Bus Stats Ch15 PDF

Exam Chapter 15
TRUE/FALSE. Write 'T' if the statement is true and 'F' if the statement is false.
1) If in the multiple linear model the slope coefficient βi is negative, it suggests an inverse T or F
(negative) relationship between the explanatory variable xi and the response variable y.
Answer: True False
2) The test statistic for testing the individual significance is assumed to follow the t T or F
distribution with n – k – 2 degrees of freedom, where n is the sample size and k is the
number of explanatory variables.
Answer: True False
3) The restricted model is a reduced model where we estimate the coefficients that are T or F
restricted under the null hypothesis.
Answer: True False
4) The term BLUE stands for Best Linear Unbiased Estimator. T or F

Answer: True False
5) When some explanatory variables of a regression model are strongly correlated, this T or F
phenomenon is called serial correlation.
Answer: True False
6) A crucial assumption in a regression model is that the error term is not correlated with any T or F
of the explanatory variables.
Answer: True False
7) A wavelike movement in residuals suggests positive serial correlation. T or F

Answer: True False
8) The term multicollinearity refers to the condition when the variance of the error term, T or F
conditional on x1, x2, …, xn, is the same for all observations.
Answer: True False
9) The alternative hypothesis for the test of joint significance is specified as: HA: At least T or F
two βj ≠ 0.
Answer: True False
Exam Chapter 15 pg-1

SHORT ANSWER. Write the word or phrase that best completes each statement or answers the
question.
10) The F test can be applied for any number of linear restrictions; the resulting test 10)
is often referred to as the ________ F test.
Answer: partial
11) The ________ model is a complete model that imposes no restrictions on the 11)
coefficients.
Answer: unrestricted
12) With the partial F test, we basically analyze the ratio of (SSER - SSEU) to 12)
________.
Answer: SSEU
13) If the confidence interval does not contain the value zero, then the ________ 13)
variable associated with the regression coefficient is significant.
Answer: explanatory
14) It is common to refer to the interval estimate for an individual value of y as the 14)
________ interval.
Answer: prediction
15) One of the required assumptions of regression analysis states that the error term 15)
has a(an) ________ value of zero.
Answer: expected
16) Conditional on x1, x2, ..., xk, the error term ε is ________ distributed. 16)
Answer: normally
17) ________ plots can be used to detect common violations, and they can be used 17)
to detect outliers.
Answer: Residual
MULTIPLE CHOICE. Choose the one alternative that best completes the statement or answers the
question.
18) Consider the following simple linear regression model: y = β0 + β1x + ε. When 18)
determining whether x significantly influences y, the null hypothesis takes the form
________.
A) H0:β1 = 0 B) H0:β1 = 1 C) H0:b1 = 0 D) H0:b1 = 1
Answer: A

determining whether there is a one-to-one relationship between x and y, the null
hypothesis takes the form ________.
A) H0:β1 = 0 B) H0:β1 = 1 C) H0:b1 = 0 D) H0:b1 = 1
Answer: B
determining whether there is a positive linear relationship between x and y, the
alternative hypothesis takes the form ________.
A) HA:β1 = 0 B) HA:β1 > 1 C) HA:β1 < 0 D) HA:b1 > 1
Answer: B
determining whether there is a negative linear relationship between x and y, the
alternative hypothesis takes the form ________.
A) HA:β1 = 0 B) HA:β1 > 0 C) HA:b1 < 0 D) HA:b1 > 0
Answer: C
22) Excel and virtually all other statistical packages report the p-value ________. 22)
A) for a two-tailed test that assesses whether the regression coefficient differs from
one
B) for a right-tailed test that assesses whether the regression coefficient is greater
than zero
C) for a two-tailed test that assesses whether the regression coefficient differs from
zero
D) for a left-tailed test that assesses whether the regression coefficient is less than
zero
Answer: C
23) Which of the following is the 95% confidence interval for the regression coefficient β1, 23)
if df = 30, b1 = −2 and
A) [−5.00, −1.00] B) [−7.88, 3.88] C) [−7.09, 3.09] D) [−8.13, 4.13]
Answer: D

24) The accompanying table shows the regression results when estimating y = β0 + β1x + ε. 24)
Coefficients Standard Error t-stat p-value

Intercept 0.083 3.56 0.9822
x 1.417 0.63 0.0745
Which of the following is the value of the test statistic when testing whether x significantly
influences y?
A) 0.66 B) 1.42 C) 1.96 D) 2.25
Answer: D

Intercept 0.083 3.56 0.02 0.9822
x 1.417 0.63 2.25 0.0745
Is x significantly related to y at the 5% significance level?

A) Yes, because the p-value of 0.0745 is greater than 0.05.
B) No, because the p-value of 0.0745 is greater than 0.05.
C) Yes, because the slope coefficient of 1.417 is less than the test statistic of 2.25.
D) No, because the slope coefficient of 1.417 is less than the test statistic of 2.25.
Answer: B

Intercept 0.083 3.56 0.9822
x 1.417 0.63 0.0745
When testing whether the slope coefficient differs from 1, the value of the test statistic is
________.
A) 0.66 B) 1.42 C) 1.96 D) 2.25
Answer: A

27) Given the following portion of regression results, which of the following conclusions is 27)
true with regard to the F test at the 5% significance level?
df SS MS F Significance F
Regression 2 7,562 3,781 8.04 0.0028
Residual 20 9,395 470
Total 22 16,957
A) Neither of the explanatory variables is related to the response variable.

B) Both of the explanatory variables are related to the response variable.
C) Exactly one of the explanatory variables is related to the response variable.
D) At least one of the explanatory variables is significantly related to the response
variable.
Answer: D
28) Given the following portion of regression results, which of the following is the value of 28)
the F2,20 test statistic?
Regression 2 7,562 3,781 0.0028
Residual 20 9,395 470
Total 22 16,957
A) 1.96 B) 3.49 C) 8.04 D) 10

Answer: C
29) Refer to the portion of regression results in the accompanying table. 29)
Regression 2 7,562 3,781 0.0028
Residual 20 9,395 470
Total 22 16,957
When testing the overall significance of the regression model at the 5% level given a critical
value of F0.05,(2.20) = 3.49, the decision is to ________.
A) reject H0 and conclude that the explanatory variables are jointly significant.
B) not reject H0 and conclude that the explanatory variables are jointly significant.
C) reject H0 and conclude that the explanatory variables are not jointly significant.
D) not reject H0 and conclude that the explanatory variables are not jointly
significant.
Answer: A

30) Given the following portion of regression results, which of the following conclusions is 30)
true with regard to the F test at the 5% significance level?
Regression 2 3,500 1,750 0.1000
Residual 20 13,500 675
Total 22 17,000
A) Neither of the explanatory variables is significantly related to the response

variable.
B) Both of the explanatory variables are significantly related to the response variable.
C) Exactly one of the explanatory variables is significantly related to the response
variable.
D) At least one of the explanatory variables is significantly related to the response
variable.
Answer: A
31) Given the following portion of regression results, which of the following is the value of 31)
the F(2,20) test statistic?
Regression 2 3,500 1,750 0.1000
Residual 20 13,500 675
Total 22 17,000
A) 1.96 B) 2.59 C) 3.49 D) 10

Answer: B

32) Refer to the portion of regression results in the accompanying table. 32)
Regression 2 3,500 1,750 0.1000
Residual 20 13,500 675
Total 22 17,000
When testing the overall significance of the regression model at the 5% level given a critical
value of F0.05,(2,20) = 3.49, the decision is to ________.
A) reject H0 and conclude that the explanatory variables are jointly significant
B) not reject H0 and conclude that the explanatory variables are jointly significant
C) reject H0 and conclude that the explanatory variables are not jointly significant
D) not reject H0 and conclude that the explanatory variables are not jointly significant
Answer: D
33) The accompanying table shows the regression results when estimating y = β0 + β1x1 + 33)
β2x2 + β3x3 + ε.
Regression 3 453 151 5.03 0.0030
Residual 85 2,521 30
Total 88 2,974
Intercept 14.96 3.08 4.80 0.0000
x1 0.87 0.29 3.00 0.0035
x2 0.46 0.22 2.09 0.0400
x3 0.04 0.34 0.12 0.9066
At the 5% significance level, which of the following explanatory variable(s) is(are) individually
significant?
A) Only x1 B) Only x3 C) x1 and x2 D) x2 and x3
Answer: C

34) A sample of 200 monthly observations is used to run a simple linear regression: Returns 34)
= β0 + β1 Leverage + ε. A 5% level of significance is used to study if leverage has a
significant influence on returns. The value of the test statistic for the regression
coefficient of Leverage is calculated as t198 = –1.09, with an associated p-value of
0.2770. The correct decision is to ________.
A) reject the null hypothesis and conclude that leverage significantly explains returns
B) reject the null hypothesis and conclude that leverage does not significantly explain
returns
C) not reject the null hypothesis and conclude that leverage does not significantly
explain returns
D) not reject the null hypothesis and conclude that leverage significantly explains
returns
Answer: C
35) When estimating y = β0 + β1x1 + β2x2 + β3x3 + ε, you wish to test H0: β1 = β2 = 0 35)
versus HA: At least one βi = 0. The value of the test statistic is F(2,20) = 2.50 and its
associated p-value is 0.1073. At the 5% significance level, the conclusion is to
________.
A) reject the null hypothesis and conclude that x1 and x2 are jointly significant
B) reject the null hypothesis and conclude that x1 and x2 are not jointly significant
C) not reject the null hypothesis and conclude that x1 and x2 are jointly significant
D) not reject the null hypothesis and conclude that x1 and x2 are not jointly
significant
Answer: D
36) When testing r linear restrictions imposed on the model y = β0 + β1x1 + ... + βkxk + ε, 36)
the test statistic is assumed to follow the F(df1, df2) distribution with ________.
A) df1 = k and df2 = n – k – 1 B) df1 = k – 1 and df2 = n – k – 1
C) df1 = r and df2 = n – k D) df1 = r and df2 = n – k – 1
Answer: D

