Self-efficacy inertia: The role of competency beliefs and academic burden in achievement

The effort to translate self-efficacy theory into statistical models has been argued to model out the complex, dynamic nature of the interaction between the person and the environment. This study aimed to understand how self-efficacy (belief in one's abilities) and academic burden (the external challenges students face in their studies) relate to academic performance over time and whether modeling both between and within subject variance components provides a more comprehensive perspective that is better aligned with theory. Self-efficacy and academic burden were collected at five time points, one month apart from undergraduate students (N = 443) enrolled in an online biology class. The data were fit to four models: 1) a standard cross lagged panel model (CLPM), 2) a random intercept cross-lagged panel model (RI-CLPM), 3) an RI-CLPM in which grade was regressed on the random-intercepts, and 4) an RI-CLPM in which grade was regressed on the random-intercepts and the within-person fluctuations. The RI-CLPM was a better fit to the data over the CLPM, indicating that separating effects that are attributed to individual differences from within-person effects appears to better capture the reciprocal relationships between self-efficacy and academic burden. Further, when only the general tendencies of self-efficacy and academic burden were specified to predict final grade, there was a significant positive relationship between self-efficacy and grades. However, when within-person variations over time were added as predictors in addition to the between-person differences, this general relationship lost significance. These findings suggest that gaining self-efficacy momentum in a class is perhaps more predictive of academic achievement than having a general tendency towards confidence in one's abilities relative to peers.