ABSTRACT
This chapter addresses the analysis of longitudinal studies in the context of HRQoL investigations. In the previous chapter, event-or condition-driven and time-driven designs were described. These designs have corresponding analytic models: a repeated measures model and a growth curve model. In a repeated measures model, time is conceptualized as a categorical variable. Each assessment must be assigned to one category. In the NSCLC study (Example 2), these categories are “Pre-therapy,” “6 weeks,” “12 weeks,” and “6 months.” In a growth curve model, time is conceptualized as a continuous variable. In the NSCLC study, the variable is days since randomization.
