Designing Research Studies on Psychosocial Interventions in Autism

Abstract

To address methodological challenges in research on psychosocial interventions for autism spectrum disorder (ASD), a model was developed for systematically validating and disseminating interventions in a sequence of steps. First, initial efficacy studies are conducted to establish interventions as promising. Next, promising interventions are assembled into a manual, which undergoes pilot-testing. Then, randomized clinical trials test efficacy under controlled conditions. Finally, effectiveness studies evaluate outcomes in community settings. Guidelines for research designs at each step are presented. Based on the model, current priorities in ASD research include (a) preparation for efficacy and effectiveness trials by developing manuals for interventions that have shown promise and (b) initial efficacy studies on interventions for core features of ASD such as social reciprocity.

Toward Effective and Preferred Programming: A Case for the Objective Measurement of Social Validity with Recipients of Behavior-Change Programs

“…We are confident that EIBI based on the principles and procedures of behavior analysis works. We still have a lot of refining to do with regard to the most efficacious aspects of EIBI, and we still have not identified all of the moderators of the effects of EIBI (Herbert, Sharp, & Gaudiano, 2002;Rogers & Vismara, 2008;Smith et al, 2006), but the intervention improves the developmental trajectories of those who receive it (Campbell, 2003;Cohen, Amerine-Dickens, & Smith, 2006;Eldevik, Eikeseth, Jahr, & Smith, 2006;Helt et al, 2008;Howard, Sparkman, Cohen, Green, & Stanislaw, 2005;Lovaas, 1987;Morris, 2009, National Institute of Mental Health, 2007Odom et al, 2003;Smith, Groen, & Wynn, 2000). We know much less, however, about the acceptability of various components of EIBI to the children experiencing them (e.g., most-to-least or least-to-most prompting, amount and pace of prompting, amount of teacher directedness during instruction, types of error correction procedures, highly structured versus more naturalistic instructional conditions, types of motivational systems).…”

ASD intervention research in real world contexts: Refining single case designs

“…While SCDs are sometimes criticized because small numbers can limit generalizability (Smith, et al, 2007), external validity can be strengthened through some relatively simple procedures: including multiple participants (at least 3 replications across participants), clearly describing participants, operationalizing the context where intervention is provided, and replicating the study across sites with independent research groups (R. Horner, et al, 2005). As external validity improves so also does the quality of the evidence base for ASD interventions using SCDs.…”

“…The individual treatment (dependent variable) response story is not well told by group designs where participant data are aggregated to determine effectiveness, masking much of the variability that is often present (R. Horner, et al, 2005;Smith, et al, 2007). SCDs focus on the individual as the unit of concern, and report specific and detailed information on the response of individual participants on repeated occasions throughout the phases of the research.…”

“…Nonetheless, the high regard of RCTs has led to an undervaluing of other research designs and, consequently, fewer studies of the wide variety of interventions in current use in the ASD field (Smith, et al, 2007).…”

“…Testing of interventions in controlled experimental environments prior to application in real world settings (i.e., the efficacy-to-effectiveness pipeline) is viewed widely as rigorous research process (Smith, et al, 2007). However, the attempt to follow rigorous practices may lead researchers to extract narrow versions of interventions used in practice and therefore divorce the intervention from a real-world context.…”