Hello,
I would like to apply a regression in SAS using the Newey-West t-stat.
Below I have used a basic regression applying the Proc Reg statement. Is it possible to apply the Newey-West correction of standard errors (for heteroscedasticity and autocorrelation) in this context or do I have to use the Proc Model statement?
My problem is that I currently only have access to the University edition which does not support Proc Model, hence I would like to use Proc Reg.
Many thanks for any help,
Tim
proc reg data=earnings;
model earn_w = L1_earn_w L1_negE_w L1_NExE_w;
where 2000 <= year <= 2010;
ODS OUTPUT parameterestimates = par_est;
run; quit;