37) When testing linear restrictions, the test statistic is assumed to follow the y = β0 + β1x1 37)
+ β2x2 + β3x3 + ε, you wish to test H0: β1 = β2 versus HA: β1 ≠ β2 The value of the
test statistic is F(1,30) = 7.75 and its associated p-value is 0.0092. At the 1%
significance level, the conclusion is to ________.
A) reject the null hypothesis and conclude that the influence of x1 and x2 on y is the
same
B) reject the null hypothesis and conclude that the influence of x1 and x2 on y is not
the same
C) not reject the null hypothesis and conclude that the influence of x1 and x2 on y is
the same
D) not reject the null hypothesis and conclude that the influence of x1 and x2 on y is
not the same
Answer: B
38) In regression, the predicted values concerning y are subject to ________. 38)
A) kurtosis B) skewness
C) mean deviation D) sampling variation
Answer: D
39) In regression, the two types of interval estimates concerning y are called ________. 39)
A) significance interval and prediction interval
B) confidence interval and significance interval
C) confidence interval and prediction interval
D) significance interval and frequency interval
Answer: C
40) For a given confidence level, the prediction interval is always wider than the confidence 40)
interval because ________.
A) the confidence interval is for a particular value of y rather than for the expected
value E(y)
B) the prediction interval is for a particular value of y rather than for the expected
value E(y)
C) the confidence interval is for a particular value of x rather than for the expected
value E(x)
D) the prediction interval is for a particular value of x rather than for the expected
value E(x)
Answer: B

41) In regression, multicollinearity is considered problematic when two or more explanatory 41)
variables are ________.
A) not correlated B) rarely correlated
C) highly correlated D) moderately correlated
Answer: C
42) Multicollinearity is suspected when ________. 42)

A) there is a low R2 coupled with significant explanatory variables
B) there is a high R2 coupled with significant explanatory variables
C) there is a low R2 coupled with insignificant explanatory variables
D) there is a high R2 coupled with insignificant explanatory variables
Answer: D
43) One of the assumptions of regression analysis is ________. 43)

A) there is perfect multicollinearity among some explanatory variables
B) there is no perfect multicollinearity among any explanatory variables
C) there is some degree of multicollinearity among some explanatory variables
D) None of these choices is correct.
Answer: B
44) Which of the following violates the assumptions of regression analysis? 44)
A) The error term is normally distributed.
B) The error term has a zero mean.
C) The error term is correlated with an explanatory variable.
D) The error term has a constant variance.
Answer: C
45) When confronted with multicollinearity, a good remedy is to ________ if we can justify 45)
its redundancy.
A) add one more collinear variable
B) drop one of the collinear variables
C) remove both the collinear variables
D) add as many collinear variables as possible
Answer: B
46) If the variance of the error term is not the same for all observations, we ________. 46)
A) get estimators that are biased
B) can perform tests of significance
C) cannot conduct tests of significance
D) Both get estimators that are biased and cannot conduct tests of significance are
correct.
Answer: C

47) Serial correlation is typically observed in ________. 47)
A) outliers B) sparse data
C) time series data D) cross-sectional data
Answer: C
48) A researcher gathers data on 25 households and estimates the following model: 48)
Expenditure = β0 + β1 Income + ε. A residual plot of the estimated model is shown in
the accompanying graph.
Which of the following can be inferred from the residual plot?

A) The residuals have little variation.
B) The residuals do not exhibit any pattern.
C) The residuals seem to fan out across the horizontal axis.
D) The residuals are evenly distributed across the horizontal axis.
Answer: C
49) Serial correlation occurs when the error term is ________. 49)
A) dispersed evenly B) randomly distributed
C) correlated across observations D) uncorrelated across observations
Answer: C

50) A marketing analyst wants to examine the relationship between sales (in $1,000s) and 50)
advertising (in $100s) for firms in the food and beverage industry and so collects
monthly data for 25 firms. He estimates the model: Sales = β0 + β1 Advertising + ε. The
following table shows a portion of the regression results.

Intercept 40.10 14.08 2.848 0.0052
Advertising 2.88 1.52 −1.895 0.0608
Which of the following are the competing hypotheses used to test whether Advertising is
significant in predicting Sales?
A) H0:b1 = 0; HA:b1 ≠ 0. B) H0:b1 = 2.88; HA:b1 ≠ 2.88.
C) H0:β1 = 0; HA:β1 ≠ 0. D) H0:β0 = 0; HA:β0 ≠ 0.
Answer: C
monthly data for 25 firms. He estimates the model:
Sales = β0 + β1 Advertising + ε. The following table shows a portion of the regression

results.

Intercept 40.10 14.08 2.848 0.0052
Advertising 2.88 1.52 −1.895 0.0608
When testing whether Advertising is significant at the 10% significance level, the conclusion is
to ________.
A) reject H0; advertising is significant
B) not reject H0; advertising is significant
C) reject H0; advertising is not significant
D) not reject H0; advertising is not significant
Answer: A

Sales = β0 + β1 Advertising + ε. The following table shows a portion of the regression results.

Intercept 40.10 14.08 2.848 0.0052
Advertising 2.88 1.52 −1.895 0.0608
Which of the following are the competing hypotheses used to test whether the slope coefficient
differs from 3?
A) H0:b1 = 3; HA:b1 ≠ 3 B) H0:b1 = 2.88; HA:b1 ≠ 2.88
C) H0:β1 = 3; HA:β1 ≠ 3 D) H0:β1 = 2.88; HA:β1 ≠ 2.88
Answer: C

results.

Intercept 40.10 14.08 2.848 0.0052
Advertising 2.88 1.52 –1.895 0.0608
When testing whether the slope coefficient differs from 3, the value of the test statistic is
________.
A) t23 = –1.895 B) t23 = –0.079 C) t23 = 0.079 D) t23 = 1.895
Answer: B


results.

Intercept 40.10 14.08 2.848 0.0052
Advertising 2.88 1.52 −1.895 0.0608
When testing whether the slope coefficient differs from 3, the critical values at the 5%
significance level are −2.069 and 2.069. The conclusion to the test is to ________.
A) reject H0; the slope coefficient differs from 3
B) not reject H0; the slope coefficient differs from 3
C) reject H0; the slope coefficient does not differ from 3
D) not reject H0; the slope coefficient does not differ from 3
Answer: D

55) A sports analyst wants to exam the factors that may influence a tennis player's chances 55)
of winning. Over four tournaments, he collects data on 30 tennis players and estimates
the following model:
Win = β0 + β1 Double Faults + β2Aces + ε, where Win is the proportion of winning,

Double Faults is the percentage of double faults made, and Aces is the number of aces. A
portion of the regression results are shown in the accompanying table.
df SS MS F Significance
F
Regression 2 1.24 0.620 41.85 5.34E-09
Residual 27 0.40 0.015
Total 29 1.64
Coefficients Standard t-stat p-value Lower 95% Upper
Error 95%
Intercept 0.451 0.080 5.646 5.4E-06 0.287 0.614
Double −0.007 0.0024 −2.875 0.0078 −0.012 −0.002
Faults
Aces 0.015 0.003 4.65 7.8E-05 0.008 0.023
Is the relationship between Win and Aces significant at the 5% significance level?
A) No, because the relevant p-value is less than 0.05.
B) Yes, because the relevant p-value is less than 0.05.
C) No, because the relevant p-value is greater than 0.05.
D) Yes, because the relevant p-value is greater than 0.05.
Answer: B

Win = β0 + β1 Double Faults + β2Aces + ε, where Win is the proportion of winning,
F
Regression 2 1.24 0.620 41.85 5.34E-09
Residual 27 0.40 0.015
Total 29 1.64
Error 95%
Intercept 0.451 0.080 5.646 5.4E-06 0.287 0.614
Double −0.007 0.0024 −2.875 0.0078 −0.012 −0.002
Faults
Aces 0.015 0.003 4.65 7.8E-05 0.008 0.023
Excel shows that the 95% confidence interval for β1 is [−0.12, −0.002]. When determining
whether or not Double Faults is significant at the 5% significance level, he ________.
A) rejects H0: β1 = 0, and concludes that Double Faults is significant
B) does not reject H0: β1 = 0, and concludes that Double Faults is significant
C) rejects H0: β1 = 0, and concludes that Double Faults is not significant
D) does not reject H0:β1 = 0, and concludes that Double Faults is not significant
Answer: A

Win = β0 + β1 Double Faults + β2 Aces + ε, where Win is the proportion of winning,

F
Regression 2 1.24 0.620 41.85 5.34E-09
Residual 27 0.40 0.015
Total 29 1.64
Error 95%
Intercept 0.451 0.080 5.646 5.4E-06 0.287 0.614
Double −0.007 0.0024 −2.875 0.0078 −0.012 −0.002
Faults
Aces 0.015 0.003 4.65 7.8E-05 0.008 0.023
When testing whether the explanatory variables jointly influence the response variable, the null
hypothesis is ________.
A) H0:β1 = β2 = 0 B) H0:β1 + β2 = 0
C) H0:β0 = β1 = β2 = 0 D) H0:β0 + β1 + β2 + 0
Answer: A

Win = β0 + β1 Double Faults + β2 Aces + ε, where Win is the proportion of winning,

F
Regression 2 1.24 0.620 41.85 5.34E-09
Residual 27 0.40 0.015
Total 29 1.64
Error 95%
Intercept 0.451 0.080 5.646 5.4E-06 0.287 0.614
Double −0.007 0.0024 −2.875 0.0078 −0.012 −0.002
Faults
Aces 0.015 0.003 4.65 7.8E-05 0.008 0.023
When testing whether the explanatory variables are jointly significant at the 5% significance
level, he ________.
A) rejects H0, and concludes that the explanatory variables are jointly significant
B) does not reject H0, and concludes that the explanatory variables are jointly
significant
C) rejects H0, and concludes that the explanatory variables are not jointly significant
D) does not reject H0, and concludes that the explanatory variables are not jointly
significant
Answer: A

