The DHS Program User Forum      
Discussions regarding The DHS Program data and results
Home » Countries » Tanzania » Constructing indexes using Principal Component Analysis/Factor Analysis
Constructing indexes using Principal Component Analysis/Factor Analysis [message #4281] Thu, 30 April 2015 12:48 Go to next message
Robert is currently offline  Robert
Messages: 18
Registered: April 2015
Location: Dar Es Salaam, Tanzania


In my study, I am expected to pool data from DHS surveys in Tanzania (1992-2010) and I am aware that wealth index may not be used throughout due to variance in items used. However, I do not know how to go about it especially with usage of PCA or factor analysis....I would like to know how to interpret the coefficients just in case they are there. How many items should be included in determining the appropriate index? I am using STATA program. Thanks



[Updated on: Thu, 30 April 2015 12:50]

Report message to a moderator

Re: Constructing indexes using Principal Component Analysis/Factor Analysis [message #4285 is a reply to message #4281] Thu, 30 April 2015 18:34 Go to previous message
user-rhs is currently offline  user-rhs
Messages: 132
Registered: December 2013
Senior Member
Robert, the DHS program has extensive documentation on how the wealth index is constructed. Read the general comparative report (link here: rative-reports.cfm).

Variables used for specific countries and survey years can be found here: onstruction.cfm

There's a document by Shea Rutstein with the steps on creating the wealth index on that page, including some code in CSPro. My colleagues and I have used it before to guide us in creating a wealth index on some primary data that we collected. The "translation" to Stata is fairly straightforward.
Previous Topic: Variable for children education/ enrollment ratio/grade attainment
Next Topic: How to compute child mortality of different ages
Goto Forum:

Current Time: Sun Jan 21 13:26:58 Eastern Standard Time 2018