“…In testing hypothesis H 0 : T =0 versus H 1 : T = 0, where T = 2 2 − 2 1 , the advantage of using factor (k=T (1 − k=T )) 1=2− with 0 6 6 1=2 is that V k is sensitive with respect to the contiguous alternative T ∼ T −1=2 (see Corollary 2.2 below, or Cs orgő and Horvà ath, 1997; Antoch and Hu skovà a, 1995). We can use both statistics, V k andk( ), to deal with tests and estimates of change points.…”