The timing of data collection is a critical element of the protocol. A cross-sectional survey is simpler, and far less costly to perform, than a longitudinal study, but the longitudinal approach is capable of providing important information about time course of fatigue, and insights about causality may be possible by examining the temporal relationships between variables. Changes over time in the incidence or prevalence of a phenomenon can be addressed through a methodology that longitudinally evaluates a specific patient sample over time, or assesses different samples over time from a defined population.

Other issues of timing also must be carefully considered. Incidence data (newly fatigued patients) must be framed by a time period of interest. The period must be long enough to yield a group of affected patients capable of being evaluated. Prevalence data (all fatigued patients) usually refers to a specific point in time (point prevalence), in which case the inquiry must focus on "fatigue right now" or fatigue during a relatively brief period (period prevalence), such as "during the past day (or week)." Some surveys that query period prevalence assess a longer interval (e.g. fatigue since a specific treatment, e.g. radiation or chemotherapy, was given).

Some information about demographics, the disease and its treatments, and drug therapies usually is recorded routinely in surveys. Additional information about potential etiologies or relevant comorbidities, which may be obtained from the medical record, elicited through questionnaire or interview, or collected through specific tests, should be pursued if it can clarify the nature of the clinical problem or the population studied, or will be used in statistical analyses. Respondent burden, reliability of data collection, and cost also influence the decision to include this other information.