Premium
Skills Diagnosis Using IRT‐Based Latent Class Models
Author(s) -
Roussos Louis A.,
Templin Jonathan L.,
Henson Robert A.
Publication year - 2007
Publication title -
journal of educational measurement
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.917
H-Index - 47
eISSN - 1745-3984
pISSN - 0022-0655
DOI - 10.1111/j.1745-3984.2007.00040.x
Subject(s) - equating , item response theory , reliability (semiconductor) , class (philosophy) , latent class model , computer science , variety (cybernetics) , artificial intelligence , psychometrics , machine learning , psychology , natural language processing , power (physics) , clinical psychology , developmental psychology , physics , quantum mechanics , rasch model
This article describes a latent trait approach to skills diagnosis based on a particular variety of latent class models that employ item response functions (IRFs) as in typical item response theory (IRT) models. To enable and encourage comparisons with other approaches, this description is provided in terms of the main components of any psychometric approach: the ability model and the IRF structure; review of research on estimation, model checking, reliability, validity, equating, and scoring; and a brief review of real data applications. In this manner the article demonstrates that this approach to skills diagnosis has built a strong initial foundation of research and resources available to potential users. The outlook for future research and applications is discussed with special emphasis on a call for pilot studies and concomitant increased validity research.