PSCI 3325 Lecture Notes - Lecture 6: Patient Protection And Affordable Care Act, Counterfactual Conditional, Scientific Control
Document Summary
Problems with experimental methods: pre-test/post-test experimental design. Measure the dependent variable of interest before the experiment starts. Randomly assign subject n a test and control group. Compare post/pre-test data: governments may not have pre-test data, there is no post-test data since there are no control/test group because the law applies to everyone. Case studies: counterfactuals and multiple causes: policy in place and look at before and after, assuming at minimum there is data from before, as for causation, it is hard to prove at the end of one. Obamacare data shows that more people have health insurance - not necessarily causation. What would have happened if obamacare was never adopted. This can"t be measured, can"t measure something that never: there can be multiple actions/causes occurred. Tendency to try to do multiple things and get results that way. Administrators, officers are interested in getting results and not providing clean information.