Browse By Day
Browse By Time
Browse By Panel
Browse By Session Type
Browse By Topic Area
Research on foster youth has shown that these youth face higher risks for cumulative disadvantage, such as school readiness difficulty, an increase in both internalizing and externalizing behaviors, and lower human and social capital overall (Nurius, Prince, & Rocha, 2015). Recent efforts have started to focus on protective factors for at-risk youth, specifically, on how academic achievement can be improved (McClelland, Cameron, & Alonso, in preparation). In the general literature on school-aged children, academic self-efficacy has been positively related to academic achievement (Lane & Lane, 2001). Yet it is unclear how the highly relevant component of behavioral engagement, a proxy for self-regulation (Fredricks, Blumenfeld, Friedel, & Paris, 2005), in school is related to the academic outcomes, such as academic self-efficacy, of youth in foster care and vice versa. This study sought to answer the questions: (1) is academic self-efficacy affected by behavioral engagement across time? and (2) is behavioral engagement affected by academic self-efficacy across time?
Quantitative data were gathered from a metropolitan Oregon sample of 328 youth sibling dyads. About half of the sample identified as female (n=81; 48%); with 51% being older siblings (n=84) and 49% being younger siblings (n=80). About half of the sample (n=86; 51%) identified as non-white, and about a fifth were Hispanic (n=30; 18%). In regards to outcome variables, youth completed the School Engagement Scale (YSES; Fredericks, Blumenfeld, Friedel & Paris, 2005) and the Self-Efficacy Questionnaire (YSEQ; Muris, 2001). Scores ranged from 1 to 5 (M=3.5, SD=0.7), and from 1 to 5 on the YSES (total score M=3.5, SD=0.7), and the YSEQ academic self-efficacy subscale (M=3.9, SD=0.8).
For the first model a two-factor CFA was fit to create two latent factors: one of academic self-efficacy and one of behavioral engagement at the first wave of data collection (i.e., baseline). After removing the third item from the behavioral engagement subscale and items 16 and 19 from the academic self-efficacy subscale the two-factor model demonstrated good model fit (See Table 1). The second model included running an autoregressive cross-lagged panel model across all four waves of data collection (fit reported in Table 1). Results for each autoregressive path tested in this model were statistically significant, with a standardized estimate of 0.49 or higher (p < .0001; See Figure 1).
Based on the current understanding of the importance of improving behavioral factors in the classroom to increase academic success (McClelland et al., 2007), it is important to acknowledge these relationships for children in the foster care system based on their higher likelihood of facing cumulative disadvantage (Nurius, Prince, & Rocha, 2015). Consistent with previous research (Linnenbrink & Pintrich, 2003), results from this study indicated that behavioral engagement at one time point was related to higher academic self-efficacy at the subsequent time point across all four waves of data and vice versa. These results add additional evidence that behavioral components, such as behavioral engagement or behavioral self-regulation, should be both measured and intervened in within classroom settings.