Written evidence submitted by Professor Emma Smith (WSC077)
This submission presents a summary of recent research findings into gendered patterns of participation in post-compulsory STEM education in the UK. It examines the trajectory of learning which takes students from A-level qualifications, through undergraduate work and into early employment or further study. It also takes a long term view to monitor participation and attainment over an extended period of time. The findings suggest that almost three decades of initiatives to increase participation in STEM subjects have had little noticeable impact on the recruitment data and gendered patterns of participation persist in several subject areas. This is despite more women entering HE and little gender difference in the entry qualifications for STEM subjects. While more women are studying science, as broadly conceived, than ever before; recruitment to key areas, namely physics and engineering remains stagnant. However for those women who do remain in the ‘science stream’, patterns of employment in graduate careers and further study appear relatively equitable for most of the main subject groups, however concerns about early employment trajectories among the engineering and computer sciences are still apparent.
Data is presented on long term trends in the participation of men and women in post-compulsory science programmes. The datasets used for this analysis are of the highest quality available and they represent many hundreds of thousands of individuals (see appendix for further information on the datasets used). They enable us to take a long term view on how the participation of male and female students in post-compulsory STEM subjects has been shaped by an expanding HE sector and allow us to evaluate the long term impact of policies and initiatives which have aimed to encourage more young women, in particular, to study the sciences. The results presented here span three phases of post-compulsory STEM training: attainment at A-level, uptake of STEM subjects at undergraduate level and the early career destinations of recent STEM graduates. They address the following research questions:
3.1 Participation
A consideration of long term trends in participation in A-level pure science courses (biology, chemistry and physics) suggests that not one of these subjects appears to be thriving. In physics there are now fewer male students studying the subject than in the 1960s, with numbers in decline since the early 1990s (Graph 1). The stability in the female participation figures comes despite initiatives (such as GIST and WISE) that focus on encouraging girls to study physics after the age of 16. While there is no evidence to suggest that female candidates are being turned away from studying A-level physics in greater numbers than previously, there is nothing to suggest that these initiatives have had a notably positive effect on recruitment either.
Graph 1: Participation in A-level physics, male and female candidates 1961-2011
Source: DES, DfES, DCSF, QCA, JCQ, AQA, Edexcel, IoP
Until relatively recently A-level chemistry was a male dominated subject (Graph 2). From the early 1990s there has been a gradual narrowing of the participation gap. This has largely been due to a decline in the number of male candidates, although numbers are starting to increase. The introduction of compulsory science in the late 1980s appears in itself to have had no noticeable impact on A-level recruitment to chemistry, for either men or women. Although we have seen a steady feminisation of the subject to the extent to which, at present, participation in A-level chemistry is largely gender neutral.
Graph 2: Participation in A-level chemistry, male and female candidates 1961-2011
Source: DES, DfES, DCSF, QCA, JCQ, AQA, Edexcel, IoP
3.2 Attainment
In order to examine how gendered patterns of attainment have varied over the history of the A-level we examined changes in the relative attainment of male and female students at A-level grade C or higher[1]. What this historical overview shows is that the gender achievement gap in the pure sciences has hardly varied over the last five decades providing no evidence that male students have ever outperformed females at this level. For example, in physics there has been some shifting of the achievement gap in favour of female candidates since the mid-1990s (Graph 3). However the gap is negligible and overall it can be said that attainment at the higher grade levels is gender neutral. The same is true for chemistry. Until the late 1990s there was no difference in the proportion of male and female candidates achieving a grade C or higher in A-level chemistry, more recently the achievement gap has moved to be slightly in favour of female candidates. As with physics, this appears to coincide with a decline in the number of men entered for the subject, which may suggest that chemistry is losing some of its higher attaining male entrants. However, the achievement gap between men and women is still relatively small suggesting that attainment in chemistry at this level is gender neutral.
A positive value indicates overall higher achievement among female candidates. Values of less than 4% are not considered to represent a noticeable gap.
4.1 The expansion of the UK Higher Education system over the last twenty years or so is reflected in a quadrupling of the number of candidates offered places to study at undergraduate level. This increase is particularly apparent among previously under-represented groups: in 1986, 47,696 women accepted places on UCCA and PCAS-administered HND and first degree programmes; by 2012 this had risen to over 225,000. The early 1990s saw the end of long established trends of HE courses recruiting more male candidates. Now more female candidates apply and are accepted, a rate of increase since 1994 which has been faster than that for men (Graph 4).
