摘要
对于正态样本中上下异常值的同时检脸，大多数文献仅处理了下异常值的个数ｋ１和上异常值个数ｋ２已知的情况，当ｋ１和ｋ２未知时．由于受ｍａｓｋｉｎｇ或ｓｗａｍｐｉｎｇ效应，检验问题变得十分困难与复杂．本文成功地解决了这一问题．
For the test of both upper and lower outliers in normal samples. most literature only dealt with the case in which the number of lower outliers k1 and that of upper outliers k2 are known. When the k1 and k2 are unknown, the problem becomes very complicated because of masking or swamping effect. In this paper. we solve this problem successfully.
出处
《经济数学》
2001年第2期62-67,共6页
Journal of Quantitative Economics
