Confidence intervals for quantile estimation using jackknife techniques (Q626236)

Population quantile estimation is considered by a finite sample when some auxiliary information is available. Ratio, regression and differences estimators are discussed. A jackknife technique is described for bias correction and variance estimation of quantile estimators. Conditions for asymptotic normality of these estimates are given. Confidence intervals are constructed for quantiles based on the asymptotic normality and jackknife variance estimates. Numerical results are presented for simulated samples from real life finite populations.