Years ago a USC study was sent for review. Looked at how white people reacted (fMRI study, so no actual questions were asked) when they viewed a series of pictures. Black men, various ages, various garbs; white men, same set of ages, same garb. So 18-year-old black (and white) men in the same suit/buttonshirt/t-shirt/hoodie.
As expected, the white men and women reacted quickly with fear or worry about the young black men, esp. in "gang" clothes (whatever that meant in the late 1990s around USC).
Obvious conclusion was white racism.
Except a review said not so fast--you need a control.
So they reran the same stimuli, same setup, with controls. Black people, from the area. Same mix of ages and male/female.
They reached the same conclusion, which took some work. Because they couldn't tell the data apart. The black men and women reacted the same way, to the same extent, seeing black men of varying ages in various outfits, as the white people did. If you didn't know the labels for the data sets you wouldn't see a difference. The researchers worked hard to explain this, and they concluded that what was *really* shown was that the black men and women had been so exposed to white-oriented representations of black men that they'd internalized the white viewpoint and prejudices perfectly. There are other, simpler, explanations.
The actual "correct" finding was simply that the responses were the same--everything after that was political and social commentary so that the researchers could feel "relevant"; the research itself had no data beyond that and so nothing beyond that was based on evidence. The fMRI data don't lie.