Women in science: fixing the leaky pipeline

Hannah Thetford looks at the challenges faced by female scientists and considers strategies for working towards complete equality.

“Whatever women do they must do twice as well as men to be thought half as good. Luckily, this is not difficult.” I first encountered this famous quotation from Canadian feminist and politician Charlotte Whitton on the wall of my history classroom when I was 13 years old, and at the time I found the inequality it suggested laughable. I was from a well-off family and studying at a single sex secondary school, and I saw no reason why my gender would make it harder for me to succeed in life. Unfortunately for women wanting to pursue a career in STEM (Science, Technology, Engineering and Mathematics), Whitton’s words have proved more true than my 13-year-old self could ever have imagined.

Whilst the prospects for women in science have undoubtedly improved since western institutions first began to offer education to women in the 19th century, studies as recent as 2014 have demonstrated that female scientists are still faced with numerous barriers that their male counterparts are not. The cumulative effect of these barriers results in the well documented ‘leaky pipeline’ effect, a term describing how the proportion of women in STEM subjects decreases as you advance through the profession.

The first barrier many women encounter is the association of STEM subjects with men. At GCSE level girls outperform or do equally well as boys in STEM subjects1 but implicit association tests (IATs) show that a strong association of men with science is still present within western societies. Harvard University’s ‘Project Implicit’ IAT for gender and science was taken 299,298 times between July 2000 and May 2006, and in that time 54% of responses suggested a strong or moderate association of males with science and females with liberal arts2. The data is almost a decade old, but implicit associations are not easily changed and it is likely that the results still reflect the current public perception of science.


Results of Harvard University’s IAT for gender and science. Data is from 299,298 tests taken between July 2000 and May 2006.

This male bias has saturated our culture, appearing in everything from the way science and scientists are portrayed by the media to the toys we give children and the subjects we encourage young people to pursue. Each particular encounter with bias is a small event, but the cumulative effect of these events can have a significant influence over young women and the choices they make about their futures. In 2011 the percentage of girls who progressed to physics A-level in single sex government funded schools was 2.4 times greater than at co-ed government funded schools3, unquestionably demonstrating that the single sex environment has a positive influence on the number of girls taking A-level physics. I believe this is due to a reduction in the gender bias present in classrooms, as girls at single sex schools cannot be passed over in favour of male classmates and are less likely to feel pressured to stick to traditionally ‘female’ subjects.

The association of men with science has also been shown to impact hiring decisions in controlled experiments. In a study from 2012, members from the physics, chemistry and biology departments of 6 US universities were asked to evaluate the application materials of a student applying for a laboratory manger position4. The student’s name was randomly assigned to be male (John) or female (Jennifer), but the applications were otherwise identical. Despite this, John was rated as significantly more competent, offered a higher starting salary and considered more worthy of career mentoring from faculty members than Jennifer was. The gender of the faculty members had no influence on their reactions to the students, but an accompanying IAT showed that an existing subtle bias against women in science was correlated with less support for Jennifer (but was unrelated to reactions to John).

In another experiment performed in 2014 ‘employers’ were provided with varying amounts of information about two candidates before ‘hiring’ one of them to complete a simple arithmetic task5. The experiment was set up such that all participants completed one task and two candidates were then randomly selected from the group. The rest of the group became employers, who were financially rewarded if they hired the candidate who performed best on a second task of a similar nature. The random selection of candidates resulted in both mixed and single gender pairs, but only cases with one male and one female candidate were analysed.

When employers made hiring decisions with no information about the candidates besides their physical appearance they exhibited a large bias in favour of men over women. Although a clear demonstration of the bias against women in mathematical subjects, this says little about real-world hiring decisions as real world employers have additional information about candidates. Access to additional information was simulated by providing employers with either a prediction from each candidate about how well they expected to do on the second task or information about their performance on the first task.

