Hello...
I'm running a multinomial logit with a three-category DV. I want to
compare the relative strength of each IV's effect on the DV, and since
my IVs are of various types (some ordinal, some continuous, and some
dichtomous), I used the "listcoef" function to generate the standardized
odds ratios. However, I'd prefer to present predicted probabilities,
not odds ratios, when discussing the results. Thus, I used the
"prchange" command to generate predicted probabilities as each IV
changes from min to max, as well one standard deviation. The one
standard deviation changes in probability follow the exact pattern as
the standardized odds ratios in ranking each continuous and ordinal IV's
effect on the DV.
The problem is that "prchange" will not generate one standard deviation
changes in probabilities for the dichotomous IVs. I cannot seem to
figure out how to do this. Am I missing something, or is there an
alternative method? I need to be able to do this because I need to be
able to compare the effects of all IVs using the same standard. If I
present one standard deviation probability changes for all ordinal and
continuous IVs and min to max changes for the dichtomous IVs, the
ranking of the relative effect of each IV does not coincide with what
the standardized odds ratios are telling me.
Thank you,
David Quinn
Department of Government and Politics
University of Maryland, College Park
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
