Validity and Reliability of the Behavior Problems Inventory the Aberrant Behavior Checklist and the Repetitive Behavior Scale-Revised among Infants and Toddlers at Risk for Developmental Disabilities: A Multi-Method Assessment Approach

摘要

Individuals with developmental disabilities (DD) are at a heightened risk of developing aberrant behaviors during the course of their lives. Reliable and valid assessment of such behaviors is an important element in empirically verifying successful prevention and intervention. Only a few instruments exist which were developed for behavior problems in children of a young age. The purpose of the current longitudinal study was to examine the performance of three behavior-rating instruments for individuals with developmental disabilities that have been proven useful and psychometrically sound in older populations: the Behavior Problems Inventory (BPI-01), the Aberrant Behavior Checklist (ABC), the Repetitive Behavior Scale – Revised (RBS-R). Data were analyzed for 180 children at risk for a developmental disability in Lima, Peru who were involved in active behavioral treatment for behavior problems. Cronbach’s alpha of all three measures showed variable internal consistency across the subscales of the three instruments. BPI-01 test-retest reliability for its three subscales over three times of measurement within approximately a 12 months period showed ICC coefficients ranging between .68 to .77 for frequency ratings and .65 to .80 for severity ratings. Using a multitrait-multimethod matrix approach for the constructs self-injurious behavior (SIB), stereotyped behavior, and aggressive/destructive behavior, we found high levels of convergent and discriminant validity across the three instruments. Sensitivity of the SIB and the Stereotyped Behavior subscales of the BPI-01, the RBS-R, and the ABC showed that while there was some overlap between the three instruments, each scale contributed unique information. This was especially true for stereotyped behavior. In addition, since each one of the three instruments contains subscales that are not represented by the others, it merits to consider using all three scales together.