Diagnostic Item Analysis: A Comparison of Methods in Detecting Schools’ Strengths and Weaknesses

• In Event: Poster Session 10
  In Poster Session: Diverse Topics in Psychometrics and Educational Measurement

Abstract

This paper illustrates and compares different methods to produce item analysis graphics that would be useful to diagnose strengths and weaknesses at an aggregate level (e.g., schools). These diagnostic item analysis graphics can assist teachers and curriculum leaders in the analysis of test results at the item or standard level. The data is from a state-wide G8 science assessment for a large Midwestern school district having seven middle schools. The results include graphical displays of the diagnostic item analysis graphics which are compared and contrasted across the robust effect-size and Rasch residual methods. In addition, all WinBUGS and R code are included for each robust effect-size method, Rasch residuals analysis, and ggplot2 graphics that developed the diagnostic item analysis graphics.

Authors

• John N. Denbleyker, University of Iowa

• Hyo Jeong Shin, University of California - Berkeley

• Shuqin Tao, Data Recognition Corporation