Construct Validity of the Learning Behaviors Scale and the Academic Competence Evaluation Scales

Abstract

This research examined the convergent and discriminant validity of the Academic Competence Evaluation Scales (ACES; DiPerna & Elliott, 2000) and the Learning Behaviors Scale (LBS; McDermott, Green, Francis, & Stott, 2001). The Adjustment Scales for Children and Adolescents (ASCA; McDermott, Stott, & Marston, 1993) was compared with the ACES and LBS to examine discriminant validity. Pearson product moment correlations were obtained to examine convergent and discriminant validity. Paired samples t-tests were conducted on the ACES and LBS total and subscale scores to compare the mean scores. The ACES Academic Enabler (ACES-AE) total score was significantly, positively correlated with the LBS total score (r =0.88) and shared 77% variance. Paired samples t-test analyses indicated that the ACES-AE Total T score (M = 46.83, SD = 10.63) was significantly higher than the LBS Total T score (M = 42.18, SD = 13.81), t(97) = 5.47, p < .001, d = .38. However, although teacher ratings on the ACES-AE were significantly higher than the LBS, the effect size was small and likely not meaningful. Both the ACES-AE and the LBS Total score were moderately, negatively correlated with the ASCA Overactivity score (r = -0.43 and r = -0.55, respectively) with 18% and 30% shared variance and the ASCA Underactivity score (r = -0.42 and r = -0.32, respectively), with 18% and 10% shared variance. The ACES and LBS demonstrated convergence (they measured similar constructs) while they each demonstrated discriminant validity when compared with the ASCA (these correlations were mostly lower than ACES/LBS correlations). Thus, the current study found construct validity support for the ACES and LBS.

Date of Award	2015
Original language	American English
Awarding Institution	Eastern Illinois University
Supervisor	Gary Canivez (Supervisor)