Right choice is (b) Predictor-corrector method
The best I can explain: To get the time derivative at the average time between t and t+Δ t, the MacCormack’s technique uses the Predictor-corrector method. This is used as we do not know the value for the time derivative at the time-step t+Δ t.