I'm of the impression that most racial bias research depends on the implicit associations test, which hasn't been shown to actually measure bias and is apparently plagued by several other weaknesses including low reproducibility. I'm curious to know more.

A lot of what I've seen is based on tests of behavior toward others in scenarios where factors other than indicators race (skin tone in photos or illustrations, race-stereotypical names in case files, etc.) are identical. These are direct measures of bias that manifests in discrimination, bypassing any problems with the IAT.

Can you recall any identifying information about those studies that might make them easier to Google for? Of course, links are also appreciated!

Here's a 2015 NYTimes article with summaries of and links to a number of studies of this type:

https://mobile.nytimes.com/2015/01/04/upshot/the-measuring-s...

Awesome! Thanks very much. I'll dig in ASAP.