![[ Visit Client Website ]](images/banner.JPG)
Friday, May 16, 2008
Champagne Terrace/Bordeaux (Novotel London West)
Background: Although ideal behavior-analytic interventions include socially important goals; methods; and outcomes, consideration of these components is difficult due to their multiple dimensions and the use of non-standardized assessment instruments. Contextual sensitivity of both evaluative criteria and evaluator involvement increases such imprecision, thus preventing sufficient social validation of behavioral treatments for autism.
Objectives: To identify influential evaluations of behavior-analytic treatments for autism and to examine social-validity trends in this sample of recent research in applied behavior analysis.
Methods: Research articles involving at least one participant with autism and the empirical evaluation of behavior-analytic procedures were selected from volumes of JABA and JADD published between 1992 and 2007. Characteristics of a study's participants and intervention were coded. A multidimensional framework (Schlosser, 1999) was used to identify conceptually distinct aspects of social-validity components (i.e., methods, outcomes) and assessments (e.g., agent, format).
Results: The 249 eligible articles from JABA included 17% with an evaluation of treatment acceptability or treatment outcome and 5% that assessed both of these components. Social validation was more common in naturalistic (29%) than in analog (15%) settings. Among 114 eligible articles from JADD, 75% assessed one component of social validity and 42% evaluated acceptability and outcome. Social validation was comparable between naturalistic (77%) and analog (71%) settings. Evaluation of outcomes instead of methods was more common in treatments based on manipulation of antecedents and/or consequences than in skill-acquisition programs, which were based on behavioral principles but often used scales of diagnostic criteria or general behavior as outcome measures. Further analysis shows differences between journals across various study characteristics.
Conclusions: Summary data replicate results of prior JABA reviews and show more-but-different social-validity activity in the smaller set of JADD studies. These trends may reflect increased interest in general measures, instead of discrete target behaviors, for the evaluation of autism-intervention outcomes.
Objectives: To identify influential evaluations of behavior-analytic treatments for autism and to examine social-validity trends in this sample of recent research in applied behavior analysis.
Methods: Research articles involving at least one participant with autism and the empirical evaluation of behavior-analytic procedures were selected from volumes of JABA and JADD published between 1992 and 2007. Characteristics of a study's participants and intervention were coded. A multidimensional framework (Schlosser, 1999) was used to identify conceptually distinct aspects of social-validity components (i.e., methods, outcomes) and assessments (e.g., agent, format).
Results: The 249 eligible articles from JABA included 17% with an evaluation of treatment acceptability or treatment outcome and 5% that assessed both of these components. Social validation was more common in naturalistic (29%) than in analog (15%) settings. Among 114 eligible articles from JADD, 75% assessed one component of social validity and 42% evaluated acceptability and outcome. Social validation was comparable between naturalistic (77%) and analog (71%) settings. Evaluation of outcomes instead of methods was more common in treatments based on manipulation of antecedents and/or consequences than in skill-acquisition programs, which were based on behavioral principles but often used scales of diagnostic criteria or general behavior as outcome measures. Further analysis shows differences between journals across various study characteristics.
Conclusions: Summary data replicate results of prior JABA reviews and show more-but-different social-validity activity in the smaller set of JADD studies. These trends may reflect increased interest in general measures, instead of discrete target behaviors, for the evaluation of autism-intervention outcomes.