|From||David Airey <email@example.com>|
|Subject||re: st: interactions|
|Date||Thu, 26 Jun 2003 10:58:45 -0500|
> >> What's the easiest way to generate all [n-way] interactions
> >> when you have many continuous variables to predict a continuous
> >> response?
On Thu, 26 Jun 2003, David Airey wrote:
> I have not been able to use xi to create continuous*continuous
> interactions on the fly. I don't think xi does this from the
> description. It does cat*cont though.
Do cont*cont interactions make sense? It is often an act of faith that a
unit increase in X_i produces a Beta_i increase in Y and that the Betas
are additive. It seems wildly optimistic that a k increase in X_i
together with an l increase in X_j should give the same additional
increase as an l increase in X_i with a k increase in X_j.
This seems in general an approach based on having columns of figures so a
computer will always do the arithmetic, whether or not the sum is