Identifying Young Students At Risk for Antisocial Behavior: The Utility of “Teachers as Tests”

Abstract

The purpose of this study was to determine whether teachers are equally as effective in differentiating between students at risk and those who are not at risk at an early age when the discrepancies between current and desired levels of performance are more subtle. Two groups of first-grade students described by their teachers as being at risk for antisocial behavior or as typically developing students who were not at risk were compared on 12 measures in academic, behavioral, and social domains using multivariate procedures. Results indicate that teachers can differentiate between these types of students in each domain, with 8 of the original 12 variables being useful in differentiating group membership. In general, the typically developing group was characterized as being more socially and academically competent, with fewer behavior problems than the at risk youngsters. Data from the descriptive and classification discriminant function analyses suggest that a relatively larger percentage of the variance between at risk and typically developing groups could be explained. This information is particularly useful in differentiating between at risk and typically developing learners. However, the rather high error estimate (27.92%; n = 14) should be addressed. Limitations of the study are addressed, and implications for early intervention research are discussed.