> Quibble about my word choice all you'd like, but use a direct IQ test at your own peril.
Insofar as its true that such a peril exists, the same peril applies for any "test that acts as a good proxy" for an IQ test, since if it does so it will have the same unequal racial impact (because if it didn't, it wouldn't be a good proxy) and, therefore, require the exact same evidence of predictive power with regard to job performance as a "direct IQ test". So the substitution you suggest achieves nothing.
Protecting against the "peril" associated with a "direct IQ test" isn't a matter of choosing a "good proxy" instead (since that does nothing to reduce the "peril"), its a matter of documenting the business justification -- the evidence that the test is a good predictor of job performance.
Insofar as its true that such a peril exists, the same peril applies for any "test that acts as a good proxy" for an IQ test, since if it does so it will have the same unequal racial impact (because if it didn't, it wouldn't be a good proxy) and, therefore, require the exact same evidence of predictive power with regard to job performance as a "direct IQ test". So the substitution you suggest achieves nothing.
Protecting against the "peril" associated with a "direct IQ test" isn't a matter of choosing a "good proxy" instead (since that does nothing to reduce the "peril"), its a matter of documenting the business justification -- the evidence that the test is a good predictor of job performance.