. Suppose a multivariate regression resulted in R2 = 0.71. What can we say about the sums of

squares in this regression?

ExpSS < RSS < TSS

RSS < ExpSS < TSS

RSS < TSS < ExpSS

ExpSS < TSS < RSS 2. What is true about the adjusted R2?

It can be greater than the regular R2.

It can be negative.

It can be greater than one.

It will always increase if a regressor is added to the model. 3. Consider the model . In which of the following cases will we be unable to compute the OLS estimators

,

, and

a. Corr[ y , u ] = 1

b. Corr[ y , x ] = 1

c. Corr[ x , u ] = 1

d. Corr[ x , z ] = 1 ? 4. Suppose we are trying to decide whether to use the model

the model

seen are R2, or . Three model selection devices that we have

, and information criteria. Which of these devices may be used here?

a. All of them work fine.

b. We should not use R2, but and information criteria work fine. c. We should not use R2 or

, but information criteria work fine.

d. None of them should be used here. 5. Consider the model

. What needs to be true in order for the

total effect of x on y to be equal to the partial effect of x on y?

a. and Corr[ x , z ] = 0. b. , it does not matter what Corr[ x , z ] is. c. Corr[ x , z ] = 0, it does not matter what is. d. One of

and Corr[ x , z ] needs to be zero, it does not matter which one.

6. Which of the following is NOT always equal to zero in the multivariate

regression model?

The correlation between the fitted values and the

residuals.

The correlation between the residuals and any of the

regressors.

The correlation between the fitted values and any of

the regressors.

The sum of all residuals. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.

. Suppose a multivariate regression resulted in R2 = 0.71. What can we say about the sums of

squares in this regression?

ExpSS < RSS < TSS

RSS < ExpSS < TSS

RSS < TSS < ExpSS

ExpSS < TSS < RSS 2. What is true about the adjusted R2?

It can be greater than the regular R2.

It can be negative.

It can be greater than one.

It will always increase if a regressor is added to the model. 3. Consider the model . In which of the following cases will we be unable to compute the OLS estimators

,

, and

a. Corr[ y , u ] = 1

b. Corr[ y , x ] = 1

c. Corr[ x , u ] = 1

d. Corr[ x , z ] = 1 ? 4. Suppose we are trying to decide whether to use the model

the model

seen are R2, or . Three model selection devices that we have

, and information criteria. Which of these devices may be used here?

a. All of them work fine.

b. We should not use R2, but and information criteria work fine. c. We should not use R2 or

, but information criteria work fine.

d. None of them should be used here. 5. Consider the model

. What needs to be true in order for the

total effect of x on y to be equal to the partial effect of x on y?

a. and Corr[ x , z ] = 0. b. , it does not matter what Corr[ x , z ] is. c. Corr[ x , z ] = 0, it does not matter what is. d. One of

and Corr[ x , z ] needs to be zero, it does not matter which one.

6. Which of the following is NOT always equal to zero in the multivariate

regression model?

The correlation between the fitted values and the

residuals.

The correlation between the residuals and any of the

regressors.

The correlation between the fitted values and any of

the regressors.

The sum of all residuals. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.

. Suppose a multivariate regression resulted in R2 = 0.71. What can we say about the sums of

squares in this regression?

ExpSS < RSS < TSS

RSS < ExpSS < TSS

RSS < TSS < ExpSS

ExpSS < TSS < RSS 2. What is true about the adjusted R2?

It can be greater than the regular R2.

It can be negative.

It can be greater than one.

It will always increase if a regressor is added to the model. 3. Consider the model . In which of the following cases will we be unable to compute the OLS estimators

,

, and

a. Corr[ y , u ] = 1

b. Corr[ y , x ] = 1

c. Corr[ x , u ] = 1

d. Corr[ x , z ] = 1 ? 4. Suppose we are trying to decide whether to use the model

the model

seen are R2, or . Three model selection devices that we have

, and information criteria. Which of these devices may be used here?

a. All of them work fine.

b. We should not use R2, but and information criteria work fine. c. We should not use R2 or

, but information criteria work fine.

d. None of them should be used here. 5. Consider the model

. What needs to be true in order for the

total effect of x on y to be equal to the partial effect of x on y?

a. and Corr[ x , z ] = 0. b. , it does not matter what Corr[ x , z ] is. c. Corr[ x , z ] = 0, it does not matter what is. d. One of

and Corr[ x , z ] needs to be zero, it does not matter which one.

6. Which of the following is NOT always equal to zero in the multivariate

regression model?

The correlation between the fitted values and the

residuals.

The correlation between the residuals and any of the

regressors.

The correlation between the fitted values and any of

the regressors.

The sum of all residuals. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.

squares in this regression?

ExpSS < RSS < TSS

RSS < ExpSS < TSS

RSS < TSS < ExpSS

ExpSS < TSS < RSS 2. What is true about the adjusted R2?

It can be greater than the regular R2.

It can be negative.

It can be greater than one.

