|
| Examining Measurement and Analysis Issues on the Validity of Single Subject Design Research |
| Sunday, May 29, 2005 |
| 10:30 AM–11:50 AM |
| Astoria (3rd floor) |
| Area: TPC; Domain: Applied Research |
| Chair: Maureen Conroy (University of Florida) |
| Discussant: Joseph H. Wehby (Vanderbilt University) |
| Abstract: Over the past twenty years, there is an increasing trend to implement behavioral interventions in naturalistic settings (e.g., home, classrooms), where natural contingencies occur. As researchers increase their investigations of behavioral principles across a variety of natural settings, a number of threats to the internal and external validity have occurred. For example, the variability of data in natural settings may be in creased, because we tend to find less predictable patterns of behaviors in natural settings (Wahler & Fox, 1981). The purpose of this symposium is to discuss and present data on issues related to measurement, data analysis, and the internal and external validity of single subject design research findings conducted in natural settings. Three papers will be presented, followed by a discussion by Dr. Joseph Wehby. First, Drs. Conroy and Stichter will present a critical analysis of measurement and validity issues related to the research conducted in the field of EBD. Next, Dr. Olive will present an empirical examination of effect size calculations and percentage of non-overlapping data. Finally, Dr. Fox and colleagues will present treatment fidelity data, analyses, and results from an ongoing treatment fidelity study. |
| |
| A Critical Analysis of the Role of Measurement on the Validity of Research |
| MAUREEN CONROY (University of Florida), Janine Peck Stichter (University of Missouri-Columbia) |
| Abstract: There is an ever-growing need to conduct high quality research and translate our research findings to practice by emphasizing high quality, scientifically based intervention strategies for teachers to use when addressing problem behaviors demonstrated by children and youth with emotional/behavioral disorders (EBD). This presentation critically analyzes the research trends in the field of emotional/behavioral disorders (EBD) with a specific emphasis on the role of measurement in single subject design methodology in the field of EBD. Specifically, a critical analysis and examination of the current measurement practices and the effect these practices have on the internal, external, and social validity of the research within the field of emotional/behavioral disorders (EBD) will be presented. In particular, an examination of how the shifts in EBD research, conceptually as well as practically, have impacted measurement systems, ultimately affecting the validity of the research and the availability of scientifically based practices will be discussed. In addition, a proposal for extending the empirical rigor and generalization of our research findings with the goal of improving the validity of the research findings and the availability scientifically based practices for children, families, and teachers is presented. |
| |
| Alternative Strategies for Determining Intervention Effectiveness |
| MELISSA L. OLIVE (University of Texas, Austin) |
| Abstract: This paper will present results from a series of assessment-based intervention studies conducted with 6 children who engaged in challenging behavior. Next results will be presented in terms of effect size calculations (i.e., standard mean difference and regression analysis), and percentage of non-overlapping data (PND; Scruggs, Mastropieri, & Casto, 1987). The author will discuss issues regarding effec t size calculations for single subject designs. |
| |
| Multiple Measures to Assess Student Behavior Change and Treatment Fidelity: Issues, Methods and Preliminary Results |
| JAMES J. FOX (East Tennessee State University), Leia D. Blevins (East Tennessee State University), Angela Spranger (East Tennessee State University), Cheri Kyzer (East Tennessee State University), Christi Carr (East Tennessee State University), Stacy Foster (East Tennessee State University) |
| Abstract: Experimental and case study reports show that functional behavior assessment (FBA) interventions effectively reduce student challenging behaviors when implemented or supervised by researchers (see, for example, Sasso, Conroy, Stichter, & Fox, 2001). Two large scale studies produced conflicting results, one finding that personnel trained in FBA reduced student behavior problems (Chandler, Dahlquist, Repp & Feltz,1999), the other reporting few differences between FBA based- and more traditional behavior management interventions (Schill, Kratochwill, & Elliot, 1998). One factor that may have contributed to these conflicting outcomes is the failure of classroom personnel to faithfully implement FBA-based interventions. Despite a clear need for such studies (e.g., Gable, Hendrickson, & Van Acker, 2001), "treatment fidelity" studies that assess both the student's and teacher's behaviors are lacking. This presentation will discuss: (a) issues in treatment fidelity analyses; (b) multi-method procedures for accomplishing such an analysis; and, (c) results from an ongoing treatment fidelity study of students with disabilities conducted in various school systems in northeast Tennessee. Data were collected using inexpensive palm top computers and commercially available software. |
|
| |