Folks,
An interesting one.
We are all familiar with identifying a suitable sample size for hypothesis testing. The methods of identifying this number are clear.
But
What happens when you are about to do regression to establish the level of association that exists between an output variable (the Y) and the input variables (the Xs).
Any thoughts?
Im plotting the data, getting process experts comfortable with the patterns visible (seasonal or caused by known things happening) and getting them to waggle their heads in the right direct regarding the datas average but that doesnt seem too exhaustive.
What would you do?
Lorax
.
