POLSCI 3N06 Lecture 8: Political Science 3N06 2014 Semester II Lecture 8a .doc
5 Apr 2014
School
Department
Course
Professor
Document Summary
Political science 3n06 2014 lecture 8a: two-sample hypothesis testing. To answer this question, we follow the same logic of inference as in our one- sample hypothesis test of means (and proportions) Following healy and prus" 5 step model: Step 1: make assumptions and meet test requirements. The samples must both have been selected according to epsem principles. The samples must also be independent (at least for the formulas being used here) Independent random sampling is required (if not, another test is performed) B) assumptions about the level of measurement. If using difference in means, the variables must be interval-ratio. If using difference in proportions, the variables can be nominal level. C) assumptions about the shape of the sampling distribution. The sampling distribution in the case of two sample hypothesis testing is the sampling distribution of the difference in sample means (for means) This sampling distribution has a mean = 0 and a standard deviation (standard error) =
