Need help with aggragating data at individual level to region level

[Jonathan Muringani]

Please can someone help




I have observations collected at individual levels in a number of regions in 23 different country.




I need help on how to aggregate individual data to regional level.




The original survey data has these three variables



Iweight – individual weights based on gender and age compared with region’s actual population
Dweight – the design weight
PSweight – the post stratification weight
Pweight – the population weight

My understanding is that I should use PS weights and population weights.
I tried this command
*Calculating mean regional scores
collapse (first) Pweight PSweight (mean) corruption impartiality qaulity q_o_g, ///
by(nuts nuts_name country nuts_c year)
I am not confident that this is the right command becuase Pweight refers to national population and Psweights is a includes design weight but also I am not sure of the collapse command.

Kind regards

Jonathan

[Phil Bromiley]

You'll increase your chances of a useful answer by following the FAQ on asking questions - provide Stata code in code delimiters, readable Stata output, and sample data using dataex.

As a general rule, the best way to check this kind of thing is to do it on a very few observations and check that it is giving you the numbers you think you should get. Sometimes, creating arbitrary numbers in a trial data set can make this easier. I'm not really qualified to check your code.

[Jonathan Muringani]

Hi Phil
Thank you for your response. I managed to crack the problem.