I have a longitudinal (even if not properly so, since my individuals are followed only for two waves) dataset, and I want to run a xtlogit, re model since my dependent variable is binary, and I am interested in estimating the effect of an interaction between gender (Xf) and my main explanatory variable (dummy) on my dependent.
Below an example of my data:
Code:
* Example generated by -dataex-. To install: ssc install dataex clear input float dip long Xa double Xb long Xc float Xd long Xf 0 2 49 2 1 1 0 2 49 2 1 1 0 3 60 1 2 2 0 3 60 1 2 2 0 3 33 1 1 2 0 3 34 1 2 2 0 3 25 1 2 1 0 3 25 1 2 1 0 3 29 5 2 1 0 3 28 5 2 1 1 2 58 1 1 2 0 3 58 1 1 2 1 3 50 1 2 1 1 3 50 1 2 1 0 3 42 1 2 2 0 3 43 1 2 2 1 3 51 2 1 2 . . 51 2 . 2 1 3 53 1 2 2 1 3 53 1 2 2 end label values Xa F004 label def F004 2 "Secondary", modify label def F004 3 "Tertiary", modify label values Xc emp_stat label def emp_stat 1 "Employee", modify label def emp_stat 2 "Self-employed", modify label def emp_stat 5 "Other", modify label values Xd partner_present label def partner_present 1 "Yes", modify label def partner_present 2 "No", modify label values Xf B002 label def B002 1 "Male", modify label def B002 2 "Female", modify
Code:
xtlogit dip i.dummy##i.Xf i.Xa i.country i.wave Xb i.Xc i.Xd, re
Code:
| Delta-method | Margin Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- dummy | 0 | .1257196 .0282364 4.45 0.000 .0703773 .1810619 1 | -2.235678 .0522 -42.83 0.000 -2.337988 -2.133368 | Xf | Male | -1.008945 .0496997 -20.30 0.000 -1.106354 -.9115351 Female | -.4854246 .0306359 -15.84 0.000 -.5454698 -.4253793 | dummy#Xf | 0#Male | -.241434 .0532371 -4.54 0.000 -.3457769 -.1370912 0#Female | .2807271 .0334563 8.39 0.000 .215154 .3463001 1#Male | -2.605775 .0931325 -27.98 0.000 -2.788312 -2.423239 1#Female | -2.079428 .0589252 -35.29 0.000 -2.194919 -1.963936
0 Response to Margins of an interaction after xtlogit, re
Post a Comment