Two Applications of Complex Sum Scores for Diagnostic Assessment

• In Event: Poster Session 1
  In Poster Session: Division D Poster Session 4

Abstract

To evaluate student learning using cognitive diagnostic assessment(CDA), scoring methods for determining student mastery of a set of cognitive skills can be divided into two approaches: sub-score reporting and probabilistic modeling. An alternative to these two approaches, named complex sum scores, has received little attention since its introduction. With the process of developing model-based diagnostic assessments becoming increasingly complex, this study demonstrate an application of CSS with two CDA development tasks: illustrating skill differences within a cognitive model, and partial mastery scoring using model-based distracters. The goal of this study is to promote CSS as a low-stake CDA scoring method, and demonstrate how CSS can be used to score CDA in an accessible manner.

Authors

• Hollis Lai, University of Alberta

• Oksana Babenko, University of Alberta

• Mark J. Gierl, University of Alberta