Suck It Dudes, Science Proves Women Are Better Coders

Suck It Dudes. Science Proves Women Are Better Coders.

Guys may loudly rule the dark corners of the internet and be considered the "computer nerds", but according to a new paper from Cal Poly and North Carolina State University, it's the ladies who are kicking arse and taking names when it comes to coding. The paper, still awaiting peer review and thus subject to all the criticism dudes will no doubt hurl at it, examined the actions of four million people who logged onto GitHub in a single day. Using SCIENCE the researchers were able to determine the gender of 1.4 millions users. They found that 78.6 per cent of pull requests made by women were accepted versus the 74.6 per cent accepted by men. That means that the code women contribute to GitHub is more likely to be accepted by their peers.

One of the reasons suggested? "Women disproportionately make contributions that projects need more urgently."

However those requests, urgent or otherwise, are only accepted when the women don't outright display their gender. Women with gender neutral profiles found their work accepted 71.8 per cent of the time, and that number dropped to just 62.5 per cent when their gender became readily identifiable.

According to the paper, "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."

The paper goes on to conclude what we're all thinking: "Our results suggest that although women on Github may be more competent overall, bias against them exists nonetheless."

So there you go. Women are statistically better programmers then dudes, but the gender bias, and the systemic sexism inherent in the industry, are keeping them down.

Thanks guys.

[PeerJ via BBC]


Comments

    Github pull requests?? Really? Is that how you judge developers?

    Does this mean women are better coders than men? Wow

    Let me ask you something. If the results were the opposite, i.e., 78% for men and 74% for women, would you write an article saying "Suck it girls, science proves guys are better coders?"

    “Women disproportionately make contributions that projects need more urgently.” =/=
    "Women are statistically better programmers then dudes"

      I agree. I think women in programming is fantastic and know a couple who are way better coders than me. However this article takes some statistical analysis and makes projected hypothesis that the data does not actually represent at all. It _may_ be true, but this data doesn't prove it.

      One of the reasons suggested? That is just a gut feel random response and has no data to back it up.
      The actual data shows they are statistically better but only slightly. What I find more interesting is that there is a 10% difference between women depending on if they are identified as women or not.

        The actual data shows they are statistically better but only slightly.
        Statistically better how? If you mean better at coding as this article implies, that's nothing more than conjecture.
        The paper never shows, implies or states any such thing.

          They are 4% better at having commits accepted.
          If you judge success as having your work approved by peers then that is better. There are any number of reasons for this but the evidence shows that have a higher percentage of accepted commits

            The problem is there are too many uncontrolled variables. The only thing the study reliably shows is that women on Github are 4% more likely to get accepted.

              Correct. The rest is clickbait.
              To me the only interesting thing here is that gender identification dropped the acceptance rate. But than in and of itself doesn't show anything, it could be sample size bias, the type of projects that women identifier their gender, a different amount of talent between those who list gender and those who don't.

    Judging by the listing image, women are also pretty good at doing data modelling in reverse.

      She's writing on a mirror....

      They do it slower in reverse so as not to intimidate the men folk

      Last edited 15/02/16 10:35 am

    I dont understand the stats, if theunidentifiable is 71.8% acceptance rate and ifentifiable is 62.5% how did the overall get to 78.6%? What is the segment pushing the average up to 78.6%????

    Firstly I am a guy and secondly these studies are a waste of time.
    There have been a thousand studies before this one that claim men are better at science and math then women and a thousand more that claim women are better in the same fields so in the end who really cares. There are that many conflicting "professional" opinions on the subject that it has become meaningless.

    Also you haven't linked the actual paper which would have been nice but by your write up it doesn't seem that the paper hasn't given a ratio of men to women. I'm under the impression that programming is still a male dominated industry so if men made 70% of all the numbers of course the average is going to drop that's basic statistics. e.g. If you do a study on height and you have 5000 thousand men and 2000 women the accuracy on the average height of the group is going to be much tighter with the men then the women. As in a few exceptionally tall women will drag the average up for the women but you would need to have lots more tall men to drag their average up.

