POLS 208 Lecture Notes - Lecture 16: Test Statistic, Dependent And Independent Variables, Absentee Ballot
5 Dec 2018
School
Department
Course
Professor
Document Summary
If there"s no jump at the cutoff, then it"s probably a placebo effect. Look at the columns (1, 2, 3) Hypothesis tests give you an estimate that is the (test statistic - The 3 columns on the right are giving you the estimates of the difference and the standard error. Should not be sorting - should not be a jump in treatment and control in terms of observations. There should be nothing else driving the difference at the cutoff. Some kind of self selection into one side of the cutoff bias. For fujiwara, you"re not responsible for anything after part 3 (pg 411) Case study - an in depth analysis of one case. Can be a nation, a neighborhood, an organization. A case is a temporarily and spatially bounded instance of a specified phenomenon (you should be asking yourself what that is) In statistics, we have a problem when the # of explanatory factors approaches the # of observations .
