance has been defined in terms of demonstrating excellence in Tripos examinations. As a result of the detailed analysis of all the comments made by the students in the cohort study, Dr Mann's main conclusion is that men and women may position these examinations rather differently within their vision of the Tripos overall. Her main conclusion is that there is a tendency for:

(i) women to see the Tripos as an opportunity to increase their understanding and to focus on the subject area 'in itself'. In this view examination performance is a by-product of learning and personal development; and
(ii) for men to be alert to 'performance' aspects of Tripos examinations from an early stage and to tailor their intellectual development to public success. In this view examination performance is the target of learning.8

These attitudes have created performance patterns in which many women invest a great deal of themselves in their academic work and many men take a more pragmatic approach to their academic work to ensure that they obtain good results. This has implications for examination preparation. There is a tendency for:

(i) women to prepare for examinations by working hard; their aim is to show the extent of their understanding; and
(ii) men to prepare for examinations by combining work with examination techniques; their aim is to produce an excellent performance.

These tendencies may be exacerbated by the effects of social class and school background.

(i) Women from single sex schools show a tendency to work towards showing 'perfection' rather than high competence; state school women tend to see hard work as their only route to success. In both cases women may suffer from trying too hard to do well, on the one hand, and fear of failure on the other.
(ii) Men have more confidence in both their innate ability and their skill in using examination techniques to good effect. Some independent school men, in all subject areas, have developed sophisticated examination strategies while at school and seem to use these as second nature within the Cambridge system.

As we shall see, these differences between men and women are more acute in some subject areas than in others. In addition, in some subjects women are beginning to move closer to the 'male' instrumental approach to Tripos and this is often associated with improving results.

2. Gender differences in examination preparation within arts subjects

Within the arts subjects many of the students felt there was a tension between the aim of a liberal education (emphasizing understanding and development) and preparation for examinations. Should they regard their work as an exploration of issues in itself or as preparing examination material? In Modern and Medieval Languages, women who interpret the Tripos as liberal education and take advantage of its rich and diverse options for study may undermine their examination performance. Men who focus on examination performance by consolidating their efforts in a more specialized way reap greater benefits. In addition, in English, Modern and Medieval Languages, and Geography, various formulations of the 'Cambridge Answer' were felt by some students to be the benchmark for excellence. However, as we discuss below, the 'Cambridge Answer' is arguably a form of examination technique, used instrumentally by some students, which may disadvantage those who do not feel comfortable with consciously using technique to excel. Classics, by rewarding a wider range of essay writing approaches, may have moved closer to liberal education ideals.

Law stands apart from the other subjects, as both men and women are highly instrumental in Tripos examinations. Successful law students find means of coping with vast amounts of material and view options such as the dissertation in terms of possible implications for performance. If arguments take a 'personal stance', this is as likely to be seen as much as an intellectual exercise as it is a personal conviction. There is remarkable consensus between male and female law students about the importance and validity of using examination strategies. This pragmatism was directly associated with the (partly) vocational nature of the degree. The consensual view may also reflect the fact that neither men nor women would have had the chance to develop examination techniques specifically relevant to Law before admission (in contrast to many other subjects); it would not be surprising if the approaches of all students new to a subject 'mirrored' each other.

3. Gender differences in examination preparation within science subjects

Students who succeed in Mathematics, Engineering, and Physics are generally those who have taken an instrumental approach to Tripos examinations. Women and men who excel have tended to focus on limited areas of the syllabus, learning targeted work in depth, and have practised Tripos questions until they can solve problems at speed under pressure. Women mathematicians, however, are much less at ease than men with being 'trained' in problem solving in order to excel in Tripos examinations. Women in Engineering and Physics also have reservations about training themselves to perform in this way but generally adopt a more pragmatic attitude. In Mathematics male students see the use of exam strategies as evidence of the drive to excel and many male students (and some male staff - especially in Mathematics) seem to interpret any other approach to the Tripos as time wasting and/or evidence of lack of ability. In contrast, Chemistry places a great emphasis on conceptual understanding, in both teaching and examining. Hence, to those chemists (often women) who see understanding as 'the point' of the Tripos, examinations in this subject are an appropriate progression.

We shall now consider in more detail the ways in which a (predominately female) emphasis on understanding and a (predominately male) emphasis on performance may 'play out' in the context of, first, arts subjects (looking at the 'Cambridge Answer') and, secondly, in sciences subjects (looking at differences in technique based and concept based teaching).

