Date of Award
2003
Document Type
Dissertation
Degree Name
Doctor of Education
Department
EdD Program for Educational Leaders
Advisor
David Shellman
Committee Member
Marty Ward
Committee Member
Barbara Packer
Committee Member
Claudia Santin
Keywords
demographic information, educational assessment, ethnicity, gender, HLM 5, hierarchical linear models, North Carolina, North Carolina End-of-Course test, socioeconomic status, state EOC test, student achievement, student assessment, student learning, teacher effectiveness, teacher evaluation
Abstract
This applied dissertation was designed to examine new ways to evaluate teachers. The extant North Carolina model placed primary emphasis on the ability of the teacher to demonstrate the certain key practices. These practices centered around the activity of the teacher. The extant evaluation system failed to take into account student learning. This resulted in a system in which over 95% of teachers were rated at standard or better. The high rating of teachers contrasted to the passing rate of students on the North Carolina End-of-Course test. Their passing rate was closer to 60%.
The writer developed a statistical model that examined teacher effectiveness in light of student achievement. Student test data and demographic information were collected and analyzed using the latest Hierarchical Linear Modeling software, HLM 5 (Raudenbush, Bryk, Cheong, & Congdon, 2001). These data were used to determine the effectiveness of individual teachers when certain factors, such as ethnicity, gender, and socioeconomic status were controlled.
An analysis of the data revealed that a new evaluation model was possible. This model could be used to measure teacher effectiveness in light of student outcomes on the state EOC test. The model also adjusted for such factors as ethnicity, gender, socioeconomic status, and other relevant factors. There were some limitations to the data, and further research was deemed necessary before full implementation of the evaluation model.