《中国兽医杂志》1987年10期53页刊登刘际强同志《建议科技工作者都应学点统计学》一文,读后笔者亦有同感,但觉有不足之处,提出几点意见,以作补充。(一) 样本含量(变数31)不一定是合理的或较大的“本试验”样本。对于样本含量大小相等的非配对试验样本含量(n)的确定:以可信度为95%为例,(见高校试用教材,贵州农学院主编,生物统计附试验设计,P159) “Chinese Veterinary Journal” 1987 年 10 期 53 pages published in Comrade Liu Jiqiang Comrade “recommended science and technology workers should learn some statistics,” a text, the author is also agreeable to read, but feel inadequate, make a few comments, to make up for . (A) The sample size (variable 31) is not necessarily a reasonable or large “test” sample. For the sample size (n) is equal to the size of unpaired samples to determine: a 95% confidence level as an example (see the experimental textbook of colleges and universities, Guizhou Agricultural College, biostatistic experimental design, P159)