Some of the measures commonly used to describe association between two ordinal categorical variables, controlling for the effects of other ordinal categorical variables, are compared in their closeness to a corresponding conditional measure for an underlying jointly continuous distribution. Partial gamma and Kendall's partial tau are observed to be poorer than a weighted average of the Kendall's τ b 's from separate control levels, according to this criterion. The asymptotic sampling distribution of a particular weighted average measure is derived, and the behavior of the mean and asymptotic variance of such a measure is analyzed as the categorical variables are more finely measured.
No takes yet. Share an insight, caveat, or question.
Alan Agresti (1977) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: