The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Countries » Bangladesh » Regarding categories s823
Regarding categories s823 [message #14173] Thu, 01 March 2018 05:23 Go to next message is currently offline
Messages: 1
Registered: March 2018
I have a sample of currently married women aged 15-25 in my study. One of my outcome variables are s283 "who has the final say on childs Health". First question-does this variable include children in the household or only the respondents own child.

One of the response categories are "other" for this variable which a large number of women who reported "0 number of everborn children" have reported but also women who reported several number children everborn. Is it right to assume that this category then include answer that were not applicable as well as maybe that no response choice were applicable?

Re: Regarding categories s823 [message #14183 is a reply to message #14173] Fri, 02 March 2018 05:49 Go to previous message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 1763
Registered: February 2013
Senior Member
Following is a response from Senior DHS Stata Specialist, Tom Pullum:

Based on the files you have downloaded, you are apparently referring to the Bangladesh 2014 survey. Forum users are asked to say what survey their questions refer to. The variable is s823 (not s283). A more standard age interval would be 15-24, rather than 15-25. Now to your questions! This is asked during the interview with women, about each of the woman's own children born in the past five years, regardless of whether the child is still alive (see b5) or is living with the mother (see b9). I will list the category labels below. Out of 7,886 such children, 40 have the response "other" and 111 have a Not Applicable code ("." In Stata). To get the unweighted frequencies in Stata, enter "tab s823 b9,m". I agree with you that the interpretation of these two codes is problematic. Ideally there would be no children with either one. If I were going to analyze these data, I would probably combine categories 5 and 6, as "someone else", and then drop (omit) the cases coded ".". Fortunately the numbers are small and the effect of doing this or something else will be negligible.

. label list S823
1 respondent
2 respondent & husband, partner jointly
4 husband, partner
5 someone else
6 other

Previous Topic: Query regarding Bangladesh Community data file
Next Topic: Children from same household
Goto Forum:

Current Time: Sun Mar 29 14:28:51 Eastern Daylight Time 2020