st: Relative Importance of predictors in regression

[Joseph Luchman <jluchman@gmail.com>]

Hi Nikos,

  The Stata module - domin - (SSC) decomposes the R2 metric using
dominance analysis (which produces equivalent results as - shapley -
by Stas Kolenikov on SSC for the general dominance weights - but also
produces several other statistics that may be of use).

  I'll note that the "epsilon" option offered by - domin - actually
utilizes the relative weight analysis procedure developed by Johnson
(2000) if you'd prefer that method.

- joe

Johnson, J. W. (2000). A heuristic method for estimating the relative
weight of predictor variables in multiple regression. Multivariate
Behavioral Research,35(1), 1-19.


Dear Statalisters,

I have a multiple regression in Stata and want to concisely show the
relative contribution of each predictor to the final model. I would
have thought this is easy but I'm finding it difficult to figure out
how to go about it.
I used
regress Y x1 x2 x3...x7, beta vce(robust)

I think, however, that the beta coefficients do not really tell the whole story.

I found this document by Nathans et al that I thought does a great job
of discussing the problem:
http://pareonline.net/pdf/v17n9.pdf

I really like the idea of repartitioning the overall model R2 between
the predictors by Relative Weight Analysis. I looked around and found
there's a package for this for R but not Stata.

I would be most grateful for your recommendations.

Many thanks

Nikos

PS: Dr Marcus Fischer previously posted on this about a year ago but I
could not find a follow-up so I thought I'd try asking the group
again.

Joseph Nicholas Luchman, M.A.
----
Behavioral Statistics Lead | Fors Marsh Group
Email: jluchman@forsmarshgroup.com
forsmarshgroup.com
----
Doctoral Candidate
Industrial Organizational Psychology
George Mason University
https://www.researchgate.net/profile/Joseph_Luchman/
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/