India 2019-2021 Missing Data [message #29014] |
Wed, 10 April 2024 10:31 ![Go to next message Go to next message](/theme/default/images/down.png) |
behayes4
Messages: 24 Registered: February 2020
|
Member |
|
|
For many of the demographic variables - education, religion, married, ae, number of children - there is absolutely no missing data in a sample of 724,000+ respondents. Is this correct?
|
|
|
Re: India 2019-2021 Missing Data [message #29022 is a reply to message #29014] |
Wed, 10 April 2024 15:28 ![Go to previous message Go to previous message](/theme/default/images/up.png) ![Go to next message Go to next message](/theme/default/images/down.png) |
Bridgette-DHS
Messages: 3151 Registered: February 2013
|
Senior Member |
|
|
Following is a response from Senior DHS staff member, Tom Pullum:
These variables do not have any cases with a dot (.), the code for Not Applicable, because they applied to everyone. Most of them do have a few "don't know" cases.
A few variables, such as age and number of children, are not allowed to have "don't know" responses, because that would essentially mean that the case would be thrown out, unusable, wasted.
There are indeed some cases in which a household, woman, or man is selected but is dropped from the sample because they cannot be located or simply refuse to be interviewed. Surprisingly few. All DHS final reports describe the response rate in Chapter 1. In the NFHS-5 report, see section 10 of Chapter 1, page 12.
|
|
|
|
Re: India 2019-2021 Missing Data [message #29749 is a reply to message #29746] |
Tue, 30 July 2024 14:58 ![Go to previous message Go to previous message](/theme/default/images/up.png) |
Bridgette-DHS
Messages: 3151 Registered: February 2013
|
Senior Member |
|
|
Following is a response from Senior DHS staff member, Tom Pullum:
In the HR file, which has one record per household, there is a variable hv044, "household selected for domestic violence module". It is incorrectly coded "1" for EVERY HOUSEHOLD. This variable is defective. I believe that hv027, "household selected for male interview" was actually used to select for the DV module as well as the male interview.
You are looking at v044, in the IR file, which has the same label as hv044. It has the following categories:
. label list V044
V044:
0 woman not selected
1 woman selected and interviewed
2 woman selected, but privacy not possible
3 woman selected, but not interviewed
Code 3, as you say, has 420,149 cases. I believe these cases arise from the defective coding of hv044. Someone thought that hv044 was correctly labeled and coded, and during data processing they used it to determine eligibility. However, those women are in households that should have been coded 0 on hv044 and NA on v044.
There is no problem here, just a coding error.
|
|
|