back to article Women devs – want your pull requests accepted? Just don't tell anyone you're a girl

A new study has found that women are more likely than men to have their open-source software contributions accepted – but only when their gender is hidden from project leaders. The study from North Carolina State University and Cal Poly examined code committed by more than 1.4 million GitHub users and their contributions to …

  1. choleric

    '"The frequent refrain that open source is a pure meritocracy must be reexamined."'

    That's always been false.

    I'm pretty sure it's a misrepresentation of the original argument. Making something open source obviously won't magically make it meritorious. However, making something open source can be a huge step towards becoming truly meritorious.

    The very fact that this study could be done and this bias exposed is because the projects studied are open source. We don't have similar studies for Microsoft, Oracle, IBM, etc because the source is closed and the commit history even more so.

    Open source isn't perfect and no ibe should claim it is, but it's a more transparent system than the alternatives. And that's a good thing.

    1. Pascal Monett Silver badge
      Flame

      Transparency doesn't matter

      I can't believe how stupid we still are.

      Code is code. It doesn't matter whose fingers typed it, or whether or not they were perfumed. What matters is that it works.

      People (okay, MEN) who are capable of refusing working code on gender basis have no business heading a project and should go back to shovelling dirt and whistling to the sight of a well-shaped behind in a fitting dress because that's their level of intelligence.

      1. Bronek Kozicki
        Coat

        Re: Transparency doesn't matter

        "Code is code. It doesn't matter whose fingers typed it, or whether or not they were perfumed. What matters is that it works."

        Almost agree, but it also matters how the code reads (unless it is "write once and forget" kind of code). Which actually might give slight advantage to women, because in other studies they have demonstrated to be more emphatic than men. Which is important mental quality when writing readable code.

        1. Tom 38
          Headmaster

          Re: Transparency doesn't matter

          Which actually might give slight advantage to women, because in other studies they have demonstrated to be more emphatic than men.

          I think actually men are more emphatic usually, women are thought to be more empathetic.

          I've seen shit code from every gender tbh.

          1. MyffyW Silver badge

            Incompetance is an Equal Opportunity Employer

            @Tom_38 I've seen shit code from every gender

            Very true hun, there is no evidence that wearing a frock makes you better or worse.

          2. FuzzyWuzzys
            Happy

            Re: Transparency doesn't matter

            "...from every gender tbh"

            So men, women and.....?

            1. Tom 38

              Re: Transparency doesn't matter

              So men, women and.....?

              Gender is not binary.

              1. Fibbles

                Re: Transparency doesn't matter

                Gender is a social construct whereby people can be assigned to one of two sets. That certain people don't fit perfectly into one set or the other is an indication of how broken the concept of gender is. Gender is still binary though, despite recent attempts to subvert the terminology.

                1. Anonymous Coward
                  Anonymous Coward

                  alt.binaries.gender

                  If gender is socially constructed (and that is indeed how the term is defined today), then it's exactly as binary as we want it to be. We can change it, and indeed we do, all the time.

                  There are three genders in the German language for example, so there is nothing about the word "gender" which dictates only two possibilities. We don't need to remain trapped with a set of two, and there's plenty of evidence that slavishly echoing this narrow range causes all kinds of problems.

                  We could start by accepting the scientific fact that there are at least a dozen different biological sexes, so that we're not overly hung up on 'reflecting the two natural sexes in culture' or some other essentialist nonsense. (Intersex births are approximately as common as red hair, about 1 in 100). If sex is not necessarily a matter of only two possibilities, why on earth are people so anxious to propagate that fallacy into culture? So anxious to insist that people who (for whatever reason) fall outside this two-way split are abominations, or just need to "get over it". Have you seen the suicide statistics for transgender and intersex folks? We can do better.

                  Besides, gender norms are already fluid. My kid tells me that purple is a girls' colour and yellow is a boys' colour. It was certainly the other way round when I was a nipper. Astonishing as it may seem, it was pink dresses for boys right up until the 1920s. Most of the first programmers were women too. More hours are spent playing computer games by women than by men (true since at least 2003). A lot of assumptions about gender turn out to be either unreliable, or quite false on closer analysis. This research reflects that quite well, despite the best efforts of some commentards here to see it as some kind of feminist/SJW conspiracy.

                  FWIW I identify as 'non-binary', despite some discomfort about the nomenclature*. No I wont discuss my genitals, but I am happy to talk shop about software development any time.

                  *I dislike the way the term "binary" is used in the context of gender discussions, because, as any programmer knows, faced with variety limitations, you just add more bits. You can have all the richness you need if you go beyond a one bit system. "If a system is to be stable, the number of states of its control mechanism must be greater than or equal to the number of states in the system being controlled" - as W.R. Ashby so beautifully put it.