59) A researcher analyzes the factors that may influence amusement park attendance. She 59)
estimates the following model: Attendance = β0 + β1 Price + β2 Temperature + β3
Rides + ε, where Attendance is the daily attendance (in 1,000s), Price is the gate price
(in $), Temperature is the average daily temperature (in °F), and Rides is the number of
rides at the amusement park. A portion of the regression results is shown in the accompanying
table.
F
Regression 3 29,524.41 9,841.47 2.18E-14
Residual 26 2,564.37 98.63
Total 29 32,088.78
Error 95%
Intercept 27.328 40.254 0.5032 −55.415 110.071
Price −1.201 0.294 0.0004 −1.805 −0.598
Temperature 0.008 0.208 0.9693 −0.419 0.435
Rides 3.621 0.364 2.32E-10 2.874 4.369
When testing whether Price is significant in explaining Attendance, the value of the test
statistic is ________.
A) −4.085. B) −1.201. C) 1.201. D) 4.085.
Answer: A

table.
F
Regression 3 29,524.41 9,841.47 2.18E-14
Residual 26 2,564.37 98.63
Total 29 32,088.78
Error 95%
Intercept 27.328 40.254 0.5032 −55.415 110.071
Price −1.201 0.294 0.0004 −1.805 −0.598
Temperature 0.008 0.208 0.9693 −0.419 0.435
Rides 3.621 0.364 2.32E-10 2.874 4.369
When testing whether Temperature is significant at the 5% significance level, she

________.
A) rejects H0: β2 = 0, and concludes that Temperature is significant
B) does not reject H0: β2 = 0, and concludes that Temperature is significant
C) rejects H0:β2 = 0, and concludes that Temperature is not significant
D) does not reject H0:β2 = 0, and concludes that Temperature is not significant
Answer: D

table.
F
Regression 3 29,524.41 9,841.47 2.18E-14
Residual 26 2,564.37 98.63
Total 29 32,088.78
Error 95%
Intercept 27.328 40.254 0.5032 −55.415 110.071
Price −1.201 0.294 0.0004 −1.805 −0.598
Temperature 0.008 0.208 0.9693 −0.419 0.435
Rides 3.621 0.364 2.32E-10 2.874 4.369
When testing whether Rides is significant at the 1% significance level, she ________.
A) rejects H0:β3 = 0, and concludes that Rides is significant
B) does not reject H0:β3 = 0, and concludes that Rides is significant
C) rejects H0:β3 = 0, and concludes that Rides is not significant
D) does not reject H0:β3 = 0, and concludes that Rides is not significant
Answer: A

(in $), Temperature is the average daily temperature (in oF), and Rides is the number of
table.
F
Regression 3 29,524.41 9,841.47 2.18E-14
Residual 26 2,564.37 98.63
Total 29 32,088.78
Error 95%
Intercept 27.328 40.254 0.5032 −55.415 110.071
Price −1.201 0.294 0.0004 −1.805 −0.598
Temperature 0.008 0.208 0.9693 −0.419 0.435
Rides 3.621 0.364 2.32E-10 2.874 4.369
Which of the following is the value of the test statistic for testing the joint significance of the
linear regression model?
A) −4.09 B) 0.04 C) 9.95 D) 99.78
Answer: D

estimates the following model:Attendance = β0 + β1 Price + β2 Temperature + β3
table.
F
Regression 3 29,524.41 9,841.47 2.18E-14
Residual 26 2,564.37 98.63
Total 29 32,088.78
Error 95%
Intercept 27.328 40.254 0.5032 −55.415 110.071
Price −1.201 0.294 0.0004 −1.805 −0.598
Temperature 0.008 0.208 0.9693 −0.419 0.435
Rides 3.621 0.364 2.32E-10 2.874 4.369
level, the researcher ________.
significant
C) rejects H0, and concludes that the explanatory variables are not jointly significant
significant
Answer: A

table.
F
Regression 3 29,524.41 9,841.47 2.18E-14
Residual 26 2,564.37 98.63
Total 29 32,088.78
Error 95%
Intercept 27.328 40.254 0.5032 −55.415 110.071
Price −1.201 0.294 0.0004 −1.805 −0.598
Temperature 0.008 0.208 0.9693 −0.419 0.435
Rides 3.621 0.364 2.32E-10 2.874 4.369
Which of the following are the hypotheses to test if the explanatory variables Temperature
Rides are jointly significant in explaining Attendance?
A) H0:β2 = β3 = 0; HA: At least one of the coefficients is nonzero.
B) H0:β2 = β3 = 0; HA: At least one of the coefficients is zero.
C) H0:β0 = β1 = 0; HA: At least one of the coefficients is nonzero.
D) H0:β0 = β1 = 0; HA: At least one of the coefficients is zero.
Answer: A

table.
F
Regression 3 29,524.41 9,841.47 2.18E-14
Residual 26 2,564.37 98.63
Total 29 32,088.78
Error 95%
Intercept 27.328 40.254 0.5032 −55.415 110.071
Price −1.201 0.294 0.0004 −1.805 −0.598
Temperature 0.008 0.208 0.9693 −0.419 0.435
Rides 3.621 0.364 2.32E-10 2.874 4.369
When testing whether the explanatory variables Temperature and Rides are jointly
significant, the error sum of squares for the restricted model is SSER = 12,343.78.
Which of the following is the value of the test statistic when conducting this test?
A) −4.09 B) 25.33 C) 49.58 D) 99.78
Answer: C

table.
F
Regression 3 29,524.41 9,841.47 2.18E-14
Residual 26 2,564.37 98.63
Total 29 32,088.78
Error 95%
Intercept 27.328 40.254 0.5032 −55.415 110.071
Price −1.201 0.294 0.0004 −1.805 −0.598
Temperature 0.008 0.208 0.9693 −0.419 0.435
Rides 3.621 0.364 2.32E-10 2.874 4.369
When testing whether the explanatory variables Temperature and Rides are jointly significant,
the p-value associated with the test is 0.0000. At the 1% significance level, the
researcher ________.
A) rejects H0, and concludes that at least one of the explanatory variables,
Temperature and Rides, is significant in explaining Attendance
B) does not reject H0, and concludes that at least one of the explanatory variables,
C) rejects H0, and concludes that neither of the explanatory variables, Temperature
and Rides, is significant in explaining Attendance
D) does not reject H0, and concludes that neither of the explanatory variables,
Answer: A

67) A researcher analyzes the factors that may influence amusement park attendance and 67)
estimates the following model: Attendance = β0 + β1 Price + β2 Rides + ε, where
Attendance is the daily attendance (in 1,000s), Price is the gate price (in $), and Rides is
the number of rides at the amusement park. The researcher would like to construct
interval estimates for Attendance when Price and Rides equal $85 and 30, respectively.
The researcher estimates a modified model where Attendance is the response variable
and the explanatory variables are now defined as
Price* = Price – 85 and Rides* = Rides – 30. A portion of the regression results is shown in
the accompanying table.
Coefficients Standard t-stat p-value Lower Upper

Error 95% 95%
Intercept 34.41 4.06 8.48 4.33E-09 26.08 42.74
Price* −1.20 0.28 −4.23 0.0002 −1.79 −0.62
Rides* 3.62 0.36 10.15 1.04E-10 2.89 4.35
According to the modified model, which of the following is the predicted value for Attendance
when Price and Rides equal $85 and 30, respectively?
A) 25,670 B) 34,410 C) 40,910 D) 55,600
Answer: B

estimates the following model:
Attendance = β0 + β1 Price + β2 Rides + ε, where Attendance is the daily attendance (in

1,000s), Price is the gate price (in $), and Rides is the number of rides at the amusement
park. The researcher would like to construct interval estimates for Attendance when
Price and Rides equal $85 and 30, respectively. The researcher estimates a modified
model where Attendance is the response variable and the explanatory variables are now
defined as
Price* = Price – 85 and Rides* = Rides – 30. A portion of the regression results is shown in

Error 95% 95%
Intercept 34.41 4.06 8.48 4.33E-09 26.08 42.74
Price* −1.20 0.28 −4.23 0.0002 −1.79 −0.62
Rides* 3.62 0.36 10.15 1.04E-10 2.89 4.35
According to the modified model, which of the following is a 95% confidence interval for
expected Attendance when Price and Rides equal $85 and 30, respectively? (Note that
t0,025,27 = 2.052.)
A) [12,740, 56,780] B) [16,330, 53,450]
C) [26,080, 42,740] D) [28,900, 41,500]
Answer: C

estimates the following model: Attendance = β0 + β1 Price + β2 Rides + ε, where
Attendance is the daily attendance (in 1,000s), Price is the gate price (in $), and Rides is
the number of rides at the amusement park. The researcher would like to construct
interval estimates for Attendance when Price and Rides equal $85 and 30, respectively.
The researcher estimates a modified model where Attendance is the response variable
and the explanatory variables are now defined as
Price* = Price – 85 and Rides* = Rides – 30. A portion of the regression results is
shown in the accompanying table.
Regression Statistics
Multiple R 0.96
R Square 0.92
Adjusted R Square 0.91
Standard Error 9.75
Observations 30

Error 95% 95%
Intercept 34.41 4.06 8.48 4.33E-09 26.08 42.74
Price* −1.20 0.28 −4.23 0.0002 −1.79 −0.62
Rides* 3.62 0.36 10.15 1.04E-10 2.89 4.35
According to the modified model, which of the following is a 95% prediction interval for
Attendance when Price and Rides equal $85 and 30, respectively? (Note that t0.025,27
= 2.052.)
A) [12,740, 56,080] B) [16,330, 53,450]
C) [26,080, 42,740] D) [28,900, 41,500]
Answer: A

β2x2 + β3x3 + ε.
Regression 3 453 151 5.03 0.0030
Total 88 2,974
Intercept 14.96 3.08 4.86 0.0000
x1 0.87 0.29 3.00 0.0035
x2 0.46 0.22 2.09 0.0400
x3 0.04 0.34 0.12 0.9066
level, the conclusion is to ________.
A) reject H0, and conclude that the explanatory variables are jointly significant
B) not reject H0, and conclude that the explanatory variables are jointly significant
C) reject H0, and conclude that the explanatory variables are not jointly significant
D) not reject H0, and conclude that the explanatory variables are not jointly
significant
Answer: A
β2x2 + β3x3 + ε.
Regression 3 453 151 5.03 0.0030
Total 88 2,974
Intercept 14.96 3.08 4.86 0.0000
x1 0.87 0.29 3.00 0.0035
x2 0.46 0.22 2.09 0.0400
x3 0.04 0.34 0.12 0.9066
significant?
A) Only x1 B) Only x3 C) x1 and x2 D) x2 and x3
Answer: C