4. Performance and understanding in the arts - the 'Cambridge Answer'

Earlier research (particularly from the Faculty of History) has suggested that male and female undergraduate writing may be 'gendered' to some extent, with male academic work being typically bold, argumentative, and confident, whilst the female written style is more cautious, conscientious, and thorough.9 As Clarke et al. suggested, 'the marks of what we would identify as a 'women's style seem to us to be a preference for cautious, discursive and synthetic approaches, a willingness to consider a wide range of views and a strong personal investment in 'getting it right''.10 As we see, an emphasis on understanding has been associated with women in previous research. In contrast, what seemed to be valued in History, and by extension other arts subjects, was: 'an argumentative and self-assertive approach to questions, the bold affirmation of a particular view and confident dismissal of others'.11 The latter approach, often referred to as the 'Cambridge Answer', is usually associated with men, and may be interpreted by examiners as either being highly original and of excellent quality (first class), or being completely misplaced (third class). Thus, to write in this way is considered 'risky'. Only a minority of students felt that the 'Cambridge Examination Answer' came naturally to them, 46% of men and 31% of women in the sample. It is interesting to note, however, that more men than women reported feeling that this type of answer came 'naturally' to them.12

These debates about 'gendered' examination writing styles are familiar to arts undergraduates in this study, particularly in English and Modern and Medieval Languages. Women regularly affirmed their belief that the exam procedure favoured men. They felt that examiners construed 'well-constructed arguments' not as subtle, discursive, 'balanced' argumentation (of the kind women were seen to favour), but as the assertive (even aggressive), reductive arguments often characterized as 'male'. However, students often characterized examination preparation and performance in polarized ways which could disguise differences in the actual ability of candidates. On the one hand, to work hard across the year was frequently associated with a rather laborious and uninspired approach to study. As a result, students who worked hard because they were struggling intellectually or could not organize their time, were rarely distinguished by these students from those whose term work reached an impressive level of expertise as a result of effort, commitment, and interest. Similarly, descriptions of the kind of answers that might be produced by 'hard working' students rarely distinguished between those which were 'cautious' or 'balanced' because students were doubtful about their own ideas, and those which showed an admirable restraint in acknowledgement of the complexity of the material under review. On the other hand, value-laden terms were frequently used about the 'slick', 'aggressive', 'opinionated and pompous' Cambridge Answer - without unpacking what was actually being described. There is a difference between students who are able to produce a clear, strong, line of argument or to take 'an original and/or daring perspective' in an exam because they are on top of the material - and those who have taken an approach to examinations so instrumental that it is, in effect, a bluff; an illusion of understanding and/or originality.

The problem is that 'love of subject', and the desire to develop an authentic personal viewpoint, may lead to extensive investigations that spurn short-cuts. Students who emphasize understanding and seek a creative exploration of multiple texts can stockpile vast quantities of complex material. The challenge is then to 'pull everything together' for the examinations. This approach may give impressive original responses backed by wide reading. Unfortunately, it more often squanders energy, leaving writers with little time to distil their knowledge and order their thoughts. Students may find themselves lost in a sea of details from which it is difficult to extricate a clear argument. As much subtlety of thought is lost in timed examinations, work methods which begin with a wide review of material also tend to reduce to an apparently pedantic 'argue with, then against, the essay title and summarize' writing style which can be dismissed as 'a mere mosaic of sources' by some staff. It is ironic that students who have striven most fervently to embrace the liberal ideal may be those who come across as having the least depth, sophistication, or originality.

In contrast, more successful students may have never sought 'a vast understanding' from which to argue. Even those students who have taken a more expansive approach to a Tripos at the beginning of the course would hesitate to see this as a sensible way of producing an effective performance in examinations. It seemed clear from the data that those students who acted instrumentally were more free to manipulate material to impress examiners than those who thought Tripos examinations sought heartfelt exposition.

Is the 'performance' game worth playing in terms of results? It may be that women in some arts subjects (such as English and Modern and Medieval Languages) might prefer an honourable defeat (in terms of sacrificing a first class result) to the greater loss of giving up the personal ideal of a voyage into understanding.

