1) Suppose we estimate a simple linear (slope-intercept) model in which the dependent variable is the number of analysts following a stock (Y) and the independent variables is the size of the firm (X). The coefficient of determination for this regression is 0.81, and we know that these variables are positively correlated. What is the correlation coefficient for these variables?

A. Corr(Y, X) = 0.64

B. Corr(Y, X) = 0.90

C. Corr(Y, X) = -0.64

D. Corr(Y, X) = -0.90

2) If we multiply the dependent variable in a regression model by positive constant k, then the estimated regression coefficients and their estimated standard errors are multiplied by k. For example, suppose we estimate a model of the price achieved by a stock during its initial public offering (IPO), and the price is stated in dollars per share. Then, we multiply the dependent variable by k=100 to form a model of the IPO price in cents per share. What impact does this transformation have on the t-statistics in which the hypothesized coefficient values are zero, and what happens to the confidence intervals based on these coefficient estimates?

A. The t-statistics and the confidence interval bounds remain unchanged

B. The confidence interval bounds remain unchanged, but the t-statistics are k=100 times larger after the transformation.

C. The confidence interval bounds remain unchanged, but the t-statistic values are 1/100 of their previous value.

D. The t-statistics remain unchanged, but the midpoint and outer boundaries of the confidence interval are multiplied by k=100

3) What happens to the width of a prediction interval if we increase the confidence level from 95 percent to 99 percent?

A. Interval width does not change

B. Interval becomes narrower

C. Interval become wider

D. Interval become wider for positive dependent variables and narrower for negative dependent variables

4) Suppose we have estimated a linear regression model with four independent variables, and we add a fifth variable to the model. What are the possible outcomes for our measures of model goodness-of-fit?

A. Both R2 and adjusted-R2 may decline

B. R2 does not decline and adjusted-R2 must increase

C. R2 does not decline and adjusted-R2 may increase or decrease

D. Both R2 and adjusted-R2 must increase

5) Commodity futures exchanges may impose limits on the daily change in price for a particular commodity contract. For example, the Chicago Board of Trade currently maintains a 30 cent per gallon limit move for its ethanol contract, so the futures price cannot move up or down by more than 30 cents per gallon from the closing price on the previous day. You want to build a regression model of the factors that influence limit-move days, so you form a binary dependent variable that equal one for limit-move days and equals zero for non-limit-move days. True or false -- a logit or probit model would be appropriate for this research project.

A. True

B. False

6) Which test would you use to check for conditional heteroskedasticity in your fitted regression model?

A. Breusch-Pagan test

B. Durbin-Watson test

C. Adjusted-R2 test

D. F-test of overall regression fit