Regression Checklist

Hey fellow stat heads,

I am curious to hear if you have any checklist that you use before actually doing a regression. (i.e what do you ask yourself before performing the analysis, what kind of tests do you perform).

My goal is to assemble some basic checklist that everyone must go through before performing a regression analysis.



Less is more. Stay pure. Stay poor.
Any particular type of regression?

I first start by framing the question and defining the purpose: exploratory, inferential, or prediction. Then make sure I have the relationships between variables mapped out and I can ascertain those variables. Then I draw out the model and proceed to the statistical assumptions and data steps.