Hi.
I'm using weights from the European social Survey which have three different weights: design, population and post estimation weights. I really don't know what kind of weights they are (if fw, or aw or pw). I simply tried to graph (histogram) a variable ( worry about climate change) using weights but stata said: "may not use noninteger frequency weights". I'm aware of that would mean that my weights are not integer and that I need frequency weight, but the question is: can i transform my non integer weights in integer ones?
i attach a brief doc that explains the ESS database's weights. (http://www.europeansocialsurvey.org/...ing_data_1.pdf)
Thank you very much
Gab
Related Posts with Non integer weights - problem
Running regression while controlling for industry and yearsI'm investigating the influence of patenting (pat_yr -> patents per year) on the returnoninvestem…
Correlation between dummy variablesI am working on STATA 14 and have a panel data- long view with cross sectional data of Indian non-fi…
Odd scalar behaviourDear statalisters, I am running the following code to cut up a very large file (>15g) into small…
generate difference within groupHi, I had a firm-level survey in different countries. I want to calculate the change in "score" (di…
Creating 2 dummy variables in StataHi everyone I am back with another question to help me out I have a table output containing lots o…
Subscribe to:
Post Comments (Atom)
0 Response to Non integer weights - problem
Post a Comment