              2. Anonymous Coward
                Anonymous Coward

                Re: Transparency doesn't matter

                So men, woman and File Not Found

      2. caffeine addict

        Re: Transparency doesn't matter

        People (okay, MEN) who are capable of refusing working code on gender basis

        Sadly, your assumption it's only men who do this isn't true - there's lots of evidence of women who get to higher levels of male dominated industries not trusting the work of other women. :/

      3. Anonymous Coward
        Anonymous Coward

        @Pascal

        "I can't believe how stupid we still are."

        That is of course assuming that the study results were on the mark, and I have some doubts. Because the part which I'm missing in this study is the quality of code sent in. It doesn't talk about code differences, only about profiles.

        So couldn't it be possible that the ladies who shared more info on their profile also send in more crappy code?

        Because that park irks me with studies like these: they draw conclusions (gender really matters in this case) without giving us the full picture. Unless they actually asked the project leader(s) for the reason of the code pull then they're in no position to draw conclusions like these.

        1. Jason Bloomberg Silver badge

          Re: @Pascal

          So couldn't it be possible that the ladies who shared more info on their profile also send in more crappy code?

          It would be reasonable to expect code from women where gender is not known would be roughly as good or as bad as code from those where gender is known.

          It is possible that women whose gender is known write crappier code but that seems unlikely.

          1. Anonymous Coward
            Anonymous Coward

            Re: @Pascal

            > It would be reasonable to expect code from women where gender is not known would be roughly as good or as bad as code from those where gender is known.

            False.

            I expect people who go around advertising they are vegans to be worse people than those who are vegan without lecturing others, for example. If you're running around online saying LOOK AT ME I HAVE BOOBIES then chances are you're going to have much worse code than people who don't care to expose their profile information.

            1. brotherelf
              Coat

              Re: @Pascal

              This was not about mammaries in userpics, or dingdongs for that matter. This was about the fact that -- given cultural familiarity -- you can tell my gender by my first name.

              Icon: yes, you could also stick your hand in my pocket to make inquiries.

        2. John Hughes

          Re: @Pascal

          That is of course assuming that the study results were on the mark, and I have some doubts.
          Yes, of course you do, dear.

        3. Tom 13

          Re: @Pascal

          they're in no position to draw conclusions like these.

          Particularly given their own study calls into question their conclusions: the differences disappear when the contributors are well known to the inner circle.

          If it's purely a gender thing, it would exist across all lines.

      4. Anonymous Coward
        Anonymous Coward

        Re: Transparency doesn't matter

        > Code is code. It doesn't matter whose fingers typed it, or whether or not they were perfumed. What matters is that it works.

        Where in the study did they analyze whether or not the code that was rejected even worked? You have no basis for making this statement.

        Let's do a line by line teardown of the code accepted and rejected and see if the best code was what was accepted.

      5. Daggerchild Silver badge

        Re: Transparency doesn't matter

        Best coder I ever knew was a giggly little 5 foot thing with a huge uncontrollable grin.

        And when she switched to code mode, she produced things with clean flat sides that very politely declined to even entertain the possibility of having flaws. Made my job as her tester a boring nightmare. (And I'd made such evil, evil tests too..)

        A colleague went on holiday and I found he'd not read the spec properly, and a deadline was coming up, and our little lady had to step in, grumbled a little, then replaced all of his work in a frankly improperly short amount of time..

        Yeah, on average I would actually say ladycode is probably superior.

    2. choleric

      Hmm. Typed my post on mobile and didn't get round to typo checking until the edit window had expired.

      I meant "no one", not "no ibe". Oh for hardware keyboards!

      1. James 51

        The BB classic has an excellent keyboard or if like myself you value the replaceable battery the Q10 is an excellent alternative.

      2. BinkyTheMagicPaperclip Silver badge

        Motorola Droid 4, off ebay, or if you're willing to either get out the soldering iron or bung some people in Bulgaria, Spain or Korea cash to modify/supply a Motorola Photon Q to have a SIM slot added. That's the latest Android landscape keyboard phone, though.

    3. This post has been deleted by its author

    4. stevel

      Not equal

      Meritocratic <> meritorious.

      Doesn't mean it's any truer, though.

    5. Anonymous Coward
      Anonymous Coward

      > The study, Gender bias in open source: Pull request acceptance of women versus men was written by Josh Terrell, Andrew Kofink, Justin Middleton, Clarissa Rainear, Emerson Murphy-Hill​, and Chris Parnin.

      As you can tell from the title of the study, the 'researchers' found the outcome they wanted.

      The conclusion actually supported by the data here is that coders who choose to include their gender are weaker coders with less skill, whether male or female, and therefore are accepted less often. It is kind of like how people with bumper stickers, even ones about peace and love, are far more likely to engage in road rage than the general population.

      1. LionelB Silver badge

        As you can tell from the title of the study, the 'researchers' found the outcome they wanted.

