主观评分数据的稳健化处理初探

统计应用领域中,统计量的稳健性的重要性在于数据有少量异常值情况下,仍能使统计处理保持较好的稳定性.体育科研中数据的收集过程受多因素影响致使所收集的数据常常混杂异常值,从而影响了统计结果准确性,甚至会误导分析者.通过对竞技体育中的主观评分数据现行处理办法的稳健性进行分析,指出该处理过程稳健性差的原因在于统计量本身缺乏稳健性,并从变换的角度提出了非参数变换办法增强统计处理过程稳健性,实证研究表明,对主观评分数据的实施非参数稳健化变换在保证结果有效性的同时具有可行性与可操作性.

The Research of Robustness In Subjective Grading Data

In the statistical application domain,the importance of statistical steady lie in that it still could make statistical processing to maintain the good stability when it has a few unusual value situation in the sports seientifie research ,the data collection process eaused by multi - faetor influenee the data whieh collects to combine frequently the assayer.Through the analysis to athletics sports subjective grading data,present processing means robustness and proposed from the transformation angle non - parameter transfomation means enhancement statistics treating processes robustness,the real diagnosis research indicated , has the feasibility and the feasibility, to subjective grading data implementation non - parameter steady transformation during guarantee result valid.