Victorian Certificate of Education: mathematics, science and gender.

Forgasz, Helen J.

Gender differences in participation and performance at 'high stakes' examinations have received much public attention, which has often focused on mathematics and science subjects. This paper describes the innovative forms of assessment introduced into mathematics and science subjects within the Victorian Certificate of Education (VCE) system. Results from these subjects are examined for patterns of gender differences in participation and performance over the period 1994-1999. A larger proportion of males than females studied all the VCE science and mathematics subjects except Biology and Psychology over this period. Based on study scores, females, on average, out-performed males in almost all VCE science and mathematics subjects in nearly every year from 1994-1999. As exceptions to the patterns, males out-performed females in Chemistry and Mathematical Methods. Results from a general ability test are used to question the legitimacy of gender comparisons in subjects in which enrolment is no longer compulsory. The data do not support simplistic conclusions about gender differences in participation and performance.

Keywords

academic achievement

gender issues

mathematics

sciences

sex differences

student participation

Introduction

Historically mathematics has been viewed as the preserve of white, middle-class males. However, over the past three decades in particular, there have been stringent efforts in many different countries to re-dress this perception. Intervention programs aimed at improving female participation rates and attaining equity in levels of achievement have flourished and, to some extent, succeeded (Leder, Forgasz, & Solar, 1996). In Australia, achieving gender equity has been a high priority. To this end, legislation has been put in place to deal with discriminatory practices in fields as diverse as education, the law, employment, and welfare. State and federal governments have published reports on girls' education that have identified specific school-related factors linked to the perpetuation of inequities (e.g. Commonwealth Schools Commission, 1975; Ministerial Advisory Committee on Women and Girls, 1991; Ministry of Education Western Australia, 1991). More recently, concerns have been expressed about problems experienced by boys (e.g. House of Representatives Standing Committee on Education and Training, 2002; O'Doherty, 1994). Again this trend is not unique to Australia (for the United Kingdom---see, for example, Warrington & Younger, 2000; Weiner, Arnot, & David, 1997; for the United States--see Kimmel, 2000; more generally, see the Organisation for Economic Cooperation and Development, 2001). Further explorations of gender differences in educational performance thus seem warranted. This paper draws on a range of data from the Victorian Certificate of Education (VCE) to examine patterns of gender differences in mathematics and science subjects.

Australian context

The Blackburn report (Ministerial Review, 1985) served as a focal point in Victoria for questioning curriculum content and assessment approaches, particularly in the postcompulsory years of schooling. In other states across Australia, similar issues were being raised. During the 1980s and 1990s, widespread course structure changes were introduced at the upper secondary school level, in part to cater for the different needs of an increasingly diverse student population remaining at school to Year 12.

Male and female enrolments in Australian upper secondary schools have increased substantially from 1970 to the present. Since 1976, there has been a greater percentage of females than males in Year 12--the final year of schooling (Allen & Bell, 1996; Cortis & Newmarch, 2000; Dekkers, De Laeter, & Malone, 1991), with the gap between the proportional participation of males and females increasing to approximately 10 per cent in 1998 (Marks, Fleming, Long, & McMillan, 2000). Year 12 retention rates in Australia in 1999 were reported to be 66.4 per cent for males and 78.5 per cent for females (Collins, Kenway, & McLeod, 2000). However, this gap of over 10 per cent in male and female enrolments in Year 12 has been shown to decrease if the numbers include overall participation in education--that is, if students enrolled in the Vocational Education and Training (VET) areas are included (Cortis & Newmarch, 2000).

At the Year 12 level the overall mathematics enrollment trends Australia-wide have been well documented (Dekkers et al., 1991; Dekkers & Malone, 2000; Lydeamore, 1993; Teese, 1994). These researchers have reported that more students, both males and females, have remained at school to complete Year 12, and that the proportion of students enrolling in mathematics was keeping pace with the increasing enrolments in postcompulsory schooling (Teese, 1994). However Dobson and Calderon (1999), who studied Australia-wide science and mathematics enrolments at Year 12, have reported a proportional decline in science enrolments in Australia. They found that although the proportion of Year 12 students enrolled in mathematics has increased slightly from 17.9 per cent in 1989 to 18.1 per cent in 1997, corresponding figures for science enrolments have fallen from 20.5 per cent in 1989 to 17.1 per cent in 1997. They have also shown that, over this period, proportional and absolute enrolment numbers in Biology, Chemistry, Geology, Physics and Science at Year 12 have declined, with Psychology the only science subject going against this trend (Dobson & Calderon, 1999). Dekkers and De Laeter (1997, 2001) have published similar findings: declining proportional enrolments in Chemistry, Biology and Physics, and increasing proportional enrolments in alternative science subjects (e.g. General Science, Environmental Science, Science for Life and Psychology) in which 64 per cent of the students were female. In their report, Who is studying science, the Australian Council of Deans of Science (1999) stated that 'at the very least, these statistics raise real questions about the adequacy and support within the secondary school system for science and mathematics teaching and careers geared towards high technology' (p. 15).

Dobson and Calderon (1999) noted that the percentage of females enrolling in the areas of science and mathematics has increased over time. They reported that there was a greater percentage of females than males enrolled in Biology (65 per cent and 35 per cent respectively) and Psychology (79 per cent and 21 per cent respectively), an almost equivalent percentage of females and males in Chemistry (49 per cent and 51 per cent respectively), but a much smaller percentage of females than males enrolled in Physics (30 per cent and 70 per cent respectively) and in Geology (37 per cent and 63 per cent respectively). Similar results have been reported by Allen and Bell (1996) in Queensland, Keightley (1999) in South Australia, and MacCann (1995) in New South Wales. Internationally, similar findings of greater proportions of females than males enrolled in Biology, similar proportions in Chemistry, and a greater proportion of males than females in Physics have been reported by Stobart, Elwood, and Quinlan (1992) for GCSE students in the United Kingdom, by Haggerty (1991) for Grade 12 students in Canada, and by Bae, Choy, Geddes, Sable, and Snyder (2000) for senior high school students in the United States.

