Using Low Budget Classroom Exercises to Teach 利用低成本的课堂练习来教.pptVIP

Using Low Budget Classroom Exercises to Teach Sampling Techniques By Richard Summers, PhD Reinhardt College Eastern Regional Competency Based Educational Consortium Students Relate Better to Data That Relate to Them Confidence intervals are more relevant, if the data compare the class to the entire student body In class activities can suggest student projects In class activities emphasize the relevance and applicability of statistics Hypothesis Test for a Proportion Have available the percent male/female enrollment for the college Count the number of male and female students in the class Obtain the confidence level at which there is a significant difference between the population and sample proportions Discuss whether or not the class represents a good sample of the overall student body The Central Limit Theorem (Idea Due to Judith M. Tanur, SUNY) Have students list their birth month and the birth months of their two closest relatives Plot all of the above data on a dot plot Central Limit Theorem (continued) Ask students to find the mean of the three birth months they listed, rounding to the nearest integer Make a dot plot of the above averages – the data will usually be more symmetrically distributed Random Sampling Provide a list of classes taught in your school together with class enrollment and mean class enrollment Use a random number generator to make random choices from the list and note the class size for each choice Take the mean of the class sizes recorded above Random Sampling(Continued) Start with a sample of 5 classes and take the mean class size Add 5 more classes and take the mean of the resulting sample of 10 class sizes Continue to increase the sample size to emphasize that the sample mean tend to the population mean as the sample size becomes larger Systematic Sampling Ask every third student as they are seated in the classroom to give their birth month Take the mean of the above results and compare with the population mean Do the sa