概率及统计第七章.pdfVIP

Statistical Intervals Hypothesis Tests Probability and Mathematical Statistics (概率与数理统计) Xi ZHANG College of Engineering Outline • Confidence interval ( 区间估计) – Introduction – Confidence interval on mean of normal distribution – Confidence interval on variance of normal distribution – Confidence interval on population proportion • Hypotheses testing （假设检验） – Introduction – Tests on mean of normal distribution – Tests on variance of normal distribution – Tests on population proportion – Contingency table tests • Statistical inference for two samples （双样本统计推断） – Inference on difference in means of two normal distributions – Paired t-test – Inference on variances of two normal distributions – Inference on two population proportions Interval Estimation • Provides Range of Values – Based on Observations from 1 Sample • Gives Information about Closeness to Unknown Population Parameter – Stated in terms of Probability • Example: Unknown attendance mean lies between 50 70 with 95% confidence Point estimate Confidence interval Confidence limit (lower) Confidence limit (upper) Confidence Interval • Confidence interval (置信区间，CI) is an interval esti mate for a population parameter from samples • A confidence estimate for is an interval with end-poi nts of random variables L and U (置信上限和下限) so t hat   P L U 1  • Once we have selected the sample, the values of L and U can be estimated, say l and u, then resulting confide nce interval for is l    u • l and u are lower- and upper-confidence limits;