The student perceptions and reflections on the 'Cambridge Answer' reported by Dr Mann do raise a number of issues, particularly taking into consideration the differences in performance between subjects. For example, what is happening in those subjects (Classics, Law, Geography) where there is no significant gender gap in performance but where essays produced under examination conditions are the main form of examination assessment? Is it the case that both men and women write the 'Cambridge Answer' or is it that such an answer is not actually valued in those subjects, or at least is not the only kind of answer that is valued? Is the 'Cambridge Answer' something that is only valued in English and Modern and Medieval Languages? Finally the question needs to be asked - is the very strong perception of some students, particularly women, that such an answer is actually sought and valued based on reality or is it a myth? These questions can only really be answered by further consideration of this issue in the Faculties and Departments concerned.

5. Performance and understanding in the sciences

In subjects such as Mathematics, Physics, and Chemistry there is evidence that, while all students reach understanding through a process which combines coming to grips with fundamental concepts and working through examples which apply these concepts, some students may prefer one approach over the other. Evidence from some staff and both male and female students suggested that there may be gender differences in the approach to learning - with women preferring to understand things at a more basic level so they could build up their knowledge while men tend to 'jump straight into the technical stuff' and to 'perform' right away.

These different approaches may have a bearing on the fact that in the Mathematical Tripos, for example, a number of women in the sample expressed feelings of discomfort about their learning experiences, which may in their perception have contributed to their relatively poorer performance in examinations. The Mathematical Tripos was described by one member of staff as 'a kind of competition you train for'; and a certain kind of problem solving forms the focus of the examination papers. Examination expertise is seen, above all, as having the ability to perform at speed under pressure, and mathematics teaching is accordingly 'fast-paced and technique-orientated', to quote from a description of the way the teaching of the subject in secondary schools is often experienced. A study of undergraduate mathematicians at Cambridge dating from 1994 found that for the great majority the preferred learning methods were rote memorization or doing examples (although understanding would gradually develop from learning proofs and working through questions). It was clear many of the women were unhappy with this approach. While some 'loved' maths and - as one put it - 'spent many, many happy hours problem-solving' or - as another said - 'gained much satisfaction from being able to solve problems on the example sheets even if it takes me many hours', another felt that 'women want to really understand mathematics rather than just crashing through the examples'. Some women felt that they could not move forward or 'learn efficiently' until they had a stronger grasp of the fundamental principles or conceptual frames which informed the formal manipulation of symbols presented in the lectures. As one woman who changed to another university explained, she wanted a course which would 'slowly open up the interconnections in mathematics logically, so that you don't have to take things on trust'.

These two different approaches have implications for supervisors. If some students (often men) accept basic principles in the expectation that all will become clear in the process of doing problems - while others (often women) ask a lot of questions to help their understanding, they are pulled two ways. Should they work on problems to keep the supervision moving along or should they stop and address fundamental questions that may need lengthy answers. If they adopt the first strategy this may help 'performance' oriented students at the expense of those who want a deeper understanding; if they adopt the second strategy this would help those who want to really understand basic principles but annoy those who want to just 'get on with it'. Furthermore, some women felt that, if it is they who ask the fundamental questions, this path to understanding might be seen in a negative light by male mathematicians and that, as they are the majority, this may inhibit the female minority from active participation.

Evidence which underlines the consistency of the finding that women favour understanding over perform-ance, is provided by the teaching in Chemistry - a subject where women were awarded significantly more first class degrees than men in 2000. Chemistry is now organized in ways that encourage students to understand things from first principles. Thus the teaching culture developing in Chemistry over recent years has made it acceptable for women to pursue a more conceptual approach to learning. What is interesting is they appear to have done this without causing a negative reaction, at least in this sample, of the more 'performance' oriented male students. It is possible that the dramatic increase in the numbers of women attaining first class results in Chemistry is associated with this opportunity to take an 'understanding' rather than a 'performance' route through Tripos.

6. The transition from school to university

In all subjects the transition from school to university was crucial. Experiences which might seem insignificant in themselves take on powerful importance for newly admitted undergraduates who, with notable exceptions, are all socially nervous and academically apprehensive on arrival. The way the first year, in particular, is negotiated may - for some students - set a seal on their eventual success. Those who arrive with low levels of confidence, or those who have reasons to lose confidence in the early part of the first year, may underperform in the first year exams. As these exams are used by the majority of students as a benchmark to prove to themselves they were not 'a mistake', the impact of these results is far-reaching in terms of confidence, motivation, and future aspiration.

