第四章 抽样与估计.pptVIP

90 90 90 4-* Confidence Intervals for the Variance A 100(1 – a)% CI is: Note the difference in the denominators! 总体方差的区间估计(图示) ?2 ?21-??? ?2??? 总体方差 1-?的置信区间 自由度为n-1的?2 4-* PHStat Example: Home Market Value Data 4-* Confidence Intervals for Population Total A 100(1 – a)% CI is: N x ? tn-1,a/2 4-* PHStat Example Sample of 20 reimbursement claims over 60 days old yielded a mean of $185 and standard deviation of $22. 4-* Confidence Intervals and Decision Making Required weight for a soap product is 64 ounces. A sample of 30 boxes found a mean of 63.82 and standard deviation of 1.05. A 95% CI is [63.43, 64.21]. What conclusion can you reach? What if the standard deviation was 0.46 and the CI is [63.65, 63.99]? 例2.抽选1300选民，有692人投给了竞选者的一人，是否意味着这位候选人很可能赢得这次选取？如果获得670人投给这一候选人呢？ 4-* Confidence Intervals and Sample Size CI for the mean, s known Sample size needed for half-width of at most E is n ? (z?/2)2(s2)/E2 CI for a proportion Sample size needed for half-width of at most E is Use p as an estimate of p or 0.5 for the most conservative estimate ? 4-* PHStat Tool: Sample Size Determination 4-* Example For the soap filling process, the sampling error associated with the CI = [63.43, 64.21] is 0.39 (half-width). What sample size is required to reduce this to 0.15? 4-* Example How many voters must be surveyed to ensure a sampling error of at most ± 2%? 4-* Additional Types of Confidence Intervals Difference in means Independent samples with unequal variances Independent samples with equal variances Paired samples Difference in proportions 两个样本均值之差的抽样分布 m 1 s 1 总体1 s 2 m 2 总体2 抽取简单随机样样本容量 n1 计算X1 抽取简单随机样样本容量 n2 计算X2 计算每一对样本 的X1-X2 所有可能样本 的X1-X2 m1- m2 抽样分布 4-* Confidence Intervals for Differences Between Means ? Population 1 Population 2 Mean m1 m2 Standard deviation s1 s2 Point estimate ? ?x1 ? ?x2 Sample size n1 n2 Point estimate for the difference in means, m1 – m2, is given by x1 - x2 4-* Independent Samples With Unequal Variance