When considering students enrolled in double mathematics courses (strongly recommended for specialised entry into tertiary mathematics, engineering and science studies), an imbalance in favour of male enrolments has been widely reported (Allen & Bell, 1996; Brinkworth, 1999; Dekkers & Malone, 2000; Lydeamore, 1993; MacCann, 1995; Teese, Davies, Charlton, & Polesel, 1995). Proportions of 30-35 per cent females in double mathematics courses were reported across all Australian states. The increase in female enrolments in mathematics, from 44.8 per cent in 1980 to 49.6 per cent in 1999, has been attributed to the increase in the non-specialist mathematics subjects being offered (Allen & Bell, 1996; Dekkers et al., 1991; Dekkers & Malone, 2000; Lydeamore, 1993; Teese et al., 1995). Theoretical and practical justifications were put forward for introducing changes in assessment practices in senior secondary science, mathematics and most other postcompulsory subjects (see e.g. Lydeamore, 1993). Following these changes, females are now often out-performing males particularly on school-based assessment that requires a greater written component (Cox, 2000; Forgasz & Leder, 2001; Lydeamore, 1993; Whitehouse & Sullivan, 1992). It has also been argued that the 'performance of girls in science examinations can be enhanced or disadvantaged according to the way the examination and its parts are constructed' (Whitehouse & Sullivan, 1992, p. 63). In a recent report, Boys: Getting it right (House of Representatives Standing Committee, 2002), it was suggested that 'the increasing literacy demands of the senior curriculum and assessment ... have been a factor in boys' declining relative performance' (pp. 21-22). Thus it appears that for both males and females performance is thought to be affected by the form of assessment used to measure achievement.

An examination of the statistics of the Year 12 results of males and females has shown that males are being out-performed by females in almost all subjects throughout Australia (Collins et al., 2000). Not only do males achieve somewhat lower average scores than females but, in most cases, the male standard deviations are higher than those of females which indicates a greater spread of male scores. We maintain, however, that comparisons of males' and females' performance should take account not only of group means and standard deviations but also of enrolment data (Cox, 1996; Rowley, Brew, & Leder, 1997).

Victorian Certificate of Education (VCE)

The VCE, a two-year program for the postcompulsory levels of schooling (Years 11 and 12), was fully introduced in 1992, replacing the one-year Higher School Certificate (HSC). A wider choice of subjects was offered to students and a broader range of assessment strategies was introduced into all VCE science and mathematics subjects to assess the performance of Years 11 and 12 students. The different assessment types included research tasks, oral communication, working independently on extended problems, and traditional examinations. The changes introduced in Victoria were mirrored in other states. The remainder of this article focuses particularly on males' and females' performance in the VCE mathematics and science subjects.

In general, Year 11 students take subjects at the Unit 1 and 2 levels (semester length units taken sequentially), and Year 12 students study Unit 3 and 4 level subjects. The VCE allows flexibility, and more capable students often undertake a Unit 3 and 4 subject in their first year of the VCE (Grade 11), and some Year 12 students take a university level mathematics subject.

The introduction of the VCE was controversial. By 1994, some changes had been made. In 2000, more radical changes were introduced (see Table 1 for an overview of the period 1992-1999).

This paper focuses on statistical data between 1994 and 1999, because of the relative stability of the assessment procedures during that time. Five science subjects and three mathematics subjects were offered over that period. Each was divided into two half-year units, namely:

* Biology Units 3 and 4 (Biology),

* Chemistry Units 3 and 4 (Chemistry),

* Physics Units 3 and 4 (Physics),

* Psychology Units 3 and 4 (Psychology),

* Science Units 3 and 4 (Science),

* Further Mathematics Units 3 and 4 (Further Mathematics),

* Mathematical Methods Units 3 and 4 (Mathematical Methods), and

* Specialist Mathematics Units 3 and 4 (Specialist Mathematics).

Between 1994 and 1999, Unit 1 and 2 subjects were assessed internally by schools and the grades for these subjects were not reported outside the school, although an S (satisfactory) or N (not satisfactory) was reported for inclusion on students' VCE certificates. At the Unit 3 and 4 level, students were assessed in each subject with a series of Common Assessment Tasks (CATs) in a variety of forms including reports, analysis tasks, projects and examinations.

Since students' VCE results were used for university selection, a balance was needed between the teachers' assessments of students' performance on school-assessed CATs and the need for statewide consistency of marking. In 1994, a new procedure was introduced to replace statewide sampling for moderation purposes. A General Achievement Test (GAT) was introduced by the Board of Studies, Victoria (now known as the Victorian Curriculum and Assessment Authority, VCAA), to be taken by all students taking Unit 3 and 4 subjects. The GAT involved examining students' achievements in three major areas--Written Communication, Mathematics/Science/Technology, and Arts/Humanities/Social Science. The GAT results provided an achievement profile for groups of students who studied each subject in each school. If school-assessed CAT scores for each subject did not match the GAT profile of the school (within a certain tolerance level) for that subject, then students' internally assessed CAT papers were reviewed by the Board of Studies. This procedure did not apply to the examination CATs which were assessed independently by external examiners.

