Validity and Diagnostic Utility of the Learning Behaviors Scale

Psychological assessment is a critical component in educational decision making and planning; therefore it is important to investigate the validity and diagnostic utility of the tools that we use. The current study examined the construct validity and diagnostic utility of the Learning Behaviors Scale (LBS; McDermott, Green, Francis, & Stott, 1999). Two hundred and seventy-three third grade students were rated by their classroom teacher using the LBS. LBS data were analyzed along with TerraNova achievement scores and In View Cognitive Skills Index. Overall, statistically significant distinct group differences were found for the LBS between the Gifted and Typical groups, as well as the Gifted Referred (students screened, but not identified as gifted) and Typical groups supporting construct validity. However, no statistically significant results were found between the Gifted and Gifted Referred groups. Moderate support was found for the incremental predictive validity for LBS scores. The LBS Total score captured statistically significant variance (1.5% to 5.3%) in academic achievement above and beyond that of the Cognitive Skills Index. Similarly, the LBS factor scores added 3.1% to 6.5% of variance in achievement scores beyond the Cognitive Skills Index. Finally, LBS scores were unable to correctly classify individual group membership for the Gifted and Gifted Referred groups. However, a critical limitation was proposed as a potential reason for the limited support and lack of discriminant validity for the use of the LBS with above average students. A ceiling effect was observed for the LBS Total and factor scores among the Gifted Referred and Gifted groups, which likely limited the range of scores and impacted correlations and variability