Journal of Statistics: Advances in Theory and ApplicationsAuthor's: Kouji Yamamoto, Hiroki Nakane and Sadao Tomizawa
Pages: [71] - [84]
Received Date: April 29, 2016
Submitted by:
DOI: http://dx.doi.org/10.18642/jsata_7100121663
For square contingency tables with ordered categories, this paper proposes a model that the cell probabilities have a similar structure of bivariate normal density function with equal marginal means and marginal variances. The proposed model has the structure of symmetry, quasi-uniform association, and normal distribution type symmetry models. The simulation studies based on the bivariate normal distribution are given. An example is also given.
bivariate normal distribution, linear diagonals-parameter symmetry model, normal distribution type symmetry model, symmetry model, symmetry plus quasi-uniform association model.
