Open AccessREPORTING TEST OUTCOMES USING MODELS FOR COGNITIVE DIAGNOSIS
Author(s) -
Davier Matthias,
DiBello Lou,
Yamamoto Kentaro
Publication year - 2006
Publication title -
ets research report series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.235
H-Index - 5
ISSN - 2330-8516
DOI - 10.1002/j.2333-8504.2006.tb02034.x
Subject(s) - cognition , latent class model , item response theory , confirmatory factor analysis , computer science , machine learning , test (biology) , completeness (order theory) , artificial intelligence , cognitive psychology , cognitive model , psychology , econometrics , psychometrics , structural equation modeling , mathematics , developmental psychology , paleontology , mathematical analysis , neuroscience , biology
Models for cognitive diagnosis have been developed as an attempt to provide more than a single test score from item response data. Most approaches are based on a hypothesis that relates items to underlying skills. This relation takes the form of a design matrix that specifies for each cognitive item which skills are required to solve the item and which are not. This report outlines one direction that developments of cognitive diagnosis models is taking. It does not claim completeness, but describes a line of models that can be traced back to Tatsuoka's seminal work on the rule space methodology and that finds its current form in models that combine features of confirmatory latent factor analysis, multiple classification latent class models, and multidimensional item response models.