        Well of course they were looking for bias - that was the whole point of the study! Scientific research in general does not (should not) proceed along the lines of "let's do some stuff and see what happens" - it goes "we have a hypothesis, let's see whether the evidence supports it".

        Their result could have been "No evidence for gender bias was found. Pull request acceptance rates were found to be the same for men and women" although, somewhat depressingly, you suspect they may not have published in that case; cf. Publication bias.

  2. Yugguy

    I expect votedowns

    From insecure blokes but it has been my experience that the truism that women usually have to work twice as hard to be seen as even half as good is still alive and well out there.

    "As a result, the researchers suggest that overall, the women contributing code to GitHub are more competent than their male counterparts,.... Despite this, the researchers also conclude that women are still seen as less competent than men "

    1. Trevor_Pott Gold badge

      Re: I expect votedowns

      "truism that women usually have to work twice as hard to be seen as even half as good is still alive and well"

      Would that this weren't the case. I honestly don't believe that the majority of men (at least in tech) think women are less competent. That said, there is a significant minority that harbour this believe - spoken or not - and those individuals seem to be pretty cognate with the cohort attracted to power.

      For my money, the five best techies I've ever met consist of three women, one man and one individual whose gender identity is still ambiguous after several years.

      Let's hope we can collectively overcome this bit of gender tribalism. We've better things to spend cycles on.

      1. Tom 13

        Re: the five best techies

        I've only met two I'd put in the class of best, both were male. The first was a programmer on level with Torvalds except without the penchant for profanity. (Yes this programmer like Torvalds actually wrote his own operating system.) He was one of those rare people who actually could do two complex tasks at once. He would often continue typing reports when you went to him to discuss a particular issue. The second was the tech who trained me for desktop support. He had a knack for knowing how something had to be configured even if he'd never seen the OS before. The only thing that slowed him down a bit was his first exposure to BSD.

        The two worst technicians I've ever dealt with were both female. The first was a paper cert MSCE working at a local private school district. The school made a deal with a local ISP to get a single dial up login account for all the students to use. Their OS was Windows 95. To hide the password so the kids wouldn't be able to use the school account on their home systems they were creating the dial up adapter from the admin account then having the kids log in. She didn't understand why the kid's account didn't have the dial up adapter credentials there.

        The second was a network admin at the company that absorbed the one I was working for when the above incident took place. Most of the time she had no opinions of her own and merely echoed whatever the consensus of the other techs was. This was when she was at her best. If she actually formed a strongly held opinion, even if you had documentation it was impossible to prove she needed to change it. At one point the company was in two locations and decided to consolidate into one location. One location was running Novell, NT 4.0, and Win95/98 desktops. The other was running strictly NT 4.0 and Win95/98. They decided that with the move they would also upgrade to 2000 Server and eliminate the Novell segment. We got all the systems migrated to the new domain. As you would expect all these changes led to an abnormally high call volume for the Service Desk. One problem in particular was perplexing. People would lose their connections to the server in the middle of the day. With access to the server lost, they'd get "Disk Full" errors when trying to save any document they had been editing. We hit upon the solution of saving it locally and rebooting the PC. With the connection newly established, they could copy the file from their local drive to the network. It wasn't everybody, but at least 20% of our users were having the problem. We built new systems for a couple of users but the problem persisted. Finally that second tech I mentioned above found the article addressing the problem on Tech Net. Because of the vagaries of the merger process, he wound up being low man on the totem pole for the network tech team. Despite having article in hand which precisely described our problem, Miss Head Bitch in Charge would NOT make the necessary change because "it involved manually editing the registry and you NEVER want to do that." We went for another full work day before the combined network team was finally able to prevail upon her to let them make the change. With the change made the problem went away. (Underlying issue was that the MS IP stack at the time was so bad, they set a time out on the Windows 2000 server. Not a problem for w Windows 2000 workstation because it would send a reconnect signal when it saw the drive was disconnected. Windows 95 didn't have that functionality, so the timeout had to be increased. The users who weren't having the problems had sufficiently frequent activity that they didn't experience the timeout.)

        Now, I will grant that on a raw basis the second woman wasn't a great deal worse than one of our other male techs. There were two things which made him tolerable. First, you could use tech articles to convince him what he was doing was wrong. When you did so, he would thank you for it. Second, he wasn't promoted above his competency level because of his sex. It is this second problem that keeps getting the sex warriors into trouble. Because they keep looking only at numbers and not quality, incompetent people keep getting promoted. And those incompetent people outdarken all the other lights that are advanced. In fact, I'll bet some of those lights avoid promotions precisely because of the incompetents.

    2. TheProf
      Devil

      Re: I expect votedowns

      "the truism that women usually have to work twice as hard to be seen as even half as good is still alive and well out there."

      Now there's something that needs to be researched. Can't have these sexist 'truism' going unchallenged can we?

