Working with survey data: using weights in svy: proportion

[Anja Heimann]

Hi,

I've recently started to work more with survey data, and discovered Stata's survey functions. However, I am having trouble with the svy:proportion method.
I have svyset my dataset and then, having collected my dependent and independent variables in local macros, I expected this following code to run:

svyset [pw=Weight],strata(Region) singleunit(certainty)
local depvar [InEmp UnEmp]
local indvars [ib1.age ib1.ethgrp ]

svy: proportion `depvar'

matrix result=r(table)

local n=e(N)

matrix `depvar'=J(1,9,.)
matrix colnames `depvar'="Percentage" "SE" "Lower95%" "Upper95%" "Difference" "SE" "Pvalue" "stars" "Base"
matrix `depvar'[1,1]=result[1,2]*100
matrix `depvar'[1,2]=result[2,2]*100
matrix `depvar'[1,3]=result[5,2]*100
matrix `depvar'[1,4]=result[6,2]*100
matrix `depvar'[1,9]=`n'

But it does not run, I get the error that "weigths are not allowed". So then I changed my first line of code, taking out the pw=Weight command, but it still won't run.

I looked at the manual, and there is probably a very simple fix, but I can't see it right now. Any ideas?

Thank you in advance

[Andrew Musau]

local depvar [InEmp UnEmp]

should be

local depvar InEmp UnEmp

Brackets in Stata syntax are read as weights.

[Anja Heimann]

Thank you! I knew it was something silly like that.