POLI 311 Lecture Notes - Lecture 10: Chi-Squared Distribution, Contingency Table, Data Analysis
Document Summary
Topic 12: examining the efects of control variables. The control variable allows us to hold constant a third variable in an analysis. This allows us to see how other factors afect the relaionship between the. Iv and the dv: cross tabulate the iv and dv by sub groups, compare percentage gaps across columns with original table/check chi square, compare measures of associaion with the original table. We"re trying to mimic the logic of the classical experimental design. In the ex-post facto quasi experimental design, staisical methods are used to control variables (instead of experimental controls) With nominal or ordinal variables with a small number of categories, we use physical methods. You can make groups of cases based on categories of the control. E. g. if there are 3 categories of the control, make a cross tabulaion for each group of cases.