    3. h4rm0ny

      Re: I expect votedowns

      Going to preface this with two facts, which are that I'm a female in the programming industry and that my impressions are just those of my experiences personally and of female friends when we've discussed this. Putting that at the start so that people can lend as much or little weight to the following as they wish and acknowledging that it's not a scientific survey. That said, anecdote is still the singular of data so a few thoughts:

      Firstly, whilst institutional sexism does exist, my encounters have always been based around a few bad eggs. Most of the people I've worked with it's not been an issue. However, those bad eggs do exist and sometimes they're in a position of power. So it happens.

      Secondly, I find that it has varied by nationality and sector. Again, I'm emphasizing this is personal experience, but I have found it markedly worse in the US than here in the UK, and in turn the UK is worse than Germany where sexism seems to have largely vanished. (I know it probably exists, but I've never seen a hint of it in working with German companies). Where I have encountered it in the UK, it's been in the financial sectors and amongst sales people. The City (as they delightfully call it) still is rife with sexism. Thankfully I don't work there and don't want to. But I have female friends who have and the impression is not good. With regards Sales, it just seems to attract a lot of male-dominated, somewhat sleazy attitudes. Maybe it's the back-room deal / old boy network culture that predominates, but the Sales teams of big companies are where I've run into some truly repugnant sexism. And the only place where someone has called me "doll" in a work-place environment or told me in a meeting to make the coffee! (That didn't end well for anyone involved in that meeting including myself who was kept out of future meetings by various means).

      In engineering and software departments and small companies, it's been fine. Including the USA on the whole. It's mainly those bad eggs rather than a sexist culture. And it's interesting that this survey is around online behaviour because in the workplace, the non-sexist majority have typically kept a lid on any sexist members of the team. Maybe it's online that they feel unrestrained enough to be open about their prejudices.

      Anyway, anecdotal evidence without an asserted conclusion. It does exist in our industry and should continue to be chipped away at. It's about equal opportunity and meritocracy. That's the end goal, imo.

      1. Tom 13

        Re: I expect votedowns

        Firstly, whilst institutional sexism does exist, my encounters have always been based around a few bad eggs.

        That's a complete non sequitur. Either it is institutional (exists everywhere and in all institutions) or it is a few bad eggs. By definition it can't be both.

        It's this sort of sloppy thinking that gets SJW types into trouble and why the anger in the backlash is growing.

        1. Tom -1
          Mushroom

          Re: I expect votedowns @Tom 13

          Evidently you think, for example, that the statement "there is institutional sexism in the N American sales division of Company X" implies that there is institutional sexism in every institution anywhere in the world? That makes it pretty clear that either you aren't aware that "institutional" in the context of that message could only mean "occurring habitually or customarily within an organization" and could not possibly mean "related to institutions in general" or you are deliberately writing nonsense in the hope of confusing some of your readers.

    4. Anonymous Coward
      Anonymous Coward

      Re: I expect votedowns

      I'd also guess the women that want to be in IT really want to be in IT (as there is so much stigma attached) as opposed to many men who just do it because they can just float along being mediocre and get away with it. I'd expect the men that really like IT are equally as good and committed.

      I also second the thing about sales, hate sales floors... horrible horrible places.

      1. h4rm0ny

        Re: I expect votedowns

        >>"I'd also guess the women that want to be in IT really want to be in IT (as there is so much stigma attached) as opposed to many men who just do it because they can just float along being mediocre and get away with it. I'd expect the men that really like IT are equally as good and committed."

        I think there's some truth to that. Not so much that men can get by being mediocre (I've known some less than stellar female programmers on rare occasions, too). But that it can feel more of a deliberate choice to get into IT as a woman. You're probably less likely to just drift into it. I don't believe there's any difference in natural aptitude between the sexes for programming. But there may be a selection bias for this reason. I certainly wasn't just going along with my friends when I picked computers as my field.

        1. Peter2 Silver badge

          Re: I expect votedowns

          As a man, might I be permitted to comment that most men also dislike the crowd of amoral and manipulative sociopaths usually recruited for sales roles and some areas of management?

          I can assure you that working with sexist, sleazy oversexed idiots is severely unpleasant for men who do not share the same attitudes towards woman.

  3. Voland's right hand Silver badge

    Peer review

    Well, I read it until I got to the "non-peer reviewed" part.

    Like it or not, there are fewer ladies in our profession than men. However, the ones that are working in our field, are usually way above average. So the submit/accept ratio for them is expected to be higher. In fact, I would expect it to be significantly higher.

    I do not buy the b*** about gender bias when "submitter identity" is known. For all high level submitters the submitter gender is known.

    So there is no way in hell to compile a blind study using real data representative of the actual contribution of real women doing real code on real open source projects.