The change in assessment methods that accompanied the introduction of the VCE in 1992 sparked much research interest in gender differences associated with different forms of achievement at Year 12 level. A substantial amount of work has been completed and published (Cox, 1996, 2000; Forgasz & Leder, 2001; Rowley et al., 1997; Teese et al., 1995). In all subjects, the general finding was that females performed better than males on the internally assessed CATs, whereas the reverse was true on the more traditional examination CATs. Rowley et al. (1997) further reported that, in mathematics, it appeared that the most significant differences between males and females were explained by students' subject choices. The researchers modelled different subject selections and combinations. As the subject selections of males and females were modelled to become more similar, the gender differences in performance became smaller or were reversed.

Present study

As argued earlier, both performance and participation data need to be considered when examining gender differences. Two sources of VCE data were used in the present study: published summary statistics and a database specifically requested and obtained from the VCAA. The latter involved gender disaggregated information for each VCE Unit 3 and 4 subject that is not publicly available including: mean (raw) scores for each CAT, and mean 'study scores' (the terminology is defined below) and standard deviations. These sources were used to probe conflicting views about gender-related educational disadvantages. This investigation is restricted to the traditional male-dominated fields of mathematics and science.

Participation and performance statistics of VCE students in science and mathematics subjects 1994-1999 The results of our explorations of the participation and performance statistics for the VCE science and mathematics subjects by gender for the period 1994-1999 have been divided into three sections: the participation statistics, the performance statistics, and the standard deviation statistics.

Participation statistics of students in VCE science and mathematics subjects 1994-1999 The percentages of females enrolled in English and all VCE science and mathematics subjects for the period 1994-1999 are presented in Figure 1. English (Units 3 and 4) was a compulsory subject for all Year 12 students enrolled in the VCE during this period. Hence English enrolments are representative of all VCE enrolments.

[FIGURE 1 OMITTED]

Throughout the period 1994-1999, the percentage of female VCE students has been constant at about 54 per cent (see English on Figure 1). It can also be seen that no VCE mathematics subject has a percentage of female students that matches that of the English cohorts. That is, there are fewer females than males studying all Unit 3 and 4 mathematics subjects than would be expected, based on the overall population percentages of females compared with males studying the VCE. For Physics and Science, the female enrolment patterns are similar to the mathematics subjects. However, for Chemistry, there is almost the expected percentage of females (about 54 per cent), and in Biology and Psychology, there is a much larger percentage of female students enrolled than would be expected--that is, greater than 54 per cent in each. The three most popular science subjects form an interesting and contrasting triptych, from Physics with fewer than expected females enrolled, to Chemistry with the expected proportion of female enrolments, to Biology with more than the expected proportion of female enrolments.

Changes in the percentage of females in each of the subjects over the period can also be seen in Figure 1. There has been a slight increase in the percentage of females studying the three mathematics subjects, and an even smaller increase in Chemistry, Biology and Physics. The percentage enrolments in Psychology have stayed fairly constant. However the data in Figure 1 ignore enrolment figures. This information is provided in Figure 2 where enrolments in mathematics and science subjects are expressed as percentages of Year 12 male and female cohorts.

[FIGURE 2 OMITTED]

From the data shown in Figure 2, it appears that the overall percentage participation of VCE students in Biology, Science and Chemistry declined over the period 1994-1999, that there was a very slight increase in the percentage participation of students in Physics, and that there was a larger increase in the percentage participation of students in Psychology. In all subjects, except Chemistry and Science, the trends were similar for males and females. In Chemistry and Science, the decline in percentage enrolments by males was greater than the decline in percentage enrolment trend data over time for females. For Psychology, the changes in trend data are more marked for females than for males. It is possible that the decline in the participation of students in the science subjects, Biology, Science and Chemistry could be caused by the matching increase in Psychology participation rates. Psychology is classified as a science subject and can be taken as the science requirement of the VCE. (Students are required to enrol in at least two mathematics, science or technology (half year length) units.) Thus it is possible that the increase in the percentage participation in Psychology could be draining Chemistry and Biology of students. The declining enrolment trend in these science subjects is consistent with the findings of Dobson and Calderon (1999) and Dekkers and De Laeter (1997, 2001) who used Australia-wide data. The two most difficult mathematics subjects (Specialist Mathematics and Mathematical Methods) appear to be on a three-year decline of proportional participation, while Further Mathematics is on a four-year trend in terms of increasing proportional participation.

The data shown in Figure 2 clearly reveal the trends regarding the 'science triptych' discussed earlier. From Biology, to Chemistry to Physics, the ratio of females to males is much higher in Biology (2:1), almost equal in Chemistry (1:1) and much lower in Physics (1:3). In Science, there is a lower ratio of females to males (approximately 1:2) and in Psychology, there is a much higher ratio (3:1) of females to males.

In all three mathematics subjects, there is a higher proportion of males than of females. The ratio of females to males changes along the 'degree of difficulty continuum' from approximately 1:1 in the least difficult subject (Further Mathematics) to 3:4 in Mathematical Methods, and finally 1:2 in the most difficult subject, Specialist Mathematics.

Performance statistics of students in VCE science and mathematics subjects 1994-1999 In each VCE subject during the period 1994-1999, students were graded on a number of individual Common Assessment Tasks (CATs). These CAT scores were then combined to form an overall 'study score'. The study scores were statistically adjusted so that, for each VCE (Unit 3 and 4 level) subject, the maximum score was 50, the mean was close to 30 and there was a standard deviation of approximately 7. (Larger variations from the figures cited occur in subjects with relatively small cohorts of students. In the present study, the only subject affected in this way was Science Units 3 and 4.) The study scores were then used to produce tertiary entrance scores (currently known as ENTERs) and used for selection into various university courses.

