The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Data » IPUMS Demographic and Health Surveys (IPUMS-DHS)  » IPUMS-DHS Tip #3: Use POPWT to estimate the number of people with a characteristic
IPUMS-DHS Tip #3: Use POPWT to estimate the number of people with a characteristic [message #15269] Fri, 22 June 2018 15:00 Go to next message
kingx025 is currently offline  kingx025
Messages: 95
Registered: August 2016
Location: Minneapolis. Minnesota
Senior Member
Most DHS users know that they need to use weights with DHS data to get reliable estimates of the proportion of people with a given characteristic. The weights included in the DHS samples adjust for factors such as oversampling small regions and weight up to the total number of people sampled by the survey.

Sometimes you'd like to know the total number (or count) of people in a country who have a given characteristic. Knowing the absolute number of people with some characteristic helps policymakers budget and plan for the future. To calculate the number of household members with a given characteristic, you can use the IPUMS-DHS variable POPWT, an expansion factor. For example, POPWT could be used as to find the number of people in Tanzania in 2015 that slept under an insecticide-treated bednet the night before the survey.

POPWT, the IPUMS-DHS expansion factor for household members, is calculated using the DHS household weight (HV005) and population figures from the United Nations Population Division. POPWT gives estimates subject to the same margin of error as the original household weight. POPWT is currently available for only household members (variables from the PR files) but will soon be available for women of childbearing age (the IR files) as well.

For more information about the construction of POPWT, see the User Note about the expansion factor:
https://www.idhsdata.org/idhs/population_weights.shtml


Miriam King


Dr. Miriam King
IPUMS-DHS Project Manager (www.idhsdata.org)
Re: IPUMS-DHS Tip #3: Use POPWT to estimate the number of people with a characteristic [message #27495 is a reply to message #15269] Tue, 22 August 2023 04:42 Go to previous messageGo to next message
gonzalo_barreix is currently offline  gonzalo_barreix
Messages: 3
Registered: August 2023
Member
Dear Miriam,
Very useful post about expansion factor.
I want to expand the Mozambique Malaria Indicators Survey 2018 for children (under 5 children) sample. I checked the link that you left that explains how the POPWT was generated for some DHS surveys but I still have doubts about how would the KIDWT would be constructed. May be you can clarifiy on this.
For expanding under5 children should I:

1)Use the KR file of the MIS 2018, and only with the alive kids compute:
KTWT= (v005/(sum of 2018 under5 v005))* (2018 Under5 Mozambique Population) ?
2)Use the PR file of the MIS 2018 and compute:
KTWT= (hv005/(sum of 2018 total population hv005))* (2018 Total Mozambique Population) ?

How is the proper way to do this?

Or maybe if it is easier to explain, it could be useful if you could clarify me how to exactly the KDWT is computed using an example as the one for POPWT shown in the link you left.


Thanks in advance
Gonzalo Barreix
Re: IPUMS-DHS Tip #3: Use POPWT to estimate the number of people with a characteristic [message #27497 is a reply to message #27495] Tue, 22 August 2023 12:42 Go to previous messageGo to next message
gcorcora is currently offline  gcorcora
Messages: 2
Registered: June 2019
Location: Minneapolis, Minnesota
Member
Dear Gonzalo Barreix,

I can answer your question for Miriam, as I wrote the code for POPWT.

For expanding children under 5 in the Mozambique MIS 2018 survey, you will want to use the first of the two methods you mentioned.

Specifically, using the MIS 2018 Mozambique KR file compute:

KTWT= (v005/(sum of 2018 under age 5 in sample))*(2018 under age 5 Mozambique population)

I hope this helps! Please respond here if you have any further questions

Gretchen Corcoran


Gretchen Corcoran IPUMS-DHS Data Analyst III (https://idhsdata.org)
Re: IPUMS-DHS Tip #3: Use POPWT to estimate the number of people with a characteristic [message #27499 is a reply to message #27497] Wed, 23 August 2023 03:34 Go to previous messageGo to next message
gonzalo_barreix is currently offline  gonzalo_barreix
Messages: 3
Registered: August 2023
Member
Thanks for the replay Gretchen!
A last question:
Also using the 2018 Mozambique MIS KT file, if I would like to expand the children population of age 0, how would be the correct way?

1- KTWTi= (v005i/(sum of 2018 under age 5 in sample))*(2018 under age 5 Mozambique population) for childen i age 0
2- KTWTi= (v005i/(sum of 2018 age 0 in sample))*(2018 age 0 Mozambique population) for childen i age 0

Best,
Gonzalo Barreix
Re: IPUMS-DHS Tip #3: Use POPWT to estimate the number of people with a characteristic [message #27515 is a reply to message #27499] Mon, 28 August 2023 12:53 Go to previous messageGo to next message
gcorcora is currently offline  gcorcora
Messages: 2
Registered: June 2019
Location: Minneapolis, Minnesota
Member
Hi Gonzalo,

Great question! We used UN population numbers to calculate the population counts, which are only available in 5 year increments. Unfortunately that means there is no number available for population under age 1, so the second method (v005i/(sum of 2018 age 0 in sample))*(2018 age 0 Mozambique population) for children i age 0 isn't feasible unless you were able to get the numbers elsewhere. Thus the 1st method, (v005i/(sum of 2018 under age 5 in sample))*(2018 under age 5 Mozambique population) for children i age 0 is the option that would work with the data available.

Let me know if the above isn't clear or you have any further questions.

Thanks!

Gretchen Corcoran


Gretchen Corcoran IPUMS-DHS Data Analyst III (https://idhsdata.org)
Re: IPUMS-DHS Tip #3: Use POPWT to estimate the number of people with a characteristic [message #27548 is a reply to message #27515] Fri, 01 September 2023 04:10 Go to previous messageGo to next message
gonzalo_barreix is currently offline  gonzalo_barreix
Messages: 3
Registered: August 2023
Member
Dear Gretchen,
Thanks again for your answer. So, in the case I finde the under 1 age population, the 2nd method would be better? Because I actually have that information, because there is also a disagregation by 1 year of age.

Thanks,
Gonzalo
Re: IPUMS-DHS Tip #3: Use POPWT to estimate the number of people with a characteristic [message #27610 is a reply to message #27548] Mon, 11 September 2023 13:14 Go to previous message
xgecemx is currently offline  xgecemx
Messages: 1
Registered: September 2023
Member
Thanks for sharing this helpful tip! Using the POPWT variable to estimate absolute numbers is a great idea that I haven't tried before. Knowing the total number of people with certain characteristics will definitely help with planning and resources allocation.

I appreciate you linking to the user note that explains how POPWT is calculated. It's useful to understand that it relies on the household weights and UN population data. Knowing it has a similar margin of error to the household weights gives me confidence in using it.

Being able to extend the analysis beyond just proportions to real counts of people will open up new possibilities for the research I can conduct with DHS data. I look forward to the upcoming expansion of POPWT to women of childbearing age. Thanks again calendar 2024 with holidays for sharing this invaluable tip!https://www.typecalendar.com/printable-yearly-calendarhttps://www.typecalendar.com/household-budget.html

[Updated on: Wed, 13 September 2023 02:13]

Report message to a moderator

Previous Topic: IPUMS-DHS Nepal district data missing?
Next Topic: Comprehension of IPUMS
Goto Forum:
  


Current Time: Sun Dec 8 01:09:23 Coordinated Universal Time 2024