Optimizing information using the EM algorithm in item response theory

详细信息    查看全文

• 作者：Alexander Weissman (1)
  
• 关键词：EM algorithm ; Item response theory ; Latent trait theory ; Statistical estimation ; Marginal maximum likelihood ; Kullback ; Leibler divergence ; Relative entropy ; Information Theory ; Model selection
• 刊名：Annals of Operations Research
• 出版年：2013
• 出版时间：July 2013
• 年：2013
• 卷：206
• 期：1
• 页码：627-646
• 全文大小：656KB
• 参考文献：1. Akaike, H. (1973). Information theory and an extension of the maximum likelihood principle. In B. N. Petrov & F. Csaki (Eds.), / Proceeding of the second international symposium on information theory (pp. 267-81). Budapest: Akademiai Kiado.
  2. Amari, S.-i. (1994). / Information geometry of the EM and em algorithms for neural networks. Tokyo: Department of Mathematical Engineering, University of Tokyo.
  3. Bain, L. J., & Engelhardt, M. (1992). / Introduction to probability and mathematical statistics. Pacific Grove: Duxbury.
  4. Baker, F. B., & Kim, S.-H. (2004). / Item response theory: parameter estimation techniques. New York: Dekker.
  5. Beal, M. J. (2003). / Variational algorithms for approximate Bayesian inference. Unpublished Doctoral dissertation, University of London.
  6. Birnbaum, A. (1968). Some latent trait models. In F. M. Lord & M. R. Novick, / Statistical theories of mental test scores. Reading: Addison-Wesley.
  7. Bock, R. D., & Aitkin, M. (1981). Marginal maximum likelihood estimation of item parameters: application of an EM algorithm. / Psychometrika, / 46(4), 443-59. CrossRef
  8. Bock, R. D., & Lieberman, M. (1970). Fitting a response model for / n dichotomously scored items. / Psychometrika, / 35(2), 179-97. CrossRef
  9. Bozdogan, H. (1987). Model selection and Akaike’s information criterion (AIC): the general theory and its analytical extensions. / Psychometrika, / 52(3), 345-70. CrossRef
  10. Cover, T. M., & Thomas, J. A. (2006). / Elements of information theory (2nd ed.). Hoboken: Wiley.
  11. Crocker, L., & Algina, J. (1986). / Introduction to classical and modern test theory. New York: Holt, Rinehart & Winston.
  12. Csiszár, I., & Tusnády, G. (1984). Information geometry and alternating minimization procedures. / Statistics & Decisions, / 1, 205-37. Supplement issue.
  13. de Boer, P.-T., Kroese, D., Mannor, S., & Rubinstein, R. (2005). A tutorial on the cross-entropy method. / Annals of Operations Research, / 134(1), 19-7. CrossRef
  14. de la Torre, J. (2009). DINA model and parameter estimation: a didactic. / Journal of Educational and Behavioral Statistics, / 34(1), 115-30. CrossRef
  15. Dellaert, F. (2002). / The expectation maximization algorithm (No. GIT-GVU-02-20). Georgia Institute of Technology.
  16. Dempster, A. P., Laird, N. M., & Rubin, D. B. (1977). Maximum likelihood from incomplete data via the EM algorithm. / Journal of the Royal Statistical Society, Series B (Methodological), / 39(1), 1-8.
  17. Hambleton, R. K., & Swaminathan, H. (1985). / Item response theory: principles and applications. Boston: Kluwer-Nijhoff.
  18. Harpaz, R., & Haralick, R. (2006). / The EM algorithm as a lower bound optimization technique (No. TR-2006001). Graduate Center, City University of New York.
  19. Harwell, M. R., Baker, F. B., & Zwarts, M. (1988). Item parameter estimation via marginal maximum likelihood and an EM algorithm: a didactic. / Journal of Educational Statistics, / 13(3), 243-71. CrossRef
  20. Ip, E. H., & Lalwani, N. (2000). Notes and Comments—A note on the geometric interpretation of the EM algorithm in estimating item characteristics and student abilities. / Psychometrika, / 65(4), 533. CrossRef
  21. Jank, W. (2006). The EM algorithm, its randomized implementation and global optimization: some challenges and opportunities for operations research. In F. B. Alt, M. C. Fu, & B. L. Golden (Eds.), / Perspectives in operations research (Vol.?36, pp. 367-92). New York: Springer. CrossRef
  22. Junker, B. W., & Sijtsma, K. (2001). Cognitive assessment models with few assumptions, and connections with nonparametric item response theory. / Applied Psychological Measurement, / 25(3), 258-72. CrossRef
  23. Kullback, S., & Leibler, R. A. (1951). On information and sufficiency. / The Annals of Mathematical Statistics, / 22(1), 79-6. CrossRef
  24. Lord, F. M., & Novick, M. R. (1968). / Statistical theories of mental test scores, with contributions by A.?Birnbaum. Reading: Addison-Wesley.
  25. McLachlan, G. J., & Krishnan, T. (2008). / The EM algorithm and extensions. Hoboken: Wiley-Interscience. CrossRef
  26. Minka, T. P. (1998). Expectation-maximization as lower bound maximization. Retrieved from http://research.microsoft.com/en-us/um/people/minka/papers/em.html. doi:10.1.1.29.8562 .
  27. Patz, R. J., & Junker, B. W. (1999). A straightforward approach to Markov chain Monte Carlo methods for item response models. / Journal of Educational and Behavioral Statistics, / 24(2), 146-78.
  28. Prescher, D. (2004). A tutorial on the expectation-maximization algorithm including maximum-likelihood estimation and EM training of probabilistic context-free grammars. Retrieved from http://arxiv.org/abs/cs/0412015. doi:10.1.1.87.8822 .
  29. SAS Institute (2008). / SAS-IML: Interactive Matrix Language (Version 9.2). Cary, NC.
  30. Spanos, A. (1999). / Probability theory and statistical inference: econometric modeling with observational data. Cambridge: Cambridge University Press.
  31. Spearman, C. (1907). Demonstration of formulae for true measurement of correlation. / The American Journal of Psychology, / 18, 161-69. CrossRef
  32. Spearman, C. (1913). Correlations of sums and differences. / British Journal of Psychology, / 5, 417-26.
  33. Weissman, A. (in press). Global convergence of the EM algorithm for unconstrained latent variable models with categorical indicators. / Psychometrika.
  34. Wets, R. J. B. (1999). Statistical estimation from an optimization viewpoint. / Annals of Operations Research, / 85(1), 79-02. CrossRef
  35. Wu, C. F. J. (1983). On the convergence properties of the EM algorithm. / The Annals of Statistics, / 11(1), 95-03. CrossRef
  