Providing employers with each candidate’s prediction did little to change the bias in favour of men, which the investigators believe was due to employers failing to fully compensate for the tendency of men to overestimate and women to underestimate their future performance. The employers most likely to believe the inflated claims of male candidates were those whose IAT results revealed a stronger implicit bias against women in science, suggesting that the implicit associations of employers are likely to induce bias in the hiring process. Of course real world employers don’t just rely on the claims of candidates, but when employers in the study were given information about candidates’ past performance the bias against women was still present (although greatly reduced). Finding bias in all three cases in this experiment and in the 2012 study suggests that women are likely to be disadvantaged in real world hiring decisions.

A different form of bias against women in the sciences was revealed in a 2013 report into UK investments in global infectious disease research6. Between 1997 and 2010 female principal investigators in infectious disease research received less founding in both absolute and relative terms than their male counterparts, a difference that remained broadly unchanged over the 14 year study period. The median funding awarded to women was lower across most infectious disease areas, by funder and by type of science, and therefore it is not unreasonable to assume that this disparity also exists in other research areas. Combined, these examples show that women in STEM face significant challenges due to the implicit association of men with science. In other words Charlotte Whitton’s statement contains a frightening amount of truth: female scientists must do better than male scientists to be thought equally good.

This means that in order to achieve true gender equality something must happen to change the way we think of science and scientists. One way to do that is to ensure that more women reach the top levels of science, which requires other reasons that women leave the profession to be identified and addressed. Some possible reasons are high stress levels, long hours, heavy workload and motherhood but in an article on attrition of women in the biological sciences7 Shelley Adamo notes that these factors don’t appear to discourage women from entering and remaining in medicine.

According to Adamo the key difference between medicine and academia is the point in a woman’s career at which the highest amount of competition occurs. In medicine this point is the entry to medical school, which occurs before many women have partners or children. In contrast, the period of greatest competition for academics is the search for fellowship awards and faculty positions that typically occurs in the late 20s and early 30s. The average age of first time mothers in the UK is 288, meaning that this period of intense competition coincides with the time that many women choose to start a family. Additionally, early career researchers are expected to demonstrate mobility by moving from lab to lab on short term contracts. Constant moves disrupt family life and therefore this requirement disadvantages those with families, forcing some women to choose between a career in science and having children.

The increasing use of simplified metrics to assess the performance of individuals is another problem. As Adamo says, “studies have repeatedly shown that women invest more time in childcare and household duties than do men”, which leaves married women at a disadvantage because metrics often take no account of maternity leave or other family considerations. Another issue is that maternity policies for early career researchers are generally less supportive than those for faculty members, penalising women who want to have children at an age comparable to the national average.

Adamo believes that the most effective way to improve the retention of female scientists would be to increase job security by training fewer graduate students and therefore reducing competition. However, this would require co-ordination from various different bodies and individuals involved in offering and funding PhDs. Such a radical change seems unlikely to happen any time soon, but in the meantime the position of women could be improved by offering supportive maternity policies at all stages of scientific careers and abolishing the use of metrics to assess individuals (or at the very least including allowances for those with family obligations).

Another relatively simple and inexpensive change that can be made is to take steps to tackle sexist behavior and sexual harassment on campuses. A 2010 NUS report revealed that 68% of 2058 female students who participated in an online survey reported being verbally or physically harassed in some way in and around their current institution9. This harassment included behavior such as wolf whistling, groping and sexual comments as well as serious sexual and physical violence, and the amount of women reporting each type of incident is chilling. For example, 1 in 7 respondents had experienced a serious physical or sexual assault and 16% had experienced unwanted kissing, touching or molesting.

Whilst a large amount of this behaviour can be avoided by not attending social events, social life is a large part of the university experience and one that female students should not be excluded from due to fears of harassment. Furthermore, harassment can undermine female students’ confidence, damage their mental health and adversely affect their studies. 1 in 4 victims of serious sexual assault said that their studies had been affected by the incident, and 13% considered leaving their course.

