Real Data is complex

That’s all great, but most datasets contain many different types of predictors. How do we know which ones are most important? The scales of the output statistics are often different. We can often convert to a p-value and go from there (and we do).

A p-value is the chances of there being no association (p ≤ 0.05 or ≤ 5%).