National discussions of health-care delivery systems have renewed the
effort to identify empirically validated psychotherapies. Although the
logic of efficacy studies in psychotherapy, which are used to empiric
ally establish the efficacy of psychotherapy, is relatively simple, th
e validity of inferences from these studies are ambiguous and the focu
s on outcome obscures other important aspects of psychotherapy. Proble
ms related to (a) common versus specific confounds, (b) difficulties i
n disproving the uniform efficacy supposition, and (c) hazards in stan
dardizing treatments are discussed. These problems make developing cri
teria for establishing empirically validated treatments difficult; a c
urrently proposed set of criteria for proving treatment efficacy is cr
itiqued.