[R] interaction of continuous terms

Prof Brian Ripley ripley at stats.ox.ac.uk
Wed Nov 28 13:58:21 CET 2007


It's meaningful, but not under the name 'interaction'.  The model is the 
same as

y ~ x1 + x2 + I(x1*x2)

that is, 'including a product term'.  However, needing product terms often 
indicates a missed transformation of y (e.g. it might be better to use 
log(y)).

On Wed, 28 Nov 2007, Irene Mantzouni wrote:

> Hi all!
>
> this is a rather statistical question:
> is it meaningful to consider an interaction effect between 2 continuous covariates?
> for example: lm(y~x1+x2+x1:x2)
> Should one of continuous x1, x2 be "transformed" to a categorical variable, i.e. be classified into groups?
> Is it easier to interpret the effect if 1 or both are centered to the mean or z-transformed?

Depends on the context, but usually not.

-- 
Brian D. Ripley,                  ripley at stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595



More information about the R-help mailing list