And although most harassment occurs outside the classroom, 16% of reported verbal harassment incidents occurred in a learning environment such as a lecture theatre or a library and more than 1 in 10 respondents reported feeling uncomfortable as a result of comments with a sexual overtone being made in a learning environment. A separate study of women students’ experiences of ‘lad culture’ in higher education10 warns that “Class discussions can be intimidating environments for many students, who sometimes refrain from participating in discussions due to feeling embarrassed or uncomfortable.” Participants mentioned incidents in which professors or fellow students had used sexist language or expressed sexist views in class discussions, and some reported their opinions being dismissed is a sexist manner. They also felt that the ‘pack’ element of lad culture encouraged men to be loud and disrespectful in class.

Class discussions dominated by male students can leave women feeling inferior, and the culture of harassment produces an environment in which many women feel uncomfortable. The gender imbalance in many STEM subjects means that female students are often surrounded by men, which can make sexist behaviour feel even more threatening. In a module I took recently I was one of 3 women in a class of 27, and although I did not encounter any sexist behaviour I was constantly aware of my gender and often felt tense waiting for lectures to begin and during breaks. Women who decide to pursue a career in STEM know that the proportion of women around them will decrease further as they advance in their careers, and for female students who have learned to associate large groups of men with harassment this can be an intimidating prospect. Furthermore, men who express sexist views at work create an unwelcoming work environment which could contribute to the number of women leaving science.

Addressing sexism in universities by adopting a zero tolerance policy for harassment of all kinds in and around university buildings would go a long way towards making the learning and working environment more pleasant for both sexes. Universities can also educate students on consent (as many Oxford and Cambridge colleges did during freshers week in 201411) to reduce incidents of harassment outside the learning environment. Information about how to report harassment and the support available for victims would also be beneficial.

Finally, a 2014 report by Casadevall and Handelsman showed that the participation of women in scientific symposia at two large microbiology conferences was greatly increased if the team of convenors for the symposia included at least one woman12. The report states that “participation in meetings as a speaker is a factor of great importance for academic advancement” and that seeing female speakers can benefit women at the beginning of their career. Having at least one woman on each team of convenors is a change that can be easily implemented, and an increase in the number of women speaking at scientific events will hopefully start to filter into the public perception of scientists.

We have come a long way in the last 200 years, but we still have a lot of work to do to achieve complete equality. Initiatives like the Athena SWAN awards are helping to erode the barriers faced by female scientists, but changes must continue to be made if we want to fix the leaky pipeline and achieve a world where naïve teenagers can think that their gender won’t make a career in science any harder for them – and be right.



[1] Office for National Statistics, Statistics GCSE (key stage 4) collection, (last updated 2014)

[2] Harvard University Implicit Associations gender-science test results. Results can be accessed by completing the gender-science IAT at https://implicit.harvard.edu/implicit/. A more detailed summary is available via a link on the results page.

[3] Institute of Physics, It’s Different for Girls: The influence of schools (2012)

[4] Moss-Racusin et al., Science faculty’s subtle gender biases favor male students, PNAS 109, 16474 (2012)

[5] Reuben et al., How stereotypes impair women’s careers in science, PNAS 111, 4403 (2014)

[6] Head MG, Fitchett JR, Cooke MK, et al., Differences in research funding for women scientists: a systematic comparison of UK investments in global infectious disease research during 1997–2010, BMJ Open (2013)

[7] Shelley A. Adamo, Attrition of Women in the Biological Sciences: Workload, Motherhood, and Other Explanations Revisited, BioScience 47, Vol. 63 No. 1 (2013)

[8] Office for National Statistics, Live Births in England and Wales by Characteristics of Mother 1, (2014)

[9] NUS, Hidden Marks, 2nd edition (2011)

[10] NUS, That’s what she said: Women students’ experiences of ‘lad culture’ in higher education (2010)

[11] BBC News, Oxford and Cambridge University sexual consent courses start (2014)

[12] Casadevall and Handelsman, The Presence of Female Conveners Correlates with a Higher Proportion of Female Speakers at Scientific Symposia, mBio Volume 5, Issue 1 (2014)

Women in science: fixing the leaky pipeline

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s