> Why is an increase in diagnosis an increase in awareness? Other way around, I would argue increased awareness -> increased diagnoses. Why? Because a lot more

> Why is an increase in diagnosis an increase in awareness? Other way around, I would argue increased awareness -> increased diagnoses. Why? Because a lot more people know what the words "autism" and "aspergers" mean today than 30 years ago. A feedback loop where diagnosis -> incr. awareness -> diagnosis seems plausible and has a nice simple mechanism. So I guess increased diagnoses would lead to increased awareness too. > you can draw the line between autistic and not autistic wherever you please. Agree, this is what makes psych evaluations so tricky. It's hard to have rigour for human mental complexity. My prefered view on this is that autism is defined by a series of correlated traits which are all continous variables, we define autism as someone who has 1 or more of those traits in the tail of that trait's distribution. e.g. very low social ability or very high impact sensory disorders. You can whinge all you like about where the line is drawn, but the fact is the line has to be drawn through six dimensions and none of those dimensions is easy to measure. > various entitlements received post-diagnosis Most people with an autism diagnosis (level 1) do not have access to any entitlements. This is a common argument against self-diagnosis which holds no water. > diagnostic drift also means the vast majority of research on autism is likely non-reproducible and worthless Clinical researchers routinely sub-divide groups with a diagnosis based on severity. You can do this with autism in much the same way we do it with cancer. So broadening the diagnostic criteria doesn't impact statistical effect size and thus your argument is incorrect.