Question 1

Which is not a standard criterion for assessing a regression model?

Accepted Answer

A) Logic of causation 
B) Overall fit 
C) Degree of collinearity 
D) Binary predictors

Question 2

The regression equation Salary = 25,000 + 3200 YearsExperience + 1400 YearsCollege describes employee salaries at Axolotl Corporation. The standard error is 2600. John has 10 years' experience and 4 years of college. His salary is $66,500. What is John's standardized residual?

Accepted Answer

A) -1.250 
B) -0.240 
C) +0.870 
D) +1.500

Question 3

In a regression with 60 observations and 7 predictors, there will be _____ residuals.

Accepted Answer

A) 60 
B) 59 
C) 52 
D) 6

Question 4

Which statement is incorrect?

Accepted Answer

A) Positive autocorrelation results in too many centerline crossings in the residual plot over time. 
B) The R2 statistic can only increase (or stay the same) when you add more predictors to a regression. 
C) If the F-statistic is insignificant, the t statistics for the predictors also are insignificant at the same α. 
D) A regression with 60 observations and 5 predictors does not violate Evans' Rule.

Question 5

A useful guideline in determining the extent of collinearity in a multiple regression model is:

Accepted Answer

A) Sturges' Rule. 
B) Klein's Rule. 
C) Occam's Rule. 
D) Pearson's Rule.

Question 6

If R2 and R2adj differ greatly, we should probably add a few predictors to improve the fit.

Accepted Answer

A) True 
 B)False

Question 7

The residual plot below suggests which violation(s) of regression assumptions?

Accepted Answer

A) Autocorrelation 
B) Heteroscedasticity 
C) Non-normality 
D) Multicollinearity

Question 8

The ill effects of heteroscedasticity might be mitigated by redefining totals (e.g., total number of homicides) as relative values (e.g., homicide rate per 100,000 population).

Accepted Answer

A) True 
 B)False

Question 9

If a regression model's F test statistic is Fcalc = 43.82, we could say that the explained variance is approximately 44 percent.

Accepted Answer

A) True 
 B)False

Question 10

Statisticians who work with cross-sectional data generally do not anticipate autocorrelation.

Accepted Answer

A) True 
 B)False

Question 11

In a multiple regression all of the following are true regarding residuals except:

Accepted Answer

A) their sum always equals zero. 
B) they are the differences between observed and predicted values of the response variable. 
C) they may be used to detect multicollinearity. 
D) they may be used to detect heteroscedasticity.

Question 12

When we have no prior guidance on which combination of predictors is best, we might consider:

Accepted Answer

A) Log transformations. 
B) Decimal transformations. 
C) Logistic regression. 
D) Best subsets regression.

Question 13

A high leverage observation will have:

Accepted Answer

A) an unusual value of the observed Y. 
B) unusual values of one or more X's. 
C) a large standardized residual. 
D) high correlations between the X's.

Question 14

Non-normal residuals lead to biased estimates of the coefficients in a regression model.

Accepted Answer

A) True 
 B)False

Question 15

A log transformation might be appropriate to alleviate which problem(s)?

Accepted Answer

A) Heteroscedastic residuals 
B) Multicollinearity 
C) Autocorrelated residuals

Question 16

Analyze the regression below (n = 50 U.S. states) using the concepts you have learned about multiple regression. Circle things of interest and write comments in the margin. Make a prediction for Poverty for a state with Dropout = 15, TeenMom = 12, Unem = 4, and Age65% = 12 (show your work). The variables are Poverty = percentage below the poverty level; Dropout = percentage of adult population that did not finish high school; TeenMom = percentage of total births by teenage mothers; Unem = unemployment rate, civilian labor force; and Age65% = percentage of population aged 65 and over.

Accepted Answer

The answer of Analyze the regression below (n = 50...

Question 17

If the residuals in your regression are non-normal, a larger sample size might help improve the reliability of confidence intervals for Y.

Accepted Answer

A) True 
 B)False

Question 18

A common misinterpretation of the principle of Occam's Razor is that a simple regression model (rather than a multiple regression model) is always best.

Accepted Answer

A) True 
 B)False

Question 19

The model Y = β0 + β1X + β2X2 cannot be estimated by Excel because of the nonlinear term.

Accepted Answer

A) True 
 B)False

Question 20

An observation with extreme values in one or more independent variables (predictors):

Accepted Answer

A) has reduced influence on the estimated coefficients. 
B) has increased influence on the estimated coefficients. 
C) causes an overstated estimate of the standard error. 
D) leads to high correlations between the X's.

Exam 13: Multiple Regression

Which is not a standard criterion for assessing a regression model?

The regression equation Salary = 25,000 + 3200 YearsExperience + 1400 YearsCollege describes employee salaries at Axolotl Corporation. The standard error is 2600. John has 10 years' experience and 4 years of college. His salary is $66,500. What is John's standardized residual?

In a regression with 60 observations and 7 predictors, there will be _____ residuals.

Which statement is incorrect?

A useful guideline in determining the extent of collinearity in a multiple regression model is:

If R² and R²_adj differ greatly, we should probably add a few predictors to improve the fit.

The residual plot below suggests which violation(s) of regression assumptions?

The ill effects of heteroscedasticity might be mitigated by redefining totals (e.g., total number of homicides) as relative values (e.g., homicide rate per 100,000 population).

If a regression model's F test statistic is F_calc = 43.82, we could say that the explained variance is approximately 44 percent.

Statisticians who work with cross-sectional data generally do not anticipate autocorrelation.

In a multiple regression all of the following are true regarding residuals except:

When we have no prior guidance on which combination of predictors is best, we might consider:

A high leverage observation will have:

Non-normal residuals lead to biased estimates of the coefficients in a regression model.

A log transformation might be appropriate to alleviate which problem(s) ?

If the residuals in your regression are non-normal, a larger sample size might help improve the reliability of confidence intervals for Y.

A common misinterpretation of the principle of Occam's Razor is that a simple regression model (rather than a multiple regression model) is always best.

The model Y = β₀ + β₁X + β₂X² cannot be estimated by Excel because of the nonlinear term.

An observation with extreme values in one or more independent variables (predictors) :