Hallmark: Definition of a group of subjects at the beginning of a period of follow-up.
e.g., study of cervical cancer ---- case of hysterectomy
select samples that have a relatively high incidence of the outcomes of interest.
e.g., study of predictors of hip fracture ---- elderly white women
Descriptive purpose: sample must closely resemble the target population to which the results will be applied.
Analytic: samples contain enough subjects with the major predictor characteristics and a sufficient number of outcomes during the study to allow meaningful analyses of the results.
Control (nested or double cohort design) should be probability samples.