Age group in a longitudinal survey

[Raphael George]

Hi I want to restrict my analysis sample to those individuals who were between 5 and 14 in wave 1 of a longitudinal survey. The wave 1 was conducted in 2004-05 and wave 2 was conducted in 2011-12. How can I do that?

 dataex HHPBASE SURVEY age

----------------------- copy starting from the next line -----------------------


	
	
* Example generated by -dataex-. To install: ssc install dataex
clear
input double HHPBASE byte SURVEY int age
101020101001 1 50
101020101002 1 45
101020101003 1 22
101020101005 1 13
101020101006 1  7
101020102001 2 76
101020102001 1 69
101020102003 1 35
101020102003 2 42
101020102004 2 40
101020102004 1 33
101020102005 1 14
101020102005 2 20
101020102006 2 19
101020102006 1 12
101020102007 1  6
101020102007 2 13
101020102008 2 37
101020102008 1 31
101020102009 1 26
101020102009 2 33
101020102010 2 13
101020102010 1  6
101020102011 1  1
101020102011 2  8
101020102012 1 24
101020102012 2 31
101020102013 1 19
101020102013 2 26
101020102014 1  1
101020102014 2  8
101020102015 1 59
101020102015 2 65
101020103001 1 37
101020103001 2 45
101020103002 2 43
101020103002 1 36
101020103003 2 23
101020103003 1 15
101020103004 2 20
101020103004 1 13
101020103005 1 10
101020103005 2 17
101020103006 2 14
101020103006 1  7
101020103007 1  5
101020103007 2 12
101020104001 1 50
101020104001 2 57
101020104002 1 40
101020104002 2 47
101020104003 1 13
101020104003 2 20
101020104004 2 14
101020104004 1  7
101020104005 2 13
101020104005 1  6
101020105002 2 50
101020105002 1 43
101020105003 2 38
101020105003 1 28
101020105006 1 20
101020105006 2 27
101020105007 2 23
101020105007 1 16
101020105008 1 11
101020105008 2 18
101020105009 1  9
101020105009 2 16
101020105010 1  6
101020105010 2 12
101020106001 1 65
101020106001 2 72
101020106012 2 39
101020106012 1 33
101020106013 1 30
101020106013 2 37
101020106014 1 27
101020106014 2 33
101020106015 1 20
101020106015 2 27
101020107001 1 45
101020107001 2 52
101020107002 1 40
101020107002 2 42
101020107003 1 19
101020107003 2 26
101020107004 2 24
101020107004 1 17
101020107006 1 13
101020107006 2 20
101020107007 2 18
101020107007 1 11
101020107008 2 16
101020107008 1  9
101020107009 2 15
101020107009 1  8
101020108001 1 35
101020108001 2 42
101020108002 2 37
end
label values SURVEY SURVEY
label def SURVEY 1 "IHDS1 1", modify
label def SURVEY 2 "IHDS2 2", modify
------------------ copy u

[George Ford]

egen id = group(HHPBASE)  //HHPBASE is long so this helps you see it in the data; not required.
egen agewave1 = min(age), by(id)  //could use HHPBASE rather than id in by

now you can restrict on agewave1.

sample is not balanced.

[Carlo Lazzaro]

Raphael:
as an aside to George's helpful advice, you can select the observations you're interested in via:

gen counter=1 if age >=4 & age<=14 & SURVEY==1

Then you need an individual -id- to move forward.

[Raphael George]

Thank you very much