Sequential parameter structure, conditional inference, and likelihood drop
In the broad framework of sequential testing of statistical hypotheses as exemplified by the analysis of variance procedure, the author examines the parameter structure implicit in such testing. Special attention is given to cases in which the parameters are linear in the sense of a transformation model or of an exponential model. Well-defined testing procedures for these two classes of models are determined by basic theoretical criteria, and the relationship between the two sample space structures for the procedures is explored. Some aspects of the serious computational issues that may arise are discussed.
This page was built for publication: Sequential parameter structure, conditional inference, and likelihood drop
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4721367)