Cancer Terms: Research Statistics

Approved by the Cancer.Net Editorial Board, 08/2015

Absolute risk: The difference between two risks, usually smaller than a relative risk.

Average/mean: The middle value of a set of numbers, calculated by adding all of the values and dividing by the number of values in the set.

Clinical significance: An assessment that a research finding will have practical effects on patient care.

Cohort: A group of individuals who share a common experience, exposure, or trait and who are under observation in a research study.

Confidence interval: A measure of the number of times out of 100 (similar to a percentage) that test results will be within a specified range. It is a measurement used to indicate the reliability of an estimate.

Confounding variable: A factor in a scientific study that wasn’t addressed that could affect the outcome of the study, such as smoking history in a study of people with cancer.

Control group: A group of individuals who do not receive the treatment being studied. Researchers compare this group to the group of individuals who do receive the treatment, which helps them evaluate the safety and effectiveness of the treatment.

Endpoint: The results measured at the end of a study to see whether the research question was answered.

Incidence: The number of new instances of a disease or condition in a particular population during a specific period. Learn more about statistics used to estimate risk and recommend screening.

Lifetime risk: The probability of developing a disease or dying from that disease across a person’s lifetime.

Median: The middle value in a range of measurements ordered by value.

Mortality rate: The number of deaths in a particular population during a specific time.

Odds ratio: A comparison of whether the likelihood of an event is similar between two groups; a ratio of 1 means it is equally likely between both groups.

Outcome: A measurable result or effect.