t test
Meaning of t test: A t test is a statistical method that determines whether an observed difference between the average values of two groups likely reflects a real effect.
Definition of t test
A t test is a statistical method used to determine whether there is a real difference between the means, or averages, of two groups, rather than a difference likely explained by chance variation alone. It is one of several statistical tests used to compare groups and is commonly applied in clinical research.
In a clinical trial, a t test might be used to compare an outcome, such as a change in blood pressure or a laboratory value, between a group receiving a treatment and a group receiving a placebo or a different treatment. The test produces a result that reflects how likely the observed difference would be if no real difference actually existed between the groups.
A t test is generally suited to comparing exactly two groups and assumes the data being compared follow a roughly normal distribution, and different statistical tests, such as analysis of variance, are used instead when more than two groups are being compared.
Also known as
Used in a sentence
A t test was used to compare outcomes between the two treatment arms of the clinical trial.
Good to know
A t test is designed to compare exactly two groups, so studies comparing three or more groups generally require a different statistical method, such as analysis of variance, rather than repeated t tests.