KNOWLEDGEBASE - ARTICLE #1009

Can Prism test for a significant difference between two distributions using, for example, the Kolmogorov-Smirnov test?

Distinguish two uses of the Kolmogorov Smirnov test:

Prism can test whether a column is Gaussian using the Kolmogorov-Smirnov test (and two better normality tests, starting with version 4.01).

Prism (new in Prism 6) can also  compare two distributions with each other. This is a choice on the t test diaog, as an alternative nonparametric test to the Mann-Whitney test. You enter two stacks of data, not frequency distributions.  The Mann-Whitney test ranks all the values and then compares the average rank of the two groups. The KS test works differently. It finds the value where the cumulative distribution of the two groups is furthest apart. The Mann-Whitney test essentially asks about whether the entire distribution is shifted between the two groups. The KS test asks about where the distributions are furthest apart.

Explore the Knowledgebase

Analyze, graph and present your scientific work easily with GraphPad Prism. No coding required.