抽样误差与假设检验演示文件修改版.pptVIP

计量资料的统计推断 统计推断 statistical inference 第四章 抽样误差与假设检验 主要内容 一、 抽样误差与标准误 二、 t 分布 三、 总体均数的参数估计 四、 假设检验的基本思想和步骤 一、 抽样误差 抽样误差的定义 由于存在个体变异，这种在抽样过程中产生的样本统计量与总体参数及样本统计量之间的差异，称之为抽样误差。 Because of Variation, during the sampling procedure, there are some differences between statistic of sample and parameter of population and also differences among sample statistics. These differences are called sampling error. 以均数的抽样误差为例，它涵盖了两方面的内容： 样本均数不同于总体均数 样本均数之间也存在着差异 The sampling error of mean covers two sides content: Mean of samples differ with population mean Mean of one sample differs with the one of another sample 某农村地区1999年14岁女孩的身高正态分布总体（cm） μ=143.08, σ=6.58, N＝10000例 x1~x10 142.3 148.8 142.7 144.4 144.7 145.1 143.3 154.2 152.3 142.7 x11~x20 156.6 137.9 143.9 141.2 139.3 145.8 142.2 137.9 141.2 150.6 x21~x30 142.7 151.3 142.4 141.5 141.9 147.9 125.9 139.9 148.9 154.9 x31~x40 145.7 140.8 139.6 148.8 147.8 146.7 132.7 149.7 154.0 158.2 x41~x50 138.2 149.8 151.1 140.1 140.5 143.4 152.9 147.5 147.7 162.6 x51~x60 141.6 143.6 144.0 150.6 138.9 150.8 147.9 136.9 146.5 130.6 x61~x70 142.5 149.0 145.4 139.5 148.9 144.5 141.8 148.1 145.4 134.6 x71~x80 130.5 145.2 146.2 146.4 142.4 137.1 141.4 144.0 129.4 142.8 x81~x90 132.1 141.8 143.3 143.8 134.7 147.1 140.9 137.4 142.5 146.6 .…………. xn-9~xn 134.4 133.1 145.9 139.2 137.4 142.3 160.9 137.7 142.9 126.8 以均数为例介绍 从正态分布总体N(143.08,6.58)中随机抽样，保持样本含量或样本例数不变，均为n ＝10，则每次抽出的10个个体组成一个样本，每个样本的均数称为样本均数，共抽M次，则有M个样本和M个样本均数。 Sampling from normal distribution population N(143.08,6.58), keep the number of size=10, then ten individuals form a sample. After sampling for M times, we get M samples and M mean of samples ( ). 由