This is a very interesting research you are doing Kevin.
Let me try to analyze this from a scientific point of view. The hypothesis being tested here is that gays' (natural) physical appearance is different from straights'. In order to test this hypothesis we need to choose two random and representative samples: (1) sample of gays and (2) sample of straights, and compare various features of their appearance. There are thousands of possible features we could look at: height, hair swirl, length of fingers, length and size of other body parts, etc. Once the various physical aspects were measured for each individual in sample 1 (gays) and sample 2 (straight), there are standard tests to determine if the differences between the samples are statistically significant.
Ideally, if the census also documented the sexual orientation and various measures of appearance, then we would be able to statistically test the hypothesis that gays and straights have certain differences in appearance. Given the limitations we are facing here on GA, I would try the following poll choices, to compare the fraction of left handed:
1. gay, Left-handed
2. gay, Right-handed
3. gay, Ambidextrous
4. straight, Left-handed
5. straight, Right-handed
6. straight, Ambidextrous
If I had these data, I could test the hypothesis that the fraction of left handed among gays is statistically greater than among straights.
Kevin presented an interesting challenge.
Michael.