The mean study scores and standard deviations, and mean raw scores for each CAT, disaggregated by gender, are not made available to the public. These data sets were obtained from the Board of Studies, Victoria. In Table 2, the performance data for all the VCE mathematics and science subjects from 1994-1999 are summarised. For consistency of comparisons, the mean CAT scores are presented as percentages, even though the maximum possible score on these various CATs varied greatly. The mean study scores have been left as scores out of 50.

From Table 2, it is difficult to get a clear picture of the trends. Consequently the performance data have been summarised in a series of difference graphs. A difference graph displays the relative achievements of males and females and is calculated from the difference between the mean study scores for males and for females. Thus a bar above the horizontal axis represents males performing better than females, and a bar below the axis represents females performing better than males. To examine the overall subject performance of males and females in each subject over the years 1994-1999, a difference graph for the study scores for each of the science and mathematics subjects is presented in Figure 3.

[FIGURE 3 OMITTED]

From Figure 3, it can be seen that females are out-performing males on mean study scores in almost all VCE science and mathematics subjects, in nearly all of the years from 1994-1999. The only two subjects in which there were exceptions, and the years in which males out-performed the females on mean study scores, were Chemistry (1995-1999) and Mathematical Methods (1995-1998). These findings are consistent with current VCE and Australia-wide data which reveal that females are out-performing males in almost all subjects at the Year 12 level (Collins et al., 2000).

A closer examination of the male and female CAT scores within each VCE subject, shown in Table 2, was carried out. Difference graphs were produced and are presented in Figure 4. (There was a CAT 4 for Physics, Psychology and Science in 1994 only--refer to Table 1--and these data have also been plotted to show the gender difference on this form of assessment.)

[FIGURE 4 OMITTED]

The performance data for the CATs within the triptych of Biology, Chemistry and Physics, with their varied participation proportions of males to females, show a very similar pattern to Further Mathematics, Mathematical Methods and Specialist Mathematics respectively. In Biology and Further Mathematics, females obtained higher mean scores than males on all CATs. However, within each of these subjects, the males performed relatively better on the examination CATs (CATs 1 and 3 in Biology, and CATs 2 and 3 in Further Mathematics) than on the school-assessed written tasks (CAT 2 in Biology, and CAT 1 in Further Mathematics). A similar pattern emerged in Physics, Psychology and Specialist Mathematics. Females obtained higher mean scores than males on all CATs. Again males performed relatively better on the examination CATs than on the written tasks (this pattern is more difficult to see for Physics because the CAT numbers for examinations and written CATs changed within the time frame 1994 to 1999--refer to Table 1). Science had a very small total cohort compared with the other subjects, but the same performance trends by gender were present for this subject. Chemistry, with its relatively equal proportions of males and females, shows a very similar pattern to Mathematical Methods. Males had higher mean scores than females on examination CATs (CATs 1 and 3 in Chemistry and CATs 2 and 3 in Mathematical Methods) and females had higher mean scores than males on the written tasks (CAT 2 in Chemistry and CAT 1 in Mathematical Methods).

A summary of gender patterns in performance, based on overall study scores and individual CAT scores for each subject, is presented in Table 3. What is shown is the number of years over the period 1994-1999 in which males and females achieved the higher mean CAT scores and higher overall study scores.

Based on the data presented in Table 3 and in the other figures in this section of the paper, it can be seen that, over the period 1994-1999, females were outperforming males in most forms of assessment in most of the science and mathematics subjects in the VCE. These findings are inconsistent with previous research on gendered patterns of performance in these subject areas in timed tests taken under examination conditions, often using multiple-choice formats. They are consistent, however, with findings based on other, more open-ended, or written forms of assessment, and those that are classroom based (e.g. Kimball, 1995). The data examined in this study provide further evidence that it is the form of assessment used that can influence which group will have the higher mean performance score. In the next section, the implications of gender-related differences in the standard deviations of the VCE performance measures are explored.

Standard deviation statistics of students in VCE science and mathematics subjects 1994-1999 According to Feingold (1992) 'the possibility that the sexes may differ in variability ... has been almost completely ignored by gender researchers' (p. 62). In this study, we have used the standard deviation as a measure of the spread of the male and female distributions of study scores. Differences in the standard deviations of scores for males and females in each of the VCE science and mathematics subjects for the years 1994-1999 are shown in Figure 5. For ease of presentation, the female standard deviation score has been subtracted from the male standard deviation score in each VCE Unit 3 and 4 science and mathematics subject for each year.

[FIGURE 5 OMITTED]

In Figure 5, it can be seen that the males have a higher standard deviation than the females for the study score distributions in all subjects over the period considered, except for Science in 1997 and 1998. Thus more males than females are likely to be found at the two extremes of the study score distributions. Similar findings have been widely, but not invariably, reported in the literature for well over a century (e.g. Ellis, 1894; Feingold, 1992; Hedges & Friedman, 1993; Maccoby & Jacklin, 1974; Terman, 1925; Thorndike, 1906). Reasons for the greater male variability have been widely debated and have been variously attributed--and not without challenge--to a greater intellectual variability among males, biological factors, social factors, and test artefacts (see Feingold, 1992).

Another angle

Based on the statistics and analyses above, it may seem appropriate to conclude that males are lagging behind females in mathematics and science achievement in the VCE. Some of those who are asking the question--what about the boys?--base their arguments on these kinds of statistical information. However it is important when making such conclusions to be certain that the gender comparisons being made are based on 'like' groups of males and females. Once students enter post-compulsory schooling in Australia, as in most other countries, they can select which subjects they wish to study. From the gender differences in subject choices, it can be inferred that the groups of males and females being compared are dissimilar. Accordingly no more can be concluded than to say that the groups of females who select most mathematics and science subjects are, on average, achieving better results than the groups of males who select the same subjects.