    1. Andrew Richards

      Re: Peer review

      I buy gender bias being a thing. It's more subtle than implying rampant and overt locker-room type sexism (which, deservedly gets headlines when it happens).

      Some auditions for professional orchestras use screens to hide identify (and so gender) of applicant. Number of women selected increases when this happens.

    2. Graham Marsden

      Re: Peer review

      What I'd have liked to have seen is for them to create two contributors (let's call them Alice and Bob, as is traditional) and get them to both submit the same code (with appropriate modifications so they're not obviously identical).

      Do this a few times and in random order and see which gets accepted more often.

      1. Kristian Walsh Silver badge

        Re: Peer review

        It'll be the "Bob" submissions, because I'm assuming that the project maintainer is male and of English-speaking origin. Change "Bob" to "Vlad", and Alice to "Kate" and you'll see the difference narrow slightly - it's not just gender bias at work.

        This is a phenomenon that's already well researched in the area of recruitment. When it comes down to a close-call decision, men favour male candidates and women favour females. To be clear: it's also well proven that both men and women will choose an obviously superior candidate, regardless of that candidate's sex. It's only when it's a case of choosing between two average applicants that the gender bias comes into play, but it's also worth noting that most candidates for a position will be of average ability...

        The generalised point is that people are more favourable to others they have more in common with (sex, race, nationality, sexual orientation, place of education, membership of sporting clubs, etc, etc...).

        This is why many large organisations now insist that interview panels are gender-balanced and as diverse as possible within that. It's very, very hard for someone to be completely impartial, so having a diverse range of biases and prejudices on a panel is what keeps the system fair to everyone... or equally unfair to everyone.

        1. Raumkraut

          Re: Peer review

          This is a phenomenon that's already well researched in the area of recruitment. When it comes down to a close-call decision, men favour male candidates and women favour females.

          The last report I saw indicated that both male and female managers preferred hiring men for STEM roles.

          Here's an article from '14: http://www.eetimes.com/document.asp?doc_id=1321681.

  4. This post has been deleted by its author

    1. h4rm0ny

      Not because being a "nerd". That wasn't an issue in a UK school in the nineties. More just being one of the only girls on the course. I was the only girl doing a design and technology GCSE with around twenty boys (from memory - will probably be off). I was one of two girls in my Physics class. And there were about ten of us on the Computer Science course I did with two dropping out.

      That's actually pretty hard. It's not nice being separate from most of your friends in several of your classes or being the odd one out. Thankfully, I think my case was at the extreme end. But it's definitely a very significant issue. I'm not a big fan of trying to correct the gender imbalance at the point of job application. Chiefly because I find quotas both offensive and unfair, but also because it's trying to fix a problem that happened years before. Provide more encouragement for young girls to take up technology subjects (in a non-patronizing way) and more female role-models, and that will do more to address sexism in the workplace in our sector than anything else. Prejudice against a minority is a lot safer than prejudice against an equal faction.

      1. I sound like Peter Griffin!!
        Pint

        Well...

        "...Provide more encouragement for young girls to take up technology subjects (in a non-patronizing way) and more female role-models, and that will do more to address sexism in the workplace in our sector than anything else..."

        I reckon that's a consultancy niche you could operate in (to the benefit of potential entrants) based on your experiences and expectations, to great success - I'll be the first to toast your new little side-earner

      2. Anonymous Coward
        Anonymous Coward

        In most of my engineering/database/programming classes, I was in an extreme minority of white males, while the class was predominantly a 75/25 split of male and female people from India. I was in a class where a professor once commented that he only taught the class in English because I was there, as the only white person.

        To pretend this is only a white female issue is biased.

        Also in those classes for projects, it seems the Indian females were desired as teammates. Unlike white females, they didn't seem to form cliques with each other. They often had better organizational skills than the Indian males in terms of setting meetings, keeping deadlines in mind, etc. They were typically very quiet and unassertive, unless they were close friends with one of the males in the group.

      3. Tom 13

        That's actually pretty hard. It's not nice being separate from most of your friends in several of your classes or being the odd one out.

        That's NOT the fault of the men. Your numbers aren't wrong. I was in Astro, we had one woman out of 12. In the physics classes the numbers essentially doubled. But in both instance the men were more than happy to see women in the classes and fell over themselves to help them when asked. And I was in college before you, so according to the SJW types, we Neanderthals were supposedly more numerous and less restrained.

  5. jake Silver badge

    Somehow I doubt it.

    "The non-peer-reviewed study"

    Oh. Enough said.

    1. Paul Kinsler

      Re: Somehow I doubt it.

      I just skimmed the article, and it seemed fairly carefully argued; my guess would be its status is more like "not peer reviewed yet", or "in the process of peer review", rather than "not up to peer review standards".

      This is not to say that it is beyond reproach, just that to me it shouldn't be dismissed out of hand. And since the text is freely available, you can read it and then make whatever specific technical criticism you feel is relevant.

