A parametric statistical test used to determine whether the averages of two (and only two) data sets are significantly different. Importantly, certain criteria must be met before this test can be used.

First, the individual data points must be independent (individual values do not depend on each other).

Second, the variables measured must be continuous (having an infinite number of possible values).

Third, the data must follow a normal Gaussian distribution. Fourth, the variance (spread of the data) of both data sets should be equivalent.

There are different types of t tests depending on what analysis is required: One-tailed test when looking for a difference in one direction (one average exclusively larger than the other); Two-tailed test when looking for a difference in either direction (one average either larger or smaller than the other); Unpaired test when comparing two independent data sets; Paired test when comparing two related data sets (ex. pre- and post- measurements of the same subjects).