在科学实验或调查研究中,有时出现个别特大或特小的数值,与均数相差很大;或大多数数值的出现有一定规律,个别数值偏离过大,以致对平均数和标准差影响很大或难以配好回归曲线。有的人随意舍弃此种“异常”数值,企图使实验结果更趋一致或使曲线更为平滑,这是不恰当的;也有人不管所得数值是否可靠或异常,毫不怀疑地一并进行统计分析也是不 In scientific experiments or investigations and studies, sometimes individual extraordinarily large or extraordinarily small values ​​differ from the mean; or the appearance of most of the numerical values ​​is definite, and the individual values ​​deviate too much, which has a great influence on the mean and standard deviation. Large or difficult to match the regression curve. Some people arbitrarily discard this “abnormal” value in an attempt to make the experimental results more consistent or make the curve smoother. This is inappropriate; some people will not statistically believe that the value is reliable or abnormal. Analysis is not