ABSTRACT
This chapter considers the relative merits of many consistency tests and discusses conditions under which each may provide relevant, nonredundant, and discriminating information that can enhance a taxometric investigation. It demonstrates and encourage the generation of empirical sampling distributions through analyses of bootstrap samples of taxonic and dimensional comparison data to help researchers determine which consistency checks are most appropriate for the study at hand. MAXSLOPE or MAMBAC can be conducted multiple times by using different pairs of variables as input and output indicators and by swapping these variables across input—output roles. Consistency checking is only valuable to the extent that the chosen tests increase the likelihood of reaching an accurate structural inference, over and the tests already performed. At present, close to nothing can be said about this subject on the basis of rigorous empirical investigation.
