Significance stars (was Re: R-beta: glm bug)
Peter Dalgaard BSA
p.dalgaard at biostat.ku.dk
Thu Jun 18 20:57:09 CEST 1998
Martin Maechler <maechler at stat.math.ethz.ch> writes:
> Jim Lindsey wanted a *global* option so that he can turn it off for all
> his students when they just use
> Any proposals for a NAME of the global option?
Anything wrong with signif.stars ?
> JL even looked for a possibility to turn off the P values..
> Do we want/need this as well?
Nope. (Did he?) These are much too often important when you need to
compensate for multiple testing and so forth. Even though a lot of
good things can be said about confidence intervals, they do have the
problem that the implied tests are always at a fixed level.
> NOTE: We do *NOT* advocate believing in P-values,
Well, actually, I *do* advocate that. At least in the sense that
probability exist and one needs to understand data in a probabilistic
context. This is not, however, the same as the boneheaded logic of
setting strict on/off rules based on the formal p-values. P-values are
usually fairly good approximations of a reasonably well-defined
quantity, the main problem being that it is often not that quantity
which is the relevant one.
> the main point of the sig.stars is a
> ``graphical summary'' of which coefficients might be significant
More or less my feeling too. If you're doing explorative analysis and
fitting a lot of models wit many regressors, it can be useful to be
able to get the 'highlights' at a glance. If not, you can just ignore
O__ ---- Peter Dalgaard Blegdamsvej 3
c/ /'_ --- Dept. of Biostatistics 2200 Cph. N
(*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918
~~~~~~~~~~ - (p.dalgaard at biostat.ku.dk) FAX: (+45) 35327907
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch
More information about the R-help