The DHS Program User Forum: Weighting data » Weighting district-level data

Home » Data » Weighting data » Weighting district-level data

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

Re: Weighting district-level data [message #8791 is a reply to message #8779]

Fri, 18 December 2015 14:43

amira.elshal.1@city.ac.uk is currently offline

amira.elshal.1@city.ac.uk
Messages: 52
Registered: April 2013

Senior Member

1) When I try to re-calculate the district-by-survey-round indicators again using the weights, I get the same results for each district (as in the unweighted case). I think this is probably due to the way I create/calculate the district-level indicator (below):

collapse (sum) women_using_modern_methods (count) women_interviewed, by(District)
gen %women_currently_using_modern_methods=(sum_of_women_using_mo dern_methods/count_of_women_interviewed)*100

Note: I denote women_using_modern_methods by 1 before calculating the district-level indicator.

When I use the weights as you have kindly suggested (below), the district-level indicators are the same as in the unweighted case:

gen myweight=v005/1000000
collapse (sum) women_using_modern_methods (count) women_interviewed [pweight=myweight], by(District)
gen %women_currently_using_modern_methods=(sum_of_women_using_mo dern_methods/count_of_women_interviewed)*100

I do NOT simply create the district-level indicators as means of observations of individuals in these districts.

2) You made this point quite clear. Thank you.

3) I fail to do the clustering when I run the diff-in-diff regression below:

diff district_indicator, t (treated) p (t), cluster(District)

Stata 12 gives me the return code 198 of invalid syntax. Shall the 'District' variable (or whichever variable I use to cluster) be written down in a particular way?

4) The PSU is not the district in my case. I spatially join the displaced cluster locations of women to the GIS polygon data of Egypt's district boundaries. i.e. I allocate each woman to the relevant district.

Report message to a moderator

[Message index]

		Weighting district-level data By: amira.elshal.1@city.ac.uk on Thu, 17 December 2015 04:35
		Re: Weighting district-level data By: Reduced-For(u)m on Thu, 17 December 2015 17:32
		Re: Weighting district-level data By: amira.elshal.1@city.ac.uk on Fri, 18 December 2015 11:55
		Re: Weighting district-level data By: Reduced-For(u)m on Fri, 18 December 2015 12:47
		Re: Weighting district-level data By: amira.elshal.1@city.ac.uk on Fri, 18 December 2015 14:43
		Re: Weighting district-level data By: Reduced-For(u)m on Fri, 18 December 2015 15:37
		Re: Weighting district-level data By: amira.elshal.1@city.ac.uk on Sat, 19 December 2015 06:11
		Re: Weighting district-level data By: Reduced-For(u)m on Mon, 21 December 2015 16:08
		Re: Weighting district-level data By: amira.elshal.1@city.ac.uk on Thu, 24 December 2015 04:00
		Re: Weighting district-level data By: Reduced-For(u)m on Mon, 28 December 2015 17:53
		Re: Weighting district-level data By: amira.elshal.1@city.ac.uk on Mon, 21 December 2015 05:48
		Re: Weighting district-level data By: amira.elshal.1@city.ac.uk on Mon, 21 December 2015 16:51

Previous Topic:	Collection of Strata Information
Next Topic:	Descriptives and chisq

Goto Forum:

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Thu Dec 26 14:44:08 Coordinated Universal Time 2024