1. Simulation Study
(1) Data Generating Process
Y = β1 + X2 β2 + X3 β3 + X4 β4 + X5 β5 + e, e ∼ N (0T ×1 , σ 2 IT ) (1)
Model 1: Y = β1 + X2 β2 + e
Model 2: Y = β1 + X2 β2 + X3 β3 + e
Model 3: Y = β1 + X2 β2 + X3 β3 + X4 β4 + X5 β5 + e
Model 4: Y = X2 β2 + X3 β3 + e
where e ∼ N (0T ×1 , σ 2 IT ).
(2.1) Report R̄2 , SIC and AIC for Model 1, 2 and 3.
(2.2) Do those criterions support the same model? If not, why does this happen?
(2.3) In the estimated Model 4, are estimates for β2 and β3 close to the true values? What
happens to R2 ?
(3) Repeat (1) and (2) in the case of T = 10 given the same true values.