POLS 3650 Lecture Notes - Lecture 16: Null Hypothesis, Statistical Significance, Statistical Hypothesis Testing

• School: University of Guelph
• Department: Political Science
• Course: POLS 3650
• Professor: Edward Koning

Document Summary

Used to investigate the relationship between a nominal or ordinal iv and an interval/ ratio. It can be understood as a compare means test for multiple groups. Independent variable: nominal or ordinal, k>2: dv: interval/ratio. We use anova when the independent variable is nominal because it is impossible to formulate a directional hypothesis. Also, anova is used because we cannot do that many mean tests. We would be conducting partial investigations: we want to know whether there is a relationship between these two variables, not whether there are significant differences between some of the groups, larger chance of type i error. Anova offers a solution to this problem. It investigates whether a case is more similar to another case in the same group than to another case in another group. In other words, are the inter-group differences larger than the intra-group differences: if so, we can conclude there is a relationship between the two variables.