Citation

Burke CJ. Psychol. Bull. 1951; 48(6): 496-504.

Copyright

(Copyright © 1951, American Psychological Association)

DOI

unavailable

PMID

14875580

Abstract

Two chi-square tests are derived. They may be used to establish the significance of the correlation of accident occurrences in separate time samples involving the same individuals. In one test the theoretical marginal Poisson distributions are assumed to have identical parameters; in the second, the parameters are different. The tests may be used in conjunction with the correlation coefficient advocated by Maritz (25: 2696) in the analysis of "accident proneness." Significant correlation coefficients are interpretable only if the experimental design includes random assignment and reassignment of operators to machines. (PsycINFO Database Record (c) 2012 APA, all rights reserved)

Language: en