|Title:||Psychometric Models for In-Classroom Observational Measures of Teaching with Applications to Value-Added Modeling|
|Principal Investigator:||Halpin, Peter||Awardee:||New York University|
|Program:||Statistical and Research Methodology in Education–Early Career [Program Details]|
|Award Period:||1½ years (9/1/14–2/29/16)||Award Amount:||$199,918|
|Type:||Methodological Innovation||Award Number:||R305D140035|
Value-added modeling is an approach to teacher evaluation that yields information about teacher performance as a function of student scores on standardized tests. While value-added modeling is based on student outcomes, this technique does not allow evaluators to focus on what teachers are doing or should be doing in their classrooms to influence those outcomes. Observational instruments offer the possibility of identifying in-classroom practices that are associated with student achievement. Recent research has found, however, that many of the instruments commonly used for in-classroom observation have poor reliability, and consequently they have also demonstrated poor validity against teachers’ value-added estimates. The latter problem is aggravated by the fact that the value-added estimates also contain a substantial degree of sampling error.
In this study, researchers will apply latent class analysis to improve how the in-classroom observational instruments are scored and how they are incorporated into the value-added framework. First, researchers will apply psychometric methods to attempt to improve the reliability and validity of in-classroom observational measures of teaching, using academic and non-academic outcomes to gauge and to compare concurrent and predictive validity of the resulting models. Secondly, researchers will determine whether the improvements can yield advances in current research on teaching effectiveness, especially value-added modeling, by incorporating latent class assignment into prediction equations of teacher performance. Researchers will use data from two frequently used teacher and classroom observation instruments, Danielson’s Framework for Teachers and the Classroom Assessment Scoring SystemTM (CLASSTM), in the development and testing of psychometric approaches that incorporate latent class analysis into value-added modeling.
Journal article, monograph, or newsletter
Halpin, P.F., and Kieffer, M.J. (2015). Describing Profiles of Instructional Practice: A New Approach to Analyzing Classroom Observation Data. Educational Researcher, 44(5): 263–277.