Premium
On Inferences Concerning a Common Correlation Coefficient
Author(s) -
Donner Allan,
Rosner Bernard
Publication year - 1980
Publication title -
journal of the royal statistical society: series c (applied statistics)
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.205
H-Index - 72
eISSN - 1467-9876
pISSN - 0035-9254
DOI - 10.2307/2346412
Subject(s) - statistics , mathematics , correlation coefficient , correlation , econometrics , geometry
S ummary Four different methods are compared of estimating a common correlation p from k ≥ 2 independent random samples drawn from bivariate normal populations. The results of a Monte Carlo study show that Hotelling's method and the standard‐score method, equivalent to averaging the simple correlations, are generally superior to Fisher's method and the method of maximum likelihood in small to moderate‐sized samples. Hypothesis‐testing is also discussed, and a new test of significance for the standard score estimator proposed and evaluated.