It will always increase if a regressor is added to the model. 3. Consider the model . In which of the following cases will we be unable to compute the OLS estimators

,

, and

a. Corr[ y , u ] = 1

b. Corr[ y , x ] = 1

c. Corr[ x , u ] = 1

d. Corr[ x , z ] = 1 ? 4. Suppose we are trying to decide whether to use the model

the model

seen are R2, or . Three model selection devices that we have

, and information criteria. Which of these devices may be used here?

a. All of them work fine.

b. We should not use R2, but and information criteria work fine. c. We should not use R2 or

, but information criteria work fine.

d. None of them should be used here. 5. Consider the model

. What needs to be true in order for the

total effect of x on y to be equal to the partial effect of x on y?

a. and Corr[ x , z ] = 0. b. , it does not matter what Corr[ x , z ] is. c. Corr[ x , z ] = 0, it does not matter what is. d. One of

and Corr[ x , z ] needs to be zero, it does not matter which one.

6. Which of the following is NOT always equal to zero in the multivariate

regression model?

The correlation between the fitted values and the

residuals.

The correlation between the residuals and any of the

regressors.

The correlation between the fitted values and any of

the regressors.

The sum of all residuals. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.

squares in this regression?

ExpSS < RSS < TSS

RSS < ExpSS < TSS

RSS < TSS < ExpSS

ExpSS < TSS < RSS 2. What is true about the adjusted R2?

It can be greater than the regular R2.

It can be negative.

It can be greater than one.

It will always increase if a regressor is added to the model. 3. Consider the model . In which of the following cases will we be unable to compute the OLS estimators

,

, and

a. Corr[ y , u ] = 1

b. Corr[ y , x ] = 1

c. Corr[ x , u ] = 1

d. Corr[ x , z ] = 1 ? 4. Suppose we are trying to decide whether to use the model

the model

seen are R2, or . Three model selection devices that we have

, and information criteria. Which of these devices may be used here?

a. All of them work fine.

b. We should not use R2, but and information criteria work fine. c. We should not use R2 or

, but information criteria work fine.

d. None of them should be used here. 5. Consider the model

. What needs to be true in order for the

total effect of x on y to be equal to the partial effect of x on y?

a. and Corr[ x , z ] = 0. b. , it does not matter what Corr[ x , z ] is. c. Corr[ x , z ] = 0, it does not matter what is. d. One of

and Corr[ x , z ] needs to be zero, it does not matter which one.

6. Which of the following is NOT always equal to zero in the multivariate

regression model?

The correlation between the fitted values and the

residuals.

The correlation between the residuals and any of the

regressors.

The correlation between the fitted values and any of

the regressors.

The sum of all residuals. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.

squares in this regression?

ExpSS < RSS < TSS

RSS < ExpSS < TSS

RSS < TSS < ExpSS

ExpSS < TSS < RSS 2. What is true about the adjusted R2?

It can be greater than the regular R2.

It can be negative.

It can be greater than one.

It will always increase if a regressor is added to the model. 3. Consider the model . In which of the following cases will we be unable to compute the OLS estimators

,

, and

a. Corr[ y , u ] = 1

b. Corr[ y , x ] = 1

c. Corr[ x , u ] = 1

d. Corr[ x , z ] = 1 ? 4. Suppose we are trying to decide whether to use the model

the model

seen are R2, or . Three model selection devices that we have

, and information criteria. Which of these devices may be used here?

a. All of them work fine.

b. We should not use R2, but and information criteria work fine. c. We should not use R2 or

, but information criteria work fine.

d. None of them should be used here. 5. Consider the model

. What needs to be true in order for the

total effect of x on y to be equal to the partial effect of x on y?

a. and Corr[ x , z ] = 0. b. , it does not matter what Corr[ x , z ] is. c. Corr[ x , z ] = 0, it does not matter what is. d. One of

and Corr[ x , z ] needs to be zero, it does not matter which one.

6. Which of the following is NOT always equal to zero in the multivariate

regression model?

The correlation between the fitted values and the

residuals.

The correlation between the residuals and any of the

regressors.

The correlation between the fitted values and any of

the regressors.

The sum of all residuals. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.

squares in this regression?

ExpSS < RSS < TSS

RSS < ExpSS < TSS

RSS < TSS < ExpSS

ExpSS < TSS < RSS 2. What is true about the adjusted R2?

It can be greater than the regular R2.

It can be negative.

It can be greater than one.

It will always increase if a regressor is added to the model. 3. Consider the model . In which of the following cases will we be unable to compute the OLS estimators

,

, and

a. Corr[ y , u ] = 1

b. Corr[ y , x ] = 1

c. Corr[ x , u ] = 1

d. Corr[ x , z ] = 1 ? 4. Suppose we are trying to decide whether to use the model

the model

seen are R2, or . Three model selection devices that we have

, and information criteria. Which of these devices may be used here?

a. All of them work fine.

b. We should not use R2, but and information criteria work fine. c. We should not use R2 or

, but information criteria work fine.

d. None of them should be used here. 5. Consider the model

. What needs to be true in order for the

total effect of x on y to be equal to the partial effect of x on y?

a. and Corr[ x , z ] = 0. b. , it does not matter what Corr[ x , z ] is. c. Corr[ x , z ] = 0, it does not matter what is. d. One of

and Corr[ x , z ] needs to be zero, it does not matter which one.

