This paper proposes an approach to test the spurious regression problem in the one-way panel data model whereby a hybrid Whittle approach is demonstrated for the restricted maximum likelihood (REML) estimator. The model applies the Fejér window, in which the remainder disturbance can be either a stationary or a non-stationary process, in the domain of the integrated degree d [-0.5, 1.5). This research shows the merits of combining two conventional approaches, differencing and tapering, to estimate the model without prior knowledge. Through Monte Carlo experiments, the consistency of the estimator is examined by growing the individual number N and time length T, in which the remainder disturbances are simulated by distinct models including stationary and non-stationary processes. Observation of the power tests shows that the estimators are quite successful and powerful.
Relation:
International Journal of Organizational Innovation, 2 (1): 83-105
