BUSS1020 Lecture Notes - Lecture 10: Test Statistic, Mean Absolute Difference, Confidence Interval

• School: University of Sydney
• Department: Business Studies
• Course: BUSS1020
• Professor: Parinaz Ezzati

Document Summary

Here we learn to compare statistics from samples selected from two populations. When using a 2-sample test that compares means of samples selected from 2 populations, you must establish whether the assumption that the variances in the 2 populations are equal holds. Whether or not this holds will affect the statistic method used to test the difference in means. It gets it name from the fact that the test statistic pools the 2 sample variances to compute the best estimate of the variance common to both populations under the assumption that the 2 population variances are equal. Confidence interval estimate for the difference between 2 means: T test for the difference between 2 means, assuming unequal variances: if you can assume that 2 independent populations are normally distributed, but cannot assume that they have equal variance you can use the separate-variance t test. Use integer portion of whatever v value is calculated.