Asymptotic Design-Unbiasedness of Jackknige Variance Estimators in the Presence of Missing Data
缺失数据下Jackknife方差估计量的渐近设计无偏性

In practical surveys, non-response is often unavoidable. How to estimate the population quantities of interest is an important problem in this case. Recently, Zou et al. (2002) have found that for the sample rotation method with missing data, Jackknife variance estimators they proposed are asymptotically design-unbiased under uniform response. This properly provides a solid basis to use Jackknife variance estimators. This paper aims at studying the similar property of some other Jackknife variance estimators existing in the literature. Our results show that the asymptotic design-unbiasedness for Jackknife variance estimators universally holds.