β2x2 + β3x3 + ε.
df SS MS F
Regression 3 453 151 5.03
Total 88 2,974
Intercept 14.96 3.08 4.86 0.0000
x1 0.87 0.29 3.00 0.0035
x2 0.46 0.22 2.09 0.0400
x3 0.04 0.34 0.12 0.9066
When testing whether or not x1 and x2 have the same influence on y, the null hypothesis is
________.
A) H0: β1 = β2 B) H0: β1 ≠ β2
C) H0: β1 + β2 = β3 D) H0: β1 + β2 = 0
Answer: A
β2x2 + β3x3 + ε.
df SS MS F
Regression 3 453 151 5.03
Total 88 2,974
Intercept 14.96 3.08 4.86 0.0000
x1 0.87 0.29 3.00 0.0035
x2 0.46 0.22 2.09 0.0400
x3 0.04 0.34 0.12 0.9066
When testing whether or not x1 and x2 are jointly significant, the null hypothesis is
________.
A) H0:β1 = β2 B) H0:β1 ≠ β2
C) H0:β1 + β2 = 0 D) H0:β1 = β2 = 0
Answer: D

74) Tiffany & Co. has been the world's premier jeweler since 1837. The performance of 74)
Tiffany's stock is likely to be strongly influenced by the economy. Monthly data for
Tiffany's risk-adjusted return and the risk-adjusted market return are collected for a
five-year period (n = 60). The accompanying table shows the regression results when
estimating the capital asset pricing model (CAPM) model for Tiffany's return.

Error 95% 95%
Intercept 0.0198 0.010 1.98 0.0598 −0.0008 0.0405
RM – Rf 1.827 0.191 9.58 1.494E-13 1.4456 2.2094
You would like to determine whether an investment in Tiffany's is riskier than the market. When
conducting this test, you set up the following competing hypotheses: ________.
A) H0:α = 0; HA:α ≠ 0 B) H0:β = 0; HA:β ≠ 0
C) H0:α ≤ 1; HA:α > 0 D) H0:β ≤ 1; HA:β > 1
Answer: D
estimating the CAPM model for Tiffany's return.

Error 95% 95%
Intercept 0.0198 0.010 1.98 0.0598 −0.0008 0.0405
RM – Rf 1.827 0.191 9.58 1.494E-13 1.4456 2.2094
When testing whether the beta coefficient is significantly greater than one, the value of the test
statistic is ________.
A) −1.98 B) 1.98 C) 4.33 D) 9.58
Answer: C


Error 95% 95%
Intercept 0.0198 0.010 1.98 0.0598 −0.0008 0.0405
RM – Rf 1.827 0.191 9.58 1.494E-13 1.4456 2.2094
When testing whether the beta coefficient is significantly greater than one, the relevant critical
value at the 5% significance level is t0.05,58 = 1.672. The conclusion to the test is to
________.
A) reject H0, and conclude that the return on Tiffany stock is riskier than the return
on the market
B) not reject H0, and conclude that the return on Tiffany stock is riskier than the
return on the market
C) reject H0, and conclude that the return on Tiffany stock is less risky than the
D) not reject H0, and conclude that the return on Tiffany stock is less risky than the
Answer: A

Error 95% 95%
Intercept 0.0198 0.010 1.98 0.0598 −0.0008 0.0405
RM – Rf 1.827 0.191 9.58 1.494E-13 1.4456 2.2094
To determine whether abnormal returns exist, which of the following competing hypotheses do
you set up?
A) H0:α = 0; HA:α ≠ 0 B) H0:β = 0; HA:β ≠ 0
C) H0:α ≤ 1; HA:α > 1 D) H0:β ≤ 1; HA:β > 1
Answer: A


Error 95% 95%
Intercept 0.0198 0.010 1.98 0.0598 −0.0008 0.0405
RM – Rf 1.827 0.191 9.58 1.494E-13 1.4456 2.2094
When testing whether there are abnormal returns, or whether the alpha coefficient is
significantly different from zero, the value of the test statistic is ________.
A) −1.98 B) 1.98 C) 4.33 D) 9.58
Answer: B

Error 95% 95%
Intercept 0.0198 0.010 1.98 0.0598 −0.0008 0.0405
RM – Rf 1.827 0.191 9.58 1.494E-13 1.4456 2.2094
When testing whether there are abnormal returns, the conclusion to the test is at the 5%
significance level is to ________.
A) reject H0, and conclude there are abnormal returns
B) not reject H0, and conclude there are abnormal returns
C) reject H0, and do not conclude there are abnormal returns
D) not reject H0, and do not conclude there are abnormal returns
Answer: D

80) A manager at a local bank analyzed the relationship between monthly salary (y, in $) 80)
and length of service (x, measured in months) for 30 employees. She estimates the
following model: Salary = β0 + β1 Service + ε. The following table summarizes a portion of
the regression results.
Coefficients Standard t-stat p-value Lower 95% Upper 95%

Error
Intercept 784.92 322.25 124.95 1,444.89
Service 9.19 3.20 2.64 15.74
Which of the following hypotheses will determine whether the intercept differs from zero?
A) H0:β0 = 0; HA:β0 ≠ 0 B) H0:β1 = 0; HA:β1 ≠ 0
C) H0:b0 = 0; HA:b0 ≠ 0 D) H0:b1 = 0; HA:b1 ≠ 0
Answer: A

Error
Intercept 784.92 322.25 124.95 1,444.89
Service 9.19 3.20 2.64 15.74
Which of the hypotheses will determine whether the slope differs from zero?
A) H0:β0 = 0; HA: β0 ≠ 0 B) H0:β1 = 0; HA: β1 ≠ 0
C) H0:b0 = 0; HA: b0 ≠ 0 D) H0:b1 = 0; HA: b1 ≠ 0
Answer: B

Coefficients Standard Error t-stat p-value Lower 95% Upper 95%

Intercept 784.92 322.25 124.95 1,444.89
Service 9.19 3.20 2.64 15.74
Using the 95% confidence interval, which of the following is the conclusion to the following
hypothesis test:
H0:β0 = 0; HA: β0 ≠ 0?
A) At the 5% significance level, reject H0 because the interval contains 0.
B) At the 5% significance level, do not reject H0 because the interval contains 0.
C) At the 5% significance level, reject H0 because the interval does not contain 0.
D) At the 5% significance level, do not reject H0 because the interval does not
contain 0.
Answer: C
Coefficients Standard Error t-stat p-value Lower 95% Upper 95%

Intercept 784.92 322.25 124.95 1,444.89
Service 9.19 3.20 2.64 15.74
Using the 95% confidence interval, what is the conclusion to the following hypothesis test:
H0:b1 = 0; HA: b1 ≠ 0?
A) At the 5% significance level, reject H0 and conclude that service is significant.
B) At the 5% significance level, do not reject H0 and conclude that service is
significant.
C) At the 5% significance level, reject H0 and conclude that service is not significant.
D) At the 5% significance level, do not reject H0 and conclude that service is not
significant.
Answer: A

84) A real estate analyst believes that the three main factors that influence an apartment's 84)
rent in a college town are the number of bedrooms, the number of bathrooms, and the
apartment's square footage. For 40 apartments, she collects data on the rent (y, in $), the
number of bedrooms (x1), the number of bathrooms (x2), and its square footage (x3).
She estimates the following model: Rent = β0 + β1 Bedroom + β2 Bath + β3Sqft + ε.
The following table shows a portion of the regression results.
F
Regression 3 5,694,717 1,898,239 50.88 4.99E-13
Residual 36 1,343,176 37,310
Total 39 7,037,893
Error 95%
Intercept 300 84.0 3.57 0.0010 130.03 470.79
Bedroom 226 60.3 3.75 0.0006 103.45 348.17
Bath 89 55.9 1.59 0.1195 −24.24 202.77
Sq ft 0.2 0.09 2.22 0.0276 0.024 0.39
When testing whether the explanatory variables jointly influence the response variable, the null
hypothesis is ________.
A) H0: β0 = β1 = β2 = β3 = 0 B) H0:β1 = β2 = β3 = 0
C) H0: β0 + β1 + β2 + β3 = 0 D) H0:β1 + β2 + β3 + 0
Answer: B

F
Regression 3 5,694,717 1,898,239 50.88 4.99E-13
Residual 36 1,343,176 37,310
Total 39 7,037,893
Error 95%
Intercept 300 84.0 3.57 0.0010 130.03 470.79
Bedroom 226 60.3 3.75 0.0006 103.45 348.17
Bath 89 55.9 1.59 0.1195 −24.24 202.77
Sq ft 0.2 0.09 2.22 0.0276 0.024 0.39
A) 1.59 B) 2.22 C) 3.75 D) 50.88
Answer: D

F
Regression 3 5,694,717 1,898,239 50.88 4.99E-13
Residual 36 1,343,176 37,310
Total 39 7,037,893
Error 95%
Intercept 300 84.0 3.57 0.0010 130.03 470.79
Bedroom 226 60.3 3.75 0.0006 103.45 348.17
Bath 89 55.9 1.59 0.1195 −24.24 202.77
Sq ft 0.2 0.09 2.22 0.0276 0.024 0.39
When testing whether the explanatory variables are jointly significant at the 5% level, she
________.
significant
C) rejects H0, and concludes that the explanatory variables are not jointly significant.
significant
Answer: A

F
Regression 3 5,694,717 1,898,239 50.88 4.99E-13
Residual 36 1,343,176 37,310
Total 39 7,037,893
Error 95%
Intercept 300 84.0 3.57 0.0010 130.03 470.79
Bedroom 226 60.3 3.75 0.0006 103.45 348.17
Bath 89 55.9 1.59 0.1195 −24.24 202.77
Sq ft 0.2 0.09 2.22 0.0276 0.024 0.39
When testing whether Bedroom is significant at the 5% significance level, she ________.
A) rejects H0:β1 = 0, and concludes that the number of bedrooms significantly
influences rent
B) does not reject H0:β1 = 0, and concludes that the number of bedrooms
significantly influences rent
C) rejects H0:β1 = 0, and concludes that the number of bedrooms does not
significantly influence rent
D) does not reject H0:β1 = 0, and concludes that the number of bedrooms does not
Answer: A

