The Merits of Meritocracy: Are Women Better Coders Than Men?

Let’s face it, it sucks that in 2016 we still have to talk about gender bias in the workplace. Even in the programming world we find a substantial degree of sexism. For those in the ‘know’ this won’t be surprising: gender bias in the tech community is a well-documented phenomenon. But, thanks to a study released this week, we can start to lay prejudice to rest. According to researchers from California Polytechnic State University and North Carolina State University, women are, statistically, better coders.

The study set out to discover, “To what extent does gender bias exist among people who judge GitHub pull requests?” For the laymen, GitHub is an open-source software repository where developers can convene, cooperate, and contribute to each other’s projects. Developers on GitHub will often write code for someone else’s project, called a pull request, and project leaders can decide to accept or reject these requests. In theory, whether or not pull requests are accepted depends entirely on the merit of the code. However, researchers found that pull requests written by women were accepted at, “a higher rate than code written by men, but only if the gender was not disclosed”.

More surprisingly, women outperform men in the top ten most used programming languages, and 16 of the top 20 (2 categories are statistical ties, and in 2 languages men outperform women). The significance of these findings encouraged the researchers to dig deeper. Were women writer easier codes, making their pull requests easier to accept? Were women doing less work, or taking fewer risks with their code? For all of these questions the answer is a resounding no. In fact, researchers found that, on average, women were writing longer code with more alterations despite the fact that pull requests written by women, “are less likely to fulfill an immediate need”.

In a work environment where women need to adopt an, “unreasonably aggressive” argumentative style to “justify ones own contributions”, one researcher suggest that the stark difference in performance boils down to “survivorship”. To survive in a community where men make up 88.8% of the workforce, women have had to outperform their peers in order to get noticed. Some women have said they experience little to no feeling of bias, whereas others feel the impact on a daily basis. That being said, the results of this study are clear and moving forward it would behoove the developer community to take larger steps in recognizing the efficiency and excellence women can bring to the workplace. 

Original Article

Women considered better coders – but only if they hide their gender

John Piela



Bletchley Park

Impossible to control for selection effect with women comprising such a small percentage of the sample. The women that decide to go into programming may be far superior in the neceesary skills than their female counterparts whom don't go into programming, resulting in an over-representation from the upper bins of intellect etc. Since so many men go into programming there is a better distribution of inherent skill or intellect, and the lower bins are better represented as a percentage of the total population. In short: as a percentage of their cohort, more male dummies go into programming at the moment.

That said, I wouldn't doubt it anyway! And if this encourages more women to go into programming all the better. It does present a paradox though: once there is sufficient of women to reproduce this test without confounding, it will necessarily mean that there is no longer an imbalance in gender representation, and the results will be a purely academic exercise.