> There is a logical flaw in suggesting that something that occurs with a small percentage of a population such as “detransitioning” implies anything about every member of a population.
> Child abuse exists, but doesn’t imply anything about every parent.
This is funny because that's the exact argument that transphobic opponents say about trans people themselves and the argument as to why gender fluidity or gender outside of sex doesn't exist. "Just because an extremely small number of people believe they are a different gender than their biological sex doesn't mean that gender is different from biological sex" is almost exactly the argument that transphobes use.
I think you fundamentally fail to understand what I just said. Proper unbiased random sampling allows you to create sub populations that tend to reflect the characteristics of a larger group, biased populations don’t share that relationship.
“Because some animals hibernate, all animals hibernate” is just as flawed as saying “Because only a small percentage of hibernate, no animals hibernate.” Instead the relationship is “Because some animals exist that hibernate, there exist animals that hibernate.”