The product moment correlation coefficient is often used even for ordinal data with only a few scale steps. This procedure may lead to biased results, where the bias depends on the number of scale steps and on the skewnesses of the observed variables. The polychoric correlation coefficient, which is a generalization of the tetrachoric correlation to the general case, is discussed as a possible measure of correlation for this kind of data.
Get full access to this article
View all access options for this article.
References
1.
LancasterH. O. and HamdanM. A. (1964), “Estimation of the Correlation Coefficient in Contingency Tables with Possibly Nonmetrical Characters,” Psychometrika, 29, 383–91.
2.
MartinW. S. (1973), “The Effects of Scaling on the Correlation Coefficient: A Test of Validity,” Journal of Marketing Research, 10 (August), 316–8.
3.
MartinW. S. (1978), “Effects of Scaling on the Correlation Coefficient: Additional Considerations,” Journal of Marketing Research, 15 (May), 304–8.
4.
MartinsonE. O. and HamdanM. A. (1972), “Maximum Likelihood and Some Other Asymptotically Efficient Estimators of Correlation in Two Way Contingency Tables,” Journal of Statistical Computation and Simulation, 1 (January), 45–54.
5.
MartinsonE. O. and HamdanM. A. (1975), “Algorithm AS 87: Calculation of the Polychoric Estimate of Correlation in Contingency Tables,” Applied Statistics, 24, 272–8.
6.
OlssonU. (1979a), “On the Robustness of Factor Analysis Against Crude Classification of the Observations,” Multivariate Behavioral Research, 14, 485–500.
7.
OlssonU. (1979b), “Maximum Likelihood Estimation of the Polychoric Correlation Coefficient,” Psychometrika, 44, 443–60.
8.
PetersC. C. and Van VoorhisW. R. (1940), Statistical Procedures and Their Mathematical Bases. New York: McGraw-Hill Book Company.
9.
TallisG. M. (1962), “The Maximum Likelihood Estimation of Correlation from Contingency Tables,” Biometrics, 18, 342–53.
