第十一讲 统计推论(三).ppt.Convertor(国外英文资料).docVIP

第十一讲 统计推论(三).ppt.Convertor(国外英文资料) Social Statistics Eleventh, statistical inference (three) Hypothesis test of mean and proportion Knowledge point Population mean test of 1. large samples (Z test) The overall proportion test of 2. large samples (Z test) 3. small sample population mean test (t test) Learning goal Grasp the calculation methods and steps of Z test and t test; Classification of single population hypothesis tests Basic steps of inspection 1. stating the original hypothesis H0 and the research hypothesis H1 2. select significant level a 3. select test statistics 4. select n 5. give the critical value 6. compute test statistics 7. statistical decision making 8. express decision results First, the hypothesis test of population mean Z Test Population mean test of 1.1 large samples (both ends) 1. assumed conditions The population obeys the normal distribution If the normal distribution is not obeyed, the normal distribution can be used to approximate (N350) 2., the original hypothesis is: H0: m=m0; the study hypothesis is: H1:, m, 1m0 3. use the Z statistic Example 1 A machine tool factory processing a part, according to experience to know, the processing parts of the ellipse approximately obeys the normal distribution, its overall mean is m0=0.081mm, the overall standard deviation is s = 0.025. Now a new machine tool is processed and n=200 parts are extracted. The average value of the ellipse obtained is 0.076mm. Can the mean of the ellipticity of the machined parts of the new machine be significantly different from the previous ones? (a = 0.05) Solution: H0: m m = 0.081 H1: m m 10.081 A a = 0.05 N = 200 Critical value = 1.96 Test statistic: Policy decision: Reject H0, accept H1. Conclusion: On the whole, the ellipticity of the parts machined by the new machine tools is significantly different from the previous ones 1.2 large sample population mean hypothesis test (one end) 1. assumed conditions The population obeys the normal distribution If it does not ob