ABSTRACT
The choice of a particular study design is a fundamental decision in designing a research study to answer a specific research question. In general, the study design that is chosen must be appropriate for answering the research question and must be appropriate for the setting in which it is used. In cross-over trials, subjects are randomly allocated to study groups in which they receive two or more treatments given consecutively. Selection bias is less likely to occur in trials with a preference group because patients are more likely to consent to enrol in the study. In the study of populations, cohort studies are the only type of study that can be used to accurately estimate incidence rates, to identify risk factors, or to collect information to describe the natural history or prognosis of disease. Nested case-control studies can be conducted within a cohort study.
