Table 1.


AdjustmentTo modify an estimate of association to account for systematic differences in other characteristics between groups, especially when these differences might have influenced that association. For example, because age is a powerful determinant of death, mortality estimates often are adjusted for differences in death because even small differences in age between groups can lead to appreciable differences in mortality rates.
BiasOccurs when results of a study systematically deviate from the truth because of nonrandom factors.
CensoringWhen the time to a particular outcome is of interest but has not occurred in a particular subject, the data for that subject are said to be censored at the date of last follow-up.
χ2 testA statistical significance test that most commonly is used for testing the null hypothesis that two proportions are equal.
CohortA group of individuals who have a common feature when they are assembled and who are followed in time.
ConfoundingOccurs when a factor is associated with both exposure and outcome.
EpidemiologyThe study of the distribution and the determinants of health-related states or events in certain populations and use of the findings to improve health.
ExposureA factor of interest in a research study, usually hypothesized to cause or influence the outcome of interest.
IncidenceNew cases of the outcome of interest that occur during a given period of time.
Information biasOccurs when data on exposure or outcome are systematically incorrect, either when the exposure is measured differently in people with the outcome or when the likelihood of detecting the outcome varies between exposed and unexposed participants.
InteractionOccurs when the association between an exposure and the outcome is influenced by a third characteristic by a greater extent than would be predicted by a simple (additive) model. For example, if a treatment prevents myocardial infarction to a significantly greater extent in women than in men, then one could say that there was an interaction between treatment and gender on the risk for myocardial infarction.
MatchingIdentifying subjects who are similar in terms of a particular characteristic (often a potential confounder) so that this characteristic does not influence the risk for the outcome of interest.
OutcomeThe disease, health condition, or health state of interest.
PrevalenceThe number of people in a population with the health condition or state of interest.
Selection biasOccurs when study participants are not representative of the broader population at risk for the outcome.
StratificationOccurs when a cohort or population is divided into groups on the basis of a characteristic of interest (often a potential confounder).
  • Adapted from reference (21).