论文部分内容阅读
(二)t检验。前面讲的u检验只适用于样本抽自正态总体,并且方差是已知的情况,但实践中所进行抽样的总体的方差往往是未知的,且样本容量是小于30的小样本,这时只能用样本的均方代替总体方差。即用s~2估计δ~2,这时标准化离差的分布不呈正态分布而作t分布,具有自由度df=n-1(所谓自由度是独立变量个数减去它所受的约束条件的差)
(B) t test. The u test mentioned above only applies if the samples are drawn from the normal population and the variance is known, but the overall variance of the sampling performed in practice is often unknown and the sample size is a small sample of less than 30 The mean square of the sample can only be used to replace the overall variance. That is, using s ~ 2 to estimate δ ~ 2, then the distribution of standard deviation is not normal distribution but t distribution, with degree of freedom df = n-1 (the degree of freedom is the number of independent variables minus what it receives The difference between constraints