Relationship between learning behaviors and academic achievement among school-aged American children and adolescents

Presents a nationwide, concurrent validity study of learning-related behaviors observed by classroom teachers of 1,100 youth ages 6 through 17 years. Ss comprised a nationally representative cross-sample stratified according to the U.S. Census and drawn from the standardization cohorts for the newly developed Learning Behaviors Scale (LBS; McDermott, Green, Francis, & Stott, 1996) and the Differential Ability Scales (Elliott, 1990). The relationship between learning behaviors and academic achievement, as reflected by both standardized measures and student grades, was assessed using the cross-sample and a subsample (n = 420). Bivariate correlations with cognitive ability and academic achievement revealed the unique and common variance associated with learning behavior style. Univariate hierarchical multiple regression assessed the simple, incremental, and interactive validity of learning style\u27s and cognitive ability\u27s prediction of achievement. Determined to be largely unique from cognitive ability, learning style better predicted student grades (21% to 32%, all ps $\u3c$.001) than did ability (14% to 17%) or demographics (15% to 17%). In contrast, cognitive ability (27% to 34%) surpassed learning style (11% to 14%) in predicting standardized achievement. Learning style, cognitive ability, and their interactions combined to account for 28% to 37% of the variance, and the inclusion of demographics maximized the prediction of both types of achievement measures (37% to 45%). As an objective, reliable, and valid behavior rating scale which can account for sizable proportions of variance in achievement, the LBS demonstrates substantial potential for use in developing prereferral interventions and individualized educational programs for school-aged youth