| Missing information in the variable v803 (length of Interview) [message #15566] |
Mon, 13 August 2018 08:11  |
nishant.raks@gmail.com
Messages: 1
Registered: August 2018
Location: India
|
Member |
|
|
Greetings !!
Using Individual Recode file, I was estimating the length of the interview in NFHS-4 from variable v803 (Lenght of the interview in minutes) and noticed that out of 6,99,686 weighted cases, the timing of interview is available just for 121120 cases and rest 578,566 (82.69%) cases are missing.
I looked at the recode DHS to ascertain if I am doing right or wrong. Please, could you bring some more insight to it as to why huge missing numbers are coming when the CAPI was used this time for the survey?
I would be highly grateful to your kind gesture.
Yours Sincerely,
Rakesh
[Updated on: Mon, 13 August 2018 08:14] Report message to a moderator |
|
|
|
| Re: Missing information in the variable v803 (length of Interview) [message #15592 is a reply to message #15566] |
Fri, 17 August 2018 14:52  |
fredarnold
Messages: 45
Registered: March 2014
|
Member |
|
|
|
The full woman's interview was only asked of women in a subsample of approximately 15% of households (the state module). In the remaining households, only the short questionnaire was canvassed, without the sections on sexual life, husband's background and woman's work, HIV/AIDS and other sexually transmitted infections, and household relations. v803 (length of interview) includes only women who were administered the long questionnaire (the state module). Since the time at the end of the interview is recorded for all women in Question 1140, you can calculate the length of interview for women who were administered only the short questionnaire directly from the dataset.
|
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
