论文部分内容阅读
四、实验研究、需用对象单位估计数科验实验需用的实验单位的多少是有其一定的统计学上的意义的,涉及到科研成果的可靠性。我们已知实验对象估计数,不一定以大数法则为依据,现认为精选小样本的优越性很多,在课题设计合理的前提下就能获得更为可靠的结果,它不仅能用以估计总体,且能估计样本与总体间的随机变异情况。样本单位数的单纯加大,并不能排除偏性;样本数加大,虽可增加精密度,但二者之间不是平行的,此因均值标准差与精密度有关,当样本实验单
IV. EXPERIMENTAL RESEARCH, NEED TO OBJECT THE NUMBER OF OBJECT UNITS The number of experimental units required for a laboratory experiment has its own certain statistical significance, which relates to the reliability of the scientific research results. We know that the estimation of experimental subjects is not necessarily based on the law of large numbers. Now we think that there are many advantages of the selected small sample, and we can get more reliable results under the reasonable design of the subject. It can not only be used to estimate Overall, and can estimate the sample and the overall random variation. The number of sample units simply increased, and can not rule out bias; sample number increases, although the increase of precision, but not parallel between the two, due to the standard deviation of the mean and precision, when the sample test sheet