Exporting F-Statistics and p-value after ivreg2 command

[Odmaa Naran]

Dear all,

I am using

ivreg2

command to do my analysis with Stata 15 (IC). However, I need to report the F-test of excluded instruments for my output table. I first, run the following code

ivreg2 sick (i.polluted=l_dur_value) female age schooling yhat rural yr2010 
yr2014 yr2016 yr2018 i.newsoum if distance<20, first cluster(identif)

then to get the F-statistics from the first-stage regression, I use the following

mat l e(first)

The output is as follows:

e(first)[21,1]
                 1.
          polluted
   rmse  .33124254
sheapr2  .00110657
    pr2  .00110657
      F  10.484074
     df          1
   df_r       4557
 pvalue  .00121271
    SWF  10.484074
 SWFdf1          1
 SWFdf2       4557
   SWFp  .00121271
 SWchi2  10.521312
SWchi2p  .00118006
   SWr2  .00110657
    APF  10.484074
 APFdf1          1
 APFdf2       4557
   APFp  .00121271
 APchi2  10.521312
APchi2p  .00118006
   APr2  .00110657

However, when I want to local the F-stat and the p-value with

 local fstat_base = e(first)[4(e(first),"SWF"),1]
 local fpval_base = e(first)[7(e(first),"SWFp"),1]

I get

invalid syntax
r(198);

Can someone please help me where I am getting the code wrong in the local command? Is there any other way to store the F-stat and the p-value?

Thanks a lot and your help is much appreciated!

---

[Dung Le]

Hi Odmaa,

You may want to read this post: https://www.statalist.org/forums/for...-stage-outcome. Andrew provided lots of useful illustrations there.

Hope this helps.

[Odmaa Naran]

Hi Dung Le,

Thanks very much. That's very helpful.