F
Regression 3 5,694,717 1,898,239 50.88 4.99E-13
Residual 36 1,343,176 37,310
Total 39 7,037,893
Error 95%
Intercept 300 84.0 3.57 0.0010 130.03 470.79
Bedroom 226 60.3 3.75 0.0006 103.45 348.17
Bath 89 55.9 1.59 0.1195 −24.24 202.77
Sq ft 0.2 0.09 2.22 0.0276 0.024 0.39
When testing whether Bath is significant at the 5% significance level, she ________.
A) rejects H0:β2 = 0, and concludes that the number of bathrooms significantly
influences rent
B) does not reject H0:β2 = 0, and concludes that the number of bathrooms
significantly influences rent
C) rejects H0:β2 = 0, and concludes that the number of bathrooms does not
D) does not reject H0:β2 = 0, and concludes that the number of bathrooms does not
significantly influence rent.
Answer: D

F
Regression 3 5,694,717 1,898,239 50.88 4.99E-13
Residual 36 1,343,176 37,310
Total 39 7,037,893
Error 95%
Intercept 300 84.0 3.57 0.0010 130.03 470.79
Bedroom 226 60.3 3.75 0.0006 103.45 348.17
Bath 89 55.9 1.59 0.1195 −24.24 202.77
Sq ft 0.2 0.09 2.22 0.0276 0.024 0.39
Which of the following are the hypotheses to test if the explanatory variables Bath and Sqft
jointly significant in explaining Rent?
A) H0:β2 = β3 = 0; HA: both of the coefficients are nonzero.
B) H0:β2 = β3 = 0; HA: at least one of the coefficients is zero.
C) H0:β2 = β3 = 0; HA: at least one of the coefficients is nonzero.
D) H0:β2 = β3 = 0; HA: both of the coefficients are zero.
Answer: C

F
Regression 3 5,694,717 1,898,239 50.88 4.99E-13
Residual 36 1,343,176 37,310
Total 39 7,037,893
Error 95%
Intercept 300 84.0 3.57 0.0010 130.03 470.79
Bedroom 226 60.3 3.75 0.0006 103.45 348.17
Bath 89 55.9 1.59 0.1195 −24.24 202.77
Sqft 0.2 0.09 2.22 0.0276 0.024 0.39
When testing whether the explanatory variables Bath and Sq ft are jointly significant, the
p-value associated with the test is 0.0039. At the 5% significance level, she ________.
A) rejects H0, and concludes that neither of the explanatory variables is significant in
explaining Rent
B) does not reject H0, and concludes that neither of the explanatory variables is
significant in explaining Rent
C) rejects H0, and concludes that at least one of the explanatory variables, Bath
and/or Sq ft, is significant in explaining Rent
D) does not reject H0, and concludes that at least one of the explanatory variables,
Bath and/or Sq ft, is significant in explaining Rent
Answer: C

91) An economist estimates the following model: y = β0 + β1x + ε. She would like to 91)
construct interval estimates for y when x equals 2. She estimates a modified model
where y is the response variable and the explanatory variable is now defined as x+ = x –
2. A portion of the regression results is shown in the accompanying table.
R Square 0.42
Standard Error 3.86
Observations 12

Error 95%
Intercept 24.78 2.76 8.98 4.2E-06 18.63 30.93
Bedroom 2.52 0.95 2.66 0.0237 0.41 4.63
According to the modified model, which of the following is the predicted value of y when
equals 2?
A) 2.52 B) 24.78 C) 27.30 D) 29.82
Answer: B

92) An economist estimates the following model:y = β0 + β1x + ε. She would like to 92)
where y is the response variable and the explanatory variable is now defined as x+ = x –
R Square 0.42
Standard Error 3.86
Observations 12

Error 95% 95%
Intercept 24.78 2.76 8.98 4.2E-06 18.63 30.93
Bedroom 2.52 0.95 2.66 0.0237 0.41 4.63
According to the modified model, which of the following is a 95% confidence interval for
when x equals 2? (Note that t0,025,10 = 2.228.)
A) [0.41, 4.63] B) [14.21, 35.35]
C) [18.63, 30.93] D) [20.35, 25.65]
Answer: C

93) An economist estimates the following model: y = β0 + β1x + ε. She would like to 93)
where y is the response variable and the explanatory variable is now defined as x* = x –
R Square 0.42
Standard Error 3.86
Observations 12

Error
Intercept 24.78 2.76 8.98 4.2E-06 18.63 30.93
x* 2.52 0.95 2.66 0.0237 0.41 4.63
According to the modified model, which of the following is a 95% prediction interval for
x equals 2? (Note that t0,025,10 = 2.228.)
A) [0.41, 4.63] B) [14.21, 35.35]
C) [18.63, 30.93] D) [20.35, 25.65]
Answer: B
94) A sociologist wishes to study the relationship between happiness and age. He interviews 94)
24 individuals and collects data on age and happiness, measured on a scale from 0 to
100. He estimates the following model: Happiness = β0 + β1Age + ε. The following table
summarizes a portion of the regression results.

Intercept 56.1772 5.2145 10.7732 0.0000
Age 0.2845 0.0871 3.2671 0.0035
Which of the following is the estimate of Happiness for the person who is 65 years old?
A) 75 B) 62 C) 78 D) 68
Answer: A


Intercept 56.1772 5.2145 10.7732 0.0000
Age 0.2845 0.0871 3.2671 0.0035
When defining whether age is significant in explaining happiness, the competing hypotheses are
________.
A) H0:β1 = 1; HA:β1 ≠ 1 B) H0:β1 = 0; HA:β1 ≠ 0
C) H0:β1 ≤ 0; HA:β1 > 0 D) H0:b1 ≥ 1; HA:β1 < 1
Answer: B

Intercept 56.1772 5.2145 10.7732 0.0000
Age 0.2845 0.0871 3.2671 0.0035
At the 1% significance level, which of the following is correct?

A) Do not reject the null hypothesis. At 1% significance level, we conclude that Age
is not significant in explaining Happiness.
B) Do not reject the null hypothesis. At 1% significance level, we conclude that Age
is significant in explaining Happiness.
C) Reject the null hypothesis. At 1% significance level, we conclude that Age is
significant in explaining Happiness.
D) Reject the null hypothesis. At 1% significance level, we conclude that Age is not
significant in explaining Happiness.
Answer: C


Intercept 56.1772 5.2145 10.7732 0.0000
Age 0.2845 0.0871 3.2671 0.0035
At the 1% significance level, which of the following is the correct confidence interval of the
regression coefficient β1?
A) [0.0390, 0.0530] B) [0.0190, 0.5300]
C) [0.0190, 0,0530] D) [0.0390, 0.5300]
Answer: D
98) A researcher studies the relationship between SAT scores, the test-taker's family 98)
income, and his or her grade point average (GPA). Data are collected from 24 students.
He estimates the following model: SAT = β0 + β1 GPA + β2 Income + ε. The following
table summarizes a portion of the regression results.
df SS MS F
Regression 2 141,927.9571 70,963.98
Residual 21 22,167.8762 1,055.613
Total 23 164.095.8333
Intercept 1,104.2580 54.7524 20.1682 0.0000
GPA 150.9920 15.0931 10.0040 0.0000
Income 0.0017 0.0003 6.7696 0.0000
significant?
A) Only Income. B) Both, GPA and Income.
C) Only GPA. D) Neither GPA nor Income.
Answer: A

He estimates the following model: SAT = β0 + β1 GPA + β2 Income + ε. The following
df SS MS F
Regression 2 141,927.9571 70,963.98
Residual 21 22,167.8762 1,055.613
Total 23 164.095.8333
Intercept 1,104.2580 54.7524 20.1682 0.0000
GPA 150.9920 15.0931 10.0040 0.0000
Income 0.0017 0.0003 6.7696 0.0000
Which of the following is the correct hypotheses for testing the joint significance?
A) H0:β0 = β2 = 0; HA: At least β0 ≠ 0
B) H0:β1 = β2 = 0; HA: At least one βj ≠ 0
C) H0:β0 = β1 = β2 = 0; HA: All βj ≠ 0
D) H0:β1 = β2 = 0; HA: Both βj ≠ 0
Answer: B
He estimates the following model:SAT = β0 + β1 GPA + β2 Income + ε. The following
df SS MS F
Regression 2 141,927.9571 70,963.98
Residual 21 22,167.8762 1,055.613
Total 23 164.095.8333
Intercept 1,104.2580 54.7524 20.1682 0.0000
GPA 150.9920 15.0931 10.0040 0.0000
Income 0.0017 0.0003 6.7696 0.0000
A) 6.40 B) 134.45 C) 67.23 D) 45.13
Answer: C

101) Which of the following can be used to conduct the test for individual significance of 101)
explanatory variables?
A) Only p-value approach.
B) Only confidence intervals of the regression coefficients.
C) Neither p-value approach nor confidence intervals.
D) Both, p-value approach and confidence intervals.
Answer: D
102) Test statistic for the test of linear restrictions is using all of the following, except 102)
________.
A) mean error sum of squares
B) error sum of squares of the restricted model
C) number of linear restrictions
D) number of explanatory variables in the unrestricted model
Answer: A
103) Which of the following is the correct expression for computing a 100(1 – α)% 103)
prediction interval for an individual value of y?
A) B)
C) D)
Answer: B
104) Which of the following is the correct expression for computing a 100(1 – α)% 104)
confidence interval of the expected value of y?
A) B)
C) D)
Answer: C

105) Which of the following statements in statistical terminology is equivalent to the 105)
statement: "There is no exact linear relationship among the explanatory variables"?
A) The error term is homoscedastic.
B) There is no endogeneity.
C) There is no autocorrelation.
D) There is no perfect multicollinearity.
Answer: D
SHORT ANSWER. Write the word or phrase that best completes each statement or answers the
question.
106) The accompanying table shows the regression results when estimating y = β0 + 106)
β1x + ε.
F
Regression 1 78.5900 78.5900 15.82 0.0106
Residual 5 24.8385 4.9677
Total 6 103.4286
Coefficients Standard t-stat p-value
Error
Intercept 15.6615 2.4020 6.520 0.0013
x −1.6927 0.4256 −3.977 0.0106
a. Specify the competing hypotheses to determine whether x is significant in y.

b. At the 5% significance level, is x significant? Explain.
Answer: a. H0:β1 = 0; HA:β1 ≠ 0.
b. Yes, since the p-value of 0.0106 is less than 0.05, we reject H0: β1 = 0 and
conclude that x is significant.

