chapter  12
18 Pages

Longitudinal Data I: Treatment of Post-Natal Depression

In the next two chapters, we shall discuss more formal modelling techniques that can be used to analyse longitudinal data.

12.3 Analysis Using SAS Data sets for longitudinal and repeated measures data may be structured in two ways. In the first form there is one observation per subject and the repeated measurements are held in separate variables. Alternatively, there may be separate observations for each measurement, with variables indicating to which subject and occasion it belongs. When analysing longitudinal data both formats may be

needed. This is typically achieved by reading the raw data into a data set in one format and then using a second data step to reformat it. In the example below, both types of data sets are created in the one data step.