Anemia variable in MR [message #18916] |
Tue, 17 March 2020 17:50 |
nagar002
Messages: 9 Registered: April 2018
|
Member |
|
|
Hi All,
I am working with men's Anemia variable for many DHS (across countries and time) pooled together. I have the following questions. Any help is much appreciated.
1. There are many men in the men's recode file who have missing haemoglobin information. As I understand from the description in the country report, the selection of men for interview and for anemia testing was random and the low response rate for men's anemia is not the result of some sample selection criterion. Is that correct?
2. If regression analysis (with anemia as dependent variable) is being done separately for men and women in a couple {i.e., subsample of men (from MR file) and women (from IR file) who can be matched to form a couple}, what sample weights should be used?
Thank you!
|
|
|