Then perhaps the real data point is bias about biases: operating in a profoundly neutral forum, blind to practically any "group" indication, individual "they're thinking X of me because I'm Y" bias emerges. This is something being increasingly noticed in general, but SO's study provides an interesting control experiment precisely because there is so little indication of group categorization.

I've spent years using StackOverflow. I've never seen any expressions of bigotry. One might get a hint of "group" based on linguistic style, but never have I seen bigoted comments based thereon. If someone feels unwelcome, it's either because they just feel that way without cause, or because of the content/tone of their comments - not their "group" - elicits such response.

While I do think it is true that you can only find "hints" of group bias when examining individual stackoverflow users, stackoverflow in general is known to have:

1. overwhelmingly dominant gender of population (92.9% of SO population)

2. overwhelmingly dominant race and ethnicity categorization (74.2% of SO population)

3. overwhelmingly dominant sexual orientation (93.2% of SO population).

(I left this ambiguous on purpose to ensure that my point is about the homogeneity of the population and not the specifically dominant demographic group. If you want more information, you can look for yourself for specifics about which way the numbers fall here - https://insights.stackoverflow.com/survey/2018/).

My point is that the stackoverflow population in general is very homogeneous by certain categories.

Instead of looking for evidence of victimization, imagine how easy it would be for someone belonging to a StackOverflow minority category to get discouraged by StackOverflows fierce gatekeeping, especially with the knowledge that stack overflow is so demographically unlike this person. From the point of view of the marginalized, evidence of group categorization isn't needed beyond that fact that "StackOverflow is very very X, and I know that I'm Y." Any sort of gatekeeping from this point on can feel very discouraging, even if this isn't the intention of those who are gatekeeping.

[edit for format]

What "gatekeeping"? The whole fascinating point is that there isn't any "gatekeeping" on SO. The demographics may be remarkably skewed, yes ... not for any categorical exclusion, but for self-selected choice to not engage.

If people choose to avoid a group, while that group in no way excludes them, why keep blaming the group?

By "gatekeeping", I'm referring to the efforts made to exclude certain content. By this definition (which I probably should have defined earlier), I would say that SO is founded on gatekeeping and that gatekeeping even makes SO the great resource that it is. After all, the quality of content on SO would surely go down if low quality content wasn't downvoted.

So I'm not "blaming the group", but instead I'm trying to make the point that when the current SO system is paired with a homogeneous user base will, it probably naturally become more an more homogeneous for the reasons stated above, unless an outside force enters the system.

I would argue that this isn't about fault or blame or guilt, but instead about deciding whether or not to take action to make SO a more inclusive place, especially for those who may be predisposed to feel exclusion in the context of SO.

For whatever it's worth, I'm a member of an out group on at least one of these and left the corresponding questions blank because I found them asinine in the context of that survey.

Perhaps that was a mistake; I'll reconsider next time around.

Back in the '90s after the Gulf War, many veterans started coming down with weird symptoms, eventually termed "Gulf War Syndrome." Some of these were quite real, but studies found the Gulf War veterans were, on the whole, healthier than the general population. But the guys who served there were convinced of it, and it made them pretty miserable to think they were sick and the government was covering it up.

When you tell someone they're sick, or unwelcome, or subject to bias, and especially if you're a neutral source, they're liable to believe you. The mind is incredibly suggestible.

So how much of these feelings are because we keep telling people that racism and other issues are a problem? I understand the impulse to not dismiss the problem, but it seems like feeling like you're surrounded by racists has to be almost as bad as the real thing.

If that's true, I'd be really interested in knowing why. It's pretty hard to fix a problem if you don't even know what's causing it in the first place. (Especially since we've pretty much ruled out the possibility of racial or sexual discrimination.)

People can experience the same thing and have different reactions to it. Someone that already feels inadequate or like they don't belong as a programmer are likely to have a more negative reaction to SO's nonsense.