Do Parent and Teacher Ratings of ADHD Reflect the Same Constructs? A Measurement Invariance Analysis.

Abstract

Discrepancies between parent and teacher ratings of problem behaviors related to Attention-Deficit/Hyperactivity Disorder (ADHD) are reported frequently. Previous studies have hypothesized that these discrepancies are the results of various informant biases and have evaluated whether the rating scales are measuring behaviors the same way across informants. The purpose of this study was to evaluate if two rating scales of ADHD behavior, the Strengths and Weaknesses of ADHD Symptoms and Normal Behavior Rating Scale (SWAN) and the Conners' Teacher Rating Scale-15 (CTRS-15), reflected the same underlying constructs across parent and teacher report. Measurement invariance analyses were conducted using parent and teacher report data from a sample of 1,645 preschool to fifth-grade children (age range 46 to 169 months) that was comprised of roughly equal number of boys and girls and had racial/ethnic diversity similar to the community (i.e., 61% White, 22% Black/African American; 4% Hispanic/Latino). Although it was hypothesized that both rating scales would demonstrate measurement invariance across parent and teacher report, at least partial weak measurement invariance was only supported for the CTRS-15 across all grade groups. These results indicate that the meaning of any rating discrepancies on the SWAN are unknown because it is not reflective of the same underlying constructs across parents and teachers across all of the examined grade groups. In general, these results have potentially important implications regarding research on ADHD symptoms and related behaviors, and raise questions regarding the utility and measurement of ADHD symptoms.