Value-Added Assessment Models with Continuous and Categorical Responses

• Activity Number: 59
• Type: Topic Contributed
• Date/Time: Sunday, August 1, 2010 : 4:00 PM to 5:50 PM
• Sponsor: Social Statistics Section
• Author(s): Sharon Lohr*+ and Jennifer Broatch
• Companies: Arizona State University and Arizona State University
• Keywords: generalized linear mixed model ; random effects ; maximum likelihood ; SAS code ; multivariate mixed model

Abstract

Value-added assessment (VAA) models attempt to estimate the value that an individual teacher adds to students' knowledge, through apportioning changes in student responses from a longitudinal database to their teachers. Traditional VAA models generally use responses such as scores on successive administrations of a standardized test. We develop a multi-response approach to VAA models that allows responses to be either continuous or categorical. This allows us to study the value added to real-world responses such as student graduation or future career in science or engineering, and makes the models usable in university settings where standardized tests are not given. We apply the models to data on university calculus instruction and graduation in a science or engineering field, and find a negative association for teacher contribution to the two responses.