I always liked Johan Bring's (1994. How to standardize regression coefficients. Am. Stat. 48(3):209-213.) arguments for standardizing the beta estimates in a linear model by their partial standard deviations which equates to comparing ratios of t-statistics between variables to determine their relative importance. It seems to me that this approach ought to work reasonably well for any of the generalized linear models.
Brian On Fri, Dec 14, 2012 at 8:00 AM, Michael Dewey <[email protected]>wrote: > At 08:33 14/12/2012, Jeremy Goss wrote: > >> Dear users, >> >> Does anyone have any idea how to generate standardised beta coefficients >> for a ZINB model in R to compare which explanatory variables are having >> the >> greatest impact on the dependent variable? >> > > You might like to look at Ulrike Grömping's work on relative importance in > regression. Not sure how directly this translates into the non-linear case > though. > > J Stat Soft October 2006 vol 17 number 1 > > > Thanks, >> Jeremy >> >> [[alternative HTML version deleted]] >> > > Michael Dewey > [email protected] > http://www.aghmed.fsnet.co.uk/**home.html<http://www.aghmed.fsnet.co.uk/home.html> > > ______________________________**________________ > [email protected] mailing list > https://stat.ethz.ch/mailman/**listinfo/r-help<https://stat.ethz.ch/mailman/listinfo/r-help> > PLEASE do read the posting guide http://www.R-project.org/** > posting-guide.html <http://www.R-project.org/posting-guide.html> > and provide commented, minimal, self-contained, reproducible code. > -- Brian S. Cade, PhD U. S. Geological Survey Fort Collins Science Center 2150 Centre Ave., Bldg. C Fort Collins, CO 80526-8818 email: [email protected] tel: 970 226-9326 [[alternative HTML version deleted]]
______________________________________________ [email protected] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
