ABSTRACT
Purpose
We examined the extent to which achievement goals predict reading comprehension, measured by two response formats (free recall and constructed response), and how these relations differ for students with and without reading difficulties (RD). We further explored how executive functions (working memory and semantic verbal fluency) mediate the relations between achievement goals and reading comprehension.
Method
We fit multigroup structural equation models with data from monolingual English-speaking fifth graders (n = 146 for RD; n = 109 for non-RD) in the United States.
Results
Results revealed that achievement goals predict reading comprehension as measured by the free recall but not by the constructed response format, and this pattern was moderated by RD status. For students with RD, mastery goals positively predicted performance on free recall, a relationship that was completely mediated by semantic verbal fluency, whereas performance-approach goals were negatively related to free recall. For students without RD, however, achievement goals did not predict reading comprehension as measured by either assessment format.
Conclusion
Our findings underscore the need to account for motivational differences in reading comprehension and the importance of fostering mastery goals when teaching reading comprehension, particularly for students with RD.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Supplementary material
Supplemental data for this article can be accessed online at https://doi.org/10.1080/10888438.2022.2127357
Notes
1. The CFA for achievement goals with polytomous items was conducted using a robust maximum likelihood estimator to account for deviations from normality, with the full-information maximum likelihood estimator for the unbiased estimates (Enders & Bandalos, Citation2001).
2. Paths from performance-avoidance goals to executive functions were not included in the model because of the null relationship between performance-avoidance goals and reading comprehension in the base model (ps ≥ .85) and a lack of research evidence supporting the relationship between performance-avoidance goals and executive functions. In addition, to improve model fit, we added residual covariances between two performance-approach goal items and between control variables and executive functions.