107) Consider the following regression results based on 40 observations. 107)

Intercept 427.93 66.29 6.46 0.0000
x 0.37 0.17 2.21 0.0335
a. Specify the hypotheses to determine if the slope differs from one.

b. Calculate the value of the test statistic.
c. At the 5% significance level, find the critical value(s).
d. Does the slope differ from one? Explain.
Answer: a. H0:β1 = 1; HA:β1 ≠ 1.
b. t38 = –3.71.
c. –t0.025,38 = –2.024 and t0.025,38 = 2.024.
d. Because −3.71 < −2.024, we reject H0 and conclude that the slope differs
from one.
108) When estimating a multiple regression model based on 30 observations, the following
108)
results were obtained.

Error 95% 95%
Intercept 395.96 76.95 5.15 0.0000 238.07 553.85
x1 −0.45 0.19 −2.35 0.0264 −0.84 −0.06
x2 1.26 0.65 1.94 0.0631 −0.07 2.59
a. Specify the hypotheses to determine whether x1 is linearly related to y. At the 5%

significance level, use the p-value approach to complete the test. Are x1 and y
linearly related?
b. Construct the 95% confidence interval for β2. Using this confidence interval, is
significant in explaining y? Explain.
c. At the 5% significance level, can you conclude that β1 differs from −1? Show the
relevant steps of the appropriate hypothesis test.
Answer: a. H0:β1 = 0; HA:β1 ≠ 0; because the p-value = 0.0264 < 0.05 = α, we
reject H0; x1 and y are linearly related.
b. The confidence interval for β2 is reported as [−0.07, 2.59]; we cannot
conclude that x2 is significant in explaining y because the interval contains 0.
c. H0:β1 = –1; HA:β1 ≠ –1; t27 = 2.89; critical value t0,025,27 = 2.052.
Because t27 > t0,025,27 we reject H0 and conclude that β1 differs from
–1.

109) The accompanying table shows the regression results when estimating y = β0 + 109)
β1x1 + β2x2 + β3x3 + ε.
F
Regression 3 878 293 13.31 0.0003
Residual 13 286 22
Total 16 1,164
Error
Intercept 32.85 5.10 6.45 2.18E-05
x1 0.75 0.45 1.65 0.1225
x2 0.23 0.43 0.54 0.5972
x3 −2.84 0.49 −5.77 0.0001
a. Specify the competing hypotheses to determine whether the explanatory variables are
jointly significant.
b. At the 5% significance level, are the explanatory variables jointly significant?

Explain.
c. At the 5% significance level, is x2 significant in explaining y? Explain.
d. At the 5% significance level, is the slope coefficient attached to x3 different from −2?
Answer: a. H0: β1 = β2 = β3 = 0; HA: At least one βj ≠ 0.
b. Yes, because the p-value of 0.0003 is less than 0.05, we reject H0 and
conclude that the explanatory variables are jointly significant.
c. H0:β2 = 0; HA:β2 ≠ 0; because the p-value of 0.5972 is greater than
0.05, we do not reject H0 and conclude that x2 is not significant in
explaining y.
d. H0:β3 = 2; HA:β3 ≠ – 2. We calculate the value of the test statistic as
t13 = –1.71. At the 5% significance level, the critical values are
–t0,025,13 = –2.160 and t0,025,13 = 2.160 Because −2.160 < −1.71 < 2.160,
we do not reject H0 and conclude that β3 is not significantly different from −2.

Variable Restricted Unrestricted

Intercept 412.42*(0.0000) 535.47*(0.0000)
x1 −2.41*(0.0000) −2.26*(0.0000)
x2 NA 4.75*(0.0002)
x3 NA −0.53*(0.0000)
SSE 11,013 4,355
Notes: Parameter estimates are in the main body of the table with the p-values in
parentheses; * represents significance at 5% level. The last row presents the error sum
of squares.
a. Formulate the hypotheses to determine whether x2 and x3 are jointly significant in
explaining y.
b. Define the restricted and the unrestricted models needed to conduct the test.
c. Calculate the value of the test statistic.
d. At the 5% significance level, find the critical value(s).
e. What is your conclusion to the test?
Answer: a. H0:β2 = β3 = 0; HA: At least one βj ≠ 0.
b. Restricted model: y = β0 + β1x + ε; unrestricted model: y = β0 + β1x1 +
β2x2 + β3x3 + ε.
c. F(2,26) = 19.87.
d. F0.05,(2,26) = 3.37.
e. We reject H0 because 19.87 is greater than 3.37. At the 5% level, we
conclude that x2 and x3 are jointly significant in explaining y.

= 238.33 – 0.95x + 7.13x2 + 4.76x3; SSE = 3,439
= 209.56 – 1.03x1 + 5.24(x2 + x3); SSE = 3,559
a. Formulate the hypotheses to determine whether the influences of x2 and x3 differ in

explaining y.
d. What is your conclusion to the test?
Answer: a. H0:β2 = β3; HA:β2 ≠ β3.
b. F(1,26) = 0.91.
c. F0.05(1,26) = 4.23.
d. We do not reject H0 because 0.91 is less than 3.37. At the 5% level, we
cannot conclude that the influence of x2 is different from the influence of
x3.
112) In a simple linear regression based on 30 observations, the following 112)

information is provided:
= 32.3 + 1.45x and se = 3.2. Also, evaluated at x = 20 is 1.3.
a. Construct a 95% confidence interval for E(y) if x = 20.

b. Construct a 95% prediction interval for y if x = 20.
Answer: a. [58.54, 63.86] b. [54.13, 68.27]

113) In a multiple regression based on 30 observations, the following information is 113)
provided:
= 38.6 – 0.1x1 and se = 5.4. Also, evaluated at x1 = 40 and x2 = 30 is

2.8.
a. Construct a 95% confidence interval for E(y) if x1 equals 40 and x2 equals 30.
b. Construct a 95% prediction interval for y if x1 equals 40 and x2 equals 30.
c. Which interval is narrower? Explain.
Answer: a. [34.85, 46.35]
b. [28.12, 53.08]
c. The confidence interval is narrower because it assumes that the expected
value of the error term is zero, whereas the prediction interval incorporates the
nonzero error term.
114) In a multiple regression based on 30 observations, the following information is 114)

provided:
= 469 – 2x1 + 6x2 – 0.5x3 and se = 16. Also, when x1 = 30, x2 = 10, and x3 =
65 = 37.
a. Construct a 90% confidence interval for E(y) if x1 = 30, x2 = 10, and x3 = 65.
b. Construct a 90% prediction interval for y if x1 = 30, x2 = 10, and x3 = 65.
c. Which interval is narrower? Explain.
Answer: a. [373.38, 499.62]
b. [367.73, 505.27]
c. The confidence interval is narrower because it assumes that the expected
value of the error term is zero, whereas the prediction interval incorporates the
nonzero error term.

115) A manager at a local bank analyzed the relationship between monthly salary (y, 115)
in $) and length of service (x, measured in months) for 30 employees. The following
table summarizes a portion of the regression results:

Intercept 784.92 322.25 2.44 0.02
Service 9.19 3.20 2.87 0.01
a. Specify the competing hypotheses to determine whether Service is significant in

explaining Salary.
b. At the 5% significance level, is Service significant? Explain.

Answer: a. H0:β1 = 0; HA:β1 ≠ 0.
b. Yes, since the p-value of 0.01 is less than 0.05, we reject H0:β1 = 0 and
conclude that Service is significant in explaining Salary.
116) An investment analyst wants to examine the relationship between a mutual 116)
fund's return, its turnover rate, and its expense ratio. She randomly selects 10
mutual funds and estimates: Return = β0 + β1Turnover + β2Expense + ε, where
Return is the average five-year return (in %), Turnover is the annual holdings
turnover (in %), Expense is the annual expense ratio (in %), and ε is the random error
component. A portion of the regression results is shown in the accompanying table.
F
Regression 2 93.33 46.67 4.90 0.047
Residual 7 66.69 9.53
Total 9 160.02
Intercept 30.60 4.30 7.12 0.000

Turnover 0.13 0.06 2.23 0.061
Expense 0.90 4.08 0.22 0.831
a. At the 10% significance level, are the explanatory variables jointly significant in
explaining Return? Explain.
b. At the 10% significance level, is each explanatory variable individually significant in

explaining Return? Explain.

Answer: a. Here the competing hypotheses for a joint significance test of Turnover
and Expense take the form: H0:β1 = β2 = 0 At least one βj ≠ 0Because
the p-value associated with F(2,7) = 4.90 is 0.047 and this value is less
than 0.10, we reject H0 and conclude that at least one of the explanatory
variables is significant in explaining Return.
b. The competing hypotheses for an individual significance test of Turnover take
the form: H0:β1 = 0; HA:β1 ≠ 0. Because the p-value associated with t7 =
2.23 is 0.061 and this value is less than 0.10, we reject H0 and conclude
that Turnover is significant. The competing hypotheses for an individual
significance test of Expense take the form:H0:β2 = 0; HA:β2 ≠ 0.
Because the p-value associated with t7 = 0.22 is 0.831 and this value is
greater than 0.10, we do not reject H0 and conclude that Expense is not
significant.
117) Pfizer Inc. is the world's largest research-based pharmaceutical company. 117)
Monthly data for Pfizer's risk-adjusted return and the risk-adjusted market return
are collected for a five-year period (n = 60). The accompanying table shows the
regression results when estimating the CAPM model for Pfizer's return.