In postcompulsory schooling, it is very difficult to match male and female students on prior achievement levels, for example. However, since 1994, all Victorian Year 12 students complete the General Achievement Test (GAT) (described earlier in this paper). The male and female summary statistics for each component of the GAT for 1999, 2000 and 2001 are presented in Table 4. Also included in Table 4 are calculated values of Cohen's effect size d: d = ([M.sub.1] - [M.sub.2]) / [[sigma].sub.pooled] (Cohen, 1988), a measure of how much overlap there is in the two distributions-male and female in this case--and a measure of the extent of the gender differences in the mean scores. An effect size (ES) of 0.3, medium by Cohen's (1988) definition, indicates that there is about 21 per cent non-overlap in the distributions, ES = 0.1 (small) about 8 per cent non-overlap, and ES = 0 means total overlap of the distributions (Becker, n.d.).

The data in Table 4 indicate that the gender differences in the mean scores for the Written Communication and Maths/Science/Technology components of the GAT in each year are both moderate and that there are considerable differences in the male and female distributions of the scores (ES [approximately equal to] 0.3). Interestingly the gender differences are in the traditional, stereotyped directions: females higher on the Written Communication component and males higher on the Maths/Science/Technology component. For the Arts/Humanities/Social Science component, males and females perform equally. When the mean scores for the three components of the GAT are combined into a single overall mean GAT score, the female and the male means in all three years are almost equal (see Table 5). In other words, the gender differences in the Written Communication and Maths/Science/Technology components nullify each other when summed. Collectively these data appear to challenge the arguments that males are underachieving compared with females. More needs to be known about the GAT and explanations found for the performance levels of males and females overall, and in the three specific components.

There are two interesting inferences to be drawn from the GAT data. First, this form of testing appears to produce results that are consistent with traditional patterns of male and female performance expectations in English--females outperform males--and in the Mathematics/Science/Technology fields--males outperform females. Second, the direction of the gender difference in the Mathematics/Science/Technology component of the GAT (males higher than females) is at variance with the VCE study scores in the mathematics and science subjects. Yet the two sets of scores--GAT and VCE study scores--are derived from the performances of the same cohorts of students. How can this inconsistency be reconciled? Why, too, are the GAT data used to 'verify and moderate' the school-based assessments--areas in which research evidence consistently reveals that females outperform males. Who is favoured by this method of verification and moderation? Questions such as these point to the influence of the social and political contexts in which schooling and assessment occur.

Conclusions

A larger proportion of males than of females studied all the VCE science and mathematics subjects except Biology and Psychology over the period 1994-1999. However, based on study scores, females, on average, out-performed males in almost all VCE science and mathematics subjects in nearly every year. There were only two subjects for which there were exceptions to the patterns: males out-performed females in Chemistry in 1995-1999, and in Mathematical Methods in 1995-1998.

In general males perform relatively better on examination-style assessment components in all subjects. In Chemistry and in Mathematical Methods, males actually out-performed females in the examinations. The distributions of males' study scores have larger standard deviations than those of females in all subjects in all years except for Science in 1997 and 1998. This means that the males' distributions are more spread out and more males than females are likely to be found at the two extremes of the pooled study score distributions.

Advocates of the 'what about the boys?' movement continue to use the VCE and similar Year 12 data from elsewhere to support arguments that males' school performance outcomes are in decline. If males are performing so poorly, why is it that their mean GAT scores, for example, are no different from females' scores? Clearly the form of assessment used makes a difference to the performance measures reported for the two groups--males and females. We would argue strongly that the broadened range of assessment forms used in the VCE should not be abandoned. What is needed, it seems, is to ensure that there is balance in the forms of assessment used and that each form is equally valued. With an increasingly diverse student population, a balance in the forms of assessment--and equally valuing these within subjects--allows students who are better at writing and discussion, prefer cooperative learning styles, and like examining human or social implications, to feel their interests and strengths do not exclude them from mathematics and science subjects. This may assist in encouraging students to study mathematics and science subjects and so broaden their subsequent study and career options to include these areas.

If differential male and female enrolment patterns in most of the VCE mathematics and science subjects were the explanation for females' superior performance over males (i.e. only more capable females take these subjects), then it could be argued that any comparison of male and female mean study scores in any noncompulsory subject has little meaning since dissimilar groups would be involved. The same could be said for comparing performances in compulsory subjects, as more females than males continue in mainstream schooling to complete Year 12. These comparisons will continue to be made. However extreme care is needed in their interpretation and those reporting the information should make clear the limitations of the data presented.

Table 1 Summary of the forms of, and modifications to, assessment
used in Unit 3 & 4 VCE science and mathematics subjects (1992-1999)

Subject Years CAT 1

Chemistry 1992-1999 Examination
Biology 1992-1993 Verified practical
 work
 1994-1999 Examination
Physics 1992-1994 School assessed
 practical work
 1995-1996 School assessed
 practical work
 1997-1999 Examination
Science 1992-1994 School assessed
 report
 1995-1999 School assessed
 report
Psychology 1992-1994 School assessed
 essay
 1995-1999 School assessed
 essay
All VCE
Mathematics 1991-1993 School assessed
subjects: investigative
 Space and Number, project
 Reasoning and Data,
 Change and Approximation
All VCE
Mathematics 1994-1999 School assessed
subjects: investigative
 Further Mathematics, project (or
 Mathematical Methods, problem) and
 Specialist Mathematics test

