International Conference on Applied Statistical in Medical Sciences, Ankara, Turkey, 01 August 1997, vol.4, pp.133-138
Medical and health policy research data me often most informative when presented as a multidimensionally cross-classified table of counts. Most researchers analyze multidimensional contingency tables by studying two or more variables at a time using chi(2) statistics. Although such an approach often gives great insight about the relationship among variables, it has its limitations. Hence, the log-linear models are suggested.