Anda di halaman 1dari 6

# 1. Choose the expression that represents the Least Squares Regression Line.

## A. yi = 0 + 1Xi + 2X2i + 3X3i + i

B. E(Y|X) = 0 + 1X1 + 2X2 + 3X3
C. E(Y|X) = 0 + 1X1 + 2X2 + 3X3 + i
D.

E.

## 2.Choose the expression that represents the Population Model

A. yi = 0 + 1Xi + 2X2i + 3X3i + i
B. E(Y|X) = 0 + 1X1 + 2X2 + 3X3
C. E(Y|X) = 0 + 1X1 + 2X2 + 3X3 + i
D.

E.

## 3. Choose the expression that represents the Regression Equation.

A. yi = 0 + 1Xi + 2X2i + 3X3i + i
B. E(Y|X) = 0 + 1X1 + 2X2 + 3X3
C. E(Y|X) = 0 + 1X1 + 2X2 + 3X3 + i
D.

E.

## 4. When any variable is added to the model, the SST

A. must fall
B. must rise
C. usually rises
D. usually falls
E. stays the same
5. When relevant independent variables are added to the model, the R^2
A. usually rises
B. usually falls
C. stays the same
D. must fall
E. must rise
6. When relevant independent variables are added to the model the adjusted R^2
A. usually rises
B. usually falls
C. stays the same
D. must fall
E. must rise

i

## 7. How does multiple regression find b 1?

A. Regress x2 on all other regressors. Regress y on the residuals from that model
B. Regress y on all other regressors. Regress x 2 on the residuals from that model.
C. Regress y on all other regressors. Regress x 1 on the residuals from that model
D. Regress x1 on all other regressors. Regress y on the residuals from that model
8. Interpret b1
A. The predicted value of y rises by 2.344 units for every 1 unit increase in x 1, holding x2, x3 & x4
constant.
B. The predicted value of y is 16.7 units when x 1 = x2 = x3 = x4 = 0
C. The predicted value of y rises by 16.7 units, holding x 1, x2, x3, & x4 constant
D. The predicted value of y is 2.344 units higher when x 1 = 1, holding x2, x3 & x4 constant.
9. If the model has multicollinearity, then which of these is occurring?
A. x1 is very related to Y
B. x1 is not related to Y
C. x1 is very related to x2
D. x1 is not related to x2
10. Multicollinearity from the previous problem can make x 1 appear
A. insignificant when it is actually significant
B. significant when it is actually insignificant
C. non-linear when it is actually linear
D. linear when it is actually non-linear
11. If the researcher believes the effect of x 1 on the dependent variable is curvilinear, they
should add which of these as a regressor
A. x1
B. a dummy variable
C. x2
D. x1 *x2
12. If the researcher believes the effect of x 1 on the dependent variable depends on the value of
x2, they should add which of these as a regressor
A. x1
B. a dummy variable
C. x2
D. x1 *x2
13. Add a quadratic term to the model when the data is
A. curvilinear
B. qualitative
C. quantitative
D. interacting

## 14. Add dummy variables to a model when the data is

A. curvilinear
B. qualitative

C. quantitative
D. interacting
Use the following scenario for problems 15-17. Administrators want to know if ACT
scores are different across school districts holding the parents income constant. The
four school districts are the City district, Suburban District, County District, and the
Rural District.
15. Which of these would be a correct population model?
A. ACTi = 0 + 1Income1 + 2SchoolDistrict + i
B. ACTi = 0 + 1Income1 + 2City + 3Suburban + 4County + 5Rural +
C. ACTi = 0 + 1City1 + 2Suburban + 3County + 4Rural + i
D. ACTi = 0 + 1City1 + 2Suburban + 3County + i
E. ACTi = 0 + 1Income1 + 2City + 3Suburban + 4County + i

16. Interpret 2
A. 2 is the change in the expected ACT scores for each School District, holding income constant
B. 2 is the change in the expected ACT scores for the City School District, holding income
constant.
C. 2 is the change in the expected ACT scores for the Suburban School District, holding income
constant.
D. 2 is the difference in the expected ACT scores between the Suburban School District and the
Rural School District, holding income constant.
E. 2 is the difference in the expected ACT scores between the City School District and the Rural
School District, holding income constant.
17. If we wanted to test if parents income had a larger or smaller effect on ACT score for each
different school district, which regressors would we add to the correct population model in #15?
A. income*City, income*Suburban, income*County, and income*Rural
B. income*City, income*Suburban, and income*County
C. income*SchoolDistrict
D. income^2
This model seeks to explain variation in residential home prices. The dependent
variable is the homes sale price. SQFT is the square feet of the house MEASURED IN
HUNDREDS. Baths is the number of bathrooms in the house. Houses were located in
either City school districts, County school districts or Suburban school districts,
where the suburban school district is the omitted category
18. Predict the sale price for a 2000 SQFT (SQFT = 20) with 2 baths in the City school district.
(Round to the nearest thousand).
A. \$277,000
B. \$222,000
C. \$186,000
D. \$264,000
E. \$179,000

