There's no evidence that one's implicit associations affect their behavior. Also the Implicit Associations Test used to determine your implicit associations in the first place is unreliable. Implicit bias testing and training is pseudoscience at best, especially when you consider that 95% of the professors in that field of study lean left.

> Implicit bias testing and training is pseudoscience at best, especially when you consider that 95% of the professors in that field of study lean left.

This strikes me as kind of like saying that evolution is pseudoscience because 95% of professors in the field are not evangelical Christians, or something.

There may be other reasons to believe that implicit bias is pseudoscience, but "People who believe certain things about it tend to end up with personal worldviews that are consistent with their research" doesn't seem like one.

You're right, it's not a good reason. I'm just skeptical how social sciences can come to fair conclusions when is almost no representation of the political right in their field. However you might be able to say the same about tech CEOs and underrepresented female gender, so I don't have a good argument here. As an aside I feel much less comfortable arguing this point after checking out your personal site and seeing that you write Debian packages. I really enjoy Debian, especially the reproducible builds work that's going on over there. I respect your opinions on this matter.

> I'm just skeptical how social sciences can come to fair conclusions when is almost no representation of the political right in their field.

By applying empiricism.

The fact that the political right is ideologicslly opposed to doing that in social science fields rather than accepting dogma (an attitude which also applies to an increasing number of areas of the physical sciences) is problematic, to be sure.

"If empiricism was the interest of social science, they would abandon all work on $X" is a pretty silly-sounding statement for any value of X.

I do believe that the links in this sibling comment prove you to be a bit too extreme in your opinion: https://news.ycombinator.com/item?id=16128390

Conservatives: "fair" is more important than factual or correctness.

> As an aside I feel much less comfortable arguing this point after checking out your personal site and seeing that you write Debian packages. I really enjoy Debian, especially the reproducible builds work that's going on over there. I respect your opinions on this matter.

Are you trying to mkae some sort of point?

In my argument I was more making the claim that you're going to see more research supporting the left if all your researchers lean left. Also, you could make a much stronger argument that liberals value 'fairness' over factual correctness. Regarding Debian, my point is that I respect someone who supports open source software and code that can be reliably built from source, as it protects our freedoms.

> especially when you consider that 95% of the professors in that field of study lean left.

But they do research that should be repeatable and thus it doesn't matter.

Unless you think they have implicit bias.

> You do not need to hate women to be sexist. You do not even need to be aware that you have biases for those biases to be affecting how you act.

I am aware of that, good to point out though. However I was wondering because the wording "you can still maintain the discrimination" implies awareness and perhaps conspiracy. For example if I was talking about "implicit bias" I might have said something like - "it doesn't modify existing implicit biases and institutional constraints, which have been responsible for ...". But even then it would be good to see what exactly those biases and constraints are at Google.

On the topic of "implicit bias", it's also useful to add that it doesn't have to be just a personal "implicit" bias, it could institutional as well, incentives and rules that combine to exclude certain groups more than others, in this case women. Though then it's not clear how they would fix it, given that it's implicit. Maybe hire an outside consultancy which will be able to identify it better (since they are not part of the culture and not affected by same incentives)...

It'd be nice if you could link to the studies so we can check them out. I've seen a decent amount of criticism on this topic: https://en.wikipedia.org/wiki/Implicit-association_test#Crit...

In one of the sibling comments I linked to a good page on those.

There is valid concern on how predictive the implicit bias test is in non-test conditions. But "implicit bias" is only the start of the conversation - it is one easy thing to measure in a sea of difficult to measure issues. If you want to jump into this rabbit hole, I have found the researches that work in this field to be eager to describe their more up-to-date work on addressing them (I do try to play a devil's advocate in such conversations and have gained much respect for their rigor). Regrettably, as usual in academia, the easies way to be exposed to those conversations is not that easy: going to talks given by those researchers.

Thanks, I'll take a look.

Claiming implicit bias is a very well greased slope towards a bad end because it lets you say (your opinion is that) someone else's actions show a pattern of whatever-ism and that they are not credible because of it and they cannot defend themselves because their whatever-ism is not consciously applied.

Just because brogrammers want to promote other brogrammers does not make them whatever-ist. Claiming that implicit bias is a type of whatever-ism is just BS doublespeak like "if you're not the solution you're the problem".