偏态分布资料均数两侧标准差计算方法的探讨

A Calculation Method of Standard Deviation for the Two Sides of the Mean of Skew Distribution Data

作者将偏态分布资料分为均数上、下两部分、然后分别为其设置与之完全对称的假设变量。从而形成两个完整的假设样本、再按一般方法计算其标准差，即得原样本上、下侧标准差。分别用上、下侧标准差制定的偏态参考人口资料评价标准，可以使实际分布显著地接近预期的分布。

Divided skew distribution data into upper and lower parts towards their mean , followed by setting up a group of entirely symmetric presumptive variables for each part in order to form two complete presumptive samples. Then calculate the standard deviation (SD) by common method to obtain the SDs of both sides of the original sample. While use the upper and lower side's SD respectively to prepare the appraisal criterion foe the data of skew referential population , the actual distribution will markedly approach the expective one.