Question on how to create a dummy variable for a panel data set

[Ayesha Afreen]

Hi,
This is an excerpt from my data set. I would like to create a dummy variable for agemon (age in months) separately for each round and study its cross-sectional effect. For example, I want to create an age variable for round 3 only by categorizing age into two cohorts. What would be the command for that? Kindly help.

childid	round	agemon
IN010001	1	6
IN010001	2	69
IN010001	3	100
IN010001	4	149
IN010001	5	185
IN010002	1	15
IN010002	2	67
IN010002	3	98
IN010002	4	146
IN010002	5	183
IN010003	1	11
IN010003	2	63
IN010003	3	94
IN010003	4	143

IN010003

5

179

[Mukesh Punia]

Are you using YLS data?

Try

 tab agemon if round==3, gen(agedm)

[George Ford]

What two age cohorts? Same for all rounds?

[Ayesha Afreen]

Are you using YLS data?

Try

 tab agemon if round==3, gen(agedm)

Yes I am using YLS data

[Ayesha Afreen]

Yes the cohort limit would change for all rounds but the mechanism stays the same. E.g., in Round 1 the min agemon in 6 and max is 12. So the cohort would be age=0 if agemon is 6-9 and age = 1 if agemon is 10-12

[George Ford]

but what if the agemon is different in another round? how do you cut it then? are you splitting at the mean?

[Ayesha Afreen]

Yes, at the mean

[George Ford]

egen agecut = mean(agemon), by(round)
egen agecohort = agemon > agecut