[codingdave]

Therapy is skills training, not medical treatments. Which is also why the reliability is varied - not everyone wants to learn mental health skills. Not everyone who learns them wants to do them. Not everyone who wants to do them is able to make themselves do so.

When you go to a psychiatrist, now we are talking medical treatments. That is where you get a specific diagnosis and possibly medicine to change how your body and mind work. That is where to expect objective results.

Both explore what is needed by talking to patients, often with some testing... but they are not the same thing. And neither are going to solve all problems for all people. But it helps to evaluate them properly if you understand what they are, and what they are not.

[kragen]

Some skills training is well supported by empirical evidence. People who take piano lessons observably improve their ability to play piano, on average, even though some don't, even though they often regress after they stop, and even though some people improve without taking lessons. Similar remarks apply to golf lessons, JavaScript bootcamps, calculus courses, and plumbing apprenticeships.

Most kinds of therapy, by contrast, are closer in their empirically measured effectiveness to studying geography through astral travel or studying history through past-life regressions. CBT and exposure therapy are among the few exceptions.