The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Data » Weighting data » NDHS 2011
NDHS 2011 [message #8159] Mon, 31 August 2015 02:05 Go to next message
Jeevana is currently offline  Jeevana
Messages: 4
Registered: August 2015
Location: Canberra
Member
Hi,
I am looking for antenatal care and skilled birth attendants in Nepal. Number of more recent birth five year before survey is 4079.
I ran
generate wgt=v005/1000000

I did not get

tab antenatal[iweight=wgt]
Is this for weighted percentage and number of observation?

how do I use svy syntax?

I am using mothers with most recent birth (4079), Do I need to apply subpop in my syntax?

svy, subpop .........


Please help.

Thank you
Re: NDHS 2011 [message #8169 is a reply to message #8159] Tue, 01 September 2015 12:03 Go to previous message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 3017
Registered: February 2013
Senior Member
Following is a response from Senior DHS Stata Specialist, Tom Pullum:

You do not need to use subpop as part of the svyset command. I prefer to include it within the estimation commands, for example with "if v025==2" if you want to restrict to rural women. You CAN identify the subpopulation in the svyset command, but you then run the risk of forgetting that all subsequent commands that include svy will be restricted to that subpopulation.

Information about antenatal care and skilled birth attendants comes from specific variables. In the IR files the main variables are m2* and m3*, respectively, with indices or subscripts such as _1, _2, through _6, and * is a letter, a through n. The index for the most recent birth is _1. You can either use the IR file or the KR file; in the KR file the cases are births in the last five years and the most recent birth has bidx=1.

In the Nepal 2011 survey, the data on antenatal care and skilled birth attendants are not restricted to the most recent birth. I would encourage you to use ALL births in the past five years, not just the most recent one. Please see Methodological Report #14 ( https://www.dhsprogram.com/publications/publication-MR14-Met hodological-Reports.cfm) for a description of how and why the most recent birth is not actually representative of ALL births.

I suggest that you open NPKR60FL.dta and then enter

set more off
describe m2a-m2n m3a-m3n
tab1 m2a-m2n m3a-m3n

You will see that there are many binary (yes/no) variables, one for each category of antenatal care or birth attendants. Several potential categories were not actually used in this survey. You will probably want to construct recodes. Some information about possible recodes is provided in Analytical Study #46 ( https://www.dhsprogram.com/publications/publication-AS46-Ana lytical-Studies.cfm) . For constructing recodes, checking variables, etc., it is easier not to use svy or even the weights. Those are important during analysis but for setting things up.

If you are trying to match a specific table in a DHS report, and have problems, you can ask about that.

Previous Topic: Education
Next Topic: Weighting in regional subsample
Goto Forum:
  


Current Time: Fri Mar 29 09:08:47 Coordinated Universal Time 2024