| Antenatal care [message #28685] |
Fri, 23 February 2024 00:23  |
Abid Monga
Messages: 2
Registered: February 2024
|
Member |
|
|
Hi,
I am trying to compute "among women with a live birth in the 5 years preceding the survey, percentage who received different types of antenatal care (ANC) during the pregnancy for their most recent live birth, by background characteristics, Jammu and Kashmir, 2019-20" using NFHS 5 data. For that, at the first instance, I have to compute the number of women covered under survey with recent birth. However, while calculating the same using SPSS there is inconsistency in data results and reports results in weighted figures. As the state report for Jammu and kashmir shows the number of women with recent births to be 4812 (weighted) and the figure computed from the data is 1379. There is no in-consistency in unweighted data.
The code used in SPSS is reproduced below. The results generated are also attached.
Where I am going wrong. Please help.
GET
FILE='D:\NFHS\Data\Individual Recode\IAIR7DFL.SAV'.
DATASET NAME DataSet1 WINDOW=FRONT.
USE ALL.
COMPUTE filter_$=(V024 = 1 & MIDX$1 = 1).
VARIABLE LABELS filter_$ 'V024 = 1 & MIDX$1 = 1 (FILTER)'.
VALUE LABELS filter_$ 0 'Not Selected' 1 'Selected'.
FORMATS filter_$ (f1.0).
FILTER BY filter_$.
EXECUTE.
COMPUTE WGT_LASR=V005 / 1000000.
EXECUTE.
FREQUENCIES VARIABLES=V013
/ORDER=ANALYSIS.
[Updated on: Fri, 23 February 2024 00:32] Report message to a moderator |
|
|
|
| Re: Antenatal care [message #28701 is a reply to message #28685] |
Tue, 27 February 2024 08:24  |
Bridgette-DHS
Messages: 3043
Registered: February 2013
|
Senior Member |
|
|
Following is a response from Senior DHS staff member, Tom Pullum:
Although the table is described in terms of numbers of women, it's easier to think in terms of numbers of births. Instead of the IR file, I would use the KR file, which is contains all the births in the past 5 years. The most recent birth has bidx=1.
I think the differences you are getting are due to the weights. You can calculate the number of women (a) with no weights, (b) with the national weight (v005), or (c) with the state weight (sweight).
The following lines are in Stata rather than SPSS but I think you can easily convert them to SPSS:
use "...IAKR7EFL.DTA", clear
keep if bidx==1
keep if hv024==1
tab v013
tab v013 [iweight=v005/1000000]
tab v013 [iweight=sweight/1000000]
The first tab command gives (a) 4,098 unweighted cases
The second tab command gives (b) 1,379 cases with the national weight
The third tab command gives (c) 4,812 cases with the state weight.
The state weights are the same as the national weight, multiplied by a state-specific constant to give a mean of 1 (times 1 million) in the IR file. Hope this helps.
|
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
