| Re: v467 - Getting medical help for self - label [message #18214 is a reply to message #18212] |
Tue, 15 October 2019 12:42  |
Bridgette-DHS
Messages: 3043
Registered: February 2013
|
Senior Member |
|
|
Following is a response from DHS Research & Data Analysis Director, Tom Pullum:
You are right, this is a counter-intuitive sequence. I believe it originated as follows. Look at one of the questionnaires--for example the one for the Albania 2017-18 DHS, in Appendix E of the main report on that survey, question 1231. The response codes are 1 (big problem), 2 (small problem), and 3 (no problem). This sequence makes sense during interviewing. However, in the IR file for this survey (ALIR71FL.dta), the variables v467a etc. are coded as you describe:
. label list V467A
V467A:
0 no problem
1 big problem
2 not a big problem
I think a data processing decision was made that code 3 on the questionnaire should be changed to code 0 in the data files. Most categories with "no" or "none", for other variables, are assigned the value 0. I can confirm that the 0-1-2 codes are indeed correct, even though they are not consistent with the ordinality of the labels.
|
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
