A study in 2000 appeared to show a strong biological basis to lesbianism based on a different finger length ratio for lesbians than for female heterosexuals.
Finger length study
Dr NE Whitehead
The essence of this article was published in the NARTH Bulletin, 8(2), 25, 2000.
This is significantly misleading. I report on this study because it is already in the popular press, and misinterpreted.
Williams et al. reported that the mean finger length ratio for lesbians was significantly less than for heterosexual women and did this by comparing the two ratios by a statistical test. They had a large number of interviewees, and in such circumstances, although means may be statistically different, they are often so close that it is not practically useful to say they are different. That is what has happened in the present case.
The original normal distributions can be reconstructed from their data, and the results are shown below in Figure 1 (the figure assumes in its two large overlapping curves that we are comparing an equal number of heterosexual women and lesbians). There is obviously a very large overlap in the two populations and although the means may be statistically different, the difference is only 1% which is a small effect and not diagnostically useful in any sense.
Figure 1 shows that there are large numbers of heterosexual women who have much more "masculine" finger length ratios than most lesbians, and, obviously, do not have a lesbian orientation. Williams et al. invoke the idea of very high pre-natal androgen levels (for which there is very scant evidence) to explain the difference in means which they find, but if this is indeed an explanation it obviously rarely affects sexual orientation. An explanation which involved considerably less biological extrapolation would be preferable. For example, does a slightly more masculine pattern for a hand, influence the self-image of a developing girl?
Figure 1. Distribution of finger length ratios