Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: Wish List for margins

From   David Backer <>
Subject   Re: st: Wish List for margins
Date   Wed, 1 Sep 2010 01:05:18 -0400 (EDT)

Following up on this in a related vein, is there a way to use margeff after -mim: logit- to calculate average marginal effects across multiply imputed datasets? When I try to do so, it generates the following error:

==0 invalid name

I tried the suggestion on the statalist of preceding the margeff command with -mim, storebv-, but this does not solve the problem.

If it matters any, I am using Stata 10.

On Tue, 31 Aug 2010, Richard Williams wrote:

I'm forcing myself to use the -margins- command and I am starting to like and understand it better. Some things I'd like to see (or else maybe somebody can tell me how to do them already):

* I'd like a -discrete- option that would tell margins to treat a 0/1 dichotomy as a dichotomy. This would make margins more useful with older user-written commands (like mine) that do not currently support the use of factor variables.

* Perhaps the syntax would be tricky, but I am a little surprised there isn't a generate option. The old adjust command has one. You might want to see, say, what the predicted probability for a case is given the values you have fixed using the -at- option.

* As far as I can tell, marginal effects after multiple-outcome commands are still clunky to generate. You have to give a separate margins command for each outcome. I wish margins acted like the user-written mfx2 and margeff commands, generating all the marginal effects for every outcome with a single command.

* Like I've said before, I wish the old adjust and mfx commands had been updated for Stata 11. One thing they would need added is an -asobserved- option, which is the default for margins. At 50 pages, the manual entry for margins is a little daunting. I think it would be easier to understand if there were simpler commands that focused on specific tasks.

Richard Williams, Notre Dame Dept of Sociology
OFFICE: (574)631-6668, (574)631-6463
HOME:   (574)289-5227
EMAIL:  Richard.A.Williams.5@ND.Edu

*   For searches and help try:

*   For searches and help try:

© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index