Question

If I run a simple proc reg in SAS, I have no problem to report the standard errors and rsquare. But why am I having some difficulties to report the Rsquare and Standard Errors for the intercepts and coefficients under this PROC MODELsetup?

proc model data=have;
var r1-3;
parms a1-a10 b1-b10;
r1=a1+b1*A+b4*B+b7*C; 
r2=a2+b2*A+b5*B+b8*C;
r3=a3+b3*A+b6*B+b9*C; 
fit r1-r3/sur outs=want outest=est;
test a1, a2, a3,/wald;
run;

where can I type OUTSEB and RSQUARE so the results can be reported in the est dataset? I’ve tried to type OUTSEB and RSQUARE after OUTEST or after proc model data=have etc… no success.

Answer 1

FSRSQ gives you the R2. It can be used as an option in both the PROC MODEL and FIT statements (if you use it in the former, it will become default for all subsequent FIT statements).

proc model fsrsq ...