explain what a residual is. How is it calculated? Why is it important?
2. Explain how the assumptions for simple linear regression and multiple linear regression differ.
3. Suppose that you have the following variables: X, Y, and Z. Z is correlated w/ X, but Z is not correlated
w/ Y. Explain what happens to your β1 estimate if you estimate the following model:
Yi = β0 + β1Xi + β2Zi + ui
4. Suppose that your DGP is defined as follow:
eY = Xβ · u
Where Y is the outcome, X is a random variable and u is your disturbance term.
a. Explain why you cannot estimate an OLS model in this case?
b. Write down the OLS model that you want to estimate in this case.
5. Explain what a dummy variable trap is. How do we avoid it?
6. Explain what a heteroskedasticity is. What happens to your OLS estimate if you have heteroskedasticity.
7. List 3 ways and write down the steps to test for the presence of heteroskedasticity