I'm using a fixed effects model. When dealing with heteroscedasticity and autocorrelation, I'm not sure which method to apply; vce(cluster) or vce(bootstrap)? What are the differences between the two?
I assumed that the reason behind the two approaches are the same, however they gave me quite different results...
Thank you!
I assumed that the reason behind the two approaches are the same, however they gave me quite different results...
Thank you!