PRMIA 8007 Exam Practice Test Instant Access

Question 1

Which of the following can induce a 'multicollinearity' problem in a regression model?

AA large negative correlation between the dependent variable and one of the explanatory variables

BA high positive correlation between the dependent variable and one of the explanatory variables

CA high positive correlation between two explanatory variables

DThe omission of a relevant explanatory variable

Answer : C

Question 2

A linear regression gives the following output:

Figures in square brackets are estimated standard errors of the coefficient estimates.

Which of the following is an approximate 95% confidence interval for the true value of the coefficient of ?

A[0, 1.5]

B[1, 2]

C[0, 3]

DNone of the above

Answer : C

Question 3

A linear regression gives the following output:

Figures in square brackets are estimated standard errors of the coefficient estimates. What is the value of the test statistic for the hypothesis that the coefficient of is zero against the alternative that is less than zero?

A0.125

B2.5

C-1.25

D-2.5

Answer : D

Question 4

A linear regression gives the following output:

Figures in square brackets are estimated standard errors of the coefficient estimates.

What is the value of the test statistic for the hypothesis that the coefficient of is less than 1?

A0.32

B0.64

C0.96

D1.92

Answer : B

Question 5

A simple linear regression is based on 100 data points. The total sum of squares is 1.5 and the correlation between the dependent and explanatory variables is 0.5. What is the explained sum of squares?

A0.75

B1.125

C0.3333

D0.375

Answer : D

Question 6

When the errors in a linear regression show signs of positive autocorrelation, which of the statements below is true?

AThe regression coefficient will be too high and the standard error of the regression coefficient will be understated

BThe regression coefficient will be too low and the standard error of the regression coefficient will be overstated

CThe regression coefficient will be unbiased, but the standard error of the regression coefficient will be understated

DThe regression coefficient will be unbiased, but the standard error of the regression coefficient will be overstated

Answer : D

Question 7

Which of the following is consistent with the definition of a Type I error?

AThe probability of a Type I error is 100% minus the significance level

BA Type I error would have occurred if the performance of a stock was positively correlated with the performance of a hedge fund, but in a linear regression, the hypothesis of positive correlation was rejected

CA Type I error would have occurred if the performance of a stock was positively correlated with the performance of a hedge fund, but in a linear regression, the hypothesis of no correlation was rejected

DA Type I occurs whenever data series are serially correlated

Answer : B

PRMIA Mathematical Foundations of Risk Measurement – 2015 Edition Exam Practice Test