Advertisement

Psychometrika

, Volume 54, Issue 4, pp 557–585 | Cite as

Latent variable modeling in heterogeneous populations

  • Bengt O. Muthén
Article

Abstract

Common applications of latent variable analysis fail to recognize that data may be obtained from several populations with different sets of parameter values. This article describes the problem and gives an overview of methodology that can address heterogeneity. Artificial examples of mixtures are given, where if the mixture is not recognized, strongly distorted results occur. MIMIC structural modeling is shown to be a useful method for detecting and describing heterogeneity that cannot be handled in regular multiple-group analysis. Other useful methods instead take a random effects approach, describing heterogeneity in terms of random parameter variation across groups. These random effects models connect with emerging methodology for multilevel structural equation modeling of hierarchical data. Examples are drawn from educational achievement testing, psychopathology, and sociology of education. Estimation is carried out by the LISCOMP program.

Key words

mixtures covariance structures multiple-group analysis MIMIC LISCOMP random parameters multilevel hierarchical data 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. Aitkin, M., & Longford, N. (1986). Statistical modeling issues in school effectiveness studies.Journal of the Royal Statistical Society, Series A149, 1–43.Google Scholar
  2. Bentler, P. M. (1980). Multivariate analysis with latent variables: Causal modeling.Annual Review of Psychology, 31, 19–456.CrossRefGoogle Scholar
  3. Bentler, P. M. (1983). Some contributions to efficient statistics in structural models: Specification and estimation of moment structures.Psychometrika, 48, 493–518.Google Scholar
  4. Bock, R. D. (1983). The discrete Bayesian. In H. Wainer & S. Messick (Eds.),Principals of modern psychological measurement (pp. 103–115). Hillsdale, NJ: Erlbaum.Google Scholar
  5. Bock, R. D. (1989a).Multilevel analysis of educational data. San Diego, CA: Academic Press.Google Scholar
  6. Bock, R. D. (1989b). Measurement of human variation: A two-stage model. In R. D. Bock (Ed.),Multilevel analysis of educational data (pp. 319–340). San Diego, CA: Academic Press.Google Scholar
  7. Braun, H., Jones, D., Rubin, D., & Thayer, D. (1983). Empirical Bayes estimation of coefficients in the general linear model from data of deficient rank.Psychometrika, 48, 171–181.CrossRefGoogle Scholar
  8. Browne, M. W. (1982). Covariance structures. In D. M. Hawkins (Ed.)Topics in applied multivariate analysis. Cambridge, MA: Cambridge University Press.Google Scholar
  9. Browne, M. W. (1984). Asymptotically distribution free methods for the analysis of covariance structures.British Journal of Mathematical and Statistical Psychology, 37, 62–83.PubMedGoogle Scholar
  10. Burstein, L. (1980). The analysis of multilevel data in educational research and evaluation.Review of Research in Education, 8, 158–233.Google Scholar
  11. Burstein, L., Kim, K. S., & Delandshere, G. (1988). Multilevel investigation of systematically varying slopes: Issues, alternatives, and consequences. In R. D. Bock (Ed.),Multilevel analysis of educational data (pp. 235–276). San Diego, CA: Academic Press.Google Scholar
  12. Converse, P. E. (1964). The nature of belief systems in mass publics. In D. Apter (Ed.),Ideology and discontent. London, England: The Free Press.Google Scholar
  13. Cronbach, L. J. (1976).Research on classrooms and schools: Formulation of questions, design, and analysis. Unpublished manuscript, Stanford University, Stanford Evaluation Consortium, School of Education.Google Scholar
  14. Crosswhite, F. J., Dossey, J. A., Swafford, J. O., McKnight, C. C., & Cooney, T. J. (1985).Second international mathematics study: Summary report for the United States. Champaign, IL: Stipes.Google Scholar
  15. de Leeuw, J. (1985).Path models with random coefficients. Leiden, The Netherlands: University of Leiden, Department of Data Theory.Google Scholar
  16. de Leeuw, J., & Kreft, I. (1986). Random coefficient models for multilevel analysis.Journal of Educational Statistics, 11, 57–85.Google Scholar
