Beliefs of member of group are not representative of beliefs of all members.

My dad told me a story once where he met a gay guy who told him that he only became gay because he could not convince any women to date him. Then for decades after that he thought that all gay people were gay because they all first failed with women.

So lets just focus on the fallacies my dad made and not the fallacies of the gay guy. Which specific fallacies are at work here?

The stereotype fallacy. That one is pretty obvious.

What about the fallacy of composition? This one seems similar to the stereotype fallacy, maybe to the extent that the stereotype fallacy is a set contained within the fallacy of composition.

How about this? This is the same but different. A man belonging to race X proclaims A, B, and C. The man asserts that all members of race X believe A, B, and C. The audience believes him, but man of race X is either mistaken of lying. It is true that a small number of race X believe A, B, and C, but they are in a small minority. The majority of race X believes D, E, and F.

Which fallacies are these?

The most obvious to me is the Hasty Generalization , which is drawing a conclusion based on a small sample size, rather than looking at statistics that are much more in line with the typical or average situation.

More importantly, this kind of thinking is the antithesis of scientific inquiry.

Essentially, Dad was provided an anecdotal story (opinion) that played off his own Confirmation Bias about gay people. He then projected that bias to gay men in general as a Part-to-Whole Fallacious deduction.