A note on Studentized confidence intervals for the change-point (Q2430243)

An At-Most-One-Change (AMOC) time series model with an abrupt change in the mean and dependent errors that fulfill certain mixing conditions is considered. The authors continue their investigations started in their previous paper, J. Time Ser. Anal. 29, No. 6, 947--972 (2008; Zbl 1194.62063), where bootstrap confidence intervals for the change-point in a model with dependent data were constructed. In this paper it is proved that the Studentized confidence intervals are also asymptotically valid and it is shown via a simulation study that their small sample behavior is better than that for the original ones without Studentizing. The approach of \textit{F. Götze} and \textit{H.R. Künsch} [Ann. Stat. 24, No. 5, 1914--1933 (1996; Zbl 0906.62040)] is used to obtain the asymptotic correctness of the Studentized confidence intervals. A simulation study indicates how well the proposed variance estimator based on flat-top kernels [see \textit{D.N. Politis} and \textit{J.P. Romano}, J. Time Ser. Anal. 16, No. 1, 67--103 (1995; Zbl 0811.62088)] behaves.