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

# Re: st: Obtaining rrr's of margins after mlogit

 From John Francis To statalist@hsphsun2.harvard.edu Subject Re: st: Obtaining rrr's of margins after mlogit Date Thu, 22 Mar 2012 12:03:30 +0100

```Let's see if I can clear this up a little,

I use -mlogit- with the -rrr- option, which gives me as an output:

.[cat1]
.
.edu
.1 [...]
.2 [...]
.
.edu#coh
.1 1 [...]
.2 1 [...]

Now, what do I do in order to check how much the effect has changed
between cohorts, and whether or not said change is significant? How
does the constant play into this? I think these things would be fairly
obvious to see if I could use -margins, over()-, but as it is, the
-mlogit- output is quite confusing to me...

John

On Thu, Mar 22, 2012 at 11:41 AM, Maarten Buis <maartenlbuis@gmail.com> wrote:
> Maybe I am still not understanding your question, as the solution
> appears to me to just not to use -margins- or -lincom- but directly
> specify the -rrr- option in your -mlogit- command:
>
> mlogit depvar i.edu##i.coh i.sex##i.coh, b(2) rrr
>
> This will directly give you the ratio or relative risk ratios that you
> seem to be looking for.
>
> Hope this helps,
> Maarten
>
> On Thu, Mar 22, 2012 at 11:31 AM, John Francis
>> Thank you Maarten,
>>
>> want is to first run
>> -mlogit-, say
>>
>> .mlogit depvar i.edu##i.coh i.sex##i.coh, b(2) nolog
>>
>> Then, I would first like to get the marginal effects for each outcome
>> in order to get RRRs of the interaction terms, with each of the four
>> categories in relation to category 2. The reason I want to do this is
>> so that I can use the method described in your Stata-Tip regarding the
>> interpretation of interaction effects in nonlinear models. This means
>> that I
>> would run
>>
>> .forval i = 1/4 {
>> .margins, over(i.edu i.coh) expr(exp(xb(#`i'))) post
>> .est sto m`i'
>> .}
>>
>> so that afterwards, I could use -lincom- in order to interpret the
>> interaction effects as RRRs. Same goes for the interaction of sex and
>> coh. This is how far I got so far. Now of course, Stata tells me
>> "Warning: prediction constant over observations." I take that to mean
>> that the specification of xb() is wrong, yet I don't know how I should
>> change it.
>>
>> Besides, is this the correct way to apply the procedure described by
>> you to a multinomial logit model?
>>
>> Regards,
>> John
>>
>> On Thu, Mar 22, 2012 at 9:14 AM, Maarten Buis <maartenlbuis@gmail.com> wrote:
>>> On Wed, Mar 21, 2012 at 5:37 PM, John Francis wrote:
>>>> I have a question on obtaining margins as relative risk ratios instead
>>>> of probabilities after I run the mlogit command.
>>>
>>> You can specify the -rrr- option in -mlogit-.
>>>
>>> Hope this helps,
>>> Maarten
>>>
>>> --------------------------
>>> Maarten L. Buis
>>> Institut fuer Soziologie
>>> Universitaet Tuebingen
>>> Wilhelmstrasse 36
>>> 72074 Tuebingen
>>> Germany
>>>
>>>
>>> http://www.maartenbuis.nl
>>> --------------------------
>>> *
>>> *   For searches and help try:
>>> *   http://www.stata.com/help.cgi?search
>>> *   http://www.stata.com/support/statalist/faq
>>> *   http://www.ats.ucla.edu/stat/stata/
>>
>> *
>> *   For searches and help try:
>> *   http://www.stata.com/help.cgi?search
>> *   http://www.stata.com/support/statalist/faq
>> *   http://www.ats.ucla.edu/stat/stata/
>
>
>
> --
> --------------------------
> Maarten L. Buis
> Institut fuer Soziologie
> Universitaet Tuebingen
> Wilhelmstrasse 36
> 72074 Tuebingen
> Germany
>
>
> http://www.maartenbuis.nl
> --------------------------
>
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/statalist/faq
> *   http://www.ats.ucla.edu/stat/stata/

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/
```