-------------------------------------------------------------------------------
help for svybsamp2
-------------------------------------------------------------------------------

Re-sampling for complex survey data

                
        svybsamp2 , [if exp] [in range]

        
        svybsamp2 requires that the survey design variables be identified using
>  svyset, see help svyset


Description

        
    svybsamp2 creates a new data set with _N randomly drawn observations
    (with replacement). Similarly to svybsamp (if installed), the procedure
    respects survey design, re-sampling within different combinations of
    strata and PSU. Therefore using the inflation factors (pweight, see help 
    weights) the new sample can be used to get to the original
    "super-population".

        
    svybsamp2 improves considerably the speed compared to the original
    svybsamp. However, the re-samplying procedure becomes slower as the
    number of strata in the sample increases.


Examples

        local i=1 
        while `i'<100 { 
                svybsamp2 
                qui svyreg y x1 x2 
                if `i'==1 { 
                        matrix `b'=`b`i'' 
                } 
                else { 
                        matrix `b'=`b'+`b`i'' 
                }       
        } 


Authors

Rafael E. De Hoyos, Faculty of Economics, University of Cambridge. red29@cam.ac
> .uk 

Jaime Ruiz-Tagle, Faculty of Economics, University of Cambridge. jr280@cam.ac.u
> k


Also see

Manual: [U] Bootstrap
        [SVY] svy estimators

Online: help for svybsamp (if installed), bsample, svyset, jknife