论文部分内容阅读
对衡量抑制生理或病理反应的药物,一般可用抑制率:IR=(x_1-x_2)/x_1。经典的抑制率可信限的近似公式当比数分母x_1的差异系数过大时,易得出假阳性结论。本文利用Fieller方法导出正确的抑制率可信限公式,获得的结论与两组均数比较的t检验保持一致。因此认为:当利用抑制率作数据处理时,为防止出现假阳性,以用本文介绍的可信限公式为宜。
For the measurement of inhibition of physiological or pathological drugs, generally available inhibition rate: IR = (x_1-x_2) / x_1. The approximate formula of the classical inhibition rate credible limit When the difference coefficient of the ratio denominator x_1 is too large, it is easy to draw a false positive conclusion. In this paper, Fieller method is used to derive the correct confidence limit formula of inhibition rate, and the obtained conclusion is consistent with the t-test of mean of two groups. Therefore, when using the suppression rate as data processing, in order to prevent false positives, it is appropriate to use the credible formula described in this article.