  17. Eaton, W., & Bohrnstedt, G. (1989). Introduction. Latent Variable Models for Dichotomous Outcomes: Analysis of data from the Epidemiological Catchment Area Program.Sociological Methods & Research, 18, 4–18.Google Scholar
  18. Everitt, B., & Hand, D. J. (1981).Finite Mixture Distributions. New York: Chapman and Hall.Google Scholar
  19. Gibbons, R., & Bock, R. (1987). Trend in correlated proportions.Psychometrika, 52, 113–124.CrossRefGoogle Scholar
  20. Goldstein, H. I. (1986). Multilevel mixed linear model analysis using iterative generalized least squares.Biometrika, 73, 43–56.Google Scholar
  21. Goldstein, H. I. (1987).Multilevel Models in Educational and Social Research. London: Oxford University Press.Google Scholar
  22. Goldstein, H., & McDonald, R. P. (1988). A general model for the analysis of multilevel data.Psychometrika, 53, 455–467.CrossRefGoogle Scholar
  23. Gustafsson, J. E. (1988). Hierarchical models of individual differences in cognitive abilities. In R. J. Sternberg (Ed.),Advances in the psychology of human intelligence, Vol. 4 (pp. 35–71). Hillsdale, NJ: Lawrence Erlbaum.Google Scholar
  24. Gustafsson,J. E. (in press). Broad and narrow abilities in research on learning and instruction. In R. Kanfer, P. L. Ackerman, & R. Cudeck (Eds.),Abilities, motivation, and methodology: The Minnesota symposium on learning and individual differences. Hillsdale, NJ: Lawrence Erlbaum.Google Scholar
  25. Härnquist, K. (1978). Primary mental abilites of collective and individual levels.Journal of Educational Psychology, 70, 706–716.Google Scholar
  26. Hauser, R. M., & Goldberger, A. S. (1971). The treatment of unobservable variables in path analysis. In H. L. Costner (Ed.),Sociological Methodology 1971 (pp. 81–177). San Francisco: Jossey-Bass.Google Scholar
  27. Hedeker, D., Gibbons, R. D., & Waternaux, C. (1988, June).Random regression models for longitudinal psychiatric data. Paper presented at the Annual Meeting of the Psychometric Society, Los Angeles.Google Scholar
  28. Hollis, M., & Muthén B. (1988).Incorporating “Item-specific” information into covariance structure models of political attitudes and policy preferences. Paper prepared for delivery at the 1988 Annual Meeting of the American Political Science Association.Google Scholar
  29. Johnson, N., & Kotz, S. (1972).Distribution in statistics: Continuous multivariate distributions. New York: John Wiley & Sons.Google Scholar
  30. Jöreskog, K. G. (1971). Simultaneous factor analysis in several populations.Psychometrika, 36, 409–426.Google Scholar
  31. Jöreskog, K. G. (1973). A general method for estimating a linear structural equation system. In A. S. Goldberger & O. D. Duncan (Eds.),Structural equation models in the social sciences (pp. 85–112) New York: Seminar Press.Google Scholar
  32. Jöreskog, K. G. (1978). Structural analysis of covariance and correlation matrices.Psychometrika, 43, 443–477.CrossRefGoogle Scholar
  33. Jöreskog, K. G., & Goldberger, A. S. (1975). Estimation of a model with multiple indicators and multiple causes of a single latent vaviable.Journal of the American Statistical Association, 70, 351.Google Scholar
  34. Keesling, J. W., & Wiley, D. E. (1974, March).Regression models of hierarchical data. Paper presented at the Annual Meeting of the Psychometric Society, Palo Alto, CA.Google Scholar
  35. Lindley, D. V., & Smith, A. F. M. (1972). Bayes estimates for the linear model.Journal of the Royal Statistical Society, Series B,34, 1–41.Google Scholar
  36. Longford, N. T. (1987). A fast scoring algorithm for maximum likelihood estimation in unbalanced mixed models with nested effects.Biometrika, 74, 817–827.Google Scholar
  37. Longford, N. T. (1988).A quasilikelihood adaption for variance component analysis. Princeton, NJ: Educational Testing Service.Google Scholar
  38. Long, F. M., & Novick, M. R. (1968).Statistical theories of mental test scores. Reading, MA: Addison-Wesley.Google Scholar
  39. Maddala, G. S. (1977).Econometrics. New York: McGraw-Hill.Google Scholar
  40. Maddala, G. S. (1983).Limited-dependent and qualitative variables in econometrics. Cambridge, MA: Cambridge University Press.Google Scholar
