Applying -suest- to logistic models

[WJ Wang]

Hi everyone!

I am working with a dataset containing data collected from two population groups (i.e., group=0 or 1, with everyone answered to the same questions). I have a binary dependent variable and a set of independent variables containing both binary and continuous variables. I fit a logistic regression model for each group and am trying to compare the differences in coefficients of my IVs from the two models. My understanding is that I can use the -suest- command for the comparison (as suggested in the Stata manual: chrome-extension://efaidnbmnnnibpcajpcglclefindmkaj/https://www.stata.com/manuals/rsuest.pdf), but I got feedback from other people saying that -suest- is not the appropriate command. Can somebody please provide some insights on whether -suest- is applicable to logistic coefficients and if not, what would be some reasonable alternatives?

Thank you all so much in advance!

[Carlo Lazzaro]

WJ:
example 3, -suest- entry, Stata .pdf manual is actually on -logit-.
In addition:

. use "C:\Program Files\Stata17\ado\base\a\auto.dta"
(1978 automobile data)

. logit foreig price

Iteration 0:   log likelihood =  -45.03321  
Iteration 1:   log likelihood = -44.947363  
Iteration 2:   log likelihood =  -44.94724  
Iteration 3:   log likelihood =  -44.94724  

Logistic regression                                     Number of obs =     74
                                                        LR chi2(1)    =   0.17
                                                        Prob > chi2   = 0.6784
Log likelihood = -44.94724                              Pseudo R2     = 0.0019

------------------------------------------------------------------------------
     foreign | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
       price |   .0000353   .0000844     0.42   0.676    -.0001301    .0002006
       _cons |  -1.079792   .5878344    -1.84   0.066    -2.231927    .0723419
------------------------------------------------------------------------------

. estimates store l_1

. logit foreig mpg

Iteration 0:   log likelihood =  -45.03321  
Iteration 1:   log likelihood = -39.380959  
Iteration 2:   log likelihood = -39.288802  
Iteration 3:   log likelihood =  -39.28864  
Iteration 4:   log likelihood =  -39.28864  

Logistic regression                                     Number of obs =     74
                                                        LR chi2(1)    =  11.49
                                                        Prob > chi2   = 0.0007
Log likelihood = -39.28864                              Pseudo R2     = 0.1276

------------------------------------------------------------------------------
     foreign | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         mpg |   .1597621   .0525876     3.04   0.002     .0566922     .262832
       _cons |  -4.378866   1.211295    -3.62   0.000    -6.752961   -2.004771
------------------------------------------------------------------------------

. estimates store l_2

. suest l_1 l_2

Simultaneous results for l_1, l_2                           Number of obs = 74

------------------------------------------------------------------------------
             |               Robust
             | Coefficient  std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
l_1_foreign  |
       price |   .0000353   .0000791     0.45   0.656    -.0001198    .0001904
       _cons |  -1.079792   .5537941    -1.95   0.051    -2.165209    .0056241
-------------+----------------------------------------------------------------
l_2_foreign  |
         mpg |   .1597621   .0514516     3.11   0.002     .0589188    .2606055
       _cons |  -4.378866   1.181227    -3.71   0.000    -6.694028   -2.063704
------------------------------------------------------------------------------

.

[WJ Wang]

WJ:
example 3, -suest- entry, Stata .pdf manual is actually on -logit-.
In addition:

. use "C:\Program Files\Stata17\ado\base\a\auto.dta"
(1978 automobile data)

. logit foreig price

Iteration 0: log likelihood = -45.03321
Iteration 1: log likelihood = -44.947363
Iteration 2: log likelihood = -44.94724
Iteration 3: log likelihood = -44.94724

Logistic regression Number of obs = 74
LR chi2(1) = 0.17
Prob > chi2 = 0.6784
Log likelihood = -44.94724 Pseudo R2 = 0.0019

------------------------------------------------------------------------------
foreign | Coefficient Std. err. z P>|z| [95% conf. interval]
-------------+----------------------------------------------------------------
price | .0000353 .0000844 0.42 0.676 -.0001301 .0002006
_cons | -1.079792 .5878344 -1.84 0.066 -2.231927 .0723419
------------------------------------------------------------------------------

. estimates store l_1

. logit foreig mpg

Iteration 0: log likelihood = -45.03321
Iteration 1: log likelihood = -39.380959
Iteration 2: log likelihood = -39.288802
Iteration 3: log likelihood = -39.28864
Iteration 4: log likelihood = -39.28864

Logistic regression Number of obs = 74
LR chi2(1) = 11.49
Prob > chi2 = 0.0007
Log likelihood = -39.28864 Pseudo R2 = 0.1276

------------------------------------------------------------------------------
foreign | Coefficient Std. err. z P>|z| [95% conf. interval]
-------------+----------------------------------------------------------------
mpg | .1597621 .0525876 3.04 0.002 .0566922 .262832
_cons | -4.378866 1.211295 -3.62 0.000 -6.752961 -2.004771
------------------------------------------------------------------------------

. estimates store l_2

. suest l_1 l_2

Simultaneous results for l_1, l_2 Number of obs = 74

------------------------------------------------------------------------------
| Robust
| Coefficient std. err. z P>|z| [95% conf. interval]
-------------+----------------------------------------------------------------
l_1_foreign |
price | .0000353 .0000791 0.45 0.656 -.0001198 .0001904
_cons | -1.079792 .5537941 -1.95 0.051 -2.165209 .0056241
-------------+----------------------------------------------------------------
l_2_foreign |
mpg | .1597621 .0514516 3.11 0.002 .0589188 .2606055
_cons | -4.378866 1.181227 -3.71 0.000 -6.694028 -2.063704
------------------------------------------------------------------------------

.

Hi Carlo,

Thanks so much for the response! And yes, that's exactly what I was referring to in my original post (I also just realized that the link was messed up a little - my apologies). My understanding is that since this example (based on, according to the manual, Clogg's test) is provided by the developers, it should be a pretty strong justification for my use of the test. Thanks again for your help!