ABSTRACT
In this first chapter we introduce the terminology, the main assumptions, and the sta-
tistical models most widely used in adaptive comparative experiments with v treatments. In particular the observed responses are assumed to behave according to one
of the models belonging to the exponential family− see Section 1.3 − and we make the essential assumption that at each step, given the treatment assigned to the exper-
imental unit, the response is independent of the other observations. The role of the
ensuing Fisher information in an adaptive design setting is highlighted in Section 1.4.