The application of a cognitive diagnostic model on the test of practical Chinese

Cognitive Diagnostic Models (CDMs) are frequently used in the second language proficiency exams to analyze test takers’ strengths and weaknesses. This study retrofits a CDM known as the Fusion model onto a non-diagnostic test data from the reading comprehension test of the Test of Practical Chinese (C.TEST). It focuses on the construction and modification of the Q-matrix which specifies the relationship between test items and attributes needed to answer each item correctly. There are three stages in this study: (1) identifying reading attributes, (2) constructing the initial Q-matrix, and (3) modifying the initial Q-matrix. In the first stage, six reading attributes were identified based on literature review and 3 experts’ judgment. In the second stage, test items were carefully reviewed by three experts, and then the initial Q-matrix was created by them. In the third stage, item parameters and model fit statistics were evaluated to refine the Q-matrix. Four item parameters, r11.3, c2, c6, c27, were dropped from the Fusion model. Results suggest that the refined Q-matrix is well-defined, and it can be applied to the C.TEST reading test since it has a good model fit and is able to effectively distinguish item masters from item non-masters.

Name thesis:degree_name

M.S.

Level thesis:degree_level

Thesis

Discipline thesis:degree_discipline

Educational Psychology

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2019

Author dc:creator

• Tang, Xiuxiu

Contributors dc:contributor

• Zhang, Jinming
• Anderson, Carolyn
• Kern, Justin

• CDM
• Q-matrix

Statement dc:rights

• Copyright 2019 Xiuxiu Tang

Language dc:language

en

Handle dc:identifier

http://hdl.handle.net/2142/105274

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/105274