This Article Full Text (PDF) References Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Alert me to new issues of the journal Add to Saved Citations Download to citation manager Request Permissions Request Reprints Add to My Marked Citations Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Citing Articles via Scopus Google Scholar Articles by LANGEHEINE, R. Articles by VAN DE POL, F. Search for Related Content Social Bookmarking                   What's this?

Bootstrapping Goodness-of-Fit Measures in Categorical Data Analysis

University of Kiel, Germany

JEROEN PANNEKOEK

Statistics Netherlands

FRANK VAN DE POL

Statistics Netherlands

When sparse data have to be fitted to a log-linear or latent class model, one cannot use the theoretical chi-square distribution to evaluate model fit, because with sparse data the observed cross-table has too many cells in relation to the number of observations to use a distribution that only holds asymptotically. The choice of a theoretical distribution is also difficult when model-expected frequencies are 0 or when model probabilities are estimated 0 or 1. The authors propose to solve these problems by estimating the distribution of a fit measure, using bootstrap methods. An algorithm is presented for estimating this distribution by drawing bootstrap samples from the model-expected proportions, the so-called nonnaive bootstrap method. For the first time the method is applied to empirical data of varying sparseness, from five different data sets. Results show that the asymptotic chi-square distribution is not at all valid for sparse data.

Sociological Methods & Research, Vol. 24, No. 4, 492-516 (1996)
DOI: 10.1177/0049124196024004004


CiteULike    Connotea    Del.icio.us    Digg    Reddit    Technorati    What's this?

This article has been cited by other articles:

				C. H. Elzinga, A. W. Hoogendoorn, and W. Dijkstra Linked Markov Sources: Modeling Outcome-Dependent Social Processes Sociological Methods Research, August 1, 2007; 36(1): 26 - 47. [Abstract] [PDF]

				K. M. Dunn, K. Jordan, and P. R. Croft Characterizing the Course of Low Back Pain: A Latent Class Analysis Am. J. Epidemiol., April 15, 2006; 163(8): 754 - 761. [Abstract] [Full Text] [PDF]

				N. G. Waller LCA 1.1: An R Package for Exploratory Latent Class Analysis Applied Psychological Measurement, March 1, 2004; 28(2): 141 - 142. [PDF]

				T. J. Croudace, M.-R. Jarvelin, M. E. J. Wadsworth, and P. B. Jones Developmental Typology of Trajectories to Nighttime Bladder Control: Epidemiologic Application of Longitudinal Latent Class Analysis Am. J. Epidemiol., May 1, 2003; 157(9): 834 - 842. [Abstract] [Full Text] [PDF]

				F. BASSI, J. A. HAGENAARS, M. A. CROON, and J. K. VERMUNT Estimating True Changes when Categorical Panel Data are Affected by Uncorrelated and Correlated Classification Errors: An Application to Unemployment Data Sociological Methods Research, November 1, 2000; 29(2): 230 - 268. [Abstract] [PDF]

				W. Chan, Y.-F. Yung, P. M. Bentler, and M.-L. Tang Tests of Independence for Ordinal Data Using Bootstrap Educational and Psychological Measurement, April 1, 1998; 58(2): 221 - 240. [Abstract]