Inferences on Correlation Coefficients in Some Classes of Nonnormal Distributions

Correlation coefficients have many applications for studying the relationship among multivariate observations. Classical inferences on correlation coefficients are mainly based on the normality assumption. This assumption is hardly realistic in the real world, which implies that the procedures on correlation coefficients used in many statistical software packages may not be relevant to most data sets in practice. However, we show that the classical procedures, possibly after simple corrections, are also valid in classes of distributions with large skewnesses and heterogeneous marginal kurtoses. A useful class of nonnormal distributions is identified for each of several types of correlation coefficients. The marginals of these distributions may include a variety of univariate distributions with different shapes. The results generalize the classical procedures to much larger classes of distributions than previously known and give a better understanding of the historical controversy regarding the behavior of the sample correlation coefficient. An implication is that one need not be worried so much by the nonnormality of data sets when using these classical procedures, providing simple corrections are evaluated and possibly undertaken.

[1]  T. Hayakawa.  Normalizing and variance stabilizing transformations of multivariate statistics under an elliptical population , 1987 .

[2]  A. Azzalini,et al.  The multivariate skew-normal distribution , 1996 .

[3]  R. Muirhead,et al.  Asymptotic distributions in canonical correlation analysis and other multivariate procedures for nonnormal populations , 1980 .

[4]  Ingram Olkin,et al.  MULTIVARIATE NON-NORMAL DISTRIBUTIONS AND MODELS OF DEPENDENCY , 1994 .

[5]  W. Dixon,et al.  Robustness in real life: a study of clinical laboratory data. , 1982, Biometrics.

[6]  P. Bentler,et al.  A necessary test of goodness of fit for sphericity , 1993 .

[7]  Sadanori Konishi,et al.  Normalizing transformations of some statistics in multivariate analysis , 1981 .

[8]  T. W. Anderson,et al.  Asymptotic Chi-Square Tests for a Large Class of Factor Analysis Models , 1990 .

[9]  James H. Steiger,et al.  The asymptotic distribution of elements of a correlation matrix: Theory and application , 1982 .

[10]  Sadanori Konishi,et al.  Normalizing and variance stabilizing transformations for intraclass correlations , 1985 .

[11]  M. L. Eaton,et al.  The asymptotic distribution of singular values with applications to canonical correlations and correspondence analysis , 1994 .

[12]  A. Gayen,et al.  The frequency distribution of the product-moment correlation coefficient in random samples of any size drawn from non-normal universes. , 1951, Biometrika.

[13]  K. Mardia.  Measures of multivariate skewness and kurtosis with applications , 1970 .

[14]  T. W. Anderson.  An Introduction to Multivariate Statistical Analysis , 1959 .

[15]  S. J. Devlin,et al.  Robust estimation and outlier detection with correlation coefficients , 1975 .

[16]  S. Konishi.  An approximation to the distribution of the sample correlation coefficient , 1978 .

[17]  S. Kotz,et al.  Symmetric Multivariate and Related Distributions , 1989 .

[18]  G. A. Baker.  The Significance of the Product-Moment Coefficient of Correlation with Special Reference to the Character of the Marginal Distributions , 1930 .

[19]  A. Joarder,et al.  Distribution of the correlation coefficient for the class of bivariate elliptical models , 1991 .

[20]  James R. Schott.  Eigenprojections and the equality of latent roots of a correlation matrix , 1996 .

[21]  G. T. Duncan,et al.  A Monte-Carlo study of asymptotically robust tests for correlation coefficients , 1973 .

[22]  P. R. Rider,et al.  ON THE DISTRIBUTION OF THE CORRELATION COEFFICIENT IN SMALL SAMPLES , 1932 .

[23]  T. Micceri.  The unicorn, the normal curve, and other improbable creatures. , 1989 .

[24]  J. R. Schott.  A Test for a Specific Principal Component of a Correlation Matrix , 1991 .

[25]  Robert I. Jennrich,et al.  An Asymptotic χ2 Test for the Equality of Two Correlation Matrices , 1970 .

[26]  E. S. Pearson,et al.  Further Experiments on the Sampling Distribution of the Correlation Coefficient , 1932 .

[27]  A. Amey.  Robustness of the Multiple Correlation Coefficient When Sampling From a Mixture of Two Multivariate Normal Populations , 1990 .

[28]  R. C. Bose,et al.  Essays in probability and statistics , 1971 .

[29]  T. W. Anderson,et al.  The asymptotic normal distribution of estimators in factor analysis under general conditions , 1988 .

[30]  Ke-Hai Yuan,et al.  ON NORMAL THEORY AND ASSOCIATED TEST STATISTICS IN COVARIANCE STRUCTURE ANALYSIS UNDER TWO CLASSES OF NONNORMAL DISTRIBUTIONS , 1999 .

[31]  R. Muirhead.  Aspects of Multivariate Statistical Theory , 1982, Wiley Series in Probability and Statistics.

[32]  C. Raghavendra Rao,et al.  Multivariate Analysis and Its Applications , 1984 .

[33]  E. S. Pearson.  SOME NOTES ON SAMPLING TESTS WITH TWO VARIABLES , 1929 .

[34]  J. Magnus,et al.  Matrix Differential Calculus with Applications in Statistics and Econometrics (Revised Edition) , 1999 .

[35]  C. Kowalski.  On the Effects of Non‐Normality on the Distribution of the Sample Product‐Moment Correlation Coefficient , 1972 .

[36]  Sik-Yum Lee.  Analysis of covariance and correlation structures , 1985 .

[37]  Ke-Hai Yuan,et al.  On asymptotic distributions of normal theory MLE in covariance structure analysis under some nonnormal distributions , 1999 .

[38]  M. Browne.  The analysis of patterned correlation matrices by generalized least squares , 1977 .

[39]  E. S. Pearson.  The Test of Significance for the Correlation Coefficient , 1931 .

[40]  J B S HALDANE.  A note on non-normal correlation. , 1949, Biometrika.

[41]  Robert Cudeck,et al.  Analysis of correlation matrices using covariance structure models. , 1989 .