Nepal dataset 2011 [message #21738] |
Mon, 14 December 2020 18:00 ![Go to next message Go to next message](/theme/default/images/down.png) |
vpatil
Messages: 9 Registered: March 2019
|
Member |
|
|
I am using the questions on whether men and women justify violence across different surveys. I found a consistency in numbers across 2001, 2006, and 2016 datasets but the numbers in 2011 are, abnormally, low. Was there a change in the way questions were asked or the way the answers have been compiled?
Also, the number of men who were asked the questions on justification is very small (110) for the men dataset and 54 in the couple 2011 dataset. Is there any reason why these numbers are so small?
|
|
|
|
|
Re: Nepal dataset 2011 [message #21815 is a reply to message #21807] |
Wed, 23 December 2020 06:54 ![Go to previous message Go to previous message](/theme/default/images/up.png) |
Bridgette-DHS
Messages: 3151 Registered: February 2013
|
Senior Member |
|
|
Following is a response from DHS Research & Data Analysis Director, Tom Pullum:
I can only give a generic answer, because I don't have time to go over the specific skip pattern. You can "back out" of the skips and filters to get at a larger denominator. For an analogy, if a question "have you had any births" is followed by "how many" for those who answer "yes", then the denominator would be the number who were asked the first question. If they answered it with "no" then you assign 0 to "how many". Usually, that kind of skip is absorbed during data processing, to produce a single variable, for example the number of children ever born, including 0. But sometimes it is not done during data processing and you have to do a recode yourself.
|
|
|