Evaluating Subgroup Analysis Indices and Guidelines for Automated Scoring Algorithm

• In Event: GSIC Poster Session I

Abstract

There is limited research regarding operational use of agreement indices to detect subgroup differences between automated and human scores. Our study evaluates the accuracy of this application and proposed guidelines based on results of a Monte Carlo simulation.

Authors

• Yichi Zhang, AMERICAN INSTITUTES FOR RESEARCH

• Edward W Wolfe, Pearson