survey weight and calibration svy

[Dayla Spassio]

Hi all,

I am using data from a survey for empirical work.
So far i have used the following command to consider the weights :

svyset id [pweigth=weight]

but i have read that the weights are calibrated, (here are the words of the methodological note)

the system of weights (variable WEIGHT in the archive) must be used on all 3,000 observations and not survey by survey. The system has been calibrated to the information on: gender, age group (less than 35, from 35 to 44, from 45 to 64, over 64), geographical area (North, Centre, South), qualification (degree, diploma, other).
(degree, diploma, other) and employment status (employee, self-employed, retired, unemployed, other) of the population aged 18 and over.

Now I have a doubt, can I continue to use the command I was using or should I take into account that the weights have been calibrated?
Any advice would be welcome.
Thank you very much

[David Radwin]

Weights are created and used to make a sample representative of a population, among other reasons. You might contact the data provider and ask if your approach is correct. If you are analyzing only a subset of the sample, say a particular region or age group, you might read the "Subpopulation estimation" section of help svy on how to use proper Stata syntax, or see https://www.stata.com/manuals/svysub...estimation.pdf and https://stats.oarc.ucla.edu/stata/fa...-data-in-stata.