[lbo-talk] Test your racial preference

[John Thornton]

socialismorbarbarism wrote:
> re: https://implicit.harvard.edu/implicit/demo/
>
> I picked "Skin Preference," or whatever it was called, because it was the
> first choice, and "scored" the result:
>
> "Your data suggest little to no automatic preference between Light Skin and
> Dark Skin."
>
> Big whoop.
>
> What I did find interesting: I found myself conditioned to the first choice
> of category (at a midpoint, they switch the "racial" image and
> [connotatively loaded] verbal categories), with myself always trying to
> remember not to go back to the first (rationally arbitrary) pairings. In
> other words, I found the difficulty to be in overcoming the mechanical
> conditioning of the original categories assigned by the test. (There is much
> emphasis on speed, placing the fingers exactly on certain keys, etc.)
>
> That being the case, should I worry that the test constructors first
> assigned the differently-toned images to "evil," "bad," etc.? Since if I
> look at this from the standpoint of classic conditioning, this can be seen
> as what is being "taught" to me by the experiment itself.
>   

Isn't the point to see how successfully you can overcome the initial 
conditioning?
That helps determine how automatic your preference might be.
The more easily you overcome their initial conditioning the less 
automatic your initial conditioning and vice versa.

I have no idea how accurate this is and I just did it for fun to see 
what the test was like more than anything else.
I was surprised at the overall results since I assumed the results would 
skew towards no automatic response but that isn't the case.
I don't see why test wouldn't test what it claims to. It seems a pretty 
limited claim.

John Thornton