ECN 388

Question 1

What is MLR 1?

Answer 1

We believe the relationship to be linear in nature.

Question 2

What is MLR 2?

Answer 2

We have a random sample of size n.

Question 3

What is MLR 3?

Answer 3

The sample outcomes on all the x values are not the same.

Question 4

What is MLR 4?

Answer 4

The error term, u, has an expected value of zero.

Question 5

What is MLR 5?

Answer 5

The error term, u, has the same variance given any value of the explanatory variables.

Question 6

Multicollinearity

Answer 6

When there is a strong correlation between x variables.

Question 7

Three ways to detect multicollinearity:

Answer 7

1. ) calculate correl coefficient and compare to R-squared.
2. ) if p-values are greater than R-squared
3. ) variance inflation factor

Question 8

Variance Inflation Factor

Answer 8

VIF = 1/(1-r^2)

VIF > 10

Question 9

Heteroskedasticity

Answer 9

Not a constant variance in the error term, u.

Violates MLR 5.

Question 10

Four ways to detect Heteroskedasticity?

Answer 10

1. ) Graphical (the error term against the predicted y)
2. ) Goldfeld-Quant Test
3. ) Park Test: multiplicative Heteroskedasticity
4. ) Breusch-Pagan Test

Question 11

Four ways to correct multicollinearity:

Answer 11

1. ) add data
2. ) fundamentally change the model
3. ) drop a variable that contributes to multicollinearity
4. ) do nothing and proceed with caution

Question 12

Three ways to correct Heteroskedasticity:

Answer 12

1. ) weighted least squares
2. ) weighted least squares, different weight
3. ) STATA fix: “robust”

Question 13

What is Functional Form Misspecification?

Answer 13

Missing important variables that should be included.

Question 14

How do we correct Functional Form Misspecification?

Answer 14

Add the needed variables

Question 15

What are Proxy Variables?

Answer 15

When we know we need a certain variable in the regression but we can’t get it. Use a different variable as a proxy.

Question 16

What is Measurement Error?

Answer 16

When the data that has been collected contains errors.

Question 17

What is Missing Data?

Answer 17

Observations that have “holes” in them… Drop the observation.

Question 18

What is Over-specification?

Answer 18

Adding variables to the regression that don’t belong there.

Question 19

What is Omitted Variable Bias?

Answer 19

(AKA: under-specification): leaving out variables that should be in the model.

Question 20

What is Endogeneity?

Answer 20

When an x variable is correlated with the error term.

Question 21

Interpret a Log-Log model

Answer 21

On average, a one percent change in Xi results in a Beta % change in Y, ceteris paribus.

Question 22

Interpret a Log-Lin model

Answer 22

On average, a one unit change in Xi results in a (Beta*100)% change in Y, ceteris paribus.

Question 23

Interpret a Lin-Log model

Answer 23

On average, a one percent change in Xi results in a (Beta/100) unit change in Y, ceteris paribus.