论文部分内容阅读
中国防痨通讯1987年3期刊载的《有关统计学处理问题》讨论,读后颇受启发,有几点拙见提出商榷,并就教于同道。一、讨论集中在对“初治肺结核耐药动态观察”一文中的表3(本文为表1)的两种统计处理方法上。究以何法为宜,我认为应取决于数据的含意及分析的目的。该文是“耐药动态观察”,从表1显示出不同年龄组对一线药物耐药率随年龄增长而下降。为此,对表1作统计学处理,首先须明确这种差异是抽样误差造成,还是具有实质性意义,方可判断此样本差异能否用以推论总体。这就必须作总的差异显著性检验。
China Anti-theft Communications, 1987 3 issue of “on the issue of statistical processing” discussion, quite inspired after reading, a few humble opinion to discuss, and to teach fellow. First, the discussion focused on two statistical methods of Table 3 (Table 1) in the article “Dynamic Observation of Newly Measured Tuberculosis Drug Resistance.” What is the proper way? I think it depends on the meaning of the data and the purpose of the analysis. The article is “dynamic observation of resistance”, from Table 1 shows that different age groups on first-line drug resistance rates decline with age. To this end, for the statistical treatment of Table 1, we must first clear whether this difference is caused by sampling error, or substantive, in order to determine whether the sample differences can be used to infer the overall. This must be done for the total difference in significance test.