4 ms·

> Are you saying there are no situations in which applying a systematic adjustment as a function of gender to interview scores could result in a predictable net

> Are you saying there are no situations in which applying a systematic adjustment as a function of gender to interview scores could result in a predictable net-fitness-increase in the resulting skill distribution of hired people? Yes if you don't try to match the distribution of the pool of candidates. if the pool of candidates ratio is 10/90 and your goal in the company is 10/90, I am ok with anything that helps you to get close to the 10/90. (When both distributions are normal gaussian with similar mean and average. I don't see basketball teams hiring short people to fit the ratio or the pool of candidates) > That's not really the important question though. Would you prefer being not being hired due to bias even though you are qualified, or being hired despite bias, even though you know it will make some people think you weren't really qualified? I prefer being hired because they saw my skills. I think is better in the long term. If I am qualified for a job I will find it, it could take me more time but at some point I will find it and I will be happier than in a job where they saw me hired by a diversity program. Of course if I need the job to survive wouldn't care, but I think this is not the common case (or I might be wrong but as I said in another comment this is my personal opinion, other will think differently)

tunesmith 9y ago
If the pool of candidates is 10/90, I would agree a hiring goal of 10/90 is fine assuming that the pool of candidates is not also negatively impacted by bias. The pool of candidates, however, is widely accepted to be negatively impacted by bias (which is why you see things like female-friendly coding camps, etc).
- jorgemf 9y ago
  I don't know the reason why women are not interested in tech. I think there is an important biological component and also a cultural issue. But I am not sure of the cultural issue because in countries like Norway, one of the most equality countries, the ratios are higher than other countries, which doesn't make sense unless the biological factor is what affects the most.
  - zimpenfish 9y ago
    > I don't know the reason why women are not interested in tech. Dumb frat bros writing 10 page screeds about how they're not fit for the job because they're biologically conditioned to work with people and not machines -might- have something to do with it. And, as a historical point, until the late 60s, software development was seen as "women's work". Which somewhat destroys any "biological component" nonsense.
    - tzs 9y ago
      You misunderstand the biological component argument, I think. It's not that men are massively more suited to some jobs and women are massively more suited to some jobs. It's that for some abilities the the center of the wide bell curve for men is not at the same place as the center of the wide bell curve for women. For example, for verbal abilities the female bell curve is a little to the right of the male bell curve. For spatial abilities, it is the other way around. Does this mean men cannot be excellent writers, and women cannot be excellent mechanical engineers? No. It just means that if all else were equal, we would expect writing and mechanical engineering to have gender distributions that don't match the general population.
      - exomancer 9y ago
        I think you are also misjudging the effects, and to talk about ability is harmful to the very debate. There are some differences in ability, but they seem to even themselves out (men are better at mental rotation, women have better visual memory, etc.), the total effect on general intelligence appears to be none. However, there are measurable psychological differences when it comes to preference of the role you want to take, how much risk you are willing to take, and what your priorities regarding status and family are, and we do have evidence that some of it is, at least partially, biological.
    - collyw 9y ago
      In Europe we don't (seem to) have the fratboy culture, but there are still very few women in tech. (Regarding race, certain groups seem to be overrepresented and others underrepresented) its mainly white guys, but Europe is mainly white.
- lagadu 9y ago
  I'm going to disagree here on one thing: The pool being biased isn't a problem that the hirer can or should deal with; it's out of their hands because their position in the chain is far further along of where the problem lies. So I would submit that no, an employer shouldn't have to compensate for a pool bias because they have no impact on it but in turn it can only lead to lower long term outcomes for them. Personally I suspect that there's a mix of both cultural (girls play with dolls traditionally) and biological (boys tend to be more attracted to tinkering and building from even before they learn to speak) reasons for the ratio in people studying IT-related fields but I just don't see how making hiring practices biased in any direction will help with that.

tunesmith 9y ago

If the pool of candidates is 10/90, I would agree a hiring goal of 10/90 is fine assuming that the pool of candidates is not also negatively impacted by bias. The pool of candidates, however, is widely accepted to be negatively impacted by bias (which is why you see things like female-friendly coding camps, etc).

jorgemf 9y ago
I don't know the reason why women are not interested in tech. I think there is an important biological component and also a cultural issue. But I am not sure of the cultural issue because in countries like Norway, one of the most equality countries, the ratios are higher than other countries, which doesn't make sense unless the biological factor is what affects the most.
- zimpenfish 9y ago
  > I don't know the reason why women are not interested in tech. Dumb frat bros writing 10 page screeds about how they're not fit for the job because they're biologically conditioned to work with people and not machines -might- have something to do with it. And, as a historical point, until the late 60s, software development was seen as "women's work". Which somewhat destroys any "biological component" nonsense.
  - tzs 9y ago
    You misunderstand the biological component argument, I think. It's not that men are massively more suited to some jobs and women are massively more suited to some jobs. It's that for some abilities the the center of the wide bell curve for men is not at the same place as the center of the wide bell curve for women. For example, for verbal abilities the female bell curve is a little to the right of the male bell curve. For spatial abilities, it is the other way around. Does this mean men cannot be excellent writers, and women cannot be excellent mechanical engineers? No. It just means that if all else were equal, we would expect writing and mechanical engineering to have gender distributions that don't match the general population.
    - exomancer 9y ago
      I think you are also misjudging the effects, and to talk about ability is harmful to the very debate. There are some differences in ability, but they seem to even themselves out (men are better at mental rotation, women have better visual memory, etc.), the total effect on general intelligence appears to be none. However, there are measurable psychological differences when it comes to preference of the role you want to take, how much risk you are willing to take, and what your priorities regarding status and family are, and we do have evidence that some of it is, at least partially, biological.
  - collyw 9y ago
    In Europe we don't (seem to) have the fratboy culture, but there are still very few women in tech. (Regarding race, certain groups seem to be overrepresented and others underrepresented) its mainly white guys, but Europe is mainly white.
lagadu 9y ago
I'm going to disagree here on one thing: The pool being biased isn't a problem that the hirer can or should deal with; it's out of their hands because their position in the chain is far further along of where the problem lies. So I would submit that no, an employer shouldn't have to compensate for a pool bias because they have no impact on it but in turn it can only lead to lower long term outcomes for them. Personally I suspect that there's a mix of both cultural (girls play with dolls traditionally) and biological (boys tend to be more attracted to tinkering and building from even before they learn to speak) reasons for the ratio in people studying IT-related fields but I just don't see how making hiring practices biased in any direction will help with that.