Unformatted text preview: Pol 51, Lecture 4 Evaluating Causal Relationships Most causal theories are BIVARIATE, which means they ask whether a single independent variable causes a single dependent variable. (wondering whether x causes y) y is the dependent variable. The theory might be bivariate, but the world is multivariate. Every phenomenon (or dependent variable) has several causes. We’ll call other variables ‘Z’. We are trying to determine the relationship of x and y, but you can’t be sure about the relationship between x and y, unless we consider all z’s. If we don’t control for z, then our conclusions whether x causes y might be mistaken. RESEARCH DESIGNS, strategy for drawing conclusion on how x causes y. CAUSALITY: example, the quality of lecture causes someone to either come or not come to class. Let’s say yesterday the lecture was terrible. You can’t expect her to not go to the class tomorrow because maybe she’ll go to class, despite knowing how terrible the class is, because she cares about her grade, or...
View Full Document
- Summer '08
- Causality, Single dependent variable, credible causal mechanism, legitimate causal theory