• 作者单位：Alexander Weissman (1)
  
  1. Psychometric Research, Law School Admission Council, 662 Penn Street, Box 40, Newtown, PA, 18940, USA
  
• ISSN：1572-9338

文摘

Latent trait models such as item response theory (IRT) hypothesize a functional relationship between an unobservable, or latent, variable and an observable outcome variable. In educational measurement, a discrete item response is usually the observable outcome variable, and the latent variable is associated with an examinee’s trait level (e.g., skill, proficiency). The link between the two variables is called an item response function. This function, defined by a set of item parameters, models the probability of observing a given item response, conditional on a specific trait level. Typically in a measurement setting, neither the item parameters nor the trait levels are known, and so must be estimated from the pattern of observed item responses. Although a maximum likelihood approach can be taken in estimating these parameters, it usually cannot be employed directly. Instead, a method of marginal maximum likelihood (MML) is utilized, via the expectation-maximization (EM) algorithm. Alternating between an expectation (E) step and a maximization (M) step, the EM algorithm assures that the marginal log likelihood function will not decrease after each EM cycle, and will converge to a local maximum. Interestingly, the negative of this marginal log likelihood function is equal to the relative entropy, or Kullback-Leibler divergence, between the conditional distribution of the latent variables given the observable variables and the joint likelihood of the latent and observable variables. With an unconstrained optimization for the M-step proposed here, the EM algorithm as minimization of Kullback-Leibler divergence admits the convergence results due to Csiszár and Tusnády (Statistics & Decisions, 1:205-37, 1984), a consequence of the binomial likelihood common to latent trait models with dichotomous response variables. For this unconstrained optimization, the EM algorithm converges to a global maximum of the marginal log likelihood function, yielding an information bound that permits a fixed point of reference against which models may be tested. A likelihood ratio test between marginal log likelihood functions obtained through constrained and unconstrained M-steps is provided as a means for testing models against this bound. Empirical examples demonstrate the approach.