On inference based on the one-sample sign statistic for long-range dependent data (Q2430249)

The problem of statistical inference for the location parameter of the marginal distribution of a strictly stationary long-range dependent random process is considered. The behaviour of the sign statistic for long-range dependent random processes is studied. The subsampling method [\textit{D. Politis, J. P. Romano} and \textit{M. Wolf}, Subsampling. New York, NY: Springer (1999; Zbl 0931.62035); \textit{S. N. Lahiri}, Resampling methods for dependent data. New York, NY: Springer (2003; Zbl 1028.62002)] for approximating the distribution of a sign statistic is discussed and its asymptotic properties are established. Asymptotically valid confidence intervals and hypothesis tests are constructed. A Monte Carlo simulation study is presented to reveal the small-sample properties of subsampling-based inference procedures.