A question I've been having about psychotherapy research is echoed in an editorial piece in the American Journal of Psychiatry I just found.

Said piece is by Arthur Rifkin and is available here. Le quote:

"If we approach the issue scientifically, we quickly reach the central problem of psychotherapy: What is the proper comparison group? Most randomized controlled trials of psychotherapy used a waiting list comparison group. This does not seem adequate. Imagine that you agreed to participate in a study comparing psychotherapy with a waiting list and were assigned to the waiting list. Would not you feel disappointed? Then, if the psychotherapy group did better, would that mean it was effective or that it did not influence the natural course of the illness and those disappointed subjects on the waiting list did worse than the natural course? Clearly we need a better comparison group than a waiting list."