The Politics and Statistics of Value-Added Modeling for Accountability of Teacher Preparation Programs

• Published in: Journal of teacher education Vol. 65; no. 1; pp. 24 - 38
• Main Authors: Lincove, Jane Arnold, Osborne, Cynthia, Dillon, Amanda, Mills, Nicholas
• Format: Journal Article
• Language: English
• Published: Los Angeles, CA SAGE Publications 01.01.2014; Corwin Press, Inc; SAGE PUBLICATIONS, INC
• Subjects: Accountability; Achievement Gains; Achievement Rating; Comparative Analysis; Computation; Curricula; Education reform; Educational Change; Educational Policy; Educational reform; Elementary Education; Evaluation Criteria; Evaluation Methods; Government accountability; Government regulation; Item Analysis; Laws, regulations and rules; Politics of Education; Predictor Variables; Robustness (Statistics); School Effectiveness; School Statistics; Stakeholders; State Policy; Teacher centers; Teacher centres; Teacher education; Teacher Education Programs; Teacher Effectiveness; Teachers; Test Reliability; Test Validity; Texas; Training; Validity; Texas; United States; United States > US; education reform; school/teacher effectiveness; educational policy; elementary teacher education
• ISSN: 0022-4871; 1552-7816
• DOI: 10.1177/0022487113504108

Despite questions about validity and reliability, the use of value-added estimation methods has moved beyond academic research into state accountability systems for teachers, schools, and teacher preparation programs (TPPs). Prior studies of value-added measurement for TPPs test the validity of researcher-designed models and find that measuring differences across programs is difficult. This study is the first to examine the reliability and usefulness of a value-added model for TPPs developed through a collaborative stakeholder process and mandated by state law for use in accountability. Based on the experience of developing a test-based metric for Texas TPPs, our results suggest that although value-added results are statistically robust, accountability status for individual programs is very sensitive to decisions about accountability criteria, the selection of teachers, and the selection of control variables.