Single-Case Methodology in Psychotherapy Process and Outcome Research Russell B. Hilliard Department of Psychology Vanderbilt University ABSTRACT The primary focus of this article is on delineating and explaining the essential characteristics of single-case research methodology applied within the domain of psychotherapy research. Single-case research is presented as a subclass of intrasubject research in which aggregation across Ss is avoided and the generality of one's findings is addressed through replication on a case-by-case basis. The basic ways in which single-case designs vary are also discussed, and 3 basic types of single-case research are differentiated: (a) single-case experiments, (b) single-case quantitative analyses, and (c) case studies. Furthermore, some of the major weaknesses in current single-case psychotherapy research are identified. I express deep appreciation to Hans H. Strupp, my mentor, whose unabated scientific interest in single-case methodology inspired this article. I also thank Irene Elkin, William P. Henry, Steven D. Hollon, and Andrew J. Tomarken for their insightful comments on earlier versions of this article. Russell B. Hilliard is currently at The Cambridge Hospital/Harvard Medical School, Cambridge, Massachusetts. Correspondence should be addressed to Russell B. Hilliard in care of Hans H. Strupp. Correspondence may be addressed to Russell B. Hilliard, Department of Psychology, Vanderbilt University, 301 Wilson Hall, Nashville, Tennessee, 37240. Received: August 5, 1992 Revised: October 15, 1992 Accepted: October 16, 1992 Single-case designs have suffered from association with the research design that Campbell and Stanley (1963) originally referred to as the "one-shot case study." The term was used to refer to a design that had no control group, assessed the dependent variable only once, and was not replicated. In their words, "such studies have such a total absence of control as to be of almost no scientific value. The design is introduced here as a minimum reference point" (p. 6). Although many researchers now associate this type of design with case studies, and even with single-case designs in general, Campbell ( Cook & Campbell, 1979 ) has moved away from such an evaluation of case study methodology, stating more recently, "Certainly the case study as normally practiced should not be demeaned by identification with the one-group posttest-only design" (p. 96), and furthermore, "While it is probable that many case studies professing or implying interpretation or explanation, or relating the case to theory, are guilty of these faults, it now seems to me clear that not all are, or need be" ( Campbell, 1979 , p. 57). He has gone on to write the foreword to a volume dedicated exclusively to case study methodology ( Yin, 1989 ). A similar shift is seen in the writings of a leading psychotherapy research methodologist. Kiesler (1981) Journal of Consulting and Clinical Psychology © 1993 by the American Psychological Association June 1993 Vol. 61, No. 3, 373-380 For personal use only--not for distribution. Page 1 of 1310/15/2001http://spider.apa.org/ftdocs/ccp/1993/june/ccp613373.htmlhas stated Both Hersen and Barlow (1976) and Gelso (1979) take me to task, and rightly so, for my earlier conclusion that single case study "has little place in the confirmatory aspect of scientific activity" (Kiesler, 1971, p. 66). As it stands, this statement is inadequate, contains ambiguities, and needs correction and clarification. (p. 213) More recently, he has gone further, claiming, "Studies seriously pursuing these [psychotherapy] change—process goals cannot attain them by use of traditional, rigorous experimental or nomothetic designs. Instead, what seems to be most appropriate and necessary are small N or single-case studies" ( 1983, p. 13 ). Kiesler's view is representative of the widespread resurgence of interest in single-case designs within psychotherapy research. However, little systematic attention has been given to describing what single-case research actually is. 1 The wide variety of terms used by both advocates and critics to refer to single-case research– single case, N of 1, case study, small N, idiographic, intensive, discovery-oriented, intrasubject, and time-series –have only served to confuse the underlying methodological issues involved in this type of research. The primary focus of this article is on delineating and explaining the essential characteristics of single-case research methodology. Also, some basic ways in which single-case designs may vary will be identified, and separate terms for the different basic types of single-case research will be proposed. My goal is to create greater clarity in thinking and communicating about single-case designs and thus facilitate a more productive discussion of the potential and limitations of such research in studying the psychotherapy change process. Furthermore, some of the major weaknesses in current single-case research within psychotherapy are identified here. The Basic Logic of Single-Case Research I turn first to the question, What is single-case research and why would one choose such a research strategy? Single-case research is often viewed merely in terms of its sample size; that is, as N -of-1 research. From the perspective of group research, a design with an N of 1 is at best highly limited, if not outright meaningless. However, viewing single-case research along the same lines as group research is highly misleading. Such a view reflects an inadequate appreciation of what single-case research actually is. I propose that single-case research is best viewed as a subclass of intrasubject research in which aggregation across cases is avoided and the generality of one's findings is addressed through replication on a case-by-case basis. Each of these basic characteristics of single-case research will be discussed in turn. The focus here is primarily on the conceptual reasons for pursuing this methodology, although I will consider pragmatic reasons in passing. Single-Case Research as Intrasubject Research It is necessary to clearly distinguish between two basic types of variation within psychotherapy research and other areas of psychology–namely, variation within subjects and variation across subjects. The former is commonly referred to as intrasubject variation and the latter as intersubject variation. Intrasubject variation implies that a variable is free to vary
View Full Document
Unlocking...