19. What is the estimated effect of increasing the number of baths by 1, given the house has 2
baths and 2000 square feet (SQFT = 20) (Round to the nearest hundred dollars)

## A. Predicted price rises by \$21,900

B. Predicted price rises by \$35,000
C. Predicted price rises by \$6,000
D. Predicted price rises by \$37,400
E. Predicted price rises by \$61,200
20. What is the estimated effect of increasing the square feet by 100 (increase SQFT by 1), given
the house has 2 baths and 2,000 SQFT (SQFT = 20). (Round to the nearest dollar).
A. Predicted price rises by \$21,900
B. Predicted price rises by \$35,000
C. Predicted price rises by \$6,000
D. Predicted price rises by \$37,400
E. Predicted price rises by \$61,200
21. Choose the correct null hypothesis for testing the entire model for significance.
A. Ho : x1 = x2 = x3 = x4 = x5 = x6 = x7 = 0
B. Ho : 0 = 1 = 2 = 3 = 4 = 5 = 6 = 7 = 0
C. Ho : 1 = 2 = 3 = 4 = 5 = 6 = 7 = 0
D. Ho : at least one Ho : 0 = 0
22. When testing the entire model for significance, what is the test statistic (round to 1 decimal)
A. 5.6
B. 9.1
C. 30.9
D. 26.7
E. 1.9
23. When testing the entire model for significance, what is the critical value with alpha = 0.05?
A. 2.37
B. 2.13
C. 2.20
D. 2.40
E. 2.29
24. When testing the entire model for significance with alpha = 0.05, what is your conclusion
A. Reject Ho, so the model is insignificant
B. Do not reject Ho, so the model is insignificant
C. Do not reject Ho, so the model is significant
D. Reject Ho, so the model is significant
25. If Ho : 2 = 0 is true, then
A. there is a relationship between SQFT^2 and sales price
B. there is no relationship between SQFT^2 and sales price
C. there is a relationship between SQFT(00s) and sales price
D. these is no relationship between SQFT(00s) and sales price
26. When testing Ho : 2 = 0, what is the test statistic? (round to 2 decimals)
A. 0.57
B. -2.25
C. 1.77

D. -0.44
E. none of the above
27. When testing Ho : 2 = 0 with alpha = 0.05, what is your conclusion?
A. Do not reject Ho, so the regressor is insignificantly related to Y
B. Reject Ho, so the regressor is significantly related to Y
C. Reject Ho, so the regressor is insignificantly related to Y
D. Do not reject Ho, so the regressor is significantly related to Y
28. The individual regressor SQFTxBATHS is
A. insignificant with alpha = .10 and significant with alpha = 0.05
B. insignificant with both alpha = .10 and alpha = 0.05
C. significant with alpha = .10 and insignificant with alpha = 0.05
D. significant with both alpha = .10 and alpha = 0.05
29. With alpha = 0.05, the individual regressor City School is
A. insignificant because the confidence interval does not include 0
B. insignificant because the confidence interval includes 0
C. significant because the confidence interval does not include 0
D. significant because the confidence interval does include 0
30. With alpha = 0.05, the individual regressor County School is
A. insignificant because the confidence interval does not include 0
B. insignificant because the confidence interval includes 0
C. significant because the confidence interval does not include 0
D. significant because the confidence interval does include 0
31. Rank the school districts from highest to lowest based on the predicted house sale price:
A. City, Suburban, County
B. County, Suburban, City
C. Suburban, County, City
D. City, County, Suburban
32. Calculate the adjusted R^2 (round to 1 decimal place)
A. 21.1%
B. 10.1%
C. 75.9%
D. 78.9%
E. none of the above

33. SQFT(00s) was insignificantly related to the houses sale price, probably because of this
reason
A. SQFT(00s) is highly correlated with SQFT^2 and BATHxSQFT
B. alpha was too large
C. square footage was measured in hundreds
D. there are too many regressors in the model
E. a houses square footage is typically not related to price.

## 34. Which of these are NOT affected by multicollinearity

A. t-test
B. f-test
C. standard error of the slops
D. all of the above are affected by multicollinearity
E. none of the above are affected by multicollinearity
35. Which regressor(s) should be added to the test if the effect of SQFT on sale price was
difference for each school district?
A. SQFT*City
B. SQFT*County
C. SQFT*Suburban
D. (a) and (b)
E. (a), (b) and (c)