The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Data » Dataset use in Stata » Brant test
Brant test [message #3657] Fri, 23 January 2015 10:04 Go to next message
Asnake Yohannes is currently offline  Asnake Yohannes
Messages: 4
Registered: January 2015
Location: South Africa
Member


Dear,I am working on 2011 Ethiopian DHS of Women dataset. I need to run brant test to test parallel line assumption for outcome variable that ordered in to three categories.
I run the command on STATA 13 like this:

svy: ologit outcome variable independent variables ,or
brant
output was "brant- does not work with ologit models with weights"

It works when I excluded the command svy or

ologit outcome variable independent variables ,or
brant : works

May I have another alternative way to run brant test by considering svyset? Or I am looking for any advice concerning my problem.

Kindly regards,

Asnake.



Re: Brant test [message #3663 is a reply to message #3657] Fri, 23 January 2015 12:37 Go to previous messageGo to next message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 3016
Registered: February 2013
Senior Member
Following is a response from Senior DHS Stata Specialist, Tom Pullum

This is a Stata question and is not specific to DHS data, but I will offer a brief response.

It is not unusual to get a message to the effect that some part of svyset cannot be implemented. I suggest that you repeat svyset with fewer components (in this case, without the pweight component), and repeat the model, until you get a result, in this case the brant test. You then insert a statement in your write-up saying that you had to do this.

An argument can be made that tests should be unweighted anyway. Also, DHS surveys do not usually have enormous variation in the weights, and you will usually reach the same conclusion about significance or non-significance whether or not you use the weights.

I won't be surprised if this specific limitation is removed in some future version of Stata.


Re: Brant test [message #3686 is a reply to message #3663] Tue, 27 January 2015 03:14 Go to previous message
Asnake Yohannes is currently offline  Asnake Yohannes
Messages: 4
Registered: January 2015
Location: South Africa
Member

Dear Tom, really I appreciate for valuable guidance!
Previous Topic: Understanding variables in PR & IR files
Next Topic: Individual / Women recode data sets
Goto Forum:
  


Current Time: Thu Mar 28 12:59:06 Coordinated Universal Time 2024