Subject CAT 2 CAT 3

Chemistry School assessed Examination
 report
Biology Examination School assessed
 investigation
 School assessed Examination
 investigation
Physics Examination School assessed
 report
 Examination Examination
 School assessed Examination
 prac
Science Examination School assessed
 report
 School assessed Examination
 report
Psychology Examination School assessed
 media appraisal
 Examination Examination
All VCE
Mathematics School assessed Multiple choice
subjects: challenging examination
 Space and Number, problem (CAT 2 in
 Reasoning and Data, (removed in 1993)
 Change and Approximation 1993
All VCE
Mathematics Multiple choice Extended
subjects: and short response
 Further Mathematics, answer examination
 Mathematical Methods, examination
 Specialist Mathematics

Subject CAT 4

Chemistry
Biology Examination
Physics Examination
Science Examination
Psychology Examination
All VCE
Mathematics Extended
subjects: response
 Space and Number, examination
 Reasoning and Data, (CAT 3 in
 Change and Approximation 1993
All VCE
Mathematics
subjects:
 Further Mathematics,
 Mathematical Methods,
 Specialist Mathematics

Table 2 Percentage CAT and Mean Study Scores for males and females in
all Unit 3 & 4 science and mathematics subjects 1994-1999

 CAT 1 (%) CAT 2 (%) CAT 3 (%)

Subject Year F M F M F M

Biology
 1994 44.8 43.4 77.5 72.5 35.9 35.3
 1995 43.0 42.9 69.6 65.7 46.5 43.9
 1996 39.2 39.4 68.5 64.3 46.4 44.6
 1997 49.7 49.7 69.1 65.8 43.9 42.9
 1998 48.4 47.9 70.1 66.0 43.8 42.4
 1999 45.3 45.2 70.9 67.5 40.3 38.5
Chemistry
 1994 57.9 58.9 79.6 76.2 59.0 59.4
 1995 57.5 60.7 73.8 71.5 55.3 57.9
 1996 61.5 62.9 74.6 72.4 51.8 53.9
 1997 61.8 63.4 75.9 73.4 45.6 48.5
 1998 55.7 57.3 75.8 73.6 54.3 56.1
 1999 58.8 60.6 77.1 74.5 59.1 61.2
Physics
 1994 80.2 73.6 44.0 42.0 81.7 74.5
 1995 75.1 68.6 42.9 40.6 51.6 47.5
 1996 75.7 69.3 44.8 44.7 46.5 43.4
 1997 48.9 47.7 75.4 69.3 50.8 46.9
 1998 58.2 55.6 76.3 69.5 58.6 56.2
 1999 69.7 66.4 77.0 71.0 61.3 57.1
Psychology
 1994 74.1 69.2 56.7 55.9 72.6 68.4
 1995 65.4 61.0 43.5 41.8 45.4 44.6
 1996 65.2 61.3 48.2 46.8 45.9 44.3
 1997 66.8 63.5 45.1 43.7 48.5 46.7
 1998 66.1 62.3 48.4 45.8 51.0 48.7
 1999 67.5 64.3 48.7 47.7 53.0 52.3
Science
 1994 66.1 56.1 34.4 32.7 72.9 64.2
 1995 62.4 53.4 63.1 55.4 50.4 53.1
 1996 70.0 56.3 68.5 53.3 48.9 42.2
 1997 71.6 57.7 67.1 56.9 57.6 52.5
 1998 69.2 58.2 70.3 64.2 49.4 45.4
 1999 66.4 55.6 67.2 60.3 55.2 58.5
Further Mathematics
 1994 61.5 54.2 38.8 38.1 33.2 32.9
 1995 57.6 50.7 51.6 50.0 37.4 37.9
 1996 58.4 50.0 52.2 50.0 44.5 42.6
 1997 59.9 51.2 42.6 41.4 37.0 33.6
 1998 61.7 52.6 48.6 46.1 41.6 40.6
 1999 64.6 56.4 53.5 51.7 38.2 36.6
Mathematical Methods
 1994 75.5 72.0 63.6 64.4 49.3 53.3
 1995 67.6 64.1 54.8 56.2 32.6 36.5
 1996 66.0 64.0 48.9 50.9 38.9 42.0
 1997 70.4 67.9 54.3 55.8 40.7 44.5
 1998 67.0 65.0 45.9 47.6 40.0 41.7
 1999 72.2 69.3 55.1 55.8 36.6 38.1
Specialist Mathematics
 1994 79.1 77.1 61.2 58.2 45.7 43.4
 1995 76.5 75.9 52.5 51.2 42.3 39.2
 1996 77.8 76.0 51.8 50.8 42.6 42.7
 1997 80.6 79.3 45.5 45.1 44.0 43.9
 1998 81.1 79.1 52.8 51.8 39.1 38.8
 1999 74.9 72.7 50.5 49.4 41.9 40.3

 Study score
 (150)

Subject Year F M

Biology
 1994 30.4 29.2
 1995 30.3 29.3
 1996 30.3 29.4
 1997 30.2 29.6
 1998 30.3 29.4
 1999 30.2 29.5
Chemistry
 1994 30.1 29.9
 1995 29.8 30.2
 1996 29.9 30.1
 1997 29.9 30.1
 1998 30.0 30.1
 1999 29.9 30.1
Physics
 1994 31.4 29.6
 1995 31.2 29.6
 1996 30.9 29.7
 1997 31.1 29.6
 1998 31.2 29.6
 1999 31.2 29.6
Psychology
 1994 30.5 28.4
 1995 30.2 29.1
 1996 30.2 29.2
 1997 30.2 29.3
 1998 30.3 29.0
 1999 30.2 29.4
Science
 1994 32.0 29.0
 1995 31.7 29.4
 1996 33.0 28.9
 1997 32.2 28.8
 1998 32.2 28.9
 1999 31.3 28.9
Further Mathematics
 1994 30.8 29.3
 1995 30.6 29.4
 1996 30.9 29.0
 1997 30.9 29.0
 1998 30.8 29.1
 1999 30.9 29.4
Mathematical Methods
 1994 30.2 30.1
 1995 30.0 30.0
 1996 29.9 30.1
 1997 29.9 30.1
 1998 30.0 30.0
 1999 30.1 29.9
Specialist Mathematics
 1994 30.5 29.8
 1995 30.4 29.8
 1996 30.2 29.9
 1997 30.1 29.9
 1998 30.3 29.8
 1999 30.4 29.8

