Teachers as “Tests”: Differential Validity of Teacher Judgments in Identifying Students At-Risk for Learning Difficulties

Abstract

The discriminant validity of teacher judgments in differentiating various groups of students who were at-risk for learning difficulties based on their referral to School Study Teams (SSTs) was assessed. Three groups of students defined as Learning Disabled (LD), Low-IQ, and Low Achievement (LA) were contrasted with a group of normally achieving Controls using the Academic Competence scale of the Social Skills Rating System (Gresham & Elliott, 1990). Large differences were found between the Controls and all at-risk groups. Consistent with past research contrasting mild disability groups using cognitive and achievement measures, relatively few differences were found among LD, Low IQ, and LA groups. Results from cross-validated stepwise discriminant function analyses showed that teacher judgments correctly classified 91% of the LD group (9% false negatives and 10% false positives), 100% of the Low IQ group (0% false negatives and 7% false positives), and 95% of the LA group (5% false negatives and 17% false positives). Results were discussed in terms of the implications for the continued time-consuming and expensive practice of differentiating among mild disability and low achievement groups for classification, placement, and intervention purposes.