Value-added models (VAMs) have become widely used in evaluating teacher accountability. The use of these models for high-stakes decisions making has been very controversial due to lack of consistency in classifying teachers as high performing or low performing. There is an abundance of research on the impact of various student level covariates on teacher value-added scores; however, less is known about the impact of teacher-level and school-level covariates. This study uses hierarchical linear modeling to examine the impact of including teacher characteristics, school characteristics, and student demographics aggregated at the school level on elementary mathematics and reading teacher value-added scores. Data for this study was collected from a large school district in north Texas. This study found that across all VAMs fitted, 32% of mathematics teachers and 37% of reading teachers changed quintile ranking for their value-added score at least once across all VAMs, while 55% and 65% of schools changed their quintile ranking of value-added scores based on mathematics and reading achievement, respectively. The results show that failing to control for aggregated student demographics has a large impact on both teacher level and school level value-added scores. Policymakers and administrators using VAM estimates in high-stakes decision-making should include teacher- and school-level covariates in their VAMs.
| Identifer | oai:union.ndltd.org:unt.edu/info:ark/67531/metadc1808459 |
| Date | 05 1900 |
| Creators | Allen, Lauren E. |
| Contributors | Chen, Qi (Educational psychologist), Hull, Darrell Magness, Middlemiss, Wendy, Savage, Melissa N. |
| Publisher | University of North Texas |
| Source Sets | University of North Texas |
| Language | English |
| Detected Language | English |
| Type | Thesis or Dissertation |
| Format | vi, 66 pages : illustrations, Text |
| Rights | Public, Allen, Lauren E., Copyright, Copyright is held by the author, unless otherwise noted. All rights Reserved. |
Page generated in 0.0019 seconds