刘攀, 冯长焕. 正态标准化数据无量纲处理在因子分析中的应用[J]. 内江师范学院学报, 2017, (12): 54-58. DOI: 10.13603/j.cnki.51-1621/z.2017.12.011
    引用本文: 刘攀, 冯长焕. 正态标准化数据无量纲处理在因子分析中的应用[J]. 内江师范学院学报, 2017, (12): 54-58. DOI: 10.13603/j.cnki.51-1621/z.2017.12.011
    LIU Pan, FENG Changhuan. Application of“the Normal Standardized Data” Dimensionless Processing in Factor Analysis[J]. Journal of Neijiang Normal University, 2017, (12): 54-58. DOI: 10.13603/j.cnki.51-1621/z.2017.12.011
    Citation: LIU Pan, FENG Changhuan. Application of“the Normal Standardized Data” Dimensionless Processing in Factor Analysis[J]. Journal of Neijiang Normal University, 2017, (12): 54-58. DOI: 10.13603/j.cnki.51-1621/z.2017.12.011

    正态标准化数据无量纲处理在因子分析中的应用

    Application of“the Normal Standardized Data” Dimensionless Processing in Factor Analysis

    • 摘要: 基于很多时候数据不服从正态分布,采用“正态标准化法”,先对数据进行正态化变换,使数据服从正态分布,再标准化.结果表明:正态化变换反映了数据在样本中的大小位置顺序,不需要对异常值进行处理,评价结果更全面、可靠,稳定性更强;正态标准化法不改变原始数据的大小分布特点,把处于中位数以上水平的数据变为正,中位数水平以下的变为负,使优者更优,劣者更劣,便于奖优罚劣;用正态标准化法得到的数据在进行 KMO检验时, KMO值更大,更适合因子分析.

       

      Abstract: In view of the data most often going against normal distribution, by use of“normal normalization method”,
      the data is subjected to normalization transformation treatment thus to get it in accordance with normal distribution and is then given a standardizing treatment. The result shows that the normalized transformation can reflect the order of the data size, and thus does not need to treat those abnormal values; the evaluation results are more comprehensive, reliable and stable; normal normalization method does not affect the size distribution characteristics of the original data, changing the data above the medi- an level into positive, and those below the median level into negative and thus making the big data even bigger, the small data even smaller and is thus conducive to reward the good and fine the bad; the data acquired via normal normalization method, when tested by KMO, the KMO value tends to be larger and thus fitter for factor analysis.

       

    /

    返回文章
    返回
    Baidu
    map