Successive Approximations towards Behavior Analysis: A Review of Single-Case Evaluation by Social Workers by Mansoor A.F. Kazi

Article excerpt


As readers of this journal know, the field of behavior analysis (BA) is characterized by four major features: (a) interventions are based on empirically-- supported principles of learning, (b) data are gathered and reported on individual participants and formatted using single-case research designs (SSDs), (c) data consists of direct observations of behavior, and (d) visual inspection of the data permits causal inferences to be made pertaining to a functional relationship between intervention and outcome. Studies characterized by all of these features represent 'radical' or complete behavior analysis, but from the very inception of our field reports have appeared which do not include all four defining features. Indeed, the majority of contemporary literature on so-called behavior therapy fails to include all four elements. Many reports (e.g., pilot studies) use research designs with low internal validity, precluding a complete functional analysis. Sometimes non-behavioral interventions are evaluated using single-case designs. The permutations are manifold.

In general, the more closely a discipline adheres to conducting research featuring all four of BA's defining characteristics, the greater the likelihood that it will make significant advances as a branch of natural science. However, the value of successive approximations should not be overlooked in promoting progress along these lines, which is why Mansoor A. F. Kazi's book Single-case Evaluation by Social Workers is such a noteworthy accomplishment. Mansoor Kazi is the Director of the Center for Evaluation Studies of the School of Human and Health Sciences at the University of Huddersfield in the United Kingdom (about 20 miles from Manchester, in northwest England). He holds the academic rank of Senior Lecturer, roughly equivalent to an Associate Professor within the North American system. Over the past decade he has been assiduously promoting the use of singlecase research designs within the field of social work and in related disciplines (e.g., education, probation services, rehabilitation).

Kazi has been consulting and lecturing in the United Kingdom, Europe, the United States, and Canada, describing the methodology and application of SSDs to various non-behavior analytic audiences comprised of social workers, teachers, probation and parole officers, health care workers, agency administrators, and program evaluators. His venues have been actual agencies, as well as professional conferences, journal articles and books (Kazi, 1996, 1997a, 1997b, 1998a, 1998b, Kazi & Wilson, 1996a, 1996b; Kazi, Mantysaari & Rostila, 1997). Entire agencies have been persuaded to make use of SSDs and the results of some of these endeavors comprise the gist of the present volume.

The initial two chapters present an practical overview of the general principles of using SSDs, and is followed by three successive chapters presenting the actual use of these designs in educational, probation, and adult rehabilitation settings, respectively. Chapter 6 discusses the strengths and limitations of using SSDs as an evaluation tool, and the seventh (final chapter) places the use of SSDs into the context of some contemporary philosophies of sciences much talked about in the program evaluation literatures, approaches that have been labeled as `empirical practice,' 'pragmatism or methodological pluralism,' 'critical theory,' and 'scientific realism.'

For example, Kazi conducted a project in one school district that involved training 29 social workers in the use of SSD, of whom 21 eventually actually applied these in their everyday practice, resulting in a total of 125 different applications of SSDs. Most of the cases presented in Kazi's book fall on the low end of the continuum of internally valid research designs, tending to be very heavy on the B (concurrent introduction of measurement and treatment) and A-B (baseline data followed by treatment data) designs. …