IPV and multi-level weights India data [message #24282] |
Tue, 12 April 2022 15:15 |
ndehingia
Messages: 3 Registered: January 2019
|
Member |
|
|
Hi,
I am running an analysis with India DHS data. My outcome variable is IPV/domestic violence, and my independent variable is a district-level estimate of climate-related distress. I am using a multi level model to account for the clustering of observations within districts. I have followed the guidance from DHS MR 27 to calculate the multi-level domestic violence weights. However, in India, the level-2 weights are calculated at the cluster/PSU level. Each district consists of multiple clusters/PSUs in India. Can you please provide some guidance on what weights I should be using for my models with a district level variable? And for level 1, can I use the domestic violence weight on its own (d005/1000000)? Or should I use dwt1, calculated using the procedure outlined in MR 27?
Thank you very much.
|
|
|
|