This method was based on that the depth of erosion was submitted to the extreme value of truncation and it was also based on random variable of minimum of the statistical distribution theory.

By analyzing and comparing the statistical characteristic and difference between sample mean,trimmed mean and Winsorised mean,the paper demonstrates the importance of substitutiy Winsoried mean for sample mean in item evaluating science and technology awards.

This paper based on the definition of trimmed mean,winsorised mean and trimmed normal distribution,we obtained that the means are the maximum likeihood estimation(MLE),their influence function are derived,the robustness can be analysed.

A class of strong convergence theorem for an arbitrarily adapted stochastic sequence without the non-decrease condition was established by using the truncation method of stochastic variables and the stop time.

By using truncation methods of random variables and Doob martingale convergence theorem,the properties of sequence of arbitrary random variables are studied,a class of strong limit theorems for sequences of arbitrary random variables is obtained under moment conditions, and some conclusions corresponded to these are generalized.

By use of truncation of random variables,the properties of sequences of arbitrary random variables are studied and a class of strong limit theorems for sequences of arbitrary random variables are discussed under moment conditions.