ABSTRACT
Table 2.1 shows the data set called Rat Data from an animal experiment with a 1:3 repeated measures design on the primary response variable (some laboratory data) comparing two drugs (an experimental drug and a control drug). Twelve rats were randomly assigned to either the experimental drug (n1 = 6) or the control drug (n2 = 6). Measurements of this response variable were made on four occasions, once at baseline, and three time points (1, 2 and 3 weeks) during the 3-week treatment period. The experimental drug is expected to reduce the value of the response variable more than the control drug. Figure 2.1 shows the subject-specific response profiles over time. In this chapter, we introduce three examples of standard or unsophisticated statistical analysis for this type of repeated measurement data in an experiment. These are often observed in papers published in many medical journals, and we shall point out statistical problems encountered in each of analysis plans.
