PO217 Chapter Notes - Chapter 11: Causal Inference, Statistical Power, Political Psychology
Document Summary
Involve an intervention by the researcher, who then tracks consequences for the outcome of interest (planned intervention: random assignment researcher creates two groups: experimental group and control group, may experience the hawthorn effect. Important that subjects remain unaware of randomization process and the existence of other groups exposed to different treatments: single-blind design, subjects and treatment administrators are unaware of the effect being tested, double-blind design. In a between-subject design subjects are randomly assigned to various treatment and control groups and examined post-treatment. In within-subject designs researchers evaluate subjects before and after they receive treatment relative to the before and after scores of subjects who did not receive the treatment and infer causality based on any differences observed. Laboratory experiments: subjects are recruited to a common place where the researcher exerts a relatively large degree of control over the experimental setting. Ethical issues and other limits of experimentation in political science. It is sometimes impossible to manipulate variables of interest.