Hi Statalisters,
I am hoping to get weighted correlations and the appropriate p-values using survey data. I have tried the -corr_svy- command and the correlations do not appear to differ from the correlations using the -corr- command which does not account for weighting and sampling design. I have read that it may not be possible to get obtain weighted correlations with Stata 14 (https://stats.idre.ucla.edu/stata/seminars/survey-data-analysis-with-stata-14/)but any help regarding a workaround is appreciated.
svyset [pweight= W7C17P_20], jkrweight(W7C17P_21-W7C17P_240) vce(jack) mse
corr_svy
Thanks!
-Jenn
I am hoping to get weighted correlations and the appropriate p-values using survey data. I have tried the -corr_svy- command and the correlations do not appear to differ from the correlations using the -corr- command which does not account for weighting and sampling design. I have read that it may not be possible to get obtain weighted correlations with Stata 14 (https://stats.idre.ucla.edu/stata/seminars/survey-data-analysis-with-stata-14/)but any help regarding a workaround is appreciated.
svyset [pweight= W7C17P_20], jkrweight(W7C17P_21-W7C17P_240) vce(jack) mse
corr_svy
Thanks!
-Jenn
Comment