If you are uncertain about which test to use, typically a two-tailed, unpaired t-test is what you want. An unpaired t-test tests whether two unrelated datasets, such as if cells treated with Chemical X versus cells treated with Chemical Y have results that are statistically significantly different.Īdditionally, a method using scipy with python has been added. A paired t-test is typically used for datasets that are linked to each other in some way, such as treating cells at timepoint 1, then treating the same cells at timepoint 2, and testing whether the results are statistically significant. Nearly all biological experiments should use a two-tailed t-test. Briefly, the tails for the t-test refer to how the data should be measured, meaning does the data go in one direction, or two.

There are a few different ways to run the t-test, but the methods to perform the test on each variant is similar in GraphPad Prism and R. Gosset used the pseudonym 'Student' on the paper that he published with the results, and the test that he developed became known as the 'Student's T-test', or simply a 't-test'.

The student's t-test was first created by William Gosset, while doing analyses on beer in the Guiness Brewery in Dublin. Common Statistical Tests in R and and Graphpad® Prism: Student's t-test Synopsis

Graphpad prism student price