The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Topics » Domestic Violence » Multilevel Weights
Re: Multilevel Weights [message #13522 is a reply to message #13482] Mon, 13 November 2017 17:12 Go to previous messageGo to previous message
dflood011 is currently offline  dflood011
Messages: 7
Registered: October 2017
Member
Thank you to Tom for his incredibly helpful comments about the challenges in constructing multilevel models with DHS data, on this thread and others.

If I may ask a follow-up question to Tom:

I am interested in building a model to explore the association between maternal contraception and child stunting in a single-country DHS dataset. I originally thought about constructing a mixed model using cluster, household, and individual levels, but given Tom's comments about the unavailability of disaggregated household + cluster weights, I am thinking it would be better to carry out a classic regression in Stata using the examples set forth in Heeringa's text, "Applied Survey Data Analysis."

Yet I'm still drawn to the elegance of a hierarchical model, and when I search PubMed, I find numerous examples of papers in high-quality journals using mixed models for DHS data -- often asserting PSU weight-adjustment. (See examples below.) How is this possible? Are these analysts making errors?

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5291969/
http://onlinelibrary.wiley.com/doi/10.1111/mcn.12463/full
https://www.ncbi.nlm.nih.gov/pubmed/26378858
http://journals.plos.org/plosone/article?id=10.1371/journal. pone.0037905


 
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Previous Topic: Disability and spousal violence [Nigeria DHS 2018]
Next Topic: Domestic violence
Goto Forum:
  


Current Time: Fri Nov 22 17:06:21 Coordinated Universal Time 2024