Error 95% 95%
Intercept 0.004 0.006 0.62 0.5364 −0.009 0.017
RM – Rf 0.716 0.119 6.00 1.35E-07 0.477 0.955
a. At the 5% significance level, is the beta coefficient less than one? Show the relevant
steps of the appropriate hypothesis test.
b. At the 5% significance level, are there abnormal returns? Show the relevant steps of
the appropriate hypothesis test.
Answer: a. H0:β ≥ 1; HA:β < 1. The value of the test statistic is t58 = –2.387. The
critical values at the 5% significance level are –t0.025,58 = –1.672 and
t0.025,58 = 1.672. Since −2.387 < −1.672, we reject H0:β ≥ 1 in favor of HA:
β < 1; implying that the return on Pfizer stock is less risky than the return on the
market.
b. H0:α = 0; HA: α ≠ 0. Because the p-value of 0.5364 is greater than the

significance level of 0.05, we cannot reject H0:α = 0 and thus cannot
conclude that there are abnormal returns.

118) Assume you ran a multiple regression to gain a better understanding of the 118)
relationship between lumber sales, housing starts, and commercial construction.
The regression uses Lumber Sales (in $100,000s) as the response variable with
Housing Starts (in 1,000s) and Commercial Construction (in 1,000s) as the
explanatory variables. The results of the regression are as follows:
F
Regression 2 180,770 90,385 103.3 0.0000
Residual 45 39,375 875
Total 47 220,145
Error
Intercept 5.37 1.71 3.14 0.0030
Housing Starts 0.76 0.09 8.44 0.0000
Commercial 1.25 0.33 3.78 0.0005
Construction

explaining Lumber Sales? Explain.
b. At the 5% significance level, is Commercial Construction significant in

explaining Lumber Sales? Explain.
c. At the 5% significance level, can you conclude that the slope coefficient attached to
Housing Starts differs from 1? Explain.
Answer: a. Here the competing hypotheses for a joint significance test of Housing
Starts and Lumber Sales take the form: H0:β1 = β2 = 0; HA: At least one
βj ≠ 0. Because the p-value associated with F(2,45) = 103.3 is 0.000 and
this value is less than 0.05, we reject H0 and conclude that at least one of the
explanatory variables is significant in explaining Lumber Sales.
b. The competing hypotheses for an individual significance test of Commercial

Construction take the form: H0:β2 = 0; HA:β2 ≠ 0. Because the p-value
associated with t45 = 3.78 is 0.0005 and this value is less than 0.05, we
reject H0 and conclude that Commercial Construction is significant.
c. H0:β1 = 1; HA:β1 ≠ 1. The value of the test statistic is t45 = –2.67. The
t0.025,45 = 2.014. Because −2.67 < −2.014, we reject H0:β1 = 1 and conclude
that the slope coefficient attached to Housing Starts differs from 1.

119) An analyst examines the effect that various variables have on crop yield. He 119)
estimates y = β0 + β1x1 + β2x2 + β3x3 + ε. where y is the average yield in
bushels per acre, x1 is the amount of summer rainfall, x2 is the average daily use
in machine hours of tractors on the farm, and x3 is the amount of fertilizer used per
acre. The results of the regression are as follows:
F
Regression 3 12,000 4,000 10 0.0095
Total 9 14,400
Intercept 1.6 1.0 1.6 0.1232

x1 7.5 2.5 3.0 0.0064
x2 6.0 4.0 1.5 0.1472
x3 1.0 0.5 2.0 0.0574
explaining crop yield? Explain.
b. At the 10% significance level, is fertilizer significant in explaining crop yield?

Explain.
c. At the 10% significance level, can you conclude that the slope coefficient attached to
rainfall differs from 9? Explain.
Answer: a. The competing hypotheses for a joint significance test of the three
explanatory variables take the form: H0:β1 = β2 = β3 = 0; HA: At least
one βj ≠ 0. Because the p-value associated with F(3,6) = 10 is 0.0095 and
this value is less than 0.10, we reject H0 and conclude that at least one of the
explanatory variables is significant in explaining crop yield.
b. The competing hypotheses for an individual significance test of fertilizer take
the form: H0:β3 = 0; HA:β3 ≠ 0. Because the p-value associated with t6 =
2 is 0.0574 and this value is less than 0.10, we reject H0 and conclude that
fertilizer is significant.
c. H0:β1 = 9; HA:β1 ≠ 9. The value of the test statistic is t6 = –0.6. The
t0.05,6 = 1.943. Because −1.943 < −0.6 < 1.943, we do not reject H0:β1 = 9
and cannot conclude that the slope coefficient attached to rainfall differs
from 9.

120) A sociologist estimates the regression relating Poverty (y) to Education (x1). 120)
The estimated regression equation is:
= 71.68 – 0.69x1; n = 39; SSE = 152.
In an attempt to improve the results, he adds two more explanatory variables: Median
Income (x2, in $1,000s) and the Mortality Rate (x3, per 1,000 residents). The
estimated regression equation is:
= 59.80 – 0.44x1– 0.18x2 + 0.17x3; n = 39; SSE = 60.
a. Formulate the hypotheses to determine whether Median Income and the Mortality
Rate are jointly significant in explaining Poverty.
c. At the 5% significance level, find the approximate value of the critical value(s).
d. What is the conclusion to the test?
Answer: a. H0:β2 = β3 = 0; HA: At least one βj ≠ 0
b. F(2,35) = 26.83
c. Using the F table, the critical value approximately is F0.05,(2,35) ≈ 3.32
d. Because 26.83 > 3.32, we reject the null hypothesis and conclude that Median
Income and the Mortality Rate are jointly significant in explaining the Poverty.
121) A marketing manager examines the relationship between the attendance at 121)
amusement parks and the price of admission. He estimates the following model:
Attendance = β 0 + β1 price + ε, where Attendance is the average daily number
of people who attend an amusement park in July (in 1,000s) and Price is the
price of admission. The marketing manager would like to construct interval
estimates for Attendance when Price equals $80. The researcher estimates a
modified model where Attendance is the response variable and the Price is now
defined as Price* = Price – 80. A portion of the regression results is shown in
R Square 0.62
Standard Error 21
Observations 30

Error 95% 95%
Intercept 86.8 4.2 20.85 1.4E-18 78.2 95.4
Price* −3.1 0.5 −6.69 2.9E-07 −4.0 −2.1

a. According to the modified model, what is the point estimate for Attendance when
Price equals $80?
b. According to the modified model, what is a 95% confidence interval for Attendance
when Price equals $80? (Note that t0.025,28 = 2.048.)
c. According to the modified model, what is a 95% prediction interval for Attendance
when Price equals $80? (Note that t0.025,28 = 2.048.)
Answer: a. = 86.8 or 86,800 people; in the modified model, the point estimate of the
intercept is the predicted value for Attendance when Price equals $80.
b. In the modified model, we have estimates of the intercept and its standard
error. We then find 78.2 ≤ E(y0) ≤ 95.4. Converting to the correct units of
measurement, we have [78,200, 95,400]. Excel also provides the 95% confidence
interval directly.
c. For prediction interval we have estimates of the intercept, its standard error,
and the standard error of estimate. We then find 42.9 ≤ (y0) ≤ 130.7.
Converting to the correct units of measurement, we have [42,900,
130,700].
122) A researcher analyzes the factors that may influence the poverty rate and 122)
estimates the following model: y = β0 + β1x1 + β2x2 + β3x3 + ε, where y is the
poverty rate (y, in %), x1 is the percent of the population with at least a high
school education, x2 is the median income (in $1,000s), and x3 is the mortality
rate (per 1,000 residents). The researcher would like to construct interval
estimates for y when x1, x2, and x3 equal 85%, $50,000, and 10, respectively.
The researcher estimates a modified model where poverty rate is the response
variable and the explanatory variables are now defined as
A portion of the regression
results is shown in the accompanying table.
R Square 0.86
Standard Error 1.30
Observations 39

Error 95% 95%
Intercept 14.89 0.37 39.73 1.04E-30 14.14 15.64
–0.44 0.06 –6.81 6.68E-08 –0.57 –0.31

–0.18 0.03 –6.14 4.97E-07 –0.24 –0.12
0.17 0.19 0.87 0.3898 –0.22 0.56
a. According to the modified model, what is the point estimate for the poverty rate when
x1, x2, and x3 equal 85%, $50,000, and 10, respectively.
b. According to the modified model, what is a 95% confidence interval for the expected
poverty rate when x1, x2, and x3 equal 85%, $50,000, and 10, respectively? (Note
that t0.025,35 = 2.030.)
c. According to the modified model, what is a 95% prediction interval for the poverty
rate when x1, x2, and x3 equal 85%, $50,000, and 10, respectively? (Note that
t0.025,35 = 2.030.)
Answer: a. = 86.8; in the modified model, the point estimate of the intercept is
when x1, x2, and x3 equal 85%, $50,000, and 10, respectively.
b. In the modified model, we have estimates of the intercept and its standard
error. We then find 14.14 ≤ E(y0) ≤ 15.64. Excel also provides the 95%
confidence interval directly.
c. For prediction interval we have estimates of the intercept, its standard error,
and the standard error of estimate. We then find 12.15 ≤ y0 ≤ 17.63.

123) A sociologist studies the relationship between a district's average score on a 123)
standardized test for 10th grade students (y), the average school expenditures per
student (x1in $1,000s), and an index of socioeconomic status of the district (x2).
The results of the regression are = 10.16 + 8.50x1 + 4.30x2; n = 25; SSE = 450;
He would like to determine whether the influence of expenditures on test scores differs
from the influence of the index on test scores, or β1 ≠ β2. The results of the restricted
model for this test are
= 12.25 + 6.05(x1 + x2); n = 25; SSE = 600.
a. Formulate the hypotheses to determine whether Expenditures and Index Rate have
the same influence on Scores.
d. What is the conclusion to the test?

Answer: a. H0:β1 = β2; HA:β1 ≠ β2
b. F(1,22) = 7.33
c. Using the F table, the critical value is F0.05,(1,22) = 4.30
d. Because 7.33 > 4.30, we reject the null hypothesis and conclude that β1
thus, the influence of the two variables on test scores is different.

