Laws of large numbers for bootstrapped U-statistics (Q789847)

A strong law of large numbers is obtained for bootstrapped U-statistics provided the \((1+\delta)^{th}\) moment of the kernel is finite for some \(\delta>0\). Further the jackknifed estimator of the variance of the bootstrapped U-statistic is shown to converge in probability to a constant under second moment conditions on the kernel of the estimator. If these moment conditions are strengthened to \((2+\delta)^{th}\) moments finite for some \(\delta>0\), then the convergence can be shown to be almost sure.