Date of Award

5-2018

Document Type

Dissertation

Degree Name

Doctor of Education (EdD)

First Advisor

Jenny Sabin

Abstract

This quantitative study addresses the ability of locally developed Math I benchmark assessments to predict student performance on the Math I end-of-course (EOC) state assessment for a rural county in North Carolina. Many districts in North Carolina lack funds to purchase commercially developed benchmarks and must rely on district personnel to develop these assessments. Locally developed benchmark assessments are not typically validated or considered psychometrically sound. Signal Detection Theory (SDT) was used to determine cut-off values on the locally developed Math I benchmark assessments for pass/fail grouping of students that will inform instructional interventions prior to the Math I EOC state assessment. The Math I benchmark assessments were determined to be accurate predictors of student performance on the North Carolina Math I EOC assessment, and the sensitivity and specificity of the Math I benchmark assessments were calculated to greater than 70% accuracy for each benchmark based on the identified cut-off scores.

Included in

Education Commons

Share

COinS