Download PDF
Methods Inf Med 1987; 26(03): 104-108
DOI: 10.1055/s-0038-1635493
Original Article
Schattauer GmbH

Analysis of Two-Way Contingency Tables : Log-Linear Models and Dual Scaling

Analyse von Zwei-Weg-Kontingenztafeln: Loglineare Modelle und Zweifachskalierung
M. A. A. Moussa
1   (From the Faculty of Medicine, Kuwait University)
› Author Affiliations
Further Information

Publication History

Publication Date:
16 February 2018 (online)

    Permissions and Reprints

Summary

The paper focuses upon the measurement of association in two-way contingency tables, using the log-linear models and dual scaling approaches. The former comprises [1] the use of pseudo-Bayes estimators to remove zeros, [2] fitting the resulting smoothed array to all possible configurations of log-linear models, [3] fitting the quasi-independence model to detect anomalous cells that caused deviation from the null-independence model. The latter includes [1] estimation of the optimal weights that maximize the canonical correlation between the two categorical variables by an optimization iterative method, [2] testing the discriminability of the estimated scoring scheme. The two approaches were applied to a set of real data for the study of the association between maternal age at marriage and types of reproductive wastage in a sampling survey conducted in the population of female nurses in Kuwait.

Schwerpunkt der Arbeit ist die Messung der Assoziation in Zwei-Weg-Kontingenztafeln unter Benutzung loglinearer Modelle und von Zweifachskalierungsansätzen. Erstere umfassen [1] die Anwendung von Pseudo-Bayes-Schätzern zur Eliminierung von Nullen, [2] die Anpassung der resultierenden geglätteten Tafel an alle möglichen Konfigurationen loglinearer Modelle, [3] die Anpassung des quasi-unabhängigen Modells, um anomale Zellen zu entdecken, die eine Abweichung von dem Nichtabhängigkeitsmodell verursachten. Letztere umfassen [1] die Schätzung optimaler Gewichte, welche die kanonische Korrelation zwischen den beiden kategorischen Variablen durch einen iterativen Optimierungsprozess maximieren, [2] die Prüfung der Diskriminanzfähigkeit des geschätzten Gewichtungsschemas. Die beiden Ansätze wurden auf einen Satz realer Daten zur Untersuchung der Assoziation zwischen dem Alter der Mutter bei der Verehelichung und Formen von Fehlgeburten in einer Beobachtungsreihe, durchgeführt an den Krankenschwestern in Kuwait, angewandt. .

Key-Words

Log-Linear Models - Correspondence Analysis - Eigenvalues - Analysis of Variance

Schlüssel-Wörter

Loglineare Modelle - Korrespondenzanalyse - Eigenwerte - Varianzanalyse
 

  • REFERENCES

  • 1 Bartlett M. S. Multivariate analysis. J. Roy. Statist. Soc, Suppl 1974; 9: 176-190.
    PubMedGoogle Scholar
  • 2 Bishop Y. M. M, Fienberg S. E, Holland R. W. Discrete Multivariate Analysis: Theory and Practice. Cambridge/ Mass: MIT Press; 1975
    Google Scholar
  • 3 Bock R. D. Methods and Applications of Optimal Scaling. The University of North Carolina Psychometric Laboratory Research Memorandum. 1960: 25
    Google Scholar
  • 4 Deming W. E, Stephen F. F. On a least squares adjustment of a sampled frequency table when the expected marginal totals are known. Ann. Math. Statist 1940; 11: 427-444.
    CrossrefPubMedGoogle Scholar
  • 5 Fienberg S. E. The Analysis of Cross-Classified Categorical Data. Cambridge/ Mass: MIT Press; 1981
    Google Scholar
  • 6 Fox L. An Introduction to Numerical Linear Algebra. Oxford: Oxford University Press; 1964
    Google Scholar
  • 7 Goodman L. A. The analysis of cross-classified data: independence, quasi-inde-pendence, and interactions in contingency tables with or without missing values. J. Amer. Statist. Ass 1968; 63: 1091-1131.
    PubMedGoogle Scholar
  • 8 Goodman L. A. The multivariate analysis of qualitative data: interactions among multiple classifications. J. Amer. Statist. Ass 1970; 65: 226-256.
    CrossrefPubMedGoogle Scholar
  • 9 Greenacre M. J. Theory and Applications of Correspondence Analysis. London: Academic Press; 1984
    Google Scholar
  • 10 Guttman L. The Quantification of a Class of Attributes: A,Theory and Method of Scale Construction. In: The Committee of Social Adjustment (Eds): The Prediction of Personal Adjustment. New York: Social Science Research Council; 1941: 319-348.
    Google Scholar
  • 11 Haberman S. J. The analysis of residuals in cross-classified tables. Biometrics 1973; 29: 205-220.
    CrossrefPubMedGoogle Scholar
  • 12 Hotelling H. Relations between two sets of variates. Biometrika 1936; 28: 321-377.
    CrossrefPubMedGoogle Scholar
  • 13 Ibrahim M. E, Hathout H. M, Moussa M. A. A, A/Razik M. On factors contributing to reproductive wastage – An analytical study on the nursing staff in Kuwait. J. Kuwait Med. Ass 1981; 15: 139-148.
    PubMedGoogle Scholar
  • 14 Maung K. Measurement of association in contigency table with special reference to the pigmentation of hair and eye colours of Scottish school children. Ann. Eugen 1941; 11: 189-223.
    CrossrefPubMedGoogle Scholar
  • 15 Moussa M. A. A. On analysis of two-way contingency tables with some sampling zero cell counts – Application to human reproductive wastage according to age. Meth. Inform. Med 1980; 19: 164-169.
    PubMedGoogle Scholar
  • 16 Moussa M. A. A. A program for the analysis of incomplete two-way cross classifications. Comp. Progr. Biomed 1980; 11: 185-189.
    PubMedGoogle Scholar
  • 17 Nishisato S. Analysis of Categorical Data: Dual Scaling and its Application. Toronto: University of Toronto Press; 1980
    Google Scholar
  • 18 Upton G. J. G. The Analysis of Cross-Tabulated Data. New York: John Wiley; 1978
    Google Scholar
  • 19 Van der Heijden P. G. M, de Leeuw J. Correspondence analysis used complementary to loglinear analysis. Psychome-trika 1985; 50: 429-447.
    CrossrefPubMedGoogle Scholar
  • 20 Van der Heijden P. G. M. Transition Matrices, Model Fitting and Correspondence Analysis. In Diday E. Edit. Data Analysis and Informatics IV. Amsterdam: North Holland; 1986
    Google Scholar