Inferring the Number of Attributes for the Exploratory DINA Model
- 1 March 2021
- journal article
- research article
- Published by Cambridge University Press (CUP) in Psychometrika
- Vol. 86 (1) , 30-64
Abstract
Diagnostic classification models (DCMs) are widely used for providing fine-grained classification of a multidimensional collection of discrete attributes. The application of DCMs requires the specification of the latent structure in what is known as the \documentclass[12pt]{minimal}\usepackage{amsmath}\usepackage{wasysym}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{amsbsy}\usepackage{mathrsfs}\usepackage{upgreek}\setlength{\oddsidemargin}{-69pt}$${\varvec{Q}}$$ matrix. Expert-specified \documentclass[12pt]{minimal}\usepackage{amsmath}\usepackage{wasysym}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{amsbsy}\usepackage{mathrsfs}\usepackage{upgreek}\setlength{\oddsidemargin}{-69pt}$${\varvec{Q}}$$ matrices might be biased and result in incorrect diagnostic classifications, so a critical issue is developing methods to estimate \documentclass[12pt]{minimal}\usepackage{amsmath}\usepackage{wasysym}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{amsbsy}\usepackage{mathrsfs}\usepackage{upgreek}\setlength{\oddsidemargin}{-69pt}$${\varvec{Q}}$$ in order to infer the relationship between latent attributes and items. Existing exploratory methods for estimating \documentclass[12pt]{minimal}\usepackage{amsmath}\usepackage{wasysym}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{amsbsy}\usepackage{mathrsfs}\usepackage{upgreek}\setlength{\oddsidemargin}{-69pt}$${\varvec{Q}}$$ must pre-specify the number of attributes, K . We present a Bayesian framework to jointly infer the number of attributes K and the elements of \documentclass[12pt]{minimal}\usepackage{amsmath}\usepackage{wasysym}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{amsbsy}\usepackage{mathrsfs}\usepackage{upgreek}\setlength{\oddsidemargin}{-69pt}$${\varvec{Q}}$$ . We propose the crimp sampling algorithm to transit between different dimensions of K and estimate the underlying \documentclass[12pt]{minimal}\usepackage{amsmath}\usepackage{wasysym}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{amsbsy}\usepackage{mathrsfs}\usepackage{upgreek}\setlength{\oddsidemargin}{-69pt}$${\varvec{Q}}$$ and model parameters while enforcing model identifiability constraints. We also adapt the Indian buffet process and reversible-jump Markov chain Monte Carlo methods to estimate \documentclass[12pt]{minimal}\usepackage{amsmath}\usepackage{wasysym}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{amsbsy}\usepackage{mathrsfs}\usepackage{upgreek}\setlength{\oddsidemargin}{-69pt}$${\varvec{Q}}$$ . We report evidence that the crimp sampler performs the best among the three methods. We apply the developed methodology to two data sets and discuss the implications of the findings for future research.
Keywords
Funding Information
- Division of Social and Economic Sciences (1758631)
- Division of Social and Economic Sciences (1758631)
- Division of Social and Economic Sciences (1758688)
- Division of Social and Economic Sciences (1951057)
This publication has 37 references indexed in Scilit:
- Minimum Discrepancy Estimation in Probabilistic Knowledge StructuresElectronic Notes in Discrete Mathematics, 2013
- A tutorial on Bayesian nonparametric modelsJournal of Mathematical Psychology, 2012
- The Generalized DINA Model FrameworkPsychometrika, 2011
- Defining a Family of Cognitive Diagnosis Models Using Log-Linear Models with Latent VariablesPsychometrika, 2009
- Model Evaluation and Multiple Strategies in Cognitive Diagnosis: An Analysis of Fraction Subtraction DataPsychometrika, 2008
- A general diagnostic model applied to language testing dataBritish Journal of Mathematical and Statistical Psychology, 2008
- Higher-Order Latent Trait Models for Cognitive DiagnosisPsychometrika, 2004
- Data Analytic Methods for Latent Partially Ordered Classification ModelsJournal of the Royal Statistical Society Series C: Applied Statistics, 2002
- Reversible jump Markov chain Monte Carlo computation and Bayesian model determinationBiometrika, 1995
- The Collapsed Gibbs Sampler in Bayesian Computations with Applications to a Gene Regulation ProblemJournal of the American Statistical Association, 1994