The Application of Minimum Discrepancy Estimation in Implementation of Cognitive Diagnostic Models

• In Event: 18.020 - Research on Cognitive Diagnostic Modeling

Abstract

Cognitive diagnostic models (CDMs) and their implementation in operational assessments, such as the NetPASS assessment, have been gaining increasing interest and attention among researchers and practitioners. Operational CDMs require parameter estimations to be accurate and efficient. This study aims to introduce two new estimators – minimum discrepancy (MD) and its extension the minimum discrepancy maximum likelihood (MDML) – in the context of CDMs and to compare the performance of the new methods with two traditional estimators in terms of accuracy and efficiency of CDM parameter estimation.

Authors

• Shenghai Dai, Washington State University - Pullman

• Xiaolin Wang, NBOME

• Dubravka Svetina, Indiana University - Bloomington

• Stephanie Underhill, Indiana University - Bloomington

• Yanan Feng, McKinsey & Company