Home » Countries » India » Help with fertility preference variables v627, v628 and v629 and v613
| Re: Help with fertility preference variables v627, v628 and v629 and v613 [message #6737 is a reply to message #6733] |
Mon, 06 July 2015 18:34   |
Bridgette-DHS
Messages: 3110
Registered: February 2013
|
Senior Member |
|
|
Following is a response from Senior DHS Stata Specialist, Tom Pullum:
Decisions such as this are really judgment calls. That is, there is not a single "correct" way to handle them.
In a quantitative analysis of the responses, a non-numeric response must be coded "missing" or in some other way be excluded from calculations such as a mean. However, you do need to say something about the prevalence of such responses. You could report the overall percentage of responses that are non-numeric and the percentage in major categories such as place of residence, wealth quintile, etc., and then exclude them from calculations.
If you are reporting, say, the percentages who would prefer a boy or would prefer a girl, then the percentage with no preference is important and could be a separate category. You could consolidate the non-numeric responses with that category. I suggest that you look at an old DHS report by Fred Arnold ( https://www.dhsprogram.com/publications/publication-CS23-Com parative-Reports.cfm) for some suggestions.
|
|
|
|
Current Time: Fri Jul 12 00:21:25 Coordinated Universal Time 2024
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
