© 1976 by Biometrika Trust
CORRECTIONS AND AMENDMENTS |
A generalization of a fundamental theorem of ranking and selection
I am grateful to Raymond J. Carroll and Thomas J. Santner for pointing out a difficulty in the above paper. The paper invokes a theorem stated by Mahamunulu (1967). However, that theorem requires that the statistics Tij be independent, a condition that is not always satisfied for the general class of procedures described and thus, for the general class, the veracity of the theorem remains an open question.