Examining the Criterion-Related Validity of the PDD-BI

Background: The Pervasive Developmental Disorder – Behavior Inventory (PDD-BI; Cohen, et al., 2003) is a questionnaire that was designed to aid in the diagnosis of children with Pervasive Developmental Disorders (PDDs). The PDD-BI assesses both adaptive and maladaptive behaviours associated with PDD’s. The adaptive subscales (Expressive Social Communication Abilities Composite; ESCSA/C and Receptive/Expressive Social Communication Abilities Composite; REXSCA/C) measure phonological skills, social approach behaviours, receptive language, semantic pragmatic skills, learning, and memory. In contrast, the maladaptive subscales (Repetitive, Ritualistic and Pragmatic Composite; REPRIT/C and Approach/Withdrawal Problems Composite; AWP/C), include sensory/perceptual approach behaviours, fears, arousal problems, aggression, and social pragmatic problems. Additionally, the PDD-BI provides an age-standardized autism score. Previous research has shown the PDD-BI to have criterion-related validity in a sample of children aged 3 to 6 (Cohen, et al., 2003).

Objectives: The objective of the present study is to replicate and extend previous research investigating the criterion-related validity of the PDD-BI.  We expected the Adaptive sections of the PDD-BI to be correlated with developmental scores on the Vineland Adaptive Behaviour Scales, 2^nd Edition, and the Mullen Scales of Early Learning; and the Maladaptive scores on the PDD-BI to be related to other autism diagnostic measures, the Childhood Autism Rating Scale (CARS; Schopler et al., 1988) and the DSM-IV criteria for Autism.  

Methods: This study examined previously collected data from 63 children (55 boys; 8 girls) seen for a psychological assessment as part of a larger research study. Participants ranged between 2 years 7 months and 9 years 3 months (M = 4 years 6 months). Forty-six participants had been diagnosed with Autistic Disorder and 14 had received a diagnosis of PDD-NOS. 

As part of the assessment, the Mullen was administered to each child and the Vineland-II was administered by interviewing parent(s). The CARS and DSM-IV diagnostic criteria were completed for each child by examiners conducting the assessment. Caregivers also completed the parent version of the PDD-BI. 

Results:  As predicted, the Adaptive scores on the PDD-BI were moderately correlated with the cognitive and adaptive scores. However, there were no significant correlations between the PDDI Maladaptive and Autism composite scores with the other diagnostic measures. In addition, results from independent samples t-tests indicated that PDD-BI scores did not differ by diagnostic group or gender. 

Conclusions:  The present findings are consistent with previous research confirming the criterion-validity for the Adaptive section of the PDD-BI, in that it was strongly correlated with the other cognitive and adaptive measures.  However, the Maladaptive section and Autism summary score from the PDD-BI were not related to two other well established measures of autism severity, leading to questions around the validity of this section of the PDD-BI.  Implications for the conceptualization of Autism and clinical implications will be discussed.