Many factors impact on confidence levels and staff cannot be expected to take responsibility for all aspects of the lives of young adults. Nevertheless, there is much that can be done to integrate new entrants. It is easy for staff who are focusing on academic matters to underestimate the positive impact that a cordial welcome can have on students. Faculties and Departments where particular attention was paid to providing a friendly environment (Classics, Law, Geography, and some of the smaller Departments in Modern and Medieval Languages were exemplary in this matter) reaped benefits from the enthusiasm, energy, and commitment that this evoked in students. Faculties and Departments which were able to offer a student common room also benefited from informal networking among students which could increase excitement about work in progress and diminish anxieties that were otherwise quietly festering.

Colleges have customarily provided the kind of welcome that students appreciate from their new 'home'. A useful practice that is followed by some but not all Colleges (or subjects within Colleges) is to set up 'College parent' systems in which students within disciplines look out for new recruits. This is particularly important for subjects where the year group is small and also for subjects where some students have minority status (such as women engineers).

7. Directors of Studies and Supervisors

The qualitative data in this study brought out very clearly the powerful role played by Directors of Studies and Supervisors in the Colleges in the academic success of students. A key finding of this study is that women thrive in teaching situations where they are given constructive feedback. As they are seeking understanding they value feedback about their progress, clarity about what is expected from them and clear advice about how to improve. They need reassurance that they are understanding the subject well, and if they get it, their confidence, ambition, and sense of fulfilment grows. Finally, many women who achieve excellent grades in both arts and sciences reach a point where they realize that to achieve a first class performance they will need to 'work for the exams and not for myself'. As this may demand a complete change of tactic from previous working patterns, they appreciate Supervisors who will help them focus their energies. Women who benefited from good relationships with staff who understood the ways in which they approached their degree work, and who neither underestimated nor patronized them, reported feeling able to fulfil their potential. For many, relationships with supportive and encouraging members of teaching staff were the strongest indicators of academic success. At its best, the Cambridge ideal is not only about playing to the strengths of those with clear potential for excellence, but also seeking ways to help students who are 'stuck' in their work to move with confidence to a stage where they can exercise creative, independent thought and will have the courage to challenge both 'teacher and text'. It is clear from this study that many Supervisors and Directors of Studies help their students achieve this.

Given that being a recipient of skilled supervision is such a clear indicator of success, a useful way forward would be to make some training for Supervisors a general requirement. In any voluntary system it is most likely to be those individuals who are already most aware of the responsibilities and possibilities of the role who take advantage of the courses for Supervisors made available by the University.

Conclusions

Dr Mann's discussion has speculated that women may lose ground in Tripos examinations, in the minority of Faculties and Departments where there is a large gender gap, because they have either been unable to work in the way they would have preferred throughout the degree or because the way they have chosen to work is not one that is rewarded in examinations. It is these circumstances rather than any difference in ability which may depress their results. However, Dr Mann drew further implications from the suggestion that women tend to see the Tripos as an opportunity to increase their personal understanding, while men are more inclined to tailor their intellectual development to public success. As we have seen, if these differences in approach remain beneath the surface, men and women may misunderstand each other in learning contexts and women may be misunderstood by staff who themselves may be more or less instrumental in their approach to Tripos examinations. It is also clear that in these Faculties and Departments some of the students felt there was little consensus about what constituted 'excellence', and consequently they were unsure about what is expected of them.

In Faculties and Departments where there is no statistically significant gender gap or where women do better than men, which in this study is the majority of those Faculties and Departments studied, Dr Mann identified many instances of good practice which seem to benefit all students but, in particular, women. These include a match between student motivation and methods of learning with the teaching approach within the subject; a recognition of students' prior experiences which is reflected in both the content and the approach to teaching in the first year; a welcoming environment which makes the transition from school to university easier; a clear understanding on the part of the students about what is expected of them and what constitutes 'excellence'; and clear and constructive feedback through the supervision system and help with focusing one's efforts. Under these circumstances students appear to be able to combine 'love of subject' and the desire for understanding with a sufficiently instrumental approach to produce a good examination performance. It is of interest to note that in Physics, Chemistry, and, in particular, Engineering, where previously men had outperformed women, these good practices have produced the most significant dividends.

APPENDIX

Part II Degree results for those graduating in 2001 and 2002 - Gender Differences

The degree results for those graduating in 2001 and 2002 are given in the table below, for those Part II examinations that have been the subject of this research project. The results have been condensed to show the number and percentage of women and men who gained first class degrees, while the second cell combines all other classes of degree, again giving the number and percentage for both women and men. Scrutiny of this table shows that men have got more firsts than women in some examinations and that women have done better than men in others. The table also illustrates the difficulties involved in interpreting these results.

