Exporting marginal effects (first differences and contrasts) to word

[Gebretsadik Shibre]

Dear all,
I need your help on how I would be able to export marginal effects (both first differences and second differences) of interactions from Stata to word. I fit a logit model:
svy: logit outcomvar i.v106_new i.v025 i.adolescent_age_F i.media_new i.husband_edu_new i.v190 i.region_f i.religion_f i.women_occupation i.husband_occupation i.wantedness_child i.birthor i.v106_new#i.v025 i.v106_new#i.adolescent_age_F i.v106_new#i.media_new i.v106_new#i.husband_edu_new i.v025#i.media_new i.v025#i.husband_edu_new i.v025#i.adolescent_age_F i.adolescent_age_F#i. husband_edu_new i.media_new#i.husband_edu_new i.v025#i.wantedness_child i.media_new#i.wantedness_child i.adolescent_age_F#i.wantedness_child i.v025#i.v106_new#i.media_new i.women_occupation i.husband_occupation i.wantedness_child i.birthor

Then I do this

margins v106_new#v025
then this:
margins v025, dydx (v106_new)
Finally this:
margins v025, dydx (v106_new) pwcompare (effects)
Now I want to do these codes for all interacting variables (not just for the interaction between v025 and v106_new, but for all the interaction in my model) and export the result to word in tables. I want to export the
dydx () of both first differences ( margins v025, dydx (v106_new) and second differences/ contrasts) (margins v025, dydx (v106_new) pwcompare (effects) for all interactions at once.

Your help is much appreciated.
Regards,

[Andrew Musau]

This is just a general approach using estout from SSC. No reproducible example, so this is not tested.

ssc install estout, replace
svy: logit outcomvar ...
est sto results
margins v106_new#v025, post
est sto m1
est restore results
margins v025, dydx (v106_new) post
est sto m2
est restore results
margins v025, dydx (v106_new) pwcompare (effects) post
est sto m3
esttab m1 m2 m3 using myfile.rtf, replace

[Gebretsadik Shibre]

Thank you. The code exports the results of the first two lines (margins v106_new#v025, post and margins v025, dydx (v106_new) post).
I cannot still export outputs from the 3rd code:
margins v025, dydx (v106_new) pwcompare (effects) post.

I really appreciate any help to do this. Also, I want to automate the export to word of all the interactions in my model.

[Andrew Musau]

margins v025, dydx (v106_new) pwcompare (effects) post.

This works for me:

webuse nhanes2, clear
logistic highbp sex##agegrp##c.bmi
eststo est3: margins agegrp, pwcompare(group) post
esttab est3

Res.:

. eststo est3: margins agegrp, pwcompare(group) post

Pairwise comparisons of predictive margins              Number of obs = 10,351
Model VCE: OIM

Expression: Pr(highbp), predict()

-------------------------------------------------
             |            Delta-method Unadjusted
             |     Margin   std. err.      groups
-------------+-----------------------------------
      agegrp |
      20-29  |   .2030932   .0087166
      30-39  |   .2829091    .010318
      40-49  |   .3769536   .0128744
      50-59  |   .5153439   .0136201
      60-69  |   .5641065    .009136
        70+  |   .6535679   .0151371
-------------------------------------------------
Note: Margins sharing a letter in the group label
      are not significantly different at the 5%
      level.

. 
. esttab est3

----------------------------
                      (1)   
                            
----------------------------
1.agegrp            0.203***
                  (23.30)   

2.agegrp            0.283***
                  (27.42)   

3.agegrp            0.377***
                  (29.28)   

4.agegrp            0.515***
                  (37.84)   

5.agegrp            0.564***
                  (61.75)   

6.agegrp            0.654***
                  (43.18)   
----------------------------
N                   10351   
----------------------------
t statistics in parentheses
* p<0.05, ** p<0.01, *** p<0.001

.

Also, I want to automate the export to word of all the interactions in my model.

Do you mean the coefficients? You stored them as results in the code #1, so

esttab results m1 m2 m3 using myfile.rtf, replace

[Gebretsadik Shibre]

Thank you. m1 and m2 are exported. But results of m3, the contrast, is not. I want to export the results stored in m2 and m3, for all interactions.
I kindly wait for your help to this problem. It saves me a lot of time because I have more than 10 variables in the interactions and I want to export their first and second differences automatically at one go.
Regards,

[Andrew Musau]

Use my data example in #2 and point out what results you want. Or present your data example using the dataex command. Beyond what I have shown, I cannot follow what you want.

[Gebretsadik Shibre]

Dear Dr Andrew thank you so much.

Let me explain what I want exactly.

I tried the code you advised me to do:

 
 esttab m1 m2 m3 using myfile.rtf, replace

However, the result stored in m3 is not exported. Only results in m1 and m2 are exported.

But my aim is to export m3, not in m1 and m2. Then, I want to do the same (export their second differences) for all other interactions in my model (#1).


Any effort to assist me is highly appreciated.

Regards,