The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Topics » General » Education, Turkey DHS
Re: Education, Turkey DHS [message #3455 is a reply to message #3449] Fri, 12 December 2014 18:50 Go to previous messageGo to previous message
Trevor-DHS is currently offline  Trevor-DHS
Messages: 805
Registered: January 2013
Senior Member
1) You are currently working with the HR (Household Recode) file, but if you are interested in statistics for the members of the household, and not for the household you probably should be working with the Persons Recode (PR) file. The difference as follows:

HR:
<..hhdata..><..data person for person $01..><..data person for person $02..><..data person for person $02..> ...
and there is one record per household.

PR:
<..hhdata..><..data person for person $01..>
<..hhdata..><..data person for person $02..>
<..hhdata..><..data person for person $03..>
...
and there is one record for each person in the household.

So, when you look at frequencies of HV106$01 you are looking at frequencies of the education data for the first person listed in each household, and HV106$02 provides the frequencies for the second person listed in each household, etc.
If you want frequencies for all household members together, use the PR file and use HV106 (without the $01, $02, etc.).

2) There is a mistake in the creation of HV106 for the Turkey DHS 2008. I would recommend looking at the original variables which can be found in SH16A and SH16B.

3) The Turkey 2008 dataset has some recoding problems in HV106, HV107, and HV108. However in general the variables have the following meaning:

HV106 - Highest level of education attended (note the word attended).
HV107 - Highest grade completed at the level in HV106 (note the word completed). A person can have attended a level, but never completed a grade at that level. For example, somebody who started secondary school but completed no grade at that level would be coded 2 on HV106 and 0 on HV107.
HV108 - Number of years of schooling, constructed from the information in HV106 and HV107. This combine the levels and grades to produce a total number of years of schooling, with an assumption as to the number of years of schooling received in lower levels. In general the calculation is as follows:

HV108 = HV107 if HV106==1 (Primary)
HV108 = HV107+p if HV106==2 (Secondary) (where p is a constant for the typical number of years of schooling at the level below, i.e. primary )
HV108 = HV107+p+s if HV106==3 (Higher) (where p is a constant for the typical number of years of schooling at primary level, and s is a constant for the typical number of years of schooling at secondary level).



 
Read Message
Read Message
Read Message
Read Message
Previous Topic: Absence and Employment
Next Topic: data mismatches
Goto Forum:
  


Current Time: Wed Jan 15 10:25:49 Coordinated Universal Time 2025