Multilevel modeling of cognitive diagnostic assessment

Xuelan QIU, Wen Chung WANG

Research output: Contribution to conferencePapers

Abstract

Most cognitive diagnostic models are unilevel. This study develops a multilevel version of the log-linear cognitive diagnosis model. The simulation results showed that the parameters and latent attribute profiles were recovered fairly well; and ignoring multilevel structure would underestimate the standard error and yield a poor classification rate.
Original languageEnglish
Publication statusPublished - Apr 2014
Event2014 Annual Meeting of National Council on Measurement in Education - Philadelphia, United States
Duration: 04 Apr 201406 Apr 2014

Conference

Conference2014 Annual Meeting of National Council on Measurement in Education
Abbreviated titleNCME 2014
Country/TerritoryUnited States
CityPhiladelphia
Period04/04/1406/04/14

Citation

Qiu, X.-L., & Wang, W.-C. (2014, April). Multilevel modeling of cognitive diagnostic assessment. Paper presented at the 76th Annual Meeting of the National Council on Measurement in Education, Philadelphia, PA.

Fingerprint

Dive into the research topics of 'Multilevel modeling of cognitive diagnostic assessment'. Together they form a unique fingerprint.