Scott Alexander explains how psychiatric conditions like ADHD can be both over- and underdiagnosed simultaneously, due to statistical quirks in diagnosis and population prevalence. Longer summary

Scott Alexander discusses the counterintuitive phenomenon of simultaneous over- and underdiagnosis in psychiatry, using ADHD as an example. He explains how, even with a highly accurate psychiatrist, it's possible for the majority of people diagnosed with a condition to not have it, while the majority of people who actually have the condition remain undiagnosed. This is due to the interplay between disease prevalence, self-selection for evaluation, and the inherent limitations of diagnostic accuracy. The post demonstrates this concept using a hypothetical scenario with ADHD, and then suggests that this phenomenon likely applies to other conditions such as depression and pain medication prescriptions. Shorter summary