6. Which of the following is NOT always equal to zero in the multivariate

regression model?

The correlation between the fitted values and the

residuals.

The correlation between the residuals and any of the

regressors.

The correlation between the fitted values and any of

the regressors.

The sum of all residuals. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.

squares in this regression?

ExpSS < RSS < TSS

RSS < ExpSS < TSS

RSS < TSS < ExpSS

ExpSS < TSS < RSS 2. What is true about the adjusted R2?

It can be greater than the regular R2.

It can be negative.

It can be greater than one.

It will always increase if a regressor is added to the model. 3. Consider the model . In which of the following cases will we be unable to compute the OLS estimators

,

, and

a. Corr[ y , u ] = 1

b. Corr[ y , x ] = 1

c. Corr[ x , u ] = 1

d. Corr[ x , z ] = 1 ? 4. Suppose we are trying to decide whether to use the model

the model

seen are R2, or . Three model selection devices that we have

, and information criteria. Which of these devices may be used here?

a. All of them work fine.

b. We should not use R2, but and information criteria work fine. c. We should not use R2 or

, but information criteria work fine.

d. None of them should be used here. 5. Consider the model

. What needs to be true in order for the

total effect of x on y to be equal to the partial effect of x on y?

a. and Corr[ x , z ] = 0. b. , it does not matter what Corr[ x , z ] is. c. Corr[ x , z ] = 0, it does not matter what is. d. One of

and Corr[ x , z ] needs to be zero, it does not matter which one.

6. Which of the following is NOT always equal to zero in the multivariate

regression model?

The correlation between the fitted values and the

residuals.

The correlation between the residuals and any of the

regressors.

The correlation between the fitted values and any of

the regressors.

The sum of all residuals. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.

squares in this regression?

ExpSS < RSS < TSS

RSS < ExpSS < TSS

RSS < TSS < ExpSS

ExpSS < TSS < RSS 2. What is true about the adjusted R2?

It can be greater than the regular R2.

It can be negative.

It can be greater than one.

It will always increase if a regressor is added to the model. 3. Consider the model . In which of the following cases will we be unable to compute the OLS estimators

,

, and

a. Corr[ y , u ] = 1

b. Corr[ y , x ] = 1

c. Corr[ x , u ] = 1

d. Corr[ x , z ] = 1 ? 4. Suppose we are trying to decide whether to use the model

the model

seen are R2, or . Three model selection devices that we have

, and information criteria. Which of these devices may be used here?

a. All of them work fine.

b. We should not use R2, but and information criteria work fine. c. We should not use R2 or

, but information criteria work fine.

d. None of them should be used here. 5. Consider the model

. What needs to be true in order for the

total effect of x on y to be equal to the partial effect of x on y?

a. and Corr[ x , z ] = 0. b. , it does not matter what Corr[ x , z ] is. c. Corr[ x , z ] = 0, it does not matter what is. d. One of

and Corr[ x , z ] needs to be zero, it does not matter which one.

6. Which of the following is NOT always equal to zero in the multivariate

regression model?

The correlation between the fitted values and the

residuals.

The correlation between the residuals and any of the

regressors.

The correlation between the fitted values and any of

the regressors.

The sum of all residuals. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.

squares in this regression?

ExpSS < RSS < TSS

RSS < ExpSS < TSS

RSS < TSS < ExpSS

ExpSS < TSS < RSS 2. What is true about the adjusted R2?

It can be greater than the regular R2.

It can be negative.

It can be greater than one.

It will always increase if a regressor is added to the model. 3. Consider the model . In which of the following cases will we be unable to compute the OLS estimators

,

, and

a. Corr[ y , u ] = 1

b. Corr[ y , x ] = 1

c. Corr[ x , u ] = 1

d. Corr[ x , z ] = 1 ? 4. Suppose we are trying to decide whether to use the model

the model

seen are R2, or . Three model selection devices that we have

, and information criteria. Which of these devices may be used here?

a. All of them work fine.

b. We should not use R2, but and information criteria work fine. c. We should not use R2 or

, but information criteria work fine.

d. None of them should be used here. 5. Consider the model

. What needs to be true in order for the

total effect of x on y to be equal to the partial effect of x on y?

a. and Corr[ x , z ] = 0. b. , it does not matter what Corr[ x , z ] is. c. Corr[ x , z ] = 0, it does not matter what is. d. One of

and Corr[ x , z ] needs to be zero, it does not matter which one.

6. Which of the following is NOT always equal to zero in the multivariate

regression model?

The correlation between the fitted values and the

residuals.

The correlation between the residuals and any of the

regressors.

The correlation between the fitted values and any of

the regressors.

The sum of all residuals. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.