作者: Stephen J. Haslett
DOI: 10.1007/978-1-4615-7070-7_7
关键词:
摘要: The calculation of degrees freedom for hierarchical loglinear models applied to complete contingency tables is usually straightforward. such sparse multiway however, requires checking the marginal totals sampling zeros in all configurations included model, together with an appropriate adjustment. GENSTAT, GLIM, SAS and SPSSX do not make this In paper algorithm outlined which calculates tables. It has been shown that under suitable conditions, distribution difference log likelihood ratio statistics two nested converges usual χ2 distribution, then provides a routine method calculating freedom.