Hello everyone,
I know that here we discussed deeply and several times the issue of weighting the data from DHS before using them in the analysis.
My observation is I employ the data in a regression analysis, and I did run the regression once with weighted data set and another time without, using the wriiten lines 'below' from the website ( http://dhsprogram.com/data/Using-DataSets-for-Analysis.cfm#C P_JUMP_14042)
generate wgt = v005/1000000
tab var [iweight=wgt]
But I did not find a big difference either in the regression estimators or in the standard errors
So my question is Do I really need to Weight the DHS data before using in the analysis?
I really need your comments
Thanks