Question: Because fixed effects allows arbitrary correlation between and the, while random effects does not, FE is widely thought to be a more convincing tool for
Because fixed effects allows arbitrary correlation between and the, while random effects does not, FE is widely thought to be a more convincing tool for estimating ceteris paribus effects. Let's say you are concerned about the correlation between and the key explanatory variable that is constant over time. Then what is your modeling strategy? Use FE or RE?
Step by Step Solution
There are 3 Steps involved in it
Get step-by-step solutions from verified subject matter experts
