姊妹染色单体互换实验结果统计分析前的数据变换

Data Transformation for Sister Chromatid Exchange Before Statistical Analysis

作者用IBM PC/XT微型计算机,统计分析了12次包括人和动物体内外姊妹染色体互换(SCE)的实验数据,提出每次实验应各筛选一个K值,对SCE原数据作SCE+K变换,变换后的数据一般都能满足方差分析或t检验的统计要求,较为合理地解决了SCE实验数据的统计分析问题。

Studies show that SCE data do not fit a normal distribution and there is problem of heterogeneous variance, so analysis of variance (ANOV) or t-tcst can not be directly applied. No ideal method has been established. According to the analytic results of 12 SCE tests, including human and animal in vivo or in vitro assays, we suggest that if a proper K value is found and the data are transformed byfor each SCE test, the data will satisfy the requirements for ANOV ort-test.