Association for Behavior Analysis International

The Association for Behavior Analysis International® (ABAI) is a nonprofit membership organization with the mission to contribute to the well-being of society by developing, enhancing, and supporting the growth and vitality of the science of behavior analysis through research, education, and practice.


36th Annual Convention; San Antonio, TX; 2010

Event Details

Previous Page


Symposium #160
Research Synthesis of Single-Subject Experimental Designs: A Critical Survey of Current Methodologies
Sunday, May 30, 2010
9:00 AM–10:20 AM
215 (CC)
Domain: Theory
Chair: Austin Mulloy (Meadows Center for Preventing Educational Risk)
Discussant: Richard I. Parker (Texas A & M University)
Abstract: Synthesis of research literature in an effort to identify evidence-based practices can involve a variety of challenges pertaining to selection and application of appropriate meta-analytic and other review methods. In this symposium, we discuss recent progress in the development of synthesis strategies for single-subject research, and offer illustrative examples from our meta-analytic work.
Meta-Analysis of Single-Subject Experimental Designs Using Nonregression-Based Techniques
OLIVER WENDT (Purdue University)
Abstract: This presentation will focus on the applicability of non-regression based procedures for the meta-analysis of single-subject experimental designs. Different metrics will be compared relative to their performance in detecting and aggregating treatment effect in behavior increase studies and behavior reduction studies. Data sets from recent meta-analyses of single-subject research related to augmentative and alternative communication (behavior increase) and Functional Communication Training (behavior reduction) will be used to illustrate performance differences, advantages and disadvantages of each technique and the relationship among the different metrics. Patterns will be highlighted of instances where the different metrics yield discrepancies or perfect agreement in effect size scores. A comparison to multi-level models of estimating treatment effect will reveal the fundamental differences between the two approaches. Merits and limitations relative to determining treatment effect for the individual case, testing overall effect, and exploring generality of effect across cases will be reported. Recommendations will be derived for researchers planning to conduct a systematic review or meta-analysis of single-subject research.
Multi-Level Modeling and Regression-Based Meta-Analytic Techniques for Single-Subject Data
AUSTIN MULLOY (Meadows Center for Preventing Educational Risk), Mark F. O'Reilly (The Meadows Center for Preventing Educational Risk)
Abstract: Abstract: In quantitative syntheses conducted with non-parametric summary statistics (e.g., percentage of non-overlapping data), a variety of common characteristics of single-subject experimental data (e.g., small samples of behavior, trended data) have the potential to confound results and/ or place limits on which research questions one can ask. Multi-level modeling and regression-based techniques are two parametric synthesis methods that can potentially overcome the confounds and limits encountered with non-parametric methods. This presentation addresses how to use multi-level modeling and regression-based techniques to meaningfully aggregate single-subject experimental data, and examine the impact of moderator variables. The comparative strengths and limitations of each method, as well as what types of research questions they allow one to ask, will be highlighted. Results of the procedures, as applied to a large sample of studies of treatments for self-injurious behavior, will illustrate the discussion.
Combining Qualitative and Quantitative Methods in Research Syntheses
RUSSELL LANG (University of California, Santa Barbara), April Regester (University of California, Santa Barbara)
Abstract: This presentation discusses writing systematic reviews of literature when the studies being reviewed involve a variety of research designs (e.g., single-subject and group designs). Such a review primarily involves narrative synthesis techniques, but can be strengthened by various quantitative metrics (e.g., PND). An example of this style literature review will be given by detailing the methods and results of a paper that systematically reviews the effects of gluten-free and casein-free (GFCF) diets in the treatment of autism spectrum disorders (ASD). This review involves a multi-step search procedure and independently extracted data from multiple authors. Disparate research designs dictated a qualitative approach to synthesis but quantitative metrics and inter-rater agreement measures were also utilized. Each reviewed study was analyzed and summarized in terms of: (a) participants, (b) specifics of the intervention, (c) dependent variables, (d) results, and (e) certainty of evidence. Findings of the review will also be presented.



Back to Top
Modifed by Eddie Soh