Models for Aggregate Growth and Progress Using Multiyear, Multicohort Data Sets

• In Event: AERA Poster Session 2
  In Poster Session: Poster Session 1 (San Diego)

Abstract

This paper investigates a 4-level longitudinal growth model used to summarize multiyear, multicohort student test score data for educational achievement monitoring. We illustrate use and interpretation of the model using empirical data, including estimating the reliability of the parameter estimates, and we use a small simulation study to evaluate model performance under a wider range of conditions. The paper focuses on practical issues that arise in application of the model including sensitivity of the results to the use of a vertical score scale and whether accurate parameters can be recovered when only aggregate data are available to the analyst. Based on results, the model provides a useful tool for researchers attempting to summarize and monitor trends in student achievement.

Authors

• Benjamin R. Shear, University of Colorado - Boulder

• Andrew Ho, Harvard University

• Sean F. Reardon, Stanford University