Graph 4: Male and female home acceptances to all first degree and HND programmes, 1986-2012
4.2 Since 1986, the different STEM subject groups[2] have tended to recruit a steady share of female candidates with rates only varying by around 1-2 percentage points. For example, in 1986 5% of all places accepted by female candidates were in the Physical Sciences, this has varied only slightly in the intervening years, although currently the trend is towards a slightly lower share (3%) of female candidates who are accepted to study this group of subjects. Almost half of all places on STEM undergraduate courses are now taken by women, a proportion which, as Graph 5 shows, has steadily increased since 1986 and is in line with the increasing numbers of women studying at university. While this suggests that a relatively large and stable proportion of women do wish to study science (as broadly conceived), gender stereotyped patterns of subject choice still persist - over 50% of candidates studying the biological and medical sciences are female, while the proportion of female students studying the engineering sciences has hardly varied over the period considered (Graph 5).
Graph 5: Proportion of candidates accepted to main STEM subject groups who are female, 1986-2011
4.3 The extent to which certain science subjects continue to divide students according to their sex can be seen in Table 1 below. While trends have varied for certain subjects, indeed higher proportions of women now study chemistry and mathematics, gender stereotyped patterns of participation do still persist in subjects such as physics, engineering, computing and psychology.
Table 1: Percentage of female candidates accepted to the largest-recruiting science subjects, selected years 1986-2011.
| 1986 | 1990 | 1994 | 1998 | 2002 | 2007 | 2009 | 2011 |
Medicine | 46 | 49 | 52 | 56 | 61 | 56 | 55 | 54 |
Pharmacy | 58 | 61 | 57 | 60 | 64 | 59 | 58 | 59 |
Biology | 51 | 54 | 56 | 57 | 61 | 58 | 56 | 55 |
Psychology | 72 | 74 | 74 | 80 | 80 | 81 | 80 | 79 |
Sports Science | - | - | - | 36 | 34 | 34 | 31 | 29 |
Chemistry | 31 | 35 | 35 | 39 | 42 | 41 | 40 | 40 |
Physics | 17 | 16 | 16 | 19 | 19 | 18 | 20 | 17 |
Mathematics | 32 | 35 | 37 | 37 | 39 | 39 | 40 | 40 |
Comp. Science | 12 | 20 | 18 | 18 | 15 | 13 | 13 | 12 |
Civil Eng. | 10 | 13 | 12 | 12 | 12 | 13 | 13 | 12 |
All subjects | 42 | 46 | 49 | 52 | 53 | 55 | 55 | 55 |
4.4 The stability of many of these participation figures is quite remarkable. Physics for example, despite many initiatives to increase the numbers who study the subject, continues to recruit around 3000 undergraduate students (male and female) each year. The robustness of the participation figures for female physics and engineering candidates is illustrated in Graph 6 below. Note also the gradual increase in the proportion of women who study mathematics.
Graph 6: Variation in the number of female candidates accepted to study selected STEM subjects, 1986-2011
5.1 General destinations
5.1.1 Tables 3a and 3b show the main destinations of male and female graduates in the computer and engineering sciences. The majority of students enter some form of employment, with a relatively small proportion entering further study. This trend differs from the physical, biological and mathematical sciences where higher proportions of students enter further study (around one third of physical science graduates, both male and female, undertake further study, for example). As the tables show, there is some fluctuation in the trends over the period considered but the proportions of men and women at each stage tend to be reasonably similar.
Table 3a: First destinations of UK domiciled computer science graduates, selected years
| 1994 | 1998 | 2002 | 2008 | 2010 | |||||
| M | F | M | F | M | F | M | F | M | F |
Full-time work | 78 | 74 | 82 | 75 | 66 | 65 | 61 | 59 | 68 | 65 |
Further study | 8 | 10 | 6 | 9 | 16 | 18 | 18 | 19 | 15 | 15 |
Unemployed | 10 | 11 | 7 | 10 | 13 | 10 | 18 | 16 | 15 | 15 |
Total (N) | 4620 | 1266 | 5382 | 1409 | 8590 | 2480 | 6450 | 1220 | 7095 | 1440 |
Table 3b: First destinations of UK domiciled engineering science graduates, selected years
| 1994 | 1998 | 2002 | 2008 | 2010 | |||||
| M | F | M | F | M | F | M | F | M | F |
Full-time work | 69 | 71 | 73 | 75 | 65 | 66 | 59 | 61 | 67 | 70 |
Further study | 15 | 15 | 14 | 15 | 18 | 21 | 21 | 22 | 19 | 18 |
Unemployed | 11 | 8 | 7 | 4 | 11 | 7 | 14 | 10 | 11 | 9 |
Total (N) | 11368 | 2245 | 9795 | 1884 | 8300 | 1695 | 8335 | 1640 | 10090 | 1870 |
5.1.2 Graph 7 below, shows unemployment rates for male and female computer science students compared with all graduates. Note the relatively high unemployment rates for computer science graduates, although unemployment appears to be generally slightly lower among female graduates (although see following section for data on low skilled employment).