      1. Known Hero
        Joke

        Re: Somehow I doubt it.

        We also need to make sure that the Review panel is not gender biased !! Heaven forbid this get turned down for being written (by/for?) women

      2. MissingSecurity

        Re: Somehow I doubt it.

        I think a cursory read of the comments about it would make sense, in particular, read the comments section:

        "...The more relevant problem is that in gendered considerations, insider men receive less a lower acceptance rate than women. On the other side of the graph, gendered representations show no statistically significant difference, as Randal Olson points out.

        In other words, the hypothesis is that gendered women are discriminated against, but what you actually show from looking at the graphs is that any gender identified person is discriminated against and you further show that men are more likely to be the recipient of this than women.

        You're also missing an important piece about modern software practices: Most software now uses automated tests and profiling when someone submits a pull. Some go even further and even automate the pulling process, using Rultor or similar software to automate the merging process, no humans involved. ..."

        This is a pre-print which to be honest, Reg is doing a dis-service to suggest this is anything more than an interesting possibility, which after reading the actual PDF, I came to the same conclusion as the comment on the site.

        So if Reg actually gives a shit about this, and isn't just panning for sensation, we should see a updated article when/if the researchers actually can conclude something more definitive and peer reviewed.

  6. CaptainHook

    How did the study identify women?

    If the project leads aren't able to tell if a contributor is Male or Female because the username and avatar is gender neutral, then how did the study authors work out the gender in order to work out various acceptance rate of different groups?

    1. Anonymous Coward
      Anonymous Coward

      Re: How did the study identify women?

      Using the same publicly available information as the patch recipient has access to (their other social media accounts). If someone is accepting code based on the other person's sex, then it's just as easy for them to find it out their sex as it was for the "researcher".

      How about this?

      "Women who declare they're women on github tend to write poorer code than those who don't"

      1. h4rm0ny

        Hmmmm...

        >>"Women who declare they're women on github tend to write poorer code than those who don't"

        So you're saying that if I had registered here with my real name (which is recognizably female) rather than h4rm0ny, it would be a sign that I was a poorer coder? What's the theory behind that hypothesis? And does it also apply to men who declare that they're men? Or would that be different because male is the presumed default or somesuch...?

        1. Kristian Walsh Silver badge

          Re: Hmmmm...

          And does it also apply to men who declare that they're men

          Well, as an experiment, I just used the "man select" command on my terminal. I now feel I know slightly less about network programming than I did beforehand.

          ...I think you could be on to something, there ;)

        2. Anonymous Coward
          Anonymous Coward

          Re: Hmmmm...

          it would be a sign that I was a poorer coder? What's the theory behind that hypothesis?

          It was an alternative spin based on the stats in the report. There are lies, damn lies, and statistics.

          And does it also apply to men who declare that they're men?

          Yes, according to the stats. Almost equally.

          My suggested conclusions from the stats:

          - Pull request are rejected based on whether or not the coders sex is disclosed.

          - The types of people who disclose their sex don't submit as good PRs.

          - A bit of both

          - Or, the data was convenient enough for a sensational (non-peer reviewed) report.

  7. John G Imrie

    Summery

    If given the choice, men would rather pull men.

    Did I understand that correctly?

  8. Anonymous Coward
    Anonymous Coward

    Of these project owners who have the alleged bias, how many are women themselves?

    Personally, I've experienced women tend to be harsher towards fellow women.

    1. Paul Kinsler

      Re: Of these project owners ... how many are women?

      From what I read (skimmed) of the work, that isn't addressed in the study.

  9. Anonymous Coward
    Anonymous Coward

    "women contributing code to GitHub are more competent than their male counterparts"

    True, if we ignore the number of contributions. According to the study the ratio of merged contributions between women and men is 111,011: 2,181,517. And this is while keeping in mind that they were only able to identify the gender of only 35.5% of the profiles.

    So yes, if we look at the percentages instead of the number of contributions, men are indeed less competent and are falling behind :-)

    1. h4rm0ny

      So... your argument is that what men lose in quality, they make up for in quantity?

      That seems a little harsh... ;)

      1. Anonymous Coward
        Anonymous Coward

        Well, technically, the quality of the work wasn't tested, only the 'merged' numbers. So considering that the research is based on quantity instead of quality, I believe that I'm not that harsh :-)

    2. Anonymous Coward
      Anonymous Coward

      Not only that, but women often have more locked down social media profiles than men do, so chances are a greater % of the unidentified accounts are women.

  10. Anonymous Coward
    Anonymous Coward

    Tossers.

    How much of a tool would you have to be to reject code on the grounds of the submitting name at all?

    1. Anonymous Coward
      Anonymous Coward

      Re: Tossers.

      Would you want jewkiller@hitlerwasright.com in your project's commit history?

      1. Anonymous Coward
        Anonymous Coward

        Re: Tossers.

