Lecture 6 - Data structure and Panel Data

Question 1

How might we set up a study to look at traffic fatalities and alcohol taxes?

Answer 1

Question 2

Given a positive regression coefficient should we conclude that a reduction in taxes on beer would reduce lethal car accidents? Why or why not?

Answer 2

Question 3

What will happen if we add fixed effects to the below graph? What will the new trend line be?

Answer 3

Question 4

What is the alternative to taking Fixed Effects?

Answer 4

Using difference-in-difference?

Question 5

Transform the below equation using difference-in-difference

Answer 5

Question 6

What is panel data?

Answer 6

Question 7

What are the two types of panel/longitudinal samples?

Answer 7

Question 8

What are the two ways in which unbalanced panels happen?

Answer 8

Question 9

Write out the basic model equation we use for panel data

Answer 9

Question 10

What are the 3 basic assumptions we make about panel data sets?

Answer 10

Question 11

Explain strict versus weak exogeneity? How is u^-(it) related to x?

Answer 11

Question 12

Explain strict versus weak exogeneity? How is f^-(i) related to x?

Answer 12

Question 13

Should we assume Random Effects (RE) or Fixed Effects (FE)?

Answer 13

Question 14

Explain how the random effects model works?

Answer 14

Question 15

What are the two different tupes of fixed effects model and what source of endogeneity to they deal with?

Answer 15

Question 16

Explain with Within Groups (WG) estimator

Answer 16

Question 17

Explain what the Least-Square Dummy-Variable (LSDV) estimator is

Answer 17

Question 18

Explain the First Difference Estimator

Answer 18

Question 19

Explain what a panel regression with time fixed effects is

Answer 19

Question 20

What is a high level explanation of clustering standard errors?

Answer 20