Abstract
This paper compares two methods involving the uniform association
model and the quasi-independence model. These models can be de-
scribed in terms of the association parameters for the analysis of trian-
gular contingency tables having ordered categories. A simulation study
based on 30,000 random triangular tables was performed for this com-
parison. Proportions of the rejected and accepted hypothesis under
these models were obtained. From the results of the simulation study,
the behaviour of the association parameters was discussed with respect
to their coefficient of variations. The homogeneity of the coefficients of
variance was also tested.