kenberg, on 2019-September-06, 12:17, said:
The Harvard article includes some caveats, and as I recall the Post article did also. For example I would say that there is a big difference between "Once had a homosexual experience of some sort" and "Chose a life partner of the same sex". And then there is this issue of just what this 8 to 25 means. From the Harvard article: "In fact, the team estimated that the genetic variants they studied could predict, at best, somewhere between 8 percent and 25 percent of the reported variation in the entire cohort's sexual behavior." Meaning what, exactly?
[size="3"]
This is normal statistics language. What it's describing is the degree of correlation between the genetic variation and the behavior variation.
For comparison, you've probably heard of BRCA, the so-called "breast cancer gene". There are actually two genes, called BRCA1 and BRCA2. The normal risk for a woman to get breast cancer is about 12%. But about 0.25% of the population have mutations in one of these genes (the normal genes actually protect against breast cancer -- genes are often named after the effects found in people with mutated variants), and in this case their chance of getting breast cancer is around 50%. I'm not sure exactly how this translates into the kind of percentage reported from the above study, but the point is that there's a relatively strong correlation between the genes and the disease (enough that some women who find out they have the gene get voluntary mastectomies, rather than wait to see if they get cancer -- actress Angelina Jolie did this).
I think that the 8-25% range cited above refers to the fact that there are a number of genes that are believed to be related to homosexuality, and they have different levels of correlation. Some are as low as 8%, while others are up to 25%. Similarly, my above BRCA description was simplified -- the estimated risk from the BRCA1 mutation is 55-65%, while BRCA2 is only 45%, and there's also a 3rd gene, PALB2, which has about 35% risk; so we might say that genetics predicts 23-53% of the risk (just being a woman predicts 12% of the risk).