etable for mlogit estimates - how to get rrr?

[Nina Fajs]

Hi all,

I am working with mlogit, and I do realises that the rrr option just affect how the results are displayed.
However, can I get the rrr and CI instead of coeff and se in etable?

Here is the example code I was running, I couldn't figure out if the eform is available in etable:
etable, estimates (model1 model 2) cstat(_r_b) cstat(_r_ci) cstat(_r_p) stars(0.0125 "*" 0.0025 "**") showstars showstarsnote ///
mstat(N) mstat(r2) mstat(aic) mstat(bic) export(results.docx, replace)

[Clyde Schechter]

-eform- does not have an -eform- option. But your -mlogit- command does, called -rrr-. So use the -rrr- option on -mlogit- and then run -etable-, and the latter will then table the relative risk ratios instead of the coefficients.

[Nina Fajs]

Hi Clyde,

Thank you so much for your answer. However, I still have this issue. Note that I am using a Stata kernel and it is ran from Jupyter notebook and that cannot be changed.
When I try to replicate this problem with test data, I get the RRR in etable, which is very frustrating.

When I run:

mlogit trait1 i.trait2 i.trait3, rrr
estimates store model1

This is the head of the output (the names of the actual variables are changed):

However, when I run the etable, it is still shown as coeff instead of RRR:

etable, estimates (model1) cstat(_r_b) cstat(_r_ci) cstat(_r_p) /// stars(0.0125 "*" 0.0025 "**") showstars showstarsnote mstat(N) mstat(r2) mstat(aic) mstat(bic) /// export(results.docx, replace)

[Clyde Schechter]

I cannot replicate your problem in my setup. Here I run -mlogit, rrr- followed by -etable-, and -etable- properly displays the odds ratios and their confidence intervals.

. sysuse auto, clear
(1978 automobile data)

. 
. mlogit rep78 i.foreign mpg price, rrr

Iteration 0:  Log likelihood = -93.692061  
Iteration 1:  Log likelihood = -77.396517  
Iteration 2:  Log likelihood = -75.281559  
Iteration 3:  Log likelihood = -74.998968  
Iteration 4:  Log likelihood = -74.951486  
Iteration 5:  Log likelihood = -74.939579  
Iteration 6:  Log likelihood = -74.937162  
Iteration 7:  Log likelihood = -74.936639  
Iteration 8:  Log likelihood = -74.936522  
Iteration 9:  Log likelihood = -74.936492  
Iteration 10: Log likelihood = -74.936486  

Multinomial logistic regression                         Number of obs =     69
                                                        LR chi2(12)   =  37.51
                                                        Prob > chi2   = 0.0002
Log likelihood = -74.936486                             Pseudo R2     = 0.2002

------------------------------------------------------------------------------
       rep78 |        RRR   Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
1            |
     foreign |
    Foreign  |   1.17e-06   .0025892    -0.01   0.995            0           .
         mpg |   1.000732   .1926954     0.00   0.997     .6861436    1.459555
       price |   .9995268   .0007917    -0.60   0.550     .9979764     1.00108
       _cons |   .8311469   5.465247    -0.03   0.978     2.10e-06    328696.4
-------------+----------------------------------------------------------------
2            |
     foreign |
    Foreign  |   1.00e-06    .001113    -0.01   0.990            0           .
         mpg |   .9777956   .1078435    -0.20   0.839     .7877111     1.21375
       price |    .999935   .0001621    -0.40   0.688     .9996173    1.000253
       _cons |   .6803167   1.899598    -0.14   0.890     .0028574    161.9789
-------------+----------------------------------------------------------------
3            |  (base outcome)
-------------+----------------------------------------------------------------
4            |
     foreign |
    Foreign  |   8.106832   6.672772     2.54   0.011     1.615188     40.6892
         mpg |   1.032125   .0852689     0.38   0.702     .8778303    1.213541
       price |    .999975   .0001346    -0.19   0.853     .9997112    1.000239
       _cons |   .2098572    .462293    -0.71   0.478     .0027978    15.74092
-------------+----------------------------------------------------------------
5            |
     foreign |
    Foreign  |   19.14689   20.01832     2.82   0.005     2.466926    148.6074
         mpg |   1.246044    .128245     2.14   0.033     1.018419    1.524546
       price |   1.000173   .0001826     0.95   0.344     .9998148    1.000531
       _cons |   .0002251   .0007341    -2.58   0.010     3.77e-07    .1343249
------------------------------------------------------------------------------
Note: _cons estimates baseline relative risk for each outcome.

. estimates store model1

. 
. etable, estimates (model1) cstat(_r_b) cstat(_r_ci) cstat(_r_p)  /// 
> stars(0.0125 "*" 0.0025 "**") showstars showstarsnote mstat(N) mstat(r2) mstat(aic) mstat(bic) 

------------------------------------------
                              rep78       
------------------------------------------
Car origin                                
  Foreign                          0.000  
                                 [0.000]  
                                    1.00  
Mileage (mpg)                      1.001  
                       [0.686     1.460]  
                                    1.00  
Price                              1.000  
                       [0.998     1.001]  
                                    0.55  
Intercept                          0.831  
                       [0.000  3.29e+05]  
                                    0.98  
Car origin                                
  Foreign                          0.000  
                                 [0.000]  
                                    0.99  
Mileage (mpg)                      0.978  
                       [0.788     1.214]  
                                    0.84  
Price                              1.000  
                       [1.000     1.000]  
                                    0.69  
Intercept                          0.680  
                       [0.003   161.979]  
                                    0.89  
Car origin                                
  Foreign                          8.107 *
                       [1.615    40.689]  
                                    0.01  
Mileage (mpg)                      1.032  
                       [0.878     1.214]  
                                    0.70  
Price                              1.000  
                       [1.000     1.000]  
                                    0.85  
Intercept                          0.210  
                       [0.003    15.741]  
                                    0.48  
Car origin                                
  Foreign                         19.147 *
                       [2.467   148.607]  
                                    0.00  
Mileage (mpg)                      1.246  
                       [1.018     1.525]  
                                    0.03  
Price                              1.000  
                       [1.000     1.001]  
                                    0.34  
Intercept                          0.000 *
                       [0.000     0.134]  
                                    0.01  
Number of observations                69  
AIC                               181.87  
BIC                               217.62  
------------------------------------------
** p<.0025, * p<.0125

It may be that this problem arises from your setup. I suggest you contact Stata Tech Support about this. Be sure to provide them with the details (version numbers) of your Stata installation, Jupyter, and operating system, as well as your exact code and a data example reproducing the problem.