ABSTRACT
The central theme of this book is the analysis of data in which there is a single response variable, binary in character. Yet it is clear that many of the topics and methods discussed in previous chapters can be generalized to other kinds of response variable. In the present chapter we outline some of the more immediate generalizations, namely to
paired preference data, in which each response is not ‘success’ or ‘failure’ but rather an indication of preference between two conditions or stimuli;
nominal data, where the response is one of a number of qualitatively different possibilities;
ordinal data, in which the qualitatively different possibilities are in some natural order;
multivariate binary data;
mixed binary and continuous responses. We outline below some of the methods for the analysis of such data, concentrating on possibilities fairly closely linked with those for binary data.