Clearly looking just at the actual numbers in each category is misleading because of the very different numbers of women and men taking the different examinations. In English Part II in 2002 there are 137 women but only 80 men, the position is reversed in Mathematics with 157 men taking the Part II compared with only 35 women. Using the percentage of men and women who gain a first does, to some extent, alleviate this difficulty and does provide a more accurate picture. This method of representing the data, however, also has problems, particularly where the numbers in the sample are relatively small and the number of firsts awarded is low. This is clearly illustrated by the Manufacturing Engineering Tripos Part II results for 2002. There appears to be a large discrepancy in the results for men and women with 22.9% of men getting a first compared with only 11.1% of women. However, if one more woman had got a first this would have raised the percentage of women getting a first to 22.2 making the percentage of men and women who got firsts equal. Thus the results of just one person can, in some cases, make a very big difference to the outcome.

It is for these sorts of reasons that tests of statistical significance are used to try to determine whether or not the observed variation in performance is greater than could be expected. The technique most commonly used with this type of data is the chi-square test for goodness of fit. What this test does is to compare the actual number of women and men getting firsts (the observed frequency) with the number that you would expect to get firsts, given the number in the sample and based on the assumption that women and men are equally likely to get a first class degree (the expected frequency). This was the test which was done on the 2000 examination, the results of which have already been discussed.

Results of the Analysis

Carrying out the chi-square test on the results for 2001 and 2002 shows that in only one case does the observed gender differences significantly differ from the expected and that is for the English Tripos Part II for 2001 where men did perform significantly better than expected and women significantly worse than expected (chi squared = 9.38 p< 0.01).

Although women performed slightly better than men in some cases (Electrical and Information Sciences Part II in 2001 and 2002, Manufacturing Engineering Part II in 2001, Law Part II in 2001, Classics Part II in 2001 and 2002, Chemical Engineering Parts IIA and IIB in 2002) and men in others (Engineering Parts IIA and IIB in 2001, English Part II in 2002, Modern and Medieval Languages Part II in 2001 and 2002, Geography Part II in 2001 and 2002, Law Part II in 2002, Mathematics Part II in 2001 and 2002, Chemistry in 2001 and 2002, Physics Part II 2001), none of these differences are statistically significant. In the remaining examinations (Chemical Engineering Parts IIA and IIB in 2001, Physics Part II in 2002, Engineering Parts IIA and IIB in 2002, Manufacturing Engineering Part II in 2002) both men and women performed as expected.

In summary, therefore, in 2001 there was one examination where the gender differences in the number of firsts awarded was significant and that was the English Tripos Part II, where men did better than women. In 2002 there were no significant gender differences in the number of firsts awarded to women and men.

Conclusion

The crucial question that needs to be answered is what can reasonably be deduced from looking at examination results in the target Faculties and Departments over the period 1990 to 2002 as regards gender differences. Clearly there are fluctuations in results from year to year both within the same Tripos, as for example, with the Law Tripos between 2001, where women performed slightly, but not significantly, better than men, and 2002, where men performed slightly, but not significantly, better than women, and between Triposes. Taking this into account, however, one tentative conclusion is that over that period of time the trend is one of increasing equality in performance, with women closing the gender gap in those areas where, at the beginning of the 1990s, they performed significantly less well than men. Further monitoring will obviously be needed to see if this is, in fact, the case.

The fact that the 'gender gap' in the numbers of women and men awarded first class degrees may have closed in the period 1990 to 2002 in no way detracts from the importance of the findings of this research project. The aim of the University is that all students should flourish in the learning environment provided by the various Faculties and Departments and should all be given the opportunity to do as well, academically, as they can. The great strength of the research reported here, particularly the in-depth study carried out by Dr Chris Mann, is that it has indicated the factors that are likely to enhance students' learning and allow them to be academically successful while at the same time highlighting those areas in which there is need for improvement if all students are going to thrive in the learning environment and reach their potential.

The great strength of the research reported here, looking at the work of both Dr Leman and Dr Mann is that it has identified the significant issues we need to consider in looking at students' academic performance. It is clear that the major influences on academic performance are not related to the students' previous experience in terms of schooling or social class, but the teaching and learning environment provided by Faculties, Departments, and Colleges. The sample used in the cohort study was not a representative sample of all students in the Faculties and Departments studied, they were those who volunteered to participate; nor were all the Faculties and Departments in the University included, so how generalizable the findings of Dr Mann's study are to other students, and to other Faculties and Departments is hard to quantify. However, in the traditions of grounded theory, what Dr Mann's in depth study has suggested are theories that require further investigation. The students in the study have drawn attention to those areas within the learning environment in which there is a need for improvement if they are going to thrive and reach their potential. They have also identified many aspects of the teaching and learning environment with the University and the Colleges which they believe have enhanced and supported their learning and allowed them to be academically successful.

