new reo prices and basket price, page-27

schmunzel75,

You wont need to graph to see a perceived direct positive correlation. So this would be both misleading and unfair.

What you need to do is a Regression Analysis using multiple variables. This will tell a different story. You can assign say "the TOL" or the court cases a discrete value (ie make them variables), put these into your Regression Analysis and model multiple scenarios where you artificially hold constant any or all variables and then see what has had the most effect on price.

Appreciate if you or anyone else is unfamiliar with Statistics that the above may sound like total BS but I am being quite serious. Correlation is not enough here. Multilinear Regression will tell a better story.