Table 3 Summary of the number of years in which each gender has
achieved the higher average CAT scores and overall study score
(1994-1999)

 Study
 CAT 1 CAT 2 CAT 3 score

 M F M F M F M F

Biology 1 5# 0 6# 0 6# 0 6#
Chemistry 6# 0 0 6# 6# 0 5# 1
Physics (a) 0 6# 0 6# 0 6# 0 6#
Psychology (a) 0 6# 0 6# 0 6# 0 6#
Science (a) 0 6# 0 6# 2 4# 0 6#
Further Mathematics 0 6# 0 6# 1 5# 0 6#
Mathematical Methods 0 6# 6# 0 6 0 4# 2
Specialist Mathematics 0 6# 0 6# 1 5# 0 6#

Note: Values indicated with # are numbers in columns display the gender
that gained the greater number of higher average CAT scores and overall
study scores.

(a) Excludes CAT 4 from 1994

Note: Bold numbers in columns display the gender that gained the
greater number of higher average CAT scores and overall study scores.

Table 4 Summary statistics for the Victorian Board of Studies GAT
scores by gender, 1999-2001 (based on VCAA, 2001), and calculated
Cohen's effect sizes

 No. of
 reliable Mean
Year Component Gender scores score

1999 Written Communication F 36183 22.8
 M 29043 21.2
 Maths/Science/Technology F 36183 18.5
 M 29043 20.5
 Arts/Humanities/Social Science F 36183 20.0
 M 29043 19.9
2000 Written Communication F 40254 22.5
 M 33840 20.7
 Maths/Science/Technology F 40254 18.8
 M 33840 20.6
 Arts/Humanities/Social Science F 40254 18.9
 M 33840 18.5
2001 Written Communication F 41751 22.2
 M 35311 20.3
 Maths/Science/Technology F 41751 18.8
 M 35311 20.5
 Arts/Humanities/Social Science F 41751 19.3
 M 35311 19.4

 Cohen's
 effect
Year Component Gender SD size

1999 Written Communication F 5.9
 0.26
 M 6.2
 Maths/Science/Technology F 6.4
 -0.3
 M 6.8
 Arts/Humanities/Social Science F 6.6
 0.01
 M 6.8
2000 Written Communication F 6.1
 0.28
 M 6.5
 Maths/Science/Technology F 6.5
 -0.27
 M 6.8
 Arts/Humanities/Social Science F 6.3
 0.06
 M 6.3
2001 Written Communication F 5.8
 0.31
 M 6.3
 Maths/Science/Technology F 6.3
 -0.26
 M 6.7
 Arts/Humanities/Social Science F 5.6
 -0.02
 M 5.8

Table 5 Average GAT scores for years 1999-2001 by gender

Year Gender Av. GAT score

1999 F 20.4
 M 20.5
2000 F 20.1
 M 19.9
2001 F 20.1
 M 20.0

References

Allen, J. R. & Bell, E. J. (1996). Changing populations and changing results: Gender differences in senior studies 1987-95. Brisbane: Queensland Board of Senior Secondary School Studies.

Australian Council of Deans of Science. (1999). Who is studying science? (ACDS Occasional Paper No. 1). Melbourne: Author.

Bae, Y., Choy, S., Geddes, C., Sable, J., & Snyder, T. (2000). Trends in educational equity of girls & women. Washington, DC: National Center for Education Statistics (ED).

Becker, L. A. (n.d.). Effect size (ES). Retrieved 8 November 2002 from http://web. uccs.edu/lbecker/Psy590/es.htm

Brinkworth, P. C. (1999). An exploration of gender differences in subject choice and beliefs among Year 12 students. In J. M. Truran & K. M. Truran (Eds.), Making the difference: Proceedings of the Twenty-Second Annual Conference of the Mathematics Education Research Group of Australasia (pp. 106-113). Adelaide: Mathematics Education Research Group of Australasia.

Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum Associates.

Collins, C., Kenway, J., & McLeod, J. (2000). Factors influencing the educational performance of males and females in school and their initial destinations after leaving school. Canberra: Commonwealth Department of Education, Training and Youth Affairs.

Commonwealth Schools Commission. (1975). Girls, school & society. Canberra: Author.

Cortis, N. & Newmarch, E. (2000). Boys in schools: What's happening: Manning the next millennium. Brisbane: Queensland University of Technology.

Cox, P. (1996). SEX and CATs: Findings from a detailed statistical analysis. In H. Forgasz, T. Jones, G. Leder, J. Lynch, K. Maguire, & C. Pearn (Eds.), Mathematics: Making connections (pp.354-361). Brunswick, Vic.: Mathematical Association of Victoria.

Cox, P. (2000). Regional and gender differences in mathematics achievement. Journal of Research in Rural Education, 16(1), 22-29.

Dekkers, J. & De Laeter, J. R. (1997). The changing nature of upper secondary school science subject enrolments. Australian Science Teachers Journal, 43(4), 35-41.

Dekkers, J. & De Laeter, J. R. (2001). Enrolment trends in school science education in Australia. International Journal of Science Education, 23(5), 487-500.