Degree results for 2001-02

 2001

2002
Subject Firsts Men
%
Other Men
%
Firsts Women
%
Other Women
%
Firsts Men
%
Other Men
%
Firsts Women
%
Other Women
%
English Tripos Part II 2331.55068.51813.611486.41721.36378.72316.811483.2
Modern and Medieval
Languages Tripos Part II
1728.34371.72220.88479.2173433662423.87776.2
Classics Tripos Part II 516.12683.9824.22575.8718.93081.1825.82374.2
Geography Tripos Part II 1223.53976.549.14090.9923.13076.9712.74887.3
Law Tripos Part II 1212.88287.22519.110680.91817.18782.91512.110987.9
Mathematics Tripos Part II 5834.910865.1720.62779.4553510265822.92777.1
Natural Sciences Tripos -
Chemistry Part II
2036.43563.6826.72273.31535.72764.3726.91973.1
Natural Sciences Tripos -
Physics Part II
4640696052515753431.87368.2932.11967.9
Engineering Tripos Part IIA
and Part IIB
3518.815181.2610.55189.54423.314576.7622.22177.8
Manufacturing Engineering
Tripos Part II
411.83088.2222.2777.8822.92777.1111.1888.9
Electrical and Information
Sciences Tripos Part II
1521.72478.3436.4763.61418.76181.3428.61071.4
Chemical Engineering Tripos
Part IIA and Part IIB
13273573827.52172.51022.23577.8827.52172.5

1 This information is supplied by all students on entry to the University and forms part of their student record.

2 Social class was classified according to the Registrar General's Classification of Occupations as updated by the ESRC Review of Government Social Classifications (1998) and ethnicity by using the classification approved by the Commission for Racial Equality.

3 'Other' includes Sixth Form colleges and centres, tertiary and further education colleges, and grant maintained special schools.

4 A fuller description of these findings is given in P. Leman (1999), 'The role of subject area, gender, ethnicity and school background in the degree results of Cambridge University undergraduates', The Curriculum Journal 10, pp. 231-252. Please note however, that there are mistakes in Table 4. The percentages given for black students are correct but the numbers are in the wrong order. The numbers should read:- 1st 1, 2i 19, 2ii 7, 3rd 5. The total number on 'Other Asian' students should read 123.

5 Glaser, B.C. and Strauss, A.L. (1967) The Discovery of Grounded Theory: Strategies for Qualitative Research. Aldine: Chicago.

6 Patton, M. Q. (1980) Qualitative Evaluation Methods. Sage: London.

7 Mann, C. and Stewart, F. (2000) Internet Communication and Qualitative Research: a Handbook for Researching Online. Sage: London and Thousand Oaks.

8 Within the psychological literature similar sex differences have been found at all levels of schooling. See for example Marks H. M. (2000) 'Student engagement in instructional activity: patterns in elementary, middle and high school students'. American Research Journal 37 (1) pp. 153-184; Whitehead, J. M. (1984) 'Motives for higher education: a study of intrinsic and extrinsic motivation in relation to academic attainment'. Cambridge Journal of Education 14(2) pp. 26-34; Whitehead, J. M. (1994) 'Academically successful schoolgirls: a case of sex-role transcendence'. Research Papers in Education 9(1) pp. 53-79.

9 N. G. McCrum (1996), 'Gender and social inequality at Oxford and Cambridge Universities', Oxford Review of Education 22, pp. 369-397, and M. Martin (1997), 'Emotional and cognitive effects of examination proximity in female and male students', Oxford Review of Education 23, pp. 479-486.

10 S. Clarke (1988), 'Another look at the degree results of men and women', Studies in Higher Education 13, pp. 315-331.

11 Cited in Martin (1997): 480 and in McCrum (1996): 372.

12 The Faculty of Modern and Medieval Languages' Working Party Report on Gender and Academic Performance, 1994.


< Previous page ^ Table of Contents Next page >

Cambridge University Reporter, 12 February 2003
Copyright © 2002 The Chancellor, Masters and Scholars of the University of Cambridge.