This is the last time I'm going to post on this. It is not my mission to defend anyone's research. These links were not provided to support an assertion, they were provided to answer theprestige's question if a possible hypothesis I presented had any basis in science. The answer is, yes, it does and those studies demonstrate that, regardless of any flaws they may have.
Yes, but if the effect re transgender disappears after controlling for homosexuality then the result isn't valid for transgender. It's like finding a correlation between people wearing shorts and people eating ice cream. Sure, the correlation is there, but it's not like wearing shorts causes someone to eat ice cream or vice versa, it's just that both wearing shorts and eating ice cream are correlated with it being a hot day.
There was one such study with that flaw, but that's not the flaw I'm talking about. The flaw I'm talking about is taking a group of male transgender homosexuals and comparing them with a group of male cisgender heterosexuals, and then ascribing the difference to transgender whereas the difference could be explained by sexual orientation. Basically the flaw is using an improper control group that differs on more than the variable being investigated.
This would, presumably, apply to the study on the androgen receptors.
The groups were sorted by trans/not-trans directly. The gene types were measured directly. Therefore there would be a statistically significant correlation between trans and the receptor types.
Homosexuality also having a correlation would not be an interference.
If by selecting trans-people the result is an over-representation of homosexuals, that suggests that homosexuality and transgenderism are linked. Which does not eliminate the correlation.
However, I agree that the research is not conclusive. One of the limitations of population studies is that it is impossible to control for all variables. Other variables could be hair color, eye color, geographical origin of ancestors, and any number of things. Homosexuality is just a single one.
It may also be that, if there is a gene (or set of genes) that predisposes trans-gender, it is not the one for the androgen receptor variant, but a linked gene. (Gene 1 does not cause trait A, but is highly correlated to Trait A because it is located on the chromosome in close proximity to Gene 2, which
does cause trait A.
And for further clarification: the existence of flaws does not debunk or discredit a study. It just points out that more work is needed before conclusive conclusions can be drawn. (Often, you will see this pointed out in the papers themselves.)
True, but that doesn't mean the issue of homosexuality as a confounding factor isn't present there either. Suppose we have a group of 100 transgenders with twins. We find an 80% concordance for their twins. Now suppose we control for homosexuality, that 80 of them are homosexual and 20 are heterosexual. All 80 homosexual ones are concordant with their twins and all 20 heterosexual ones are discordant with their twins. Then the correct interpretation is not that a biological basis for transgenderism is found but one for homosexuality.
Not quite. First, you are correct that the results
would suggest a biological basis for homosexuality. But the results would also suggest a biological basis for transgenderism
within the group of homosexual transgender people. In other words, such a result would suggest that homosexual transgenderism and heterosexual transgenderism are not the same thing. (Which, kind of falls within the general scope of Blanchard's ideas, I think.)
If, on the other hand, it turns out that 65 of the homosexual ones are concordant and 15 of the heterosexual ones, then the correct interpretation is indeed that a biological basis for transgenderism has been found. However we can not distinguish between these possibilities unless the study explicitly controls for sexual orientation.
Partial agreement: Controlling for sexuality may allow for finer resolution of the results: Does this apply for trans-genderism overall, or just for homosexual transgender people or just for heterosexual transgender people?
But in neither case does the correlation disappear.
There are a bunch of studies showing that about 80% of transgender youth grow up to be cisgender homosexuals, I could go look for them if you want although a quick google search should find them. Whether that means that transgenderism is a possible outcome of homosexuality, or that homosexuality is a possible outcome of transgenderism, or that there is some other unknown factor influencing both is hard to tell, but that there is a strong association between transgenderism and homosexuality is clear. You can just take a look at the studies you've been linking to as well, homosexuality is vastly overrepresented in the transgender population (something like 80% of transgenders being homosexual, as opposed to like 5% in the general population). This (the vast majority of transgenders being homosexual) combined with there already being known biological markers for homosexuality (both in brain scans and twin studies), complicates studies into biological markers for transgenderism. Hence the need to control for it.
Given the subjectivity of the diagnosis criteria, it's kind of hard to say who is/was transgender and who was mis-diagnosed as transgender.