On Wed, 27 Sep 2000, John Maindonald wrote:
From: Jim Lindsey <james.lindsey at luc.ac.be>
Bill's excellent discussion (which I do not reproduce here) describes
precisely why I was (and still am) dead against having stars printed
as the default option in lm and glm. Let's get rid of that and let
people turn them on, if they want, in those cases in which they make
sense.
I agree. When using R for teaching, they are a serious distraction,
and I find it absolutely necessary to turn them off.
There are very obtrusive. If there was a way to vary the shade of
grey in which the coefficients or t-statistics were printed, that
would be more acceptable!
I always forget that that they are turned on by default, for I have
auk% cat ~/.Rprofile
options(show.signif.stars=FALSE)
ps.options(horizontal=FALSE)
and so never see them.
In short, the out-of-the-box defaults can be over-ridden easily enough,
including the system ones for a teaching lab, say, by changing
.../library/base/R/Rprofile.