  41. Mason, W. M., Wong, G. Y., & Entwistle, B. (1984). Contextual analysis through the multi-level linear model.Sociological Methodology, San Francisco, CA: Jossey-Bass.Google Scholar
  42. McDonald, R. P., & Goldstein, H. (1988, June).Balanced versus unbalanced designs for linear structural relations in two-level data. Paper presented at the Annual meeting of the Psychometric Society, Los Angeles.Google Scholar
  43. Meredith, W. (1964). Notes on factorial invariance.Psychometrika, 29, 177–185.Google Scholar
  44. Mislevy, R. J. (1975). Estimation of latent group effects.Journal of the American Statistical Association, 80, 993–997.Google Scholar
  45. Mislevy, R. J. (1987). Exploiting auxiliary information about examinees in the estimation of item parameters.Applied Psychological Measurement, 11, 81–91.Google Scholar
  46. Mundlak, Y. (1978). Models with variable coefficients: Integration and extension.Annales de'l INSEE, 30–31, 483–509.Google Scholar
  47. Muthén, B. (1978). Contributions to factor analysis of dichotomous variables.Psychometrika, 43, 551–560.CrossRefGoogle Scholar
  48. Muthén, B. (1979). A structural probit model with latent variables.Journal of the American Statistical Association, 74, 807–811.Google Scholar
  49. Muthén, B. (1983). Latent variable structural equation modeling with categorical data.Journal of Econometrics, 22, 43–65.Google Scholar
  50. Muthén, B. (1984). A general structural equation model with dichotomous, ordered categorical, and continuous latent variable indicators.Psychometrika, 49, 115–132.CrossRefGoogle Scholar
  51. Muthén, B. (1985, July).Tobit factor analysis. Paper presented at the Fourth European Meeting of the Psychometric Society, Cambridge, England. (Forthcoming inThe British Journal of Mathematical and Statistical Psychology.Google Scholar
  52. Muthén, B. (1987).LISCOMP: Analysis of linear structural equations with a comprehensive measurement model [User's Guide]. Mooresville, IN: Scientific Software.Google Scholar
  53. Muthén, B. (1988a). Some uses of structural equation modeling in validity studies: Extending IRT to external variables. In H. Wainer & H. Braun (Eds.),Test Validity (pp. 213–238). Hillsdale, NJ: Lawrence Erlbaum.Google Scholar
  54. Muthén, B. (1988b).Instructionally sensitive psychometrics: Applications to the Second International Mathematics Study. Unpublished manuscript, University of California, Graduate School of Education, Los Angeles.Google Scholar
  55. Muthén, B. (1989a). Using item-specific instructional information in achievement modeling.Psychometrika, 54, 385–396.CrossRefGoogle Scholar
  56. Muthén, B. (1989b).Covariance structure modeling in heterogeneous populations: Mean-adjusted analysis. In preparation, University of California, Graduate School of Education, Los Angeles.Google Scholar
  57. Muthén, B. (1989c). Dichotomous factor analysis of symptom data. In Eaton and Bohrnstedt (Eds.), Latent Variable Models for Dichotomous Outcomes: Analysis of Data from the Epidemiological Catchment Area Program.Sociological Methods & Reserch, 18, 19–65.Google Scholar
  58. Muthén, B. (1989d). Multiple-group factor analysis with non-normal continuous variables.British Journal of Mathematical and Statistical Psychology, 42, 55–62.Google Scholar
  59. Muthén, B. (1989e).Covariance structure analysis of hierarchical data. Paper presented at the American Statistical Association meeting in Washington D.C.Google Scholar
  60. Muthén, B. (1989f, October).Analysis of longitudinal data using latent variable models with varying parameters. Invited paper presented at the University of Southern California Conference on Best Methods for the Analysis of Change, Los Angeles. (Forthcoming as a chapter in a book edited by J. Horn and L. Collins)Google Scholar
  61. Muthén, B. (in press). Moments of the censored and truncated bivariate normal distribution.British Journal of Mathematical and Statistical Psychology.Google Scholar
  62. Muthén, B., Burstein, L., Gustafsson, J.-E., Webb, N., Kim, S-W., Short, L. (1989).General and specific factors in mathematics achievement data. In preparation, University of California, Graduate School of Education, Los Angeles.Google Scholar
