topic Re: PROC SURVEYREG in SAS with the BRR (Fay) in Statistical Procedures

PROC SURVEYREG in SAS with the BRR (Fay)

julieegholm99 — Tue, 14 May 2024 17:37:35 GMT

proc surveyreg data=XXX varmethod=brr(fay);
model score_m=ESCS ST004D01T IMMIG ;
weight w_fstuwt;
repweights w_fsturwt;
run;

I have used the PROC SURVEYREG in SAS with the BRR (Fay) method for my analysis of the PISA 2022 survey data. I am now uncertain whether it estimates using OLS or GLS. In my thesis, I have stated that the estimation is conducted using the OLS estimator. I am concerned that this might be incorrect. Could anyone clarify whether the estimation method is indeed OLS or if it is GLS?

xx

Re: PROC SURVEYREG in SAS with the BRR (Fay)

SAS_Rob — Tue, 14 May 2024 17:41:45 GMT

PROC SURVEYREG uses elementwise regression to compute the regression coefficient estimators by generalized least squares estimation. The procedure assumes that the regression coefficients are the same across strata and primary sampling units (PSUs). To estimate the variance-covariance matrix for the regression coefficients, PROC SURVEYREG uses either the Taylor series (linearization) method or replication (like BRR) methods to estimate sampling errors of estimators, based on complex sample designs.

Re: PROC SURVEYREG in SAS with the BRR (Fay)

Season — Fri, 14 Jun 2024 10:25:43 GMT

Weighted least squares, to be exact. Because of the fact that weighted least squares is a special kind of generalized least squares, it is not wrong to say that the regression coefficients are estimated by generalized least squares in the SURVEYREG procedure.