        > Would you want jewkiller@hitlerwasright.com in your project's commit history?

        I can see where all these women are going wrong now, thanks.

  11. Anonymous Coward
    Anonymous Coward

    So what you are saying is we get better female coders by being more picky. So surely the moral of this is to get better code by setting the bar higher for everyone not opening up more to bad female coders!?

  12. Charlie Clark Silver badge

    Incorrect extrapolation

    In short, as a whole women contribute more successful submissions to GitHub than men do

    Nope, that is a logical fallacy. Just because the study contends that women are more likely to make more successful contributions does not mean that, in absolute terms, they do. I certainly many, many more PRs from men than women.

    Anyway, the logic for identifying the gender of the contributors is flawed. That aside, I can imagine a lot of men perhaps tending to be less harsh on code from women and maybe making changes post-merge or to the PR directly. Call me patronising but I know that I do behave differently towards people who I think may need a bit of help and encouragement than others. At a conference recently I saw a presentation by a woman who emphasised how important it is for women to be encouraged in their work, which would seem to fit this pattern.

    Conversely, I can be blunt, terse and even downright rude to people submitting shit.

    1. Anonymous Coward
      Anonymous Coward

      Re: Incorrect extrapolation

      "That aside, I can imagine a lot of men perhaps tending to be less harsh on code from women and maybe making changes post-merge or to the PR directly. Call me patronising but I know that I do behave differently towards people who I think may need a bit of help and encouragement than others"

      I'll only call you patronising if you automatically think someone needs more help and encouragement simply because they're a woman... Otherwise it's perfectly ok to behave differently to someone who's submitting shit code when you know they can do better than towards someone submitting shit code because they're only just starting out and don't know better.

      That said, some women at least will have been told from a fairly early age that things like IT are "too hard" or otherwise not suitable for women. Which is ironic as, iirc programming used to be considered a female job (and going even further back men were secretaries). And whilst, yes, people can and do break past early (and continuing) conditioning that they shouldn't do x, y or z because of gender or whatever; it is hard. As someone posted earlier, you need to change the culture that make people think women can't/shouldn't do certain jobs (and also that men can't/shouldn't do certain jobs).

      I'm lucky in that I had an unconventional family life to start with, and the secondary school I went to had a strong engineering and tech focus, and was an all girls school so I never really experienced any resistance to being interested in computers.

      Not posting as anon to protect my gender, but because I'm posting at work and bored with trying to sort out problems a now ex member of staff created!

      1. Charlie Clark Silver badge

        Re: Incorrect extrapolation

        I'll only call you patronising if you automatically think someone needs more help and encouragement simply because they're a woman.

        Can't be ruled out but only as an initial response – I'm the go to techie for a couple of female friends so it easily becomes a habit. Not really by inclination. Though it works both ways: there are things I ask them about because I haven't got a clue. In both situations you can get into trouble by either assuming knowledge of a subject or its absence.

        One thing I have learned, however, is always to ask before grabbing hold of keyboard or mouse.

      2. Tom 13

        Re: Incorrect extrapolation

        I'll only call you patronising if you automatically think someone needs more help and encouragement simply because they're a woman

        Well then your first targets need to be the SJW types, including the authors of this so called study, because they are the most prominent promoters of precisely that bit of sexism.

  13. banalyzer

    Fuzzy stuff

    I've found that women are, in general, far better than men at a task if there is a lump of fuzzy type logic somewhere in the problem space. Where it is generally just plain hard/tortured logic I found that men were generally better. This is with quite a large cross-over between.

    There are differences between men and women, apart from the obvious physical ones, that mean they will think and act differently. This is not always nurture, the nature word comes into it as well. This is not a bad thing. I look at it as men and women are complimentary, they can each fill deficiencies in the others method of thinking or doing things. Teams should include both sexes as it can vastly improve how a team operates.

    I have seen sexist wide-boys in my time and, initially, women would just suffer it. Over the last decade or so they have really started to assert their right to be recognised and this level of sexism is retreating. The people who treated women the worst, however, were women who had been promoted 2 or 3 levels higher. To refer to those ones as bitchy is to underestimate the disrespect they would show to other women. This I have not seen reduce by a significant amount.

    When it comes down to it, some men are just tossers and some women are real bitches. Do these people a favour and forget where the ladder is when they stop digging.

    1. Anonymous Coward
      Anonymous Coward

      Re: Fuzzy stuff

      Nuture does have a lot to answer for though. If you're told from an early age that maths is too hard, or you're not going to be good at logic, or you can't show emotions, or you have to always be strong or any of the other it can take a lot of sheer bloody mindedness to break through it.

      I wholeheartedly agree that people of either gender can be utter arseholes!

  14. Anonymous Coward
    Anonymous Coward

    nevermind the gender nonsense

    I appreciate the use (finally) of push / pull double entendreurs in relation to github! Well done Shaun Nichols!

    fnar.

