The Stability of Teacher Value-Added Rankings Across Measurement Model Assumptions and Educational Contexts

• In Event: Roundtable Session 28
  In Roundtable Session: Value-Added Methodology: Considerations and Model Comparisons

Abstract

This study will use Monte Carlo simulation to examine the impact of different measurement model assumptions on value-added teacher rankings within two hypothetical educational contexts. Measurement model assumptions will incorporate latent variable modeling with cross-classified mixed effects models (CCMEMs). CCMEMs will vary by two variable combinations (univariate and multivariate), two measurement assumptions (parallel and non-parallel indicators), and two longitudinal invariance assumptions (non-invariant and invariant). Educational contexts will vary by classroom size (n=12; n=25) and whether students change teachers each year. Our results will contribute to the body of literature on value-added assessment, particularly in terms of the assumptions made regarding the outcome measures and the generalizability of models across educational contexts.

Authors

• Leslie R. Hawley, University of Nebraska - Lincoln

• Chaorong Wu, University of Nebraska - Lincoln

• James A. Bovaird, University of Nebraska - Lincoln