ABSTRACT
A logical place to begin in developing analyses for longitudinal clinical trial data is to determine the dependent (i.e., response or outcome) variable and upon what basis it will be evaluated. These may seem like obvious decisions dictated by the design of the trial and historical precedent. However, a number of subtleties apply and these factors reinforce the need for the iterative study development process detailed in Chapter 2. Specifically, the trial design informs the analysis, but analytic consequences also need to inform the design.
