> normally distributed IQ arguably follows a normal distribution because such tests are constructed to do so. This leads to the criticism that useful measures

> normally distributed IQ arguably follows a normal distribution because such tests are constructed to do so. This leads to the criticism that useful measures of intelligence are excluded because they have a threshold effect which can't be sensibly normalized ( https://en.wikipedia.org/wiki/Intelligence_quotient#Systematic_exclusion_of_threshold_effects https://en.wikipedia.org/wiki/Intelligence_quotient#Systemat... ). Disorders typically refer to such threshold effects, where rather than expected variance, there is a distinct anomalous population. ADHD is such a case where there is a threshold, which appears to be most commonly caused by a genetic factor affecting the dopamine transmission system. Although that's not to say the distribution function isn't continuous, and a person's membership in a statistical population isn't arbitrary.