124) A simple linear regression, Sales = β0 + β1 Advertising + ε, is estimated using 124)
cross-sectional data from 10 firms. The resulting residuals e, along with the
values of the explanatory variable Advertising (x, in $100s), are shown in the
accompanying table.
x 2 3 6 8 11 15 16 21 25 30
e −3 2 −4 3 5 −6 −7 9 11 −11
a. Graph the residuals e against Advertising (x) and look for any discernible pattern.
b. Which assumption is being violated? Discuss its consequences and suggest a possible
remedy.
Answer: a.
b. Heteroscedasticity is likely a problem in this application in that the residuals

seem to fan out across the horizontal axis. We can correct the biased standard
errors using White's correction method.

125) A simple linear regression, Sales = β0 + β1 Advertising + ε, is estimated using 125)
time-series data over the last 10 years. The residuals, e, and the time variable, t,
are shown in the accompanying table.
t 1 2 3 4 5 6 7 8 9 10
e −3 −2 −1 3 5 6 3 −3 −4 −3
a. Graph the residuals e against time and look for any discernible pattern.
b. Which assumption is being violated? Discuss its consequences and suggest a possible
remedy.
Answer: a.
b. Positive serial correlation is likely a problem in this application because the

residuals form a wavelike pattern over time. We can correct the biased standard
errors using the Newey-West procedure.

Answer Key
Testname: BUS_STATS_CH15
1) TRUE
Explanation: If βi equals zero, there is no linear relationship between the explanatory variable xi and the
response variable y. Conversely, if βi does not equal zero, then the explanatory variable xi
influences the response variable. For βi < 0, the relationship is negative in the sense that if
xi increases, y tends to decrease. On the other hand, βi > 0 implies the positive relationship.
ID: Bs2Jaggi 15-1
2) FALSE
Explanation: The test statistic for a test of individual significance is assumed to follow the t distribution
with n – k – 1 degrees of freedom.
ID: Bs2Jaggi 15-2
3) FALSE
Explanation: The restricted model is a reduced model where we do not estimate the coefficients that are
restricted under the null hypothesis.
ID: Bs2Jaggi 15-3
4) TRUE
Explanation: BLUE is the acronym for Best Linear Unbiased Estimator.
ID: Bs2Jaggi 15-4
5) FALSE
Explanation: When explanatory variables within a model are correlated with each other, this
phenomenon is called multicollinearity.
ID: Bs2Jaggi 15-5
6) TRUE
Explanation: One of the required assumptions of regression analysis is that the error term ε is not
correlated with any of the explanatory variable – there is no endogeneity.
ID: Bs2Jaggi 15-6
7) TRUE
Explanation: The wavelike movement in the residuals over time creates a pattern around the horizontal
axis, and we conclude that positive serial correlation is likely to be a problem.
ID: Bs2Jaggi 15-7
8) FALSE
Explanation: The term homoskedastic refers to the condition when the variance of the error term,
conditional on x1, x2, … , xn, is the same for all observations.
ID: Bs2Jaggi 15-8
9) FALSE
Explanation: The alternative hypothesis for test of joint significance is specified as HA: At least one βj ≠
0.
ID: Bs2Jaggi 15-9

Answer Key
10) partial
Explanation: The F test is a test of k restrictions that determines whether or not all slope coefficients are
zero. The F test can be applied for any number of linear restrictions; the resulting test is
often referred to as the partial F test.
ID: Bs2Jaggi 15-10
11) unrestricted
Explanation: The unrestricted model is a complete model that imposes no restrictions on the coefficients;
therefore, all coefficients are estimated.
ID: Bs2Jaggi 15-11
12) SSEU
Explanation: With the partial F test, we basically analyze the ratio of (SSER - SSEU) to SSEU.
ID: Bs2Jaggi 15-12
13) explanatory
Explanation: If the confidence interval for the slope coefficient contains the value zero, then the
explanatory variable associated with the regression coefficient is not significant.
Conversely, if the confidence interval does not contain the value zero, then the explanatory
variable associated with the regression coefficient is significant.
ID: Bs2Jaggi 15-13
14) prediction
Explanation: It is common to refer to the interval estimate for an individual value of y as the prediction
interval. The interval estimate for the mean of y is referred to as the confidence interval.
ID: Bs2Jaggi 15-14
15) expected
Explanation: Conditional on x1, x2, ..., xk, the error term has an expected value of zero, or E(ε) = 0.
ID: Bs2Jaggi 15-15
16) normally
Explanation: Conditional on x1, x2, ..., xk, the error term ε is normally distributed.
ID: Bs2Jaggi 15-16
17) Residual
Explanation: Residual plots can be used to detect common violations, and they can be used to detect
outliers.
ID: Bs2Jaggi 15-17
18) A
Explanation: The individual significance of all explanatory variables is evaluated first. If the slope
coefficient equals zero, there is no linear relationship between xi and y. And conversely, if β
does not equal to zero, then xi influences y.
ID: Bs2Jaggi 15-18

Answer Key
19) B
Explanation: When testing whether xi significantly influences y, in general, the null hypothesis is H0:βj =
βj0, where βj0 = 0. In some situations we might wish to determine whether the slope
coefficient differs from nonzero value.
ID: Bs2Jaggi 15-19
20) B
Explanation: It can be easily specified that the one-tailed competing hypotheses for a positive linear
relationship are H0:βj ≤ βj0; HA:βj > βj0, and the one-tailed competing hypotheses for a
negative linear relationship are H0:βj ≥ βj0; HA:βj < βj0.
ID: Bs2Jaggi 15-20
21) C
Explanation: It can be easily specified that the one-tailed competing hypotheses for a positive linear
relationship are H0:βj ≤ βj0; HA:βj > βj0 or a negative linear relationship are H0:βj ≥ βj0;
HA:βj < βj0.
ID: Bs2Jaggi 15-21
22) C
Explanation: Excel automatically provides the value of the test statistic and its associated p-value for a
two-tailed test that assesses whether the regression coefficient differs from zero.
ID: Bs2Jaggi 15-22
23) D
Explanation: A 100(1 - α)% confidence interval of the regression coefficient βj is computed as βj ± t
α/2,df , where is the standard error of and βj and df = n – k – 1.
ID: Bs2Jaggi 15-23
24) D
Explanation: The test statistic for a test of individual significance is assumed to follow the tdf
distribution with df = n – k – 1 and its value is computed as

ID: Bs2Jaggi 15-24
25) B
Explanation: Using the p-value approach, the decision rule is to reject the null hypothesis if the p-value
is less than the significance level α; do not reject the null hypothesis if the p-value is greater
than the significance level α.
ID: Bs2Jaggi 15-25
26) A
distribution with df = n – k – 1 and its value is computed as If testing

whether the slope coefficient differs from 1, bj0 = 1.
ID: Bs2Jaggi 15-26

Answer Key
27) D
Explanation: To conduct the test of joint significance, we employ a one-tailed F test. The value under the
heading Significance F is the p-value. Using the p-value approach, the decision rule is to
reject the null hypothesis if the p-value is less than the significance level α; do not reject the
null hypothesis if the p-value is greater than the significance level α.
ID: Bs2Jaggi 15-27
28) C
Explanation: The test statistic for a test of joint significance is assumed to follow the F(df1, df2)
distribution with df1 = k, df2 = n – k – 1 and its value is computed as

ID: Bs2Jaggi 15-28
29) A
Explanation: Using the critical value approach, the decision rule is to reject the null hypothesis if the test
statistic F(df1,df2) is greater than F(df1,df2) otherwise, do not reject the null hypothesis.
ID: Bs2Jaggi 15-29
30) A
ID: Bs2Jaggi 15-30
31) B
Explanation: The test statistic for a test of joint significance is assumed to follow the F(df1,df2)

ID: Bs2Jaggi 15-31
32) D
statistic F(df1,df2) is greater than Fα(df1,df2); otherwise, do not reject the null hypothesis.
ID: Bs2Jaggi 15-32
33) C
ID: Bs2Jaggi 15-33
34) C
ID: Bs2Jaggi 15-34

Answer Key
35) D
ID: Bs2Jaggi 15-35
36) D
Explanation: When testing linear restrictions, the test statistic is assumed to follow the F(df1,df2)
distribution with df1 equal to the number of linear restrictions and df2 = n – k – 1.
ID: Bs2Jaggi 15-36
37) B
ID: Bs2Jaggi 15-37
38) D
Explanation: The predictions are subject to sampling variations: the predicted value is a point estimate
that ignores sampling error.
ID: Bs2Jaggi 15-38
39) C
Explanation: There are two kinds of interval estimates regarding y: a confidence interval for the expected
value of y, and a prediction interval for an individual value of y.
ID: Bs2Jaggi 15-39
40) B
Explanation: The prediction interval will be wider than the confidence interval because it also
incorporates the error term.
ID: Bs2Jaggi 15-40
41) C
Explanation: In regression, multicollinearity is considered severe if the sample correlation coefficient between
any two explanatory variables is more than 0.80 or less than −0.80.
ID: Bs2Jaggi 15-41
42) D
Explanation: The detection methods for multicollinearity are mostly informal. The presence of a high R2
coupled with individually insignificant explanatory variables can be indicative of
multicollinearity.
ID: Bs2Jaggi 15-42
43) B
Explanation: One of the assumptions of a regression analysis is there is no exact linear relationship
among any explanatory variables, or in statistical terminology, there is no perfect
multicollinearity.
ID: Bs2Jaggi 15-43

Answer Key
44) C
Explanation: There are several required assumptions of regression analysis. One of them states that the
error term is not correlated with any of the explanatory variables.
ID: Bs2Jaggi 15-44
45) B
Explanation: When confronted with multicollinearity, a good remedy is to drop one of the collinear
variables if we can justify its redundancy.
ID: Bs2Jaggi 15-45
46) C
Explanation: In the presence of heteroscedasticity (when the variance of the error term is not the same for
all observations), the OLS estimators are still unbiased, though they are no longer efficient.
Consequently, we cannot put much faith in the confidence intervals, or the standard t or F
tests, in the presence of heteroscedasticity.
ID: Bs2Jaggi 15-46
47) C
Explanation: Serial correlation or autocorrelation means that the error term is uncorrelated across
observations. In the time series data the errors are likely to exhibit positive autocorrelation.
ID: Bs2Jaggi 15-47
48) C
Explanation: The scatterplot of