The DHS Program User Forum: Weighting data » weighting issues of multilevel modelling using the DHS survey data with multiple-stage sampling

Home » Data » Weighting data » weighting issues of multilevel modelling using the DHS survey data with multiple-stage sampling

Show: Today's Messages :: Show Polls :: Message Navigator

Re: weighting issues of multilevel modelling using the DHS survey data with multiple-stage sampling [message #15713 is a reply to message #15703]

Wed, 05 September 2018 08:13

Bridgette-DHS
Messages: 3038
Registered: February 2013

Senior Member

Following is a response from Senior DHS Specialist, Tom Pullum:

I believe there is a problem with your approach, because it does not take into account that a cluster's probability of selection is not a constant within each stratum, but is proportional to the size of the cluster. Without knowing the size of the cluster, you cannot calculate its probability of selection.

We at DHS hope to prepare guidelines for dealing with the requirement for separate weights for multi-level models. I don't think the results will be very sensitive to the allocation (as your comparison suggests) but we do need something better than just wij and 1. Your approach is worth pursuing but I think the best that can possibly come out of it will be a better approximation, because we simply don't know the size of the cluster.

Report message to a moderator

[Message index]

		weighting issues of multilevel modelling using the DHS survey data with multiple-stage sampling By: YUJP on Mon, 03 September 2018 05:19
		Re: weighting issues of multilevel modelling using the DHS survey data with multiple-stage sampling By: Bridgette-DHS on Wed, 05 September 2018 08:13
		Re: weighting issues of multilevel modelling using the DHS survey data with multiple-stage sampling By: YUJP on Wed, 05 September 2018 14:32
		Re: weighting issues of multilevel modelling using the DHS survey data with multiple-stage sampling By: Bridgette-DHS on Wed, 05 September 2018 19:38
		Re: weighting issues of multilevel modelling using the DHS survey data with multiple-stage sampling By: YUJP on Thu, 06 September 2018 09:00

Previous Topic:	weighting and creating SVY command for multilevel analysis of two level of data SPA -NEPAL
Next Topic:	number of women

Goto Forum:

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Wed Apr 24 04:37:32 Coordinated Universal Time 2024