Predictors of Outcome in a Community Based Parent Training Program

Background: Research that has been conducted on predictors of parent training program effectiveness (e.g., Kaminski, Valle, Filene, & Boyle, 2008) has found predictors of success in regards to the components of the treatment package and training provided. However, there is less research on aspects of parent, child, and family characteristics that may make it more or less likely for a parent training program to be effective despite the potential utility of such information for making treatment decisions. Further, it has been suggested that in research on psychosocial interventions for children with autism spectrum disorders that sources of variation, such as family variables or individual differences (e.g., ethnicity) of children and families, should be considered (Lord et al., 2005).

Objectives: Using data collected as part of a treatment program providing education to families of children with autism spectrum disorders, possible family and individual difference predictors of treatment outcome (i.e., change in problem behaviors) will be analyzed with a relatively diverse sample of children and families using multiple regression and ANOVA statistical procedures.

Methods: Information on ethnicity and child age will be gathered.  Further, the General Maladaptive Index of the Scales of Independent Behavior–Revised (SIB-R; Bruininks, Woodcock, Weatherman, & Hill, 1996) will be used as a primary measure of behavior problems exhibited for pre- and post- treatment.  This rating scale represents a variety of problem behaviors, such as destructive, disruptive, and inattentive behaviors. Models will also explore subscale index scores (i.e., Internalized Maladaptive Index, Asocial Maladaptive Index, and Externalized Maladaptive Index). Data on parent skill acquisition as well as number of sessions attended will also be used as possible predictors of change in SIB-R scores.

Results: To begin, descriptive statistics and correlations will be explored. Subsequently, two sets of regression analyses will be conducted. Multiple regression analyses will be used to analyze predictors (e.g., age of child, ethnicity, SIB-R score at initial assessment, parent skill acquisition) of change in SIB-R General Maladaptive Index scores across treatment. Changes in the subscale scores will also be explored for those that completed the full treatment package. In addition, predictors of changes in SIB-R scores will also be considered for families that did not complete treatment, with number of sessions considered as a covariate. Finally, group differences will be explored between those who completed treatment and those families that did not.

Conclusions: Data will be interpreted in regards to the importance of including potential covariates or predictors in research on outcomes in parent training programs of children with autism spectrum disorders.  In particular, whether change in problem behavior reached levels of clinical significance will be evaluated. Implications for research and practice for parent training programs will be discussed.