Paper Summary

Detecting Item Sensitivity to Instruction: A Comparison Between Mantel-Haenszel and Logistic Regression Procedures

Fri, April 13, 12:00 to 1:30pm, Vancouver Convention Centre, Floor: First Level, East Ballroom B

Abstract

The purpose of this study is to explore the relationship between students' instructional experiences and their test scores in standardized achievement tests. Through studies of students' performance on test items in a Midwestern state interim assessment on mathematics, the study aims to detect items that are able to reflect the impact of effective instruction. This study will examine how students' performance on test items differ before and after instruction, detect what items can effectively reflect instruction on the tested content, and explore what common characteristics these items have. Results of this study will provide educators and high-level authorities with new insights and knowledge for both improving achievement measurement process and taking action to bolster the practice of instruction in schools.

Authors