# Using ANOVA with a "Repeated measures" / "Within-subjects" design

Repeated measures designs or "Within-Subjects" designs are very important in life sciences and social sciences. Unfortunately, Mathematica's ANOVA package is not able to analyze those designs. It only handles "Between-Subjects" type of analyses.

During my whole research career I haven't been able to use Mathematica to analyze most of the studies I performed, which really is a pity.

In an e-mail exchange I had with WRI's Andy Ross in the beginning of this year he confirmed that I wasn't overlooking built-in functionality.

The question is: is any one aware of third party packages that handle these types of ANOVAs?

• I don't mean to sound like a jerk, but if you had your "whole research career" to build some tools for this in Mathematica why didn't you? Is the algorithm behind it that prohibitive? Jul 1, 2012 at 6:32
• @Mr.Wizard have you read the linked article? It's sufficiently complex and error-prone that I preferred to use the other tools that I had readily available (Statistica in my case) instead of trying to come up with something myself. Jul 1, 2012 at 6:58
• Okay, just asking. It went over my head but that didn't surprise me. I guess it really is complex. :^) Jul 1, 2012 at 7:00
• Wouldn't the GeneralizedLinearModel be able to handle what ANOVA does, which after all is (more or less) a special form of linear regression?
– gwr
Sep 13, 2016 at 17:30
• @gmr ANOVA's are not used for regression, but are a methodology to compare means of data grouped into categories. However, the cited article describes an alternative approach using GLM. Sep 13, 2016 at 17:38