ESTOUT in case of interaction.

[Luca Tognoni]

Dear all, I have some troubles in using estout command for a model.
The model I have built look like this:
mixed score EUROPE_NONEU EU ROW Lescs Hescs OL IT IP male B99 B00 B01 B03 i.itfather##i.itmother insuff_ita insuff_mat nord sud_isole || schid:,mle vce(robust)


score Coefficient std. err. z P>z [95% conf. interval]

EUROPE_NONEU -8.608609 2.267492 -3.80 0.000 -13.05281 -4.164407
EU 4.132151 2.028547 2.04 0.042 .1562725 8.10803
ROW -8.749745 1.857788 -4.71 0.000 -12.39094 -5.108548
Lescs -1.768383 .5980749 -2.96 0.003 -2.940589 -.5961781
Hescs 1.90282 .6040967 3.15 0.002 .7188127 3.086828
OL -19.70843 2.059975 -9.57 0.000 -23.7459 -15.67095
IT -26.34957 1.63437 -16.12 0.000 -29.55287 -23.14626
IP -45.20013 1.92037 -23.54 0.000 -48.96399 -41.43628
male -.4215046 .5723091 -0.74 0.461 -1.54321 .7002008
B99 -6.287114 4.563022 -1.38 0.168 -15.23047 2.656245
B00 -4.36976 1.537388 -2.84 0.004 -7.382984 -1.356536
B01 -6.208741 .766699 -8.10 0.000 -7.711444 -4.706039
B03 1.995999 .8178015 2.44 0.015 .3931375 3.598861
1.itfather 9.594129 1.468848 6.53 0.000 6.71524 12.47302
1.itmother 9.147676 1.764706 5.18 0.000 5.688915 12.60644

itfather#itmother
1 1 -11.12236 2.137621 -5.20 0.000 -15.31202 -6.932705

insuff_ita -11.83531 .8733229 -13.55 0.000 -13.54699 -10.12363
insuff_mat -8.809681 .6620096 -13.31 0.000 -10.1072 -7.512166
nord 11.26301 1.470998 7.66 0.000 8.379902 14.14611
sud_isole -14.29844 1.467714 -9.74 0.000 -17.17511 -11.42177
_cons 208.763 1.855075 112.54 0.000 205.1271 212.3989



Robust
Random-effects parameters Estimate std. err. [95% conf. interval]

schid: Identity
var(_cons) 121.4309 10.19598 103.0049 143.153

var(Residual) 784.3837 11.56615 762.0389 807.3837


But when i use:
. esttab n4, p wide nostar transform(ln*: exp(2*@) 2*exp(2*@)) eqlabels("" "var(_cons)" "var(Residual)", none) varlabels(,elist(weight:_cons "{brea
> k}{hline @width}")) varwidth(13)

---------------------------------------
(1)
score
---------------------------------------
EUROPE_NONEU -8.609 (0.000)
EU 4.132 (0.042)
ROW -8.750 (0.000)
Lescs -1.768 (0.003)
Hescs 1.903 (0.002)
OL -19.71 (0.000)
IT -26.35 (0.000)
IP -45.20 (0.000)
male -0.422 (0.461)
B99 -6.287 (0.168)
B00 -4.370 (0.004)
B01 -6.209 (0.000)
B03 1.996 (0.015)
0.itfather 0 (.)
1.itfather 9.594 (0.000)
0.itmother 0 (.)
1.itmother 9.148 (0.000)
0.itfather#~r 0 (.)
0.itfather#~r 0 (.)
1.itfather#~r 0 (.)
1.itfather#~r -11.12 (0.000)
insuff_ita -11.84 (0.000)
insuff_mat -8.810 (0.000)
nord 11.26 (0.000)
sud_isole -14.30 (0.000)
_cons 208.8 (0.000)
var(_cons) 121.4 (0.000)
var(Residual) 784.4 (0.000)
---------------------------------------
N 16381
---------------------------------------
p-values in parentheses



Can someone explain why there are those null extra interaction terms? Is there a way to get rid of them?
Thank you all

[Andrew Musau]

estout is from SSC (FAQ Advice #12). You have categorical variables, so the omitted categories are referred to as base levels. If you want to suppress them, specify the option -nobaselevels-.

esttab n4, p wide nostar nobaselevels ...

[Michael Schuster]

You could improve the chance to get an answer to your question by using -help dataex- and providing code by using the #-button when you write a comment/question.

Other than that, the zeros in front of your variables are given due to the fact that you typed

i.itfather

, so Stata basically provides both dummy 0 and 1. If you type just

1.itfather

you will get just the result for the case the dummy variable is 1.

I assume you refer to those

0.itfather#~r 0 (.)
0.itfather#~r 0 (.)
1.itfather#~r 0 (.)
1.itfather#~r -11.12 (0.000)

when you write

null extra interaction terms

I assume this comes from all possible connections of i.father and i.mother combined.

[Luca Tognoni]

Thank you Andrew and Micheal for your precious help!