摘要
标准差在反映标志变异程度上存在着准确性低、误差大的致命性缺陷,并且缺乏解构性功能。从平均差的传统计量公式出发推导出具有精准性和解构性兼备的新的计量模型,阐述和证明平均差所具有的10条数学性质。它所涵盖和反映的信息量大大超越标准差,应用范围更广泛。提出新的离散系数计量模型,并运用负数转正取值的最近邻性和分母、分子对称位移性原理,有效地解决了均值X为负数和零值条件下的离散系数计量难题。
Standard deviation has fatal defect of poor accuracy and big error, which lacks de-construction function. The metrological model, which is inferred from average deviation metrological formula and which has accuracy and de-construction, verifies 10 mathematical features that average deviation possesses. The information it contains and reflects is well beyond standard deviation, and has wider range of use. This article puts forward a new metrological model of coefficient of dispersion. The application of propinquity of negative number transforming into positive number and the principle of symmetrical shifting of numerator and denominator is effectively to solve the problem of coefficient deviation metrology under the terms of negative and zero X mean value.
出处
《丽水学院学报》
2012年第2期1-13,共13页
Journal of Lishui University
关键词
标准差
平均差
标准差系数
平均差系数
average deviation
mean deviation
coefficient of standard deviation
coefficient of mean deviation
