[R] Interaction term in anova - how it should be written in a manuscript table?

[Marc Schwartz]

For a me too post, I agree with Andy's recommendation, which in turn is
supported by "How to Report Statistics in Medicine" by Lang and Secic,
ACP, 1997. There is an example table (8.2) on page 133.

HTH,

Marc Schwartz

On Sun, 2005-03-20 at 15:44 -0500, Liaw, Andy wrote:
> I'd suggest a $\times$ b, as you'd find in most stat textbook.
> 
> Andy
> 
> > From: Witold Eryk Wolski
> > 
> > Dear Rgurus,
> > 
> > Interaction terms in the linear models function lm are 
> > specified by the 
> > colon :
> > eg: x ~ a + b + a:b
> > 
> > a shortcut for the above is:
> >  x ~ a*b
> > 
> > the output if calling anova on the lm object will be the same 
> > in both cases
> > 
> > a ....
> > b ....
> > a:b ...
> > Resdiuals ...
> > 
> > What I am wondering is how the interaction term (a:b) given 
> > above should 
> > be written in a table in an manuscript?
> > 
> > a ) a*b
> > b ) a$\cdot$ b
> > c ) a:b
> > d) ....
> > 
> > Cheers Eryk.
> >