ABA Research Methods for Teaching Autistic Children

Introduction to ABA and Autism Research

When approaching the target behavior of teaching an autistic child to obey verbal commands, it is important to understand what previous research has presented about this target behavior, how it has defined autism, which treatment methodologies have been tried and shown effectiveness in the past, and which experimental designs can be utilized today to treat the behavior. This paper examines each of these points and provides a standard for assessing validity.

Defining autism in the field has mainly been the result of Applied Behavior Analysis (ABA). As Wohr and Scattoni (2013) note, "autism is a behaviorally defined disorder" (p. 5). And as Ollendick and Cerny (2010) observe, the methods used to define this behavior and to assess treatments are numerous: for example, rating scales can be utilized as well as observation methods, IQ tests, and examination of "the social-cultural-physical environment" in which the child lives (p. 33). Thus, there is no single method required for validly assessing the target behavior of this study — and in fact, numerous methods may be employed to allow the research to better evaluate and test the hypothesis.

Defining and Measuring Autism in the Field

Autism has been measured in the field in diverse ways. Stahmer, Collings, and Palinkas (2005) use a qualitative approach in a specified setting with a specified sample in order to assess the effectiveness of interventions in communities via focus groups, with self-reporting as a measuring tool. Denne, Thomas, Hastings, and Hughes (2015), on the other hand, use video ratings, in-situ observation, self-reporting, and a Test of Knowledge in order to measure ABA competency levels in education intervention methods. Thus, measurements are diverse in the field studies connected to autism and ABA evaluations.

The treatment methodologies and experimental designs utilized to treat the target behavior are equally diverse. Mohammadzaheri, Koegel, Rezaee, and Rafiee (2014) compare two ABA treatments — pivotal response treatment and structured ABA — applied to autistic children in a classroom setting. They found that pivotal response treatment (the naturalistic approach, as opposed to the structured approach) was more effective in improving target behaviors over a three-month period. The independent variables are the methods used to "produce more rapid improvements in communication" — namely, pivotal response treatment and structured ABA. The dependent variables are social skills and attentiveness. The subjects were two groups of children with autism, "matched according to age, sex and mean length of utterance," and the setting was a classroom (Mohammadzaheri et al., 2014, p. 2269). This study is helpful in showing how target behaviors are identified and how ABA can be used to improve them through the application of a specific method — in this case, the naturalistic method of pivotal response treatment.

Treatment Methodologies and Experimental Designs

Callahan and Shukla-Mehta (2010) define and validate comprehensive treatment models in autism by comparing ABA and TEACHH. By assessing the findings of a social validation survey, the researchers concluded that there was "no clear preference for either model" among special education teachers, but that elements of both methods were deemed to have a high degree of social validity (Callahan & Shukla-Mehta, 2010, p. 74). The study itself had external validity, evidenced by the relevance of its findings to special education teachers in the field. In terms of internal validity, however, the study was not providing a cause-and-effect analysis but rather performing evaluation research concerning the results of a survey conducted among special education teachers regarding their preferred approach to treating children with autism.

Studies such as this are common because there is no single, clearly defined avenue for treating the target behavior considered here. Treating autism itself has several different accepted approaches, and the most effective are often related to the personal preference and comfort level of the practitioner. However, Foxx (2008) finds that the only effective treatment approaches for autism are those grounded in Applied Behavior Analysis. As Foxx (2008) notes, "the only interventions that have been shown to produce comprehensive, lasting results in autism have been based on the principles of ABA" (p. 821).

Foxx's study has internal validity, as it identifies a cause-and-effect relationship between independent variables and dependent variables — that is, between the various methods and approaches utilized as interventions and their impact on the autistic child. The study also has external validity, as its findings have broad appeal and could be relevant to any practitioner in fields related to psychology or behavioral studies. Additionally, the study demonstrates social validity, which, according to Rapoff (2010), is rooted in the ABA tradition and "is a broader and more inclusive term than clinical significance" (p. 114). The social validity evident in Foxx's study lies in the fact that the researcher identifies the most helpful method of treating autism in children as ABA methodology, because this approach makes the strongest advancements in modifying child behavior and honing in on the target behavior.

Conclusion

Some practical benefit derived from applied research in the area of ABA and autistic children may be found in the fact that applied research takes the theory identified in pure research — such as the Giger-Davidhizar assessment — and tests it in the real world to see if it actually applies. The studies by Foxx (2008) and Callahan et al. (2010) serve as relevant examples. Applied research can also be utilized as an aid when attempting to clarify or identify the most effective treatment for the target behavior considered in this study. ABA can moreover be a practice whereby more than one theory or experimental method is used, as the study by Callahan et al. (2010) indicates. This suggests that development is achieved by combining or integrating various theories derived from pure research and using them together in a single experiment to find a solution to the problem underlying the target behavior.