Identification of latent class Markov models with multiple indicators and correlated measurement errors

  • Francesca Bassi


A necessary condition for identification of latent class models is that the number of unknown independent parameters must not be greater than the number of observed cells in the contingency table. Such condition is not sufficient at all. Verifying Goodman’s sufficient condition for local identifiability may be, for complex models, a cumbersome procedure. In any case, local identifiability does not guarantee global identifability. The paper provides rules to ascertain global identifiability of some specifications of latent class Markov models, expressing the unknown parameters as a function of the observed frequencies. In the case that not all parameters of a model are identified, the outlined rules provide hints about the restrictions to impose in order to obtain fully identified models.


Latent class models Global identifiability Sufficient condition 


  1. Bassi, F., M. Croon, J.A. Hagenaars, J.K. Vermunt (1995),Estimating latent turnover tables when data are affected by correlated and uncorrelated classification errors.WORC PAPER 95.12.25n, Tilburg University.Google Scholar
  2. Bassi F., N., Torelli U. Trivellato (1998), Data and modelling strategies in estimating labour force gross flows affected by classification errors.Survey Methodology, 24, 109–122.Google Scholar
  3. Goodman, L. A. (1974), Exploratory latent structure analysis using both identifiable and unidentifiable models.Biometrika, 79, 1179–1259.Google Scholar
  4. Henry, N. W. (1973), Measurement models for continuous and discrete variables. InStructural Equation Models in the Social Sciences (eds. A. S. Goldberger and O. D. Duncan), 441–453. New York: Seminar.Google Scholar
  5. Hagenaars, J. A. (1988), Latent structure models with direct effects between the indicators, local dependence models.Sociological Methods and Research, 16,379–405.CrossRefGoogle Scholar
  6. Hagenaars, J.A. (1990),Categorical Longitudinal Data: Log-Linear, Panel, Trend and Cohort Analysis. Newbury Park: Sage.Google Scholar
  7. Hughes J. P., P. Guttorp, S. P. Charles (1999), A Non-homogeneous Hidden Markov Model for Precipitation Occurrence.Applied Statistics, 48, 15–30.zbMATHGoogle Scholar
  8. Lazarsfeld, P., Henry N. W. (1968)Latent Structure Analysis. New York: Houghton Mifflin.zbMATHGoogle Scholar
  9. Macdonald I.L., W. Zucchini (1997)Hidden Markov and Other Models for Discretevalued Time Series. London: Chapman & Hall.Google Scholar
  10. Torelli, N., F. Bassi (1998) La stima dei flussi rispetto al lavoro in presenza di indicatori affetti da errore. InProblemi di misurazione e riflessi sulla modellistica econometrica, Annali di Statistica, Anno 127, Serie X, vol. 15, 175–192, Roma: ISTAT.Google Scholar
  11. Van de Pol, F., R. Langeheine (1990) Mixed Markov latent class models. InSociological Methodology (ed. C. Clogg), 213–247. Oxford: Blackwell.Google Scholar
  12. Van de Pol, F., R. Langeheine, W. De Jong (1991)PANMARK USER MANUAL: PANel Analysis Using MARkov Chains, Version 2.2. Voorburg: Netherlands Central Bureau of Statistics.Google Scholar
  13. Vermunt, J.K. (1997)Log-linear and Event History Analysis: A General Approach with Missing Data, Latent Variables, and Unobservable Heterogeneity. Newbury Park: Sage.Google Scholar
  14. Wiggins, L. M. (1955)Panel Analysis: Latent Probability Models for Attitude and Behavior Process. Amsterdam: Elsevier.Google Scholar

Copyright information

© Società Italiana di Statistica 1997

Authors and Affiliations

  • Francesca Bassi
    • 1
  1. 1.Statistics DepartmentUniversity of PadovaPadovaItaly

Personalised recommendations