THE STUDY OF CHANGE IN PSYCHOTHERAPY - A REEXAMINATION OF THE PROCESS-OUTCOME CORRELATION PARADIGM - COMMENT

W. B. Stiles and D. A. Shapiro (1994) present a provocative argument t hat the yield of process research has been minimal because of research ers' overreliance on the drug metaphor and its corollary, the correlat ional design. Although Stiles and Shapiro raise excellent points regar ding the possible misinterpretations of correlational research, their conclusion that the process-outcome correlation paradigm is flawed and has outlived its usefulness is questioned. The basic thesis is that S tiles and Shapiro did not provide a fair test of the correlation parad igm. The process variables used to test the paradigm do not have the s trong theoretical and empirical grounding necessary to support the ass umption that they should be related to outcome. In this article, examp les are described, of programmatic process research that has used the correlation paradigm, along with other methods of inquiry, to advance the understanding of how change occurs and to improve treatment effica cy. It is contended that the correlational method is one useful tool o f discovery and that it has contributed significantly to the advanceme nt of the field when the process variables studied are grounded in sol id theory and research.