Dekkers, J., De Laeter, J. R., & Malone, J. A. (1991). Upper secondary school science and mathematics enrolment patterns in Australia, 1970-1989 (Key Centre Monograph Number 4). Perth: Curtin University of Technology.

Dekkers, J. & Malone, J. A. (2000). Mathematics enrolments in Australian upper secondary schools (1980-1999): Trends and implications. Australian Senior Mathematics Journal, 14(2), 49-57.

Dobson, I. R. & Calderon, A. J. (1999). Trends in science education: Learning, teaching and outcomes 1989-1997. Melbourne: Australian Council of Deans of Science.

Ellis, H. (1894). Man and woman. London: Walter Scott.

Feingold, A. (1992). Sex differences in variability in intellectual abilities: A new look at an old controversy. Review of Educational Research, 62(1), 61-84.

Forgasz, H. J. & Leder, G. C. (2001). The Victorian Certificate of Education: A gendered affair? Australian Educational Researcher, 28(2), 53-66.

Haggerty, S. M. (1991). Gender and school science: Achievement and participation in Canada. Alberta Journal of Educational Research, 37(3), 195-208.

Hedges, L. V. & Friedman, L. (1993). Computing gender difference effects in tails of distributions: The consequences of differences in tail size, effect size, and variance ratio. Review of Educational Research, 63(1), 110-112.

House of Representatives Standing Committee on Education and Training. (2002). Boys: Getting it right. Report on the inquiry into the education of boys. Canberra: AGPS.

Keightley, J. V. (1999). Girls' participation in senior secondary science subjects and the implications for assessment. SASTA Journal, 99(1), 10-16.

Kimmel, Michael. (2000). What about the boys? Wellesley Centers for Women Research Report, 21(2), 6-7.

Leder, G. C., Forgasz, H. J., & Solar, C. (1996). Research and intervention programs in mathematics education: A gendered issue. In A. Bishop, K. Clements, C. Keitel, J. Kilpatrick, & C. Laborde (Eds.), International handbook of mathematics education, Part 2 (pp.945-985). Dordrecht, Netherlands: Kluwer.

Lydeamore, J. (1993). Gender equity in senior secondary school assessment (ESSSA) project. Adelaide: Department of Employment, Education and Training. (Publication restricted)

MacCann, R. (1995). Sex differences at the NSW Higher School Certificate after adjustment for the effects of differential selection. Australian Journal of Education, 39(2), 163-188.

Maccoby, E. E. & Jacklin, C. N. (1974). The psychology of sex differences. Stanford, CA: Stanford University Press.

Marks, G. N., Fleming, N., Long, M., & McMillan, J. (2000). Longitudinal surveys of Australian youth: Patterns of participation in Year 12 and higher education in Australia: Trends and issues (Research Report Number 17). Melbourne: Australian Council for Educational Research.

Ministerial Advisory Committee on Women and Girls. (1991). Working for gender justice in schools. Melbourne: Ministry of Education and Training, Victoria.

Ministerial Review of Postcompulsory Schooling (Jean Blackburn, Chair). (1985). Report: Volume 1. Melbourne: Author.

Ministry of Education Western Australia. (1991). Social justice in education: Policy guidelines for gender equity. Perth: Author.

O'Doherty, S. (1994). Inquiry into boys' education 1994: Challenges & opportunities: A discussion paper. Sydney: Ministry of Education, Training and Youth Affairs, NSW.

Organisation for Economic Cooperation and Development. (2001). Knowledge and skills for life: First results from the OECD Programme for International Student Assessment (PISA) 2000. Paris: Author.

Rowley, G., Brew, C., & Leder, G. (1997). The impact of self-selection on patterns of gender differences in mathematics achievement. Paper presented at the Annual Conference of the Australian Association for Research in Education, Brisbane.

Stobart, G., Elwood, J., & Quinlan, M. (1992). Gender bias in examinations: How equal are the opportunities? British Educational Research Journal, 18(3), 261-276.

Teese, R. (1994). Mass secondary education and curriculum access: A forty-year perspective on mathematics outcomes in Victoria. Oxford Review of Education, 20(1), 93-110.

Teese, R., Davies, M., Charlton, M., & Polesel, J. (1995). Who wins at school? Boys and girls in Australian secondary education. Canberra: AGPS.

Terman, L. M. (1925). Genetic studies of genius (Vol. 1). Stanford, CA: Stanford University Press.

Thorndike, E. L. (1906). Sex in education. Bookman, 23, 211-214.

Warrington, M. & Younger, M. (2000). The other side of the gender gap. Gender and Education, 12(4), 493-508.

Weiner, G., Arnot, M., & David, M. (1997). Is the future female? Female success, male disadvantage, and changing gender patterns in education. In A. H. Halsey, H. Lauder, & A. Stuart Wells (Eds.), Education, economy, culture and society (pp. 620-630). Oxford: Oxford University Press.

Whitehouse, H. & Sullivan, M. (1992). Girls and Year 12 science examinations: A study of the participation and performance of females in publicly examined science subjects in South Australia from 1986 to 1991 (SSABSA Research Monograph No. 1). Adelaide: SSABSA.

Authors

Peter J. Cox is a Lecturer at the Institute for Education, La Trobe University, Bendigo campus, PO Box 199, Bendigo, Victoria 3552.

Gilah C. Leder is a Professor in the Institute of Education at La Trobe University, Bundoora, Victoria 3086.

Helen J. Forgasz is a Senior Lecturer in the Faculty of Education, Monash University, Clayton, Victoria 3800.

E-mail: p.cox@bendigo.latrobe.edu.au

g.leder@latrobe.edu.au

helen.forgasz@education.monash.edu.au