UVM Theses and Dissertations
Format:
Print
Author:
Sullivan, Valerie B.
Dept./Program:
College of Education and Social Services
Year:
2012
Degree:
Ed. D.
Abstract:
The purpose of this study was to investigate the use of teacher effectiveness scores as calculated by value-added models (VAM) to help identify effective teachers. VAM are a collection of statistical approaches intended to estimate the effect of educational inputs (teacher and school) on student outcomes (achievement). This quantitative study employed correlational analyses in an attempt to determine a relationship between value-added (VA) scores based on student achievement and subjective measures of teacher effectiveness such as teaching evaluations and confidential assessments provided by the supervising principal.
VA teacher effect scores were generated for four to five outcome measures per year for two years, using two different methods for calculating VA scores, for a total of 18 teacher VA scores. The resulting scores were standardized, ranked, and placed in quintiles and then correlated with the two subjective measures of teacher effectiveness. While VA scores did vary significantly across teachers, none of the VA scores were correlated with the subjective measures of teacher effectiveness. These results suggest that, in the context of a relatively small school district, VA scores not be used in isolation for high stakes decisions and that more research on methodology, estimation, model assumptions, and application of such scores is needed.
VA teacher effect scores were generated for four to five outcome measures per year for two years, using two different methods for calculating VA scores, for a total of 18 teacher VA scores. The resulting scores were standardized, ranked, and placed in quintiles and then correlated with the two subjective measures of teacher effectiveness. While VA scores did vary significantly across teachers, none of the VA scores were correlated with the subjective measures of teacher effectiveness. These results suggest that, in the context of a relatively small school district, VA scores not be used in isolation for high stakes decisions and that more research on methodology, estimation, model assumptions, and application of such scores is needed.