    1. Charlie Clark Silver badge
      Headmaster

      Re: nevermind the gender nonsense

      double entendreurs

      What's one of them, then? Some kind of cross between an entendre and an entrepreneur with some witty reference to the use of irony in the hipster start up scene?

      If you're going to use words you're not sure how to spell then at least use a spellchecker before you post! ;-)

      1. Tom -1
        Headmaster

        Re: nevermind the gender nonsense

        If we use French words like entendre and entrepreneur there's surely nothing wrong with introducing nother one (entendreur)? So perhaps a "double entendreur" is a person who, on hearing a double entendre, instantly understands both meanings, rather than a spelling mistake?

  15. Anonymous Coward
    Anonymous Coward

    Gender bias in open source: we're being paid to write this so we had to find something

    Should've been the title of this report.

    I am seeking sponsorship for my report: Species bias in barking: Why dogs lead the way in wagging and woofing and humans must catch up to win

    and will let the internet know how I get on. Woof.

    And to the person who wrote about 'perfumed fingers': just odd.

  16. pbowl

    Only half the story

    The conclusions reported here are for "outsiders", which is only half the story. For "insiders" (owners or collaborators) it goes the other way: men and women have similar acceptance rates when gender is unidentified, but women have significantly more acceptances when identified. (From fig. 5.)

    I would encourage both genders not to worry too much though. The effects are quite small - they only become statistically significant because of the huge size of the dataset. It's possible that these differences could be dominated by systemic errors, e.g., (a) the uncertain status of open requests which could be gender-linked in some way (e.g., one gender tending to submit bigger patches, which take longer to evaluate), and (b) possible confounding effect if decision to self-identify gender is correlated with ability/experience (e.g. if "serious" women tend more often to appear ambiguously), (c) possible confounding effect if ability to infer gender (which is only managed 35% of the time) is correlated with gender or ability/experience etc.

    (There are some mysterious aspects to the paper. What is "Michael's tool"?)

    1. Tom 13
      Joke

      Re: (There are some mysterious aspects to the paper. What is "Michael's tool"?)

      I don't know. But I hear there are a couple of companies in Van Nuys, CA that want to know if it's insured.

  17. Martin Milan

    Still not evolved beyond this?

    Why the hell should it matter in this day and age if the author is female?

    Maybe we're all scared of women seeking commit -m ent...

  18. Arthur the cat Silver badge
    Joke

    I got a pull request ...

    from a really good looking female programmer once, and was very happy to accept it.

    It's our 31st anniversary in May. All romantics can now go "aahh, how sweet". :-)

  19. Anonymous Coward
    Anonymous Coward

    > The non-peer-reviewed study

    I guess some aspects of the conclusions don't surprise me but the above is always cause for concern to me.

    Lots of the problems of these kinds of studies are revealed to be issues of bias or statistical issues and good peer review would have a good chance of discovering them.

    As others have commented above, the data doesn't always necessarily show what you think it shows and it is easy to become blinded to the limit of reasonable interpretation when you're personally looking for a particular result.

  20. Old Handle

    Interesting

    Other than the main results highlighted, the most interesting detail I saw in there was that pull requests were less likely to be accepted from men who were identifiable as such too. (Although this effect was weaker than for women.) So maybe that headline should be:

    Women devs – want your pull requests accepted? Just don't tell anyone you're a girl your gender

    Why that might be is an interesting question and this study didn't really make any attempt to get into it. The two possibilities that come to mind are that people who use their real name and/or picture on github are actually slightly worse (one could further speculate about why) or that because people have all kinds of biases, your best bet over all is to let your code stand on its own and tell as little about yourself as possible.

    1. Anonymous Coward
      Anonymous Coward

      Re: Interesting

      I noticed that too and thought it odd.

      I have a hypothesis:-

      People initially sign up with their actual name, then are subjected to the abuse, stalking, security leaks that pervade the internet and so change to an anonymous name resulting in the more experienced people being anonymous.

      It's entirely possible that such a process would affect women more than men, and so they anonymise sooner, resulting in a larger proportion of the identifiably female profiles being noobs.

      As a single data point, once upon a time I used my actual name here, and the security leak a while back led me to switch to this fake ID (Yes, I know you were fooled). I don't know if Zanzibar is considered identifiably male or female, but it's probably apparent which I am from my wittering.

      So here on the Reg the identifiable me was less experienced than the unidentifiable me.

  21. pewpie

    Need a new icon..

    A tiny pink violin.

    Seriously - this is not IT news.. it's pure bollocks.

    1. Anonymous Coward
      Anonymous Coward

      Re: Need a new icon..

      Perhaps we need a "My privilege blinds me to the experience of others, so their experience does not even exist" icon too?

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Other stories you might like