  63. Muthén, B., & Christoffersson, A. (1981). Simultaneous factor analysis of dichotomous variables in several groups.Psychometrika, 46, 485–500.CrossRefGoogle Scholar
  64. Muthén, B., & Jöreskog. (1983). Selectivity problems in quasi-experimental studies.Evaluation Review, 7, 139–173.Google Scholar
  65. Muthén, B., Kao, Chih-fen, & Burstein, L. (in press). Instructional sensitivity in mathematics achievement test items: Applications of a new IRT-based detection technique.Journal of Educational Measurement.Google Scholar
  66. Muthén, B., & Satorra, A. (1989). Multilevel aspects of varying parameters in structural models. In Bock(Ed.),Multilevel Analysis of Educational Data (pp. 87–99). San Diego: Academic Press. (Invited paper for the conference onMultilevel analysis of educational data, Princeton, NJ, April 1987)Google Scholar
  67. Olsson, U. (1979). Maximum likelihood estimation of the polychoric correlation coefficient.Psychometrika, 44, 443–460.CrossRefGoogle Scholar
  68. Olsson, U., Drasgow, F., & Dorans, N. J. (1982). The polyserial correlation coefficient.Psychometrika, 37, 337–347.Google Scholar
  69. Raudenbush, S., & Bryk, A. (1988). Methodological advances in studying effects of schools and classrooms on student learning.Review of Research in Education, 1988.Google Scholar
  70. Robinson, W. S. (1950). Ecological correlations and the behavior of individuals.American Sociological Review, 15, 351–357.Google Scholar
  71. Rogosa, D. R. (1987). Causal models do not support scientific conclusions: A comment in support of Freedman.Journal of Educational Statistics, 12, 185–195.Google Scholar
  72. Rogosa, D. R., & Willett, J. B. (1985). Understanding correlates of change by modeling individual differences in growth.Psychometrika, 50, 203–228.CrossRefGoogle Scholar
  73. Rubin, D. B. (1980). Using empirical Bayes techniques in the Law School Validity Studies.Journal of the American Statistical Association, 75, 801–827.Google Scholar
  74. Rubin, D. B. (1983). Some applications of Bayesian statistics to educational data.The Statistician, 32, 55–68.Google Scholar
  75. Satorra, A., & Saris, W. E. (1985). Power of the likelihood ratio test in covariance structure analysis.Psychometrika, 50, 83–90.CrossRefGoogle Scholar
  76. Schmidt, W. H. (1969).Covariance structure analysis of the multivariate random effects model. Unpublished doctoral dissertation, University of Chicago.Google Scholar
  77. Schmidt, W., & Wisenbaker, J. (1986).Hierarchical data analysis: An approach based on structural equations (CEPSE, No. 4., Research Series). University of Michigan, Department of Counseling Educational Psychology and Special Education.Google Scholar
  78. Schuman, H., & Presser, S. (1981).Questions and answers in attitude surveys. New York: Academic Press.Google Scholar
  79. Sörbom, D. (1974). A general method for studying differences in factor means and factor structure between groups.British Journal of Mathematical and Statistical Psychology, 27, 229–239.Google Scholar
  80. Sörbom, D. (1982). Structural equation models with structured means. In K. G. Jöreskog & H. Wold (Eds.),Systems under indirect observation: Causality, structure, prediction (pp. 183–195). Amsterdam: North-Holland.Google Scholar
  81. Swamy, P. A. V. B. (1970). Efficient inference in a random coefficient regression model.Econometrica, 311–323.Google Scholar
  82. Tiao, G. C., & Tan, W. Y. (1965). Bayesian analysis of random-effect models in the analysis of variance. I. Posterior distribution of variance components.Biometrika, 52, 37–53.Google Scholar
  83. Wong, G. Y., & Mason, W. M. (1985). The hierarchical logistic regression model for multilevel analysis.Journal of the American Statistical Association, 80, 513–524.Google Scholar
  84. Yamamoto, K. (1988).A model that combines IRT and latent class models. Paper presented at the 1988 American Educational Research Association Annual Meeting.Google Scholar

Copyright information

© The Psychometric Society 1989

Authors and Affiliations

  • Bengt O. Muthén
    • 1
  1. 1.Graduate School of EducationUniversity of CaliforniaLos Angeles

Personalised recommendations