5.2 Graduates entering employment
5.2.1 Tables 4 and 5 compare the percentage of male and female STEM graduates who are employed in the largest recruiting science-related careers. The figures in parenthesis are the Standard Occupational Class groupings which represent the official UK government’s classifications for employment.
5.2.2 Well over half of engineering science students who enter employment remain in the field, indeed by far the largest recruiter of engineering graduates are the engineering professional occupations (civil engineers, production engineers and so on). However, this is an area where gender stereotypes patterns of participation persist and to some extent they remain when female engineers leave education and enter the workforce, especially when the proportions who enter professional occupations is considered. Gendered recruitment patterns among mathematical science graduates are relatively slight with a large proportion of these graduates who enter employment working in the business and finance sector. Having a computer science degree, on the other hand, appears to be more of an advantage for men, in terms of entry to graduate professions – note in particular that almost 30% of female computer science graduates are in relatively unskilled jobs six months after graduation. This later trend is one that exists to a certain degree between men and women in each of the three subject areas considered here.
Table 4: First destinations of Engineering, Mathematics and Computer Science graduates who enter employment (%), 2010/2011
| Engineering Sci. | Mathematical Sci. | Computer Sci. | |||
Occupation | Male | Female | Male | Female | Male | Female |
Managers & Senior Officials (SOC1) | 7 | 7 | 5 | 4 | 7 | 7 |
Professional occupations (SOC2) | 57 | 39 | 41 | 41 | 50 | 23 |
Engineering professionals | 47 | 31 | - | - | 2 | - |
ICT professionals | 6 | 3 | 9 | 4 | 44 | 26 |
Teaching professionals | - | - | 6 | 10 | 1 | 4 |
Business & statistical profess | 2 | 1 | 23 | 25 | 1 | 2 |
Associate Prof/Technical (SOC3) | 17 | 25 | 29 | 23 | 21 | 21 |
Science/Engineering tech | 2 | 2 | - | - | - | - |
IT Service Delivery Occ. | 1 | 0 | - | - | 8 | 7 |
Design Ass. Prof. | 2 | 6 | - | - | 4 | 4 |
Media Ass. Prof. | 3 | 2 | - | - | - | - |
Business & finance ass. prof | 3 | 2 | 20 | 16 | 2 | 4 |
Sales and related ass. prof. | 2 | 10 | 2 | 4 | 3 | 5 |
SUBTOTAL (SOC1-3) | 81 | 71 | 75 | 68 | 78 | 51 |
|
|
|
|
|
|
|
Administrative and Secretarial (SOC4) | 3 | 6 | 9 | 12 | 4 | 10 |
Skilled Trades occupations (SOC5) | 2 | 1 | 1 | - | 1 | - |
|
|
|
|
|
|
|
Personal service occupations (SOC6) | 1 | 3 | 2 | 4 | 1 | 4 |
Sales and Customer Service (SOC7) | 7 | 14 | 8 | 12 | 12 | 21 |
Process, Plant, Machine Ops (SOC8) | 1 | - | - | - | 1 | - |
Elementary occupations (SOC9) | 5 | 4 | 6 | 4 | 5 | 4 |
SUBTOTAL (SOC6-9) | 14 | 21 | 16 | 20 | 19 | 29?? |
|
|
|
|
|
|
|
Total (N) | 6820 | 1270 | 1500 | 1120 | 4675 | 820 |
5.2.3 It is also worth considering longer term trends in employment participation. Note the difference in proportions for men and women engineers who enter professional occupations (SOC2) and those who work in sales and customer services jobs (SOC7) six months after graduation.
