1: \begin{abstract}

2: This paper develops asymptotic theory of integrals of empirical quantile functions with respect to random weight functions, which is an extension of classical $L$\hyp{}statistics.

3: They appear when sample trimming or Winsorization is applied to asymptotically linear estimators.

4: The key idea is to consider empirical processes in the spaces appropriate for integration.

5: First, we characterize weak convergence of empirical distribution functions and random weight functions in the space of {\em bounded integrable} functions.

6: Second, we establish the delta method for empirical quantile functions as {\em integrable} functions.

7: Third, we derive the delta method for $L$\hyp{}statistics.

8: Finally, we prove weak convergence of their bootstrap processes, showing validity of nonparametric bootstrap.

9: \end{abstract}