Women write better code than men, but only if gender is hidden says new study

US researchers have revealed that computer code written by women is better than that written by men – but only if their gender remains hidden. Github, a program-sharing service for code review and management, found that suggested code changes known as 'pull requests' by women were more likely to be accepted than those written by men.

In total, researchers from the computer science departments at Caly Poly and North Carolina State University, assessed around four million people who used Github on 1 April 2015. The scientists were able to identify the genders of 1.4 million after they analysed users' profiles or Google + social network.

Researchers did concede that this was a risk to the privacy of users of the website, but affirmed that they would not publish the raw data to the public. They discovered that 78.6 per cent of pull requests made by females were accepted, compared with 74.6 per cent of those by men.

After they discovered the results, researchers were implored to explore different possible contributory factors. These included whether women were more likely to be responding to common issues, whether their changes were shorter, therefore easier to appraise, and which programming language they were using to make changes, but could not find a correlation to the overall figure.

One surprising factor was that the acceptance rate was lower when a woman's gender was obvious rather than if the user's gender was unidentifiable. In fact, their pull request acceptance rate dropped by 9.3 per cent, taking it below the rate for men.

The paper is awaiting peer review, but if accurate it would fuel fears that the technology world has a problem with women. In December 2014 Technojobs.co.uk revealed that 91% of IT companies see only a fifth of job applications from women.

The paper noted: "Our results suggest that although women on Github may be more competent overall, bias against them exists nontheless. For outsiders, we see evidence for gender bias: women's acceptance rates are 71.8% when they use gender neutral profiles, but drop to 62.5% when their gender is identifiable .

"There is a similar drop for men, but the effect is not as strong. Women have a higher acceptance rate of pull requests overall, but when they're outsiders and their gender is identifiable, they have a lower acceptance rate than men."