pushmedia1's Study Plan
I'm just playing, but I should be studying for finals
- derive wald statistic
- review asymptotic theory
- LLN
- CLT
- matrix calculus
- assumptions of OLS and what to do if they're violated
- E(u|X) != 0, then find an instrumental variable (i.e. a variable correlated with the regressor but not the y)
- E(uu'|X) != sigma^2 I ==> use robust std errors if just heteroskedasticity or GLS if correlated errors
- heteroskedasticity
