Testing Expert-Based Versus Student-Based Cognitive Models for a Grade 3 Diagnostic Mathematics Assessment

The purpose of this study was to evaluate the adequacy of three cognitive models, one developed by content experts and two generated from student verbal reports for explaining examinee performance on a grade 3 diagnostic mathematics test. For this study, the items were developed to directly measure...

Full description

Saved in:
Bibliographic Details
Published in:Applied measurement in education 2014-07, Vol.27 (3), p.173-195
Main Authors: Roduta Roberts, Mary, Alves, Cecilia B., Chu, Man-Wai, Thompson, Margaret, Bahry, Louise M., Gotzmann, Andrea
Format: Article
Language:English
Subjects:
Academic achievement
Canada
Cognitive Measurement
Cognitive models
Cognitive Processes
Cognitive Tests
Diagnostic Tests
Difficulty Level
Educational evaluation
Foreign Countries
Mathematics education
Mathematics Tests
Measures (Individuals)
Psychometrics
Scores
Test Items
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The purpose of this study was to evaluate the adequacy of three cognitive models, one developed by content experts and two generated from student verbal reports for explaining examinee performance on a grade 3 diagnostic mathematics test. For this study, the items were developed to directly measure the attributes in the cognitive model. The performance of each cognitive model was evaluated by examining its fit to different data samples: verbal report, total, high-, moderate-, and low ability using the Hierarchy Consistency Index (Cui & Leighton, 2009), a model-data fit index. This study utilized cognitive diagnostic assessments developed under the framework of construct-centered test design and analyzed using the Attribute Hierarchy Method (Gierl, Wang, & Zhou, 2008; Leighton, Gierl, & Hunka, 2004). Both the expert-based and the student-based cognitive models provided excellent fit to the verbal report and high ability samples, but moderate to poor fit to the total, moderate and low ability samples. Implications for cognitive model development for cognitive diagnostic assessment are discussed.
ISSN:0895-7347
1532-4818
DOI:10.1080/08957347.2014.905787