    The fact that the industry is on the misogynistic side of the scale means that women who do work in this industry need to be on top of their game to be accepted, this is supported by the drop in acceptance of they're submissions when there profile showed they are female. This could possibly improve their results.

      Wouldn't a few short women also drag their statistic down more than men as well though? Your example doesn't seem to make sense.

        Yes but that's my point. When 1 of your test groups is significantly smaller then the other test group you are comparing too then your statistics are weighted differently and your results can be heavily skewed positively or negatively. In this article I'm simply arguing that it looks to have been skewed positively.

    It's good that this research reveals an issue, but then aren't headlines and articles like 'women are better' just perpetuating the sexism problem as well?

    Some people, regardless of gender, race, age, beliefs, etc. are better/more knowledgable/more experienced/etc. at some things than others. And those who are better/etc. should help those who want to learn and improve.

    2016 do we still need to try and relate skill sets to gender?

    Kinda backwards. Woman are probably a little better in this specific context (opensource) because only the most awesome and determined women contribute - men are more regressed to the mean.

    "Another theory is that women in open source are, on average, more competent than men ... This theory is consistent with observations 1–5 ... (and 6) ... The frequent refrain that open source is a pure
    meritocracy must be reexamined
    "

    This paper is not saying that women coders are better, but that women opensource coders are better on average because the bar is higher than it should be.

    In fact only 5% of open source commits are from women, or perhaps 8% based on LOC, adjusted for women's larger commits on average based on this paper, vs 20-25% of coders who are women overall.

    Based on core committers in projects I am aware of this seems about right, WordPress being a good example, check here : https://make.wordpress.org/core/handbook/about/organization/#the-wordpress-core-team%7Csimilar (less than 10 women out of maybe 50 total "key folk")

    Definitely something to think about - women seem even more under-represented on opensource projects than IT in general, which given the core place of opensource and its positive contribution to career outcomes, is a problem.

      This is exactly what I was trying to get at above. Well worded.

    Wait!!
    acceptance of woman who showed gender 62.5%
    acceptance of woman who didnt show gender 71.8%
    acceptance of all woman combined 78.6%??????

    How does 63% and 72% combine to give 78%??

    I await your "Suck It Girls" article next time a study determines a similar but opposite finding.
    Just kidding, we both know you won't.

      My point exactly. This data seems to reinforce a politically correct view of the world, so let's write an article about it. If white men were found to best the best programmers in the world, nobody would write about it. Or maybe they would, just to call it racist.

    Who cares.

    I'm the Ideas person, a coder merely puts that Idea into machine readable format..

    Ideas keep coming, better to employ the best coder there is to "type it out" we could all do with someone who types faster than we can think...

    In the end it's about ownership of the concept... Not who gets it into the machine first.

    Does everyone on this site think that coding is the be-all and end-all, to me coding is just a means to an end, that end is to be productive in the real world. Coders are the useful sacrificial pawns.

    PS, don't forget that with the imbalance of M-F in STEM, it takes a certain type of woman to WANT to be involved. Isn't it likely that they would become fairly proficient in their chosen field (contrasted to the typical technological-male who just fell into their allotted position in society) if they can't at least gold their own they will have to get out, so dominate they must (and in many cases remain a technocrat-coder while their less adept male counterparts get moved up the chain where they can't go so much gamage at the coding-coalface, and get paid handsomely for their technologically less interesting roles (you don't really think that those endless meetings actually achieve anything do you, just like government vs the Depts, noone cares what resolution management passed, life goes on??). Ideas are what counts. lol, just TTP.

    Last edited 15/02/16 3:19 pm

      I agree on your means to an end
      I am an electronic designer and in the end all my coding in the micro's is doing is tying all the hardware together. Mind you that's not saying it's easy but it is in my industry lower down the train and everyone seems to do it.

    These articles just keep getting better! There is so much wrong with this that I don't even know where to start!

Join the discussion!

Trending Stories Right Now