ABSTRACT
In Chapter 6, some of the key ideas behind the analysis of data from randomized trials were introduced. The development was entirely in terms of a normally distributed outcome. This was because: (1) the methods used, such as
t
-tests, were likely to be widely familiar, (2) the effect to be estimated is unambiguous, namely, the difference in means, and (3) the theory behind certain aspects, such as the way baseline information is taken into account, is relatively accessible. In this chapter, some further aspects of the analysis of data from RCTs will be considered.
