Using Cognitive Diagnostic Assessment to Model the Underlying Structure of a Listening Test: A Sub-Skill-Based Approach

| December 1, 2012
Title
Using Cognitive Diagnostic Assessment to Model the Underlying Structure of a Listening Test: A Sub-Skill-Based Approach

Keywords: cognitive diagnostic assessment, fusion model, IELTS listening module

Authors
Vahid Aryadoust
National University of Singapore
elcsva@nus.edu.sg

Bio Data
Vahid Aryadoust is a lecturer at the Centre for English Language Communication of the National University of Singapore (CELC-NUS). His current research focuses on the application of latent trait / class and neuro-fuzzy models as well as neuroanatomy in assessment. His most recent book on language assessment, entitled “Building a Validity Argument for a Listening Test of Academic Proficiency”, is based on his PhD thesis and will be published by Cambridge Scholars Publishing in 2013.

Abstract
Research into the psychological and cognitive aspects of language learning, and second language (L2) learning in particular, demands new measurement tools that provide highly detailed information about language learners’ progress and proficiency. A new development in measurement models is Cognitive Diagnostic Assessment (CDA), which helps language assessment researchers evaluate students’ mastery of specific language sub-skills with greater specificity than other item response theory models. This paper discusses the tenets of CDA models in general and the fusion model (FM) in particular, and reports the results of a study applying the FM to lecture-comprehension section of a practice version the International English Language Testing System (IELTS) listening module. FM separates only two major listening sub-skills (i.e., the ability to understand explicitly stated information and make close paraphrases), likely indicating construct-underrepresentation. It also provides a master / non-mastery profile of test takers. Implications for assessing listening comprehension and IELTS are discussed.
[private] See page: 81-106

Download PDF

[/private]

Category: Quarterly Journal, Volume 14 Issue 4