Drugs, recipes, babies, bathwater, and psychotherapy process-outcome relations

In their critiques of Stiles and Shapiro's (1994) discussion of the process-outcome correlation problem, Silberschatz (1994) and Sechrest (1994) suggested that the problem is not fundamental but merely technical. Silberschatz suggested that more complex measures would solve the problem; Sechrest suggested that more complex analyses would solve the problem. Following Sechrest's multivariate suggestions, however, produced no better result. Contrary to Silberschatz's and Sechrest's suggestions, the problem is not in the measures or the analyses but in the interpretation of the results (null results as well as positive results), particularly in a failure to incorporate fully the phenomenon of responsiveness into an understanding of process-outcome relations.