Assuming you're relying on Analysis Workspace to run statistical calculations, then one challenge that you're going to face is that the data reported in Analysis Workspace is always aggregated, whereas statistical calculations like p-value use individual data.
For example: in Analysis Workspace, you can build a histogram visualisation: https://experienceleague.adobe.com/en/docs/analytics/analyze/analysis-workspace/visualizations/histo.... However, when you look at the underlying data that is used to build a histogram, you'll discover that it's actually comprised of multiple segments. So in a way, this is like working in reverse of what statisticians want. Statisticians would wonder: "Why don't you just report on the individual (hit) data that you already have?" But Analysis Workspace can't do that. It has to use data that is already aggregated across many visits/visitors, and then use segments to extract out the visits/visitors per dimension to build that histogram.
So, I think your first task is to get the raw, individual data that would be used for calculating your p-values, and this would have to be outside of Analysis Workspace.
I'm not sure if the above helps. I haven't delved much into statistics like this in Analysis Workspace, so perhaps my thinking is entirely wrong.
