Preprint #96-2
On Inconsistency of the Jackknife-After-Bootstrap
Bias Estimator for Dependent Data
by
S.N. Lahiri
Abstract
Efron (1992) introduced Jackknife-After-Bootstrap as a computationally
efficient method for estimating standard errors of bootstrap estimators.
In a recent paper (cf. Lahiri (1996)), consistency of the
Jackknife-After-Bootstrap variance estimators has been established for
different bootstrap quantities for independent and dependent data. In
this paper, it is shown that in the dependent case, the standard
Jackknife-After-Bootstrap estimator for the bias of block bootstrap
quantities is inconsistent for almost any sensible choice of the
blocking parameters. Some alternative bias estimators are proposed and
shown to be consistent.
Copies of preprints are available from the author upon request. Use
the preprint number (located at the top of the page) and
make the request directly to the author, Iowa State
University,
Department of Statistics, Snedecor Hall, Ames, IA 50011-1210.