Repeated Measures Data

Repeated measures (or repeated measurements) data are usually obtained from multiple measurements of a response variable. Such multiple measurements are carried out for each experimental unit over time (as in a longitudinal study) or under multiple conditions.

An essential statistical peculiarity of such data is dependence of the response on the experimental unit itself. This often makes variability of response between units significantly higher than variability between different conditions or time points for the same unit.

Consider, for example, a study of 3 drugs – A, B, C – to determine whether they reduce arterial blood pressure. There are 10 patients, and each patient takes all three drugs, but at widely separated times. The outcome of such a study for 10 patients (experimental units) can be represented in the following table as the magnitude of blood pressure decline 3 hours after the drug intake:

Repeated measures are not necessarily continuous variables. Suppose the above data are represented as a discrete random variable taking on three values “Smallest,” “Middle,” and “Largest” and reflecting relative values of the 3 responses for each patient.