| Re: TV exposure decline and v159 coding error? [message #25850 is a reply to message #25844] |
Mon, 19 December 2022 11:04  |
Bridgette-DHS
Messages: 3230
Registered: February 2013
|
Senior Member |
|
|
Following is a response from Senior DHS staff member, Tom Pullum:
If you want to assess media exposure, you will get more complete information from the IR files (for women) and the MR files (for men) than from the CR files, which are restricted to women and men who name each other as partners. The CR file will be useful if your main interest is in comparing the responses of women and men who are partners.
The NFHS-5 data files have the same value labels for (m)v157-159 as the NFHS-4 files. This is a mistake, because the response categories changed. In the NFHS-4 files, the label is correct:
. label list V159
V159:
0 not at all
1 less than once a week
2 at least once a week
3 almost every day
However, the NFHS-5 questionnaire omitted the last response category. In effect, categories 2 and 3 were combined into a new category 2. Category 3 should not have been included in the NFHS-5 label.
The data files show increases, for all three questions and for both women and men, in the percentage "not at all" or "less than once a week". There is an increase in the sum of the percentages with codes 2 and 3. We too would not have expected this but it appears to be correct. Perhaps this is due to increased use of cell phones? Let us know if you have other questions.
|
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
