The DHS Program User Forum: Kenya

Home » Countries » Kenya » Proportions

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

Proportions [message #10603]

Tue, 16 August 2016 06:22

Khanjila
Messages: 7
Registered: July 2016

Member

I would like to create the proportion of households in a cluster that have access to electricity.

This is what I have done:

bys hv001: gen proportion= sum(hv206)/_N
bys hv001: replace proportion= proportion[_N]

However, I realized that since there can be duplicate households in a cluster as shown below. (This is because my study is focusing on impact of electricity on school attendance and so there can multiple children in a household who attend school)

hv001	hv002
2	76
2	76
2	76
3	7
3	7
3	23
3	39
3	39
3	39
3	47
3	56
3	80
3	80

I tried using

duplicates drop household number

Before proceeding with the previous syntax. But my sample reduced drastically to around 147 from about 9000.

Is there something I am doing wrong?
Or is there a better method of getting the proportion of households in a cluster with access to electricity?

Report message to a moderator

Re: Proportions [message #10608 is a reply to message #10603]

Tue, 16 August 2016 10:40

Bridgette-DHS
Messages: 3230
Registered: February 2013

Senior Member

Following is a response from Senior DHS Stata Specialist, Tom Pullum:

I suggest that you construct the proportion of households in the cluster that have electricity with a collapse command. You could do this from the HR file or from the PR file, selecting one record per household. Then merge that back with the PR file or whatever other file you are using. The lines below show how to construct this proportion. You do not need to use the weights because the weights are constant within clusters. Let me know if this is not what you need.

use KEPR70FL.dta, clear
keep if hvidx==1
keep hv001 hv206
tab hv206
collapse (mean) hv206, by(hv001)
rename hv206 electricity
label variable electricity "Prop. of hh in cluster with elect."
sort hv001
* next merge this file with whatever other file you are using

Report message to a moderator

Previous Topic:	Missing variables in data set
Next Topic:	House hold sample weight

Goto Forum:

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Mon Oct 20 11:55:38 Coordinated Universal Time 2025