Graph 8: Employment patterns for engineering science graduates, selected occupational groups only
5.2.4 As Table 5 shows, in the physical sciences, the majority of graduates who enter employment work in managerial, professional or technical fields (SOC1-3). Gender differences are relatively small in terms of occupational destinations for chemistry and physics graduates, who were generally more likely to be working in a field related to their degree subject. Occupations in the business and statistical professional group (namely accountancy and economist jobs) were particularly appealing to recent physics graduates, while relatively high proportions of chemists were employed as science professionals or technicians. There are some differences in the types of employment sought by physics graduates, namely teaching for women and ICT roles for men, but otherwise the types of jobs taken by both groups is fairly similar. Among chemistry graduates a higher proportion of women take jobs as science professionals and or as teachers but otherwise, patterns again are relatively similar for men and women. In contrast to several other subjects examined here, relatively small proportions of physics graduates, of either sex, undertake employment in routine occupations (SOC6-9). However, there is a large difference in the proportions of physical science graduates entering professional fields (33% and 23% for men and women respectively). In the biological sciences, the largest recruiter appears to be the associate professional/technical group with relatively few graduates entering professional occupations within six months of graduation. With the exception of the relatively high proportion of male biological science graduates undertaking managerial type work (13%), gender differences in terms of occupational destination among this group appear limited.
Table 5: First destinations of Biological and Physical Science graduates who enter employment
| Biological Sci. | Biology | Physical Science | Chemistry | Physics | |||||
Occupation | Male | Female | Male | Female | Male | Female | Male | Female | Male | Female |
Managers and Senior Officials (SOC1) | 13 | 5 | 8 | 5 | 7 | 6 | 5 | 6 | 4 | 3 |
Professional occupations (SOC2) | 12 | 13 | 16 | 19 | 33 | 23 | 39 | 35 | 49 | 47 |
Science professionals | 2 | 2 | 4 | 6 | 9 | 8 | 14 | 12 | 4 | 8 |
Engineering professionals | - | - | - | - | 4 | 2 | - | - | 9 | 8 |
ICT professionals | - | - | - | - | - | - | - | - | 14 | 5 |
Teaching professionals | 4 | 5 | 4 | 5 | 4 | 4 | 5 | 5 | 6 | 13 |
Research professionals | 2 | 2 | 5 | 4 | 4 | 4 | 10 | 8 | - | - |
Business & statistical profs) | - | - | - | - | 4 | 2 | 5 | 4 | 9 | 8 |
Associate Prof/Technical (SOC3) | 31 | 27 | 25 | 27 | 25 | 28 | 28 | 28 | 23 | 21 |
Science/Engineering technicians | 2 | 3 | 6 | 9 | 6 | 7 | 12 | 10 | - | - |
Social Welfare ass. profs | 3 | 6 | - | - | - | - | - | - | - | - |
Sports & Fitness Occ. | 12 | 4 | - | - | - | - | - | - | - | - |
Business & finance ass. profs | 3 | 3 | 4 | 4 | 6 | 5 | 6 | 6 | 11 | 11 |
Sales and related ass. profs | 4 | 4 | 3 | 4 | 4 | 5 | - | - | - | - |
Conservation ass. professionals | - | - | - | - | 2 | 3 | - | - | - | - |
Public Service & other ass profs. | 3 | 3 | - | - | 2 | 3 | - | - | - | - |
SUBTOTAL (SOC1-3) | 56 | 45 | 49 | 51 | 65 | 57 | 72 | 69 | 76 | 71 |
|
|
|
|
|
|
|
|
|
|
|
Administrative and Secretarial (SOC4) | 6 | 11 | 8 | 11 | 6 | 12 | 5 | 9 | 6 | 13 |
Skilled Trades occupations (SOC5) | 2 | - | 3 | 1 | 3 | - | 1 | - | 2 | - |
|
|
|
|
|
|
|
|
|
|
|
Personal service occupations (SOC6) | 10 | 18 | 7 | 12 | 3 | 6 | 4 | 4 | 1 | 3 |
Sales and Customer Service (SOC7) | 16 | 18 | 18 | 17 | 13 | 16 | 11 | 11 | 6 | 8 |
Process, Plant, Machine Op.s (SOC8) | 1 | - | 2 | 1 | 1 | - | - | 1 | - | - |
Elementary occupations (SOC9) | 10 | 8 | 14 | 9 | 10 | 9 | 8 | 6 | 7 | 5 |
SUBTOTAL (SOC6-9) | 37 | 44 | 41 | 39 | 27 | 31 | 23 | 22 | 14 | 16 |
Total (N) | 6005 | 9450 | 775 | 1135 | 3385 | 2640 | 660 | 582 | 695 | 190 |
All figures are % (except where noted) and non-bold figures correspond to the largest recruiting groups within the main SOC category
6.1 This submission has presented findings from recent and on-going research projects that have tracked the gendered patterns of participation in STEM-related post-compulsory education in the UK. It has examined the trajectory of learning which takes students from A-level qualifications, through undergraduate work and into employment or further study. As well as looking at how levels of engagement between men and women vary at each stage, the research reported here has also taken a long term view and considered the best available evidence to monitor participation and attainment over an extended period of time.
