Developmental pathways in underachievement

Abstract

Insufficient research attention has been focused on the development of academic under-achievement and motivational beliefs. A longitudinal, person-centered approach was used to identify latent subgroups of growth in the joint development of teacher-reported underachievement and four motivational beliefs (self-concept, task importance, psychological cost value, and self-worth) from first through sixth grade. Two types of underachievement emerged: Sustained and Growing Underachievement. An Achievement class was also consistently found. Sustained underachievement was not associated with declining self-concept or task importance, but was related to moderately lower psychological cost value, self-worth, and middle school achievement. Growing underachievement was associated with relatively lower levels of self-concept and task importance that declined over time, similar to the achievement patterns. Class membership did not differ by gifted status. Gender effects emerged for the Task Importance and Self-Worth models in the hypothesized direction, but were not robust. These findings highlight heterogeneity in the development of academic underachievement.

Acknowledgments

We are very grateful to Dr. Eccles for allowing us to use these data from the Childhood and Beyond study. We also wish to thank Martha Putallaz, Harris Cooper, Kristen R. Stephens, and D. Betsy McCoach for their helpful feedback on earlier versions of this manuscript and we thank Loralyn Rudy for her assistance with dataset preparation. This research is based on a doctoral dissertation by Kate E. Snyder that was submitted to Duke University. A prior version of this research was presented at the 2013 annual meeting of the American Educational Research Association.

Notes

1 A quadratic growth term was originally included for both underachievement and task importance in the Task Importance model, but resulted in out-of-bounds values for estimated means in one-class, two-class, three-class, and four-class solutions. Out-of-bounds estimated means can occur due to poor fit, which can result from the inclusion or exclusion of a quadratic term (B. Muthén, personal communication). The quadratic term for task importance was removed for the final analysis, resulting in estimated means within normal bounds.

2 Although underachievement, self-concept, task importance, and psychological cost value are all measured with reference to the domain of reading, the descriptions throughout the Results section will refer simply to construct names (i.e., “self-concept,” rather than “reading self-concept”) for brevity.

Additional information

Funding

This work was supported by the National Institute for Child Health and Human Development (Grant HD 17553) in a grant awarded to Jacquelynne S. Eccles, Allan Wigfield, Phyllis Blumenfeld, and Rena Harold.