Reporting Gain Scores Using Multiple Methods

• In Event: AERA Poster Session 6
  In Poster Session: Division D Section 1 Poster Session 2

Abstract

More and more programs are designed to measure multiple domains in addition to measuring one content domain. Reports are usually designed to show students’ overall achievement using overall score, and detailed strengths and weaknesses at subscore level. The goal of this study is to compare the differences among different models (Multidimensional IRT vs. Classical Testing Theory and Unidimensional IRT) and their impact on institutions’ gain score calculations. Using a college level learning outcome assessment, we applied different models to conduct calibration and equating. The equating results from the three models were compared in terms of the gain scores at total score and subscore level for each institution across 2 time points.

Authors

• Zhen Wang, ETS

• Yu Sun, ETS