6.2 Despite many years of initiatives to encourage more women into STEM education and careers, women remain under-represented in many parts of the field and as the results presented here show, levels of unequal participation in key STEM subjects have varied little over the last twenty five years. This is despite more women entering HE and little gender difference in the entry qualifications for STEM subjects. So if qualification is not a barrier to participation, then why have the increased number of women studying in HE not been reflected in subjects such as physics and engineering? The answer according to much of the literature in this area appears to lie in the various socio-cultural barriers to STEM education which women experience. Indeed the data presented here shows that women are engaged with science and wish to study the subject at the highest levels. Rather, the types of science subjects they study, and psychology is a notable example here, are not necessarily those which policy makers and employers consider to be the ‘strategic, important and vulnerable’ science subjects. Women are also studying many science subjects in higher numbers than previously and yet recruitment to subjects such as physics and engineering remains stagnant. This may be a reflection of the critical mass model which suggests that female recruitment to these subjects is still too low to be self-sustaining and generate future growth. However, it is important to remember that numbers of male participants in physics and engineering subjects are also stable. In addition, for women who do remain in the ‘STEM pipeline’, patterns of employment in graduate careers and further study appear relatively equitable: over half of all chemistry and physics graduates (male and female) who enter employment remain in the field and similar proportions engage in further study. However there are some notable exceptions, especially among the engineering and computer sciences where in the later subject, relatively high proportions of graduates are unemployed or undertaking relatively low skilled jobs.
6.3 It is very difficult to provide policy recommendations that will easily remedy inequalities in the numbers of women who participate in physics, engineering and computer science in HE. Decades of well-funded and well-targeted initiatives have had limited impact and even making all young people study the sciences up to the age of 16 had little long term effect on recruitment at the next educational levels. As male and female students achieve at similar levels in gateway qualifications, there is limited evidence that attainment is a barrier to participation and for women who do study these shortage subjects evidence suggests that many are likely to remain in the field upon graduation. Whatever socio-cultural barriers are preventing the recent expansion of the HE system being reflected in female participation in physics and engineering, they appear to also apply to men, to some extent, and seem largely unaffected by gender-related widening participation initiatives.
September 2013
Appendices
A1 Sources of data
A1.1 Post-compulsory schooling: data were obtained electronically and as hard copy from a number of government departments and other official organisations, including the Department for Children, Schools and Families (and its predecessors the Department for Education and Skills and Department for Education and Science), the former Qualifications and Curriculum Authority, the Joint Council for Qualification, the Institute of Physics and the Assessment and Qualifications Alliance.
A1.2 Higher Education: Data retrieved from the Universities Central Admissions Service (UCAS) were used to investigate patterns of participation in STEM first degree and Higher National Diploma programmes at UK universities.
A1.3 Graduate destinations: Data on participation in further study and employment destinations were retrieved from the Higher Education Statistics Agency (HESA). Analysis of the UCAS data shows that the majority of STEM graduates are UK-domiciled, full-time students, so the focus here was limited to this group. The HESA first destination survey gathers information on the activities of graduates six months after they leave university. Response rates tend to be relatively high: for the Physical Sciences they are over 80%. It is recognised that this dataset only considers destinations six months after the student has left university and that career trajectories may be very different in the subsequent period. However, it tells us a great deal about the sorts of jobs that are available to science graduates.
A2 Calculation of Achievement gaps
Achievement gap = performance - entry
gap gap
GP = number of girls achieving that grade or better
BP = number of boys achieving that grade or better
GE = number of girls entered
BE = number of boys entered
A3 UCAS main subject group categories for science and science-related subjects:
|
|
|
|
|
|
|
|
[1] See appendix for a description of the method used to measure relative achievement
[2] The organisation of UCAS STEM subject groups is provided in the Appendix