Icing on the Cake: The Role of Research in Practitioner Training

Abstract

Dixon, Reed, Smith, Belisle, and Jackson (Behavior Analysis in Practice 8, 7–15, 2015) evaluated research productivity in BACB training programs. While research productivity may indeed be an important metric for future researchers, we posit that the field consists mainly of practitioners, and as such training should be focused on practice. We provide an alternative perspective on enhancing the quality of education for behavior analysts who will, by and large, go on to be practitioners.

Dixon et al. (2015) began an interesting discussion on the role of research in evaluating behavior analysis programs. Their description of behavior analysis as a “knowledge based industry” (p. 8) is both apt and useful in identifying interdisciplinary frames of reference for concerns about graduate training in our field. While behavior analysis may have originated in the departments of psychology, it is time to determine which training model best creates the desired outcomes of our students. In 1949, the APA agreed upon an approach to a graduate training, in what is now known as the Boulder Model, the goal of which was to provide rigorous training in both research and practice during a 4 year graduate degree. In the intervening years, it has been noted that psychologists trained in the Boulder Model by and large do not engage in research (Shinn 1987). Moreover, there is no convincing evidence of which we are aware to suggest that their additional training in research aids in research consumption when compared to the recent alternative, more practitioner-oriented programs, termed Vail Model programs. Contrast the Boulder Model with that of medicine. When physicians are in school, they likely receive some training in research methodology, but the bulk of their training is designed to prepare them for the naturally occurring contingencies of a career treating sick patients. In fact, the 2016 US News and World Report ranking of best medical schools provides separate lists for “Research” and “Primary Care.” There is no overlap between the top 10 schools represented on these lists. It is reasonable to assume that medical schools that emphasize research training compromise on preparation for practice, and those that emphasize practice training compromise on preparation for conducting research. This suggests that faculty research productivity may be a good indicator of a program’s capacity to prepare students for research careers, but not necessarily for practice careers.

Like physicians, the majority of practicing behavior analysts rely on research but probably spend little, if any, time conducting it. There are several reasons why many behavior analysts do not maintain active research programs, including competing contingencies, lack of easy access to research databases, lack of access to institutional review boards, and lack of expertise (in the last case, we would argue that research is likely the most difficult task a behavior analyst could undertake). If most behavior analysts do not generate research, the real question becomes: Does exposure to productive research faculty during graduate training increase one’s ability to interpret and apply research findings after graduation? While this is an empirical question, we suspect not.

If research training is not vital to would be practitioners, then—just as practice-focused medical schools presumably have done—our field must determine what kind of training is essential. Psychologists and medical doctors usually receive 4 years of graduate education. Most practicing behavior analysts, on the other hand, receive about half of that. This disparity increases the need for exacting selection of how and what we train. According to the BACB® requirements, it appears that to properly train a new behavior analyst we must, at minimum, provide a student with course work that provides specific information covering prescribed topics, and a specific amount of clinical experience. ABAI accreditation requires a thesis or thesis-like project as well. The BACB® approval process involves reviewing the curricula provided by potential programs and course sequences for breadth of content and allocation of instructional time. The BACB® has also recently begun releasing annual pass rate data that could be said to be a measure of instructional quality. However, the behavior analytic community may appreciate metrics of programs such as quality of instruction, differential quality of students accepted, and courses provided beyond minimum competencies maintained by the BACB®. Metrics of this sort may be beneficial along with BACB® pass rates to aid students and the behavior analytic community at large in evaluating programs.

Another variable worth considering is whether or not a program provides the clinical experience required for certification. Some programs provide only the BACB® required instructional course sequence and students must independently arrange for their own supervised experience. We believe that a better model would be for a program to adopt the physician training model of rotations in practical training. In medical school after students complete the first 2 years of training, the third year tends to focus on a series of rotations in various specialties to provide breadth of experience (Hirsh et al. 2007). This is repeated in the fourth year but typically with more electives designed to facilitate career selection (Lyss-Lerman et al. 2009). A system of this sort would allow future behavior analytic professionals the opportunity to work with varied populations, under numerous supervisors, and engage in all of the activities they will be expected to complete after obtaining their certification. We believe this would likely be more beneficial than extensive research training. Although it would be difficult to develop an objective means of evaluating the quality and breadth of supervised experience provided by each program, this information should be of interest to students and to prospective employers of program graduates.

In sum, the majority of behavior analysts are engaged in practice, and they need to be able to interpret and apply rather than conduct research. It is an open question whether accomplished researchers are required to teach this skill set (but we would love to see data bearing on this). The practitioners also need targeted clinical experiences with high-quality supervision and accurate behavior analytic lessons delivered in a scientifically sound manner. While Dixon et al. (2015) raise an interesting topic that may very well be germane to potential researchers, the majority of our field consists of practitioners and they should be trained as such. Of course, if we can accomplish these goals while teaching practitioners to conduct quality research then we can have our cake and eat it too.