Effects of collapsing multidimensional contingency tables.

摘要: The conditions are defined under which collapsing multidimensional contingency tables, by adding over variables, will affect the apparent interaction between remaining variables. This leads to a simple method of distinguishing those log-linear models for cell estimates may be obtained direct multiplication, from requiring iterative fitting. implications fitting over-parametrized discussed with particular reference 'partial association' model used implicitly (a) when information separate two-dimensional tables is combined test association two and (b) rates adjusted indirect standardization.