| Re: Malawi 2010: Recent Testing for HIV [message #23540 is a reply to message #23531] |
Mon, 04 October 2021 15:56   |
Bridgette-DHS
Messages: 3230
Registered: February 2013
|
Senior Member |
|
|
Following is a response from DHS Research & Data Analysis Director, Tom Pullum:
If you do "tab v781,m" you will see that ALL women have a response for v781, although for 121 women (unweighted) the code is "9", which is not valid. This code sometimes should probably be interpreted as a combination of "don't know" and "refusal". The survey-specific variable s1312 that appears to be a substitute for the usual v826 is harder to figure out. It is coded NA (a dot in Stata) for 12,117 women, including 6,143 women who said they had had an HIV test. I suggest that you look for the skip pattern in the questionnaire, which appears as an appendix in the final report. For example, page 544 in the report shows some skips related to Q1311, which is the basis of v781. It shows that all women are asked whether they had a test. There must be something similar leading to s1312, but the correspondence between s1312 and the usual v826 is not clear. There should also be some clarification in chapter 15 of the final report.
|
|
|
|
Members
Search
Help
Register
Login
Home
