Self-efficacy and reading achievement of young children in urban schools: Implications for social work

Background: A substantial body of research supports the role of self-efficacy in older student's academic achievement. The focus on self-efficacy is grounded in the assumption that the beliefs that students create, develop, and hold to be true about themselves are vital forces in their academic success (Pajares, 2003). It is not clear, however, at what age self-efficacy is sufficiently developed to impact schooling. Nor do we know whether particularized task measures as compared to global measures are better predictors of outcomes for young children (Pajares, 1996). Many young students in the United States are reading below the basic level (Lee, Grigg, & Donahue, 2007), which is linked to later school failure. If self-efficacy related to reading can be detected in a young population and, if it is related to reading achievement, we may be able to target this mechanism at a much earlier stage to improve rates of later school completion. This study examined the role of reading task self-efficacy and global reading self-concept in reading achievement of children in grades one through three.

Method: A sample of 881 first through third grade students at risk for reading failure from a large randomized trial study of a tutoring program is used. Students attended 23 urban public schools in three cities (6 schools in New York, 9 schools in Boston, 8 schools in Port Arthur, Texas). Human subjects approval was obtained from the university IRB, schools, and parents. Questions measuring reading ability (Woodcock-Johnson word attack and passage comprehension; Peabody Picture Vocabulary Test (PPVT)), reading task self-efficacy, and global reading self-concept were verbally administered to students in the fall and at the end of the school year. Other data (e.g., student demographics, classroom behavior, motivation, or special education status) were collected via teachers' surveys and school records. Missing data were imputed using Markov Chain Monte Carlo (MCMC) multiple imputation. To adjust for the clustered structure of the data (e.g., students clustered within classrooms and schools), a Generalized Estimating Equation (GEE) was used.

Findings: Reading task self-efficacy was a significant predictor of two of three reading measures (b=2.06, p<.05 for WJ-passage comprehension; b=1.15, p<.001 for PPVT), controlling for a series of covariates such as prior reading scores, demographics, motivational/behavioral measures, and parental involvement. There was no statistically significant relationship between global reading self-concept and the Woodcock-Johnson reading measures. Global reading self-concept had a weak but significant association with the PPVT (b= -0.57, p<.05). Positive teacher ratings of student behavior were also linked to improved reading achievement.

Conclusions and Implications: While replication is needed, findings from this study suggest that reading task self-efficacy may be measured and is important in achievement at a younger age than previously thought. Further study is needed to explore if task specific self-efficacy is also important in other academic and social domains. To help schools meet accountability standards, school social workers should focus on increasing self-efficacy around specific tasks as well as school behavior.