3 ms·

That's survivor bias [0]. The point isn't whether there's still at least one woman in the field, or whether there's "plenty of women" (what does that even mean,

by jxf 12y ago

That's survivor bias [0]. The point isn't whether there's still at least one woman in the field, or whether there's "plenty of women" (what does that even mean, anyway?). The point is whether anyone was stopped from entering the field at all. The ones who are still in the field are the ones most resilient to the abuse. But you miss out on the contributions of all the people who left and did something else. How much farther along might we be in software engineering, or sculpture, or architecture, or <insert your favorite field here> -- had the toxic environment not been present in the first place? [0] http://en.wikipedia.org/wiki/Survivorship_bias http://en.wikipedia.org/wiki/Survivorship_bias

yummyfajitas 12y ago
"plenty of women" (what does that even mean, anyway?). Not really sure - I don't personally view a lack of $GROUP as a problem. If you feel similarly then disregard my entire post - it's premised on the idea that the reader does view a lack of women as a problem. There are two claims here. One is that bad behavior exists. The other is that this bad behavior plays a significant causal role in keeping women out of tech. The latter is what I'm questioning. Since large number of women seem to want to enter acting in spite of it having a far worse reputation (never heard of a "casting couch" in tech), it seems that bad behavior alone is a poor explanation for the lack of women.
- purerandomness 12y ago
  Basically, you want a formal proof that "bad behavior plays a significant causal role in keeping women out of tech" is a valid hypothesis? While scientifically interesting, I don't see the utillity for that, honestly. Do we really not agree that this has to stop? Does it really matter how many people harassment would keep out of the business? While skepticism has its place, harassment is not a technical, but a social problem, and I, for one, can bring up enough empathy to not need the scientific method to want to stop harassment. As to the casting couch, I don't know how it's where you live, but here in Europe, while everyone in the acting and music community is familiar with the term, the "casting couch" exists merely as a myth, not an actual thing preventing women to begin working in acting. Besides, the acting/music community is far beyond the 1/10 (or whatever) gender split in the tech community, so even if someone would try to establish the casting couch as a de-facto gate-keeping procedure, female members of the in-group would call it out, ending this phenomenon much sooner than would be possible if the gender gap in those communities were to be as it is now in tech. Even then, you wouldn't bring up statiscics how many women were entering the acting field by "passing" the "casting couch". So I'm confused and would like to know what your point really is.
  - yummyfajitas 12y ago
    No, I'd like to see evidence of some sort that the effect is significant before concluding it plays a significant role in keeping women out of tech. If you want to tackle the problem of crazy people trolling public figures, be my guest. I don't see any reasonable steps that we could take that wouldn't be oppressive, but maybe you have more imagination than me. Just don't be surprised if you solve this problem but tech remains 25% women.
    - purerandomness 12y ago
      We're not discussing "crazy people trolling public figures" here. No one wants to do anything about your favourite sunday newspaper cartoonist and his daily Obama caricatures. We're not talking about your neighbour kid playing phone pranks on you. What we're talking about here are death threats, sexual harassment, and calling your boss to get you fired. Deliberate, organized hate. Effective and planned activities to harm an individual who doesn't represent the in-group you're happening to be in. I couldn't care less how the gender distribution remains after we've solved THAT issue. This is a showstopper bug that has to be solved BEFORE we even begin to debug the gender gap. Anything else would be ridiculous: "Yeah, we're trying to find ways to get more women in tech, but since we don't have any scientific evidence to back up that the harassment, hate and verbal abuse against women in tech is a root cause of too few women in tech, we're still investigating and solving other possible causes" If you were a woman (I suppose you aren't), would you want to work in an environment that is obviously this toxic?
      - yummyfajitas 12y ago
        If you were a woman (I suppose you aren't), would you want to work in an environment that is obviously this toxic? I don't think tech is "toxic". I've put as much of my money where my mouth is as my plumbing allows. I recently (read: past few months) did everything I could to help a girl I care about get into tech. I think she'll be happier in the field than in her various alternatives.
        12y ago
- purerandomness 12y ago
  I'll add another reply to this, because this deserves its own discussion: > I don't personally view a lack of $GROUP as a problem Do you really not see how a group's problem-solving could be much richer if more $GROUPs are added to the mix? I find this astounding. A lack of $GROUP in a design team designing products for $GROUP leads to obvious overhead where the members have to account for every $GROUP's needs, while not having a representant of this very group, leading to blindness and bias - one popular example being crash test dummies which no one even thought to test for female body types, leading to greater injuries in car accidents for women (it's not difficult to find scientific articles on this, if that's new to you) Another obvious problem is that by having $GROUP out of the mix, a member of $GROUP trying to enter the field faces doubt and skepticism ("scene whore"), just because of the fact that $GROUP is under-represented, perpetuating the situation. Let's play devil's advocate here and say that a lack of $GROUP isn't a problem in itself - how would you solve the two aforementioned problems without increasing $GROUP's representation in a field?
  - yummyfajitas 12y ago
    In at least two jobs, not to mention my current consulting project, I've been the only member of my racial group. I can't think of a time my group membership ever helped or affected the work at all. At most it was a source of fun within the office - e.g. convincing an officemate that I had two illegitimate children might have been more difficult if I were a member of the majority group. I agree that diversity among crash test dummies, nurses and security theatre pat down people is important since specific job functions depend on it. Another obvious problem is that by having $GROUP out of the mix, a member of $GROUP trying to enter the field faces doubt and scepticism ("scene whore"), just because of the fact that $GROUP is under-represented, perpetuating the situation. Your entire process is broken if irrational concerns like this are playing a significant role. This is a bigger problem, and merely throwing in some more bias to compensate is papering over the problem. The solution is to vigorously make your process systematic and objective, removing human judgement as much as possible.
    - purerandomness 12y ago
      > In at least two jobs... While anecdotal trivia of your work history didn't make you feel a lack in your work environment, you're actively downplaying and trying to make a non-issue of the lack that is viewed by your peers in the group. Why? What would you lose if more $GROUPs happened to work in your field? > make your process systematic and objective You mean the crash test dummy example? You can't automate the design of a new process, because the design of the automation in itself is the designing of a process which underlies human bias and blindness. How exactly are you going to make the process of a female being called a "scene whore" on IRC as soon as the mob find out she's female make more "systematic and objective"? I think you mixed up the different examples here.

by jxf 12y ago

yummyfajitas 12y ago

"plenty of women" (what does that even mean, anyway?). Not really sure - I don't personally view a lack of $GROUP as a problem. If you feel similarly then disregard my entire post - it's premised on the idea that the reader does view a lack of women as a problem. There are two claims here. One is that bad behavior exists. The other is that this bad behavior plays a significant causal role in keeping women out of tech. The latter is what I'm questioning. Since large number of women seem to want to enter acting in spite of it having a far worse reputation (never heard of a "casting couch" in tech), it seems that bad behavior alone is a poor explanation for the lack of women.

purerandomness 12y ago
Basically, you want a formal proof that "bad behavior plays a significant causal role in keeping women out of tech" is a valid hypothesis? While scientifically interesting, I don't see the utillity for that, honestly. Do we really not agree that this has to stop? Does it really matter how many people harassment would keep out of the business? While skepticism has its place, harassment is not a technical, but a social problem, and I, for one, can bring up enough empathy to not need the scientific method to want to stop harassment. As to the casting couch, I don't know how it's where you live, but here in Europe, while everyone in the acting and music community is familiar with the term, the "casting couch" exists merely as a myth, not an actual thing preventing women to begin working in acting. Besides, the acting/music community is far beyond the 1/10 (or whatever) gender split in the tech community, so even if someone would try to establish the casting couch as a de-facto gate-keeping procedure, female members of the in-group would call it out, ending this phenomenon much sooner than would be possible if the gender gap in those communities were to be as it is now in tech. Even then, you wouldn't bring up statiscics how many women were entering the acting field by "passing" the "casting couch". So I'm confused and would like to know what your point really is.
- yummyfajitas 12y ago
  No, I'd like to see evidence of some sort that the effect is significant before concluding it plays a significant role in keeping women out of tech. If you want to tackle the problem of crazy people trolling public figures, be my guest. I don't see any reasonable steps that we could take that wouldn't be oppressive, but maybe you have more imagination than me. Just don't be surprised if you solve this problem but tech remains 25% women.
  - purerandomness 12y ago
    We're not discussing "crazy people trolling public figures" here. No one wants to do anything about your favourite sunday newspaper cartoonist and his daily Obama caricatures. We're not talking about your neighbour kid playing phone pranks on you. What we're talking about here are death threats, sexual harassment, and calling your boss to get you fired. Deliberate, organized hate. Effective and planned activities to harm an individual who doesn't represent the in-group you're happening to be in. I couldn't care less how the gender distribution remains after we've solved THAT issue. This is a showstopper bug that has to be solved BEFORE we even begin to debug the gender gap. Anything else would be ridiculous: "Yeah, we're trying to find ways to get more women in tech, but since we don't have any scientific evidence to back up that the harassment, hate and verbal abuse against women in tech is a root cause of too few women in tech, we're still investigating and solving other possible causes" If you were a woman (I suppose you aren't), would you want to work in an environment that is obviously this toxic?
    - yummyfajitas 12y ago
      If you were a woman (I suppose you aren't), would you want to work in an environment that is obviously this toxic? I don't think tech is "toxic". I've put as much of my money where my mouth is as my plumbing allows. I recently (read: past few months) did everything I could to help a girl I care about get into tech. I think she'll be happier in the field than in her various alternatives.
      - 12y ago
purerandomness 12y ago
I'll add another reply to this, because this deserves its own discussion: > I don't personally view a lack of $GROUP as a problem Do you really not see how a group's problem-solving could be much richer if more $GROUPs are added to the mix? I find this astounding. A lack of $GROUP in a design team designing products for $GROUP leads to obvious overhead where the members have to account for every $GROUP's needs, while not having a representant of this very group, leading to blindness and bias - one popular example being crash test dummies which no one even thought to test for female body types, leading to greater injuries in car accidents for women (it's not difficult to find scientific articles on this, if that's new to you) Another obvious problem is that by having $GROUP out of the mix, a member of $GROUP trying to enter the field faces doubt and skepticism ("scene whore"), just because of the fact that $GROUP is under-represented, perpetuating the situation. Let's play devil's advocate here and say that a lack of $GROUP isn't a problem in itself - how would you solve the two aforementioned problems without increasing $GROUP's representation in a field?
- yummyfajitas 12y ago
  In at least two jobs, not to mention my current consulting project, I've been the only member of my racial group. I can't think of a time my group membership ever helped or affected the work at all. At most it was a source of fun within the office - e.g. convincing an officemate that I had two illegitimate children might have been more difficult if I were a member of the majority group. I agree that diversity among crash test dummies, nurses and security theatre pat down people is important since specific job functions depend on it. Another obvious problem is that by having $GROUP out of the mix, a member of $GROUP trying to enter the field faces doubt and scepticism ("scene whore"), just because of the fact that $GROUP is under-represented, perpetuating the situation. Your entire process is broken if irrational concerns like this are playing a significant role. This is a bigger problem, and merely throwing in some more bias to compensate is papering over the problem. The solution is to vigorously make your process systematic and objective, removing human judgement as much as possible.
  - purerandomness 12y ago
    > In at least two jobs... While anecdotal trivia of your work history didn't make you feel a lack in your work environment, you're actively downplaying and trying to make a non-issue of the lack that is viewed by your peers in the group. Why? What would you lose if more $GROUPs happened to work in your field? > make your process systematic and objective You mean the crash test dummy example? You can't automate the design of a new process, because the design of the automation in itself is the designing of a process which underlies human bias and blindness. How exactly are you going to make the process of a female being called a "scene whore" on IRC as soon as the mob find out she's female make more "systematic and objective"? I think you mixed up the different examples here.