[logicalmonster]

It's easy to hand-wave the discussion away by citing a bunch of links, but taking a gander over that Wikipedia page seems to verify much of what I stated about bad statistics. Much of what's mentioned is comparing the median of giant groups rather than making an apples to apples comparison.

That's the same bad statistical approach that lets activists claim that women make a fraction of what men make: they compare the incomes of all women to all men rather than taking into account individual choices, education, career fields, etc. When you do a good comparison (comparing women with equivalent careers, education, ages, etc to men) the wage gap between men and women is blurred.

For this question about Black loans, all I'm interested in is an apples to apples comparison. Compare loan rates of Black people in X career with X education making $XXX with XXX+ credit scores to the White, Latino, and Asian equivalent. That's the better approach to see if bias exists. Does that study exist? You tell me because I'd be interested in it.

[WorldMaker]

It's easy to claim no evidence exists if you get to be really picky about which evidence you will accept. Sociologists don't get clean room labs and from scratch experimental design for their studies and have to work with the available data.

That "perfect" study you are looking for in fact can't quite exist because you can't control for all variables with respect to any systemic issue and variables like career and education are likely too deeply connected co-factors with housing.

It doesn't sound like you have any interest in being convinced, and it sounds like you are happy being a contrarian here.

[logicalmonster]

> It's easy to claim no evidence exists if you get to be really picky about which evidence you will accept.

It's easy to claim all of the evidence in the world exists when you're willing to accept very bad statistics and interpretations.

See how that works?

I'm very simply asking for something other than very obvious bad statistics or bad statistical interpretations. I think this is very important because bad stats or interpretations can cause bad policy decisions.

> Sociologists don't get clean room labs and from scratch experimental design for their studies and have to work with the available data.

Are you saying that this data can't exist, or can't even be attempted to be found? No sociologist has thought to even attempt to come up with a sample of this data to try and analyze this correctly? If this is the case, why?

> That "perfect" study you are looking for in fact can't quite exist because you can't control for all variables with respect to any systemic issue and variables like career and education are likely too deeply connected co-factors with housing.

Hold your horses, buddy. I'm not looking for some fictional "perfect" study. Every study, whether in physics or sociology, probably has flaws. I'm not looking for perfection, I'm looking for an honest attempt to at least make a good statistical comparison.

> It doesn't sound like you have any interest in being convinced, and it sounds like you are happy being a contrarian here.

Mind-reading isn't a thing, but I'm generally a happy person, thank you.