The DHS Program User Forum: Domestic Violence » Pakistan DHS 2017-18 "number of observations"

Home » Topics » Domestic Violence » Pakistan DHS 2017-18 "number of observations"

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

Pakistan DHS 2017-18 "number of observations" [message #23969]

Sun, 23 January 2022 05:57

Mason92
Messages: 33
Registered: December 2021

Member

For Pakistan DHS 2017-18, the variable 'd110a' (ever had bruises because of husband/partners actions) the 'tab' command in Stata gives output as No=741, Yes=292, and total=1,033. But when using the 'svy: prop d110a' command, the output gives the "number of observations" as 903. Is it because of the weighting of data? But, if I am not mistaken, the number of observations provided in the output is unweighted.
Thanks

Report message to a moderator

Re: Pakistan DHS 2017-18 "number of observations" [message #23982 is a reply to message #23969]

Mon, 24 January 2022 09:07

Bridgette-DHS
Messages: 3038
Registered: February 2013

Senior Member

Following is a response from DHS Research & Data Analysis Director, Tom Pullum:

If you use "tab" without specifying a weight and without svy, you will get unweighted frequencies. If you use svyset and then svy, then the frequencies will be weighted by whatever weight (v005 for the IR file) was specified in svyset.

Report message to a moderator

Re: Pakistan DHS 2017-18 "number of observations" [message #23998 is a reply to message #23982]

Wed, 26 January 2022 08:43

Mason92
Messages: 33
Registered: December 2021

Member

Thank you.

Report message to a moderator

Previous Topic:	Number of women who were ever married/in a relationship
Next Topic:	Burkina Faso DHS 2010, 9 records are coded as '3'

Goto Forum:

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Tue Apr 23 11:37:49 Coordinated Universal Time 2024