| Pooled weights [message #12281] |
Thu, 20 April 2017 09:15  |
denisshek
Messages: 2
Registered: April 2017
|
Member |
|
|
I am doing an anlaysis using merged data set from different countries and different survey years. I understand that i have to denormalize the weights (v005) after having gone through previous post in the forum, however, i want to be sure i am doing the right thing.
Can i de-normalize such that "the weights in each survey sum up to 1" and then multiply these weights by the population of interest in this case women aged 15-49 years at the time of survey such that each country in total gets weight equal to the population size of women aged 15-49 years at the time of survey.
Thanks.
|
|
|
|
| Re: Pooled weights [message #12283 is a reply to message #12281] |
Fri, 21 April 2017 07:42  |
Bridgette-DHS
Messages: 3214
Registered: February 2013
|
Senior Member |
|
|
Following is a response from Senior DHS Stata Specialist, Tom Pullum:
Yes, you can do that. Such weights are sometimes called "inflation" weights. The main problem with that approach, rather than, say, giving the same weight to each country, is that large countries will outweigh small countries and dominate the pooled estimates. I suggest that you try this approach and then see what happens to the estimates when you add or drop various countries--that is, do a sensitivity analysis. There is also a conceptual problem, that the surveys refer to different time points and the countries almost certainly do not comprise a standard geographic region or subregion. However, this is a judgment call by the researcher.
|
|
|
|
|
|
|
|
|
|
|
|
| Re: Weights in country-year level regressions [message #16189 is a reply to message #16188] |
Mon, 19 November 2018 13:39 |
Bridgette-DHS
Messages: 3214
Registered: February 2013
|
Senior Member |
|
|
Following is another response from Senior DHS Stata Specialist, Tom Pullum:
Even if you do not collapse the surveys, you still have to deal with the variation in sample sizes. Surveys with larger samples will tend to dominate. I prefer to revise v005, multiplying by a survey-specific factor. If, say, your combined data file with k surveys has N cases, you would revise the weights so that the weighted number of cases for each survey is the same, N/k. However, even that approach is vulnerable to criticism. Stata code to do this is posted.
Here at DHS, we usually do not pool surveys. When we combine successive surveys from one country into a single data file, that's usually just to make it easier to describe trends. There have been a few times when we pooled surveys because that was the only way to get enough cases, for some rare outcome. The main reason for not pooling is that the reference population, of which the data are supposed to be representative, is too difficult to define. But other users may have a different perspective.
|
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
