[pbhjpbhj]

Re C: you could stand in a place that attracts what you consider to be an unbiased clientèle, ask women if they'll take a test, give them an IQ test (or establish highest level of education or similar). Score them on looks somehow. Compare results and test for correlation.

Or you could take a group with similar intelligence (eg a university class group) and have them rated for looks, compare the result curves with the expected results in general population and look for skew or spread variations.