A recent publication in Nature Communications on early-career informal mentorship [1] has caused scientists to take to the internet to express their discontent. Why, you ask? The authors of the paper concluded that junior female scientists benefit greatly from senior male mentorship and suffer if they have a female mentor. On the other hand, female mentors also seem to compromise their academic success by having female, instead of male, mentees. In a nutshell, you are unlikely to succeed in STEM as a woman. Many, including me, believed this paper should be retracted. The calls for retraction on social media have also led some academics to label Science Twitter as a mob. Of course, publications cannot be retracted because some people do not like their results, but a paper that resorts to an academic version of victim-blaming has no place in a scientific journal.
Methodology issues:
The study was heavily criticized on various platforms for not attributing their observations to the existing gender bias within academia and primarily using co-authored publications as a marker for mentorship. The authors analyzed historical data for mentor-mentee relationships (dating back to 1897, Source: https://bit.ly/3cHJJuC) across various STEM fields disregarding the underrepresentation of women scientists over the last century. Comparing data from the early 1900s to the present day may be useful if you are studying a trend, but it will probably produce unreliable correlation results if you don’t account for varied distribution, in this case, gender distribution. The authors use first names to determine gender, which in itself is problematic, but do not provide gender identities for the mentors in any of their datasets which brings into question how they compared their mentor to mentee gender relationships. Further, their use of unrealistic data points such as average academic age of mentors, reported as large as 213, and number of mentors, reported as high as 115, are contributing to doubts about the robustness of their analysis.
Ethical issues:
The authors seem to reinforce what should be an outdated notion – choosing your mentor based on their sex, let alone gender. While a person’s gender identity, as well as ethnicity and race, shape their life experiences, they are not a measure of intelligence or skill. Even if we attribute the methodological flaws of the paper to the pitfalls of large data analysis, the authors are perpetuating gender bias in a climate where women already experience detrimental implicit bias [2]. To suggest that the only way to overcome these biases is to work with big-shot male scientists invalidates all the hard work done by women to have their voice heard in the scientific community.
It does not escape me that the lead authors of the paper are actually women scientists, and this fact calls for some level of introspection. We should ask ourselves how often does a lab-mate’s gender influence our perception of their abilities? The flawed conclusions of the paper only underline the fact that womxn, i.e., all woman-identified persons, regardless of assigned sex at birth, must build confidence in each other, advocate for one another, and take pride in each other’s achievements.
Sumitra D. Mitra
November 30th, 2020
Sources: [1] AlShebli, B., K. Makovi, and T. Rahwan, The association between early career informal mentorship in academic collaborations and junior author performance. Nature Communications, 2020. 11(1): p. 5855. [2] Flaherty, C. Gender Bias in TA Evals: https://bit.ly/3mhMmXT. 2020.