Ask a Female Engineer: How Can Managers Help Retain Technical Women?

Interestingly, I've heard women working in those fields lament the lack of male representation. In teaching and child care, some kids just respond better to men. In nursing, it's because men are generally going to have an easier time providing care for patients who are overweight.

Which one of these e-mails do you think a boss looking to seduce his employee is more likely to send?

1. That's not the problem of anybody except the person who does not negotiate aggressively.

2. People respond differently to men and women (a fact that seems to be true across all species). We don't have a clear, testable model for why this happens.

3. Implicit bias is not scientific and not relevant. Sorry for the lengthy write-up, but this should be called out for being farcical and/or psychologically dangerous.

Implicit bias scores have no meaningful correlation with actual actions and the scores themselves will vary wildly for the same person for reasons that are unknown and unaccounted for. This lack of correlation turns it into a baseless accusation.

Lots of people accept IAT (which has a super low predictive validity) as true while rejecting IQ which has an amazing amount of predictive validity in all kinds of areas. Interestingly, higher IQ is linked to lower rates of racism and other bias. Given that programmers have on average significantly higher IQ scores relative to the general population, you would expect such things to not be the primary issues of the field.

Even if we assume implicit bias exists, does it matter if you cannot prove it has any real world effects (and such effects have not been proven). If we assume even further (against good evidence) that such implicit bias exists AND has real world effects, is there a solution?

Lots of companies force their employees to undergo implicit bias training. There are ZERO studies proving that the training works. Much more scarily, we don't know if it has other negative psychological impacts. Companies messing around with peoples heads without any supporting studies and clinical trials is EXTREMELY dangerous.

Finally, let's examine the theoretical framework for fixing implicit bias (under the assumption that it exists and that the individual does not consciously desire to be biased). We must start by either assuming that implicit outweighs explicit (radical behavioralism with no free will) or that explicit outweighs implicit.

If explicit outweighs implicit, then teach people the golden rule, tell them discrimination is bad, and call it a day because that's all you can do. Implicit bias training is sold on the basis that this is not enough and will not work. It thus carries the implication that you cannot choose to not be racist, sexist, homophobic, etc, but if you give them your money, they have the magic incantations to "pray the away".

How do you change the subconscious? The first option is new age garbage. The second is hypnosis (we can't prove hypnosis even exists). The third is operant and/or classical conditioning of the most extreme variety (to reach past the conscious and modify the subconscious -- hopefully without ripping the individual apart in the process) may be enough to do the trick (immersive re-education camps are a must in order to ensure all aspects of the subject are controlled for). Basically no other ideas exist.

TL;DR With no solid supporting science behind the idea of implicit bias, there remains only moral panic and ideological fanaticism attempting to sell either snake-oil cures or barbaric, unethical psychological experiments of a most Stalinesque variety.

This is a topic of great interest to me. In part because I love coding and think that anyone else that does should have the same opportunities. In part because though I'm a cis straight male, I've generally gotten along with women better as friends and colleagues. And in part because I'm generally socially liberal, so something like this where the TREND line is going in the wrong direction is particularly bothersome. (Racism, on the other hand, is something that is definitely wrong and definitely exists, but it's easier (though not "good") to be blase about it because the trend line is in the right direction.)

I've had two relatively recent discoveries on the topic (these may or may not be known to everyone else):

* One workplace I was at had an annual summer internship for high school women that had taken AP Comp Sci. Overwhelming the interns had a good experience, and a thing they commonly said was that they didn't expect coding to be so social. Given how U.S. women (at least) are socialized to be social and being judged by their success at being social, I can see that impression subconsciously turning away potential CS grads from ever entering the field.

* I've attended and read talks on the problem for the last decade or two, and I've seen a change in the messaging. Originally it was "women need to stand up, get mentors, and push harder". More recently, the messaging has been "Common male behaviors tend to push women away. Those places that are more diverse have been more successful. Let's reduce how much we push women to socialize like men, and perhaps teach men some of the positive ways women socialize".

That is, teach behaviors like:

* Don't shut down discussion, promote it

* Don't insult ideas

* Don't talk over people

* Make suggestions, not declarations

Now, I'm not saying that the behaviors women are taught are all good or positive, but within the realm of building consensus and being open to new ideas, these specific behaviors have been shown to work better.

Also, this change in messaging has been more useful to me: Originally, if I wanted to promote women in the field the advice was "Don't be a sexist jerk, and provide mentoring for junior coders that are women". Now the advice is "look to your own behaviors and improve them, and both men and women will benefit". It's way too early to say if this messaging is really an answer, but I feel like I've made positive changes based on it.

[ Side note: I've noticed that comments like these tend to inevitably kick off comments about how backstabbing and bitter the socialization between women is, a position often vigorously promoted by women themselves. I really can't say, other than the specific qualities I list above are positive ]

I see