lincom with three way interactions

[Karen Arulsamy]

Hi all,
I am analyzing the effect of the pandemic on health behaviours comparing women and men and would like to get a better understanding of heterogeneity. For example, in the equation below, I would like to compare the outcomes for low income women vs. high income women before and after the pandemic to see whether the effects are different by income groups. I would like to use lincom to get the relevant coefficient and the associated standard errors and confidence intervals. The coefficient I'm getting is quite large (after converting to a probability) and was wondering if I had the right lincom specification?

global controls_simp i.agecat_sec i.educ_ref i.race_main i.jbstat_ref i.cathhincome_ref i.mhvalue_ref_casenessc i.marstat_ref i.nchild015_ref c.hhsize_ref ///
i.gor_main i.imonth_final

xtreg dfruit_simp i.treatfem##i.July2020##i.cathhincome_refb $controls_simp, fe vce(cluster pidp)
lincom 1.treatfem + 1.treatfem#1.July2020 + 1.treatfem#1.cathhincome_refb + 1.cathhincome_refb#1.July2020 + 1.treatfem#1.July2020#1.cathhincome_refb, or

* Example generated by -dataex-. To install: ssc install dataex
clear
input long pidp float(dfruit_simp treatfem July2020 cathhincome_refb agecat_sec educ_ref race_main jbstat_ref cathhincome_ref mhvalue_ref_casenessc marstat_ref nchild015_ref hhsize_ref gor_main imonth_final)
   76165 1 1 0 1 0 . 1 1 3 0 1 1 3  5  4
   76165 . 1 . 1 1 . 1 1 3 0 1 1 3  5  9
   76165 . 1 . 1 1 . 1 1 3 0 1 1 3  5  3
   76165 . 1 . 1 1 . 1 1 3 0 1 1 3  5 11
   76165 1 1 . 1 1 . 1 1 3 0 1 1 3  5  1
   76165 1 1 1 1 1 . 1 1 3 0 1 1 3  5  7
 1587125 1 1 . 1 2 . 0 1 2 0 0 0 2  1  1
 1587125 . 1 . 1 2 . 0 1 2 0 0 0 2  1  9
 1587125 . 1 . 1 2 . 0 1 2 0 0 0 2  1  3
 1587125 1 1 0 1 2 . 0 1 2 0 0 0 2  1  9
 1587125 . 1 . 1 2 . 0 1 2 0 0 0 2  1 11
 1587125 0 1 1 1 2 . 0 1 2 0 0 0 2  1  7
 4849085 1 0 . 1 1 1 1 1 3 1 1 0 3 11  1
 4849085 . 0 . 1 1 1 1 1 3 1 1 0 3 11  9
 4849085 . 0 . 1 1 1 1 1 3 1 1 0 3 11  3
 4849085 . 0 . . 1 1 1 1 . 1 1 0 3 11 11
 4849085 0 0 0 1 0 1 1 1 3 1 1 0 3 11  4
 4849085 1 0 1 1 1 1 1 1 3 1 1 0 3 11  7
68002725 0 1 . 0 3 . 0 4 1 0 0 0 1  7  1
68002725 . 1 . 0 3 . 0 4 1 0 0 0 1  7  9
68002725 0 1 0 0 2 . 0 4 1 0 0 0 1  7  3
68002725 . 1 . 0 3 . 0 4 1 0 0 0 1  7  3
68002725 0 1 1 0 3 . 0 4 1 0 0 0 1  7  7
68002725 . 1 . 0 3 . 0 4 1 0 0 0 1  7 11
68008847 0 1 . . 2 0 1 1 . 0 0 0 1  1  1
end

Many thanks
Karen

[Carlo Lazzaro]

Karen:
set aside for a while my idiosyncrasy for any >2-term interaction (too difficult to disseminate), I would have coded something like_

lincom 1.treatfem + 1.July2020 +1.cathhincome_refb + 1.treatfem#1.July2020 + 1.treatfem#1.cathhincome_refb + 1.cathhincome_refb#1.July2020 + 1.treatfem#1.July2020#1.cathhincome_refb, or

[Karen Arulsamy]

Thanks Carlo Lazzaro, appreciate that.

Many thanks
Karen