Heteroscedasticity-robustness of jackknife variance estimators in linear models (Q1106594)

The asymptotic unbiasedness and consistency of three types of jackknife variance estimators in the presence of error variance heteroscedasticity in linear models are studied. The results are given in terms of the number of observations deleted and measures of imbalance of the model. The consistency of a class of the second author's [ibid. 14, 1261-1295 (1986; Zbl 0618.62072)] weighted jackknife variance estimators for nonlinear parameters is also studied. A necessary and sufficient condition is given for the asymptotic unbiasedness and consistency of the unweighted delete-1 jackknife variance estimator and \textit{D. V. Hinkley}'s [Technometrics 19, 285-292 (1977; Zbl 0367.62085)] weighted delete-1 jackknife variance estimator. This condition is m60J10 We consider sequences of tests with error exp(-nA) of the first type for some hypothesis on the transition matrix of a finite Markov chain. If the exponential rate of convergence of type II errors serves as measure of the asymptotic performance, appropriate sequences of likelihood ratio tests are shown to be uniformly optimal on a very large set of alternatives.