The asymptotic joint distribution of self-normalized censored sums and sums of squares (Q2277698)

Let \(X,X_ 1,X_ 2,..\). be i.i.d. random variables with infinite variance, and let \(\{r_ n\}\) be a real sequence satisfying \(0<r_ n\to \infty\) and \(r_ n/n\to 0\). The authors investigate empirical versions \(a_ n\) and \(\gamma_ n\) of scaling and centering constants defined by \[ \sum^{n}_{j=1}\{(X^ 2_ j/a^ 2_ n)\wedge 1\}=r_ n\text{ and } \gamma_ n=n^{-1}\sum^{n}_{j=1}(| X_ j| \wedge a_ n)sgn(X_ j), \] which are called scale estimator and empirically censored mean, respectively. Tightness criteria and subsequential limit laws for the normalized pairs \((a_ n,\gamma_ n)\) are obtained. Criteria for convergence to each of the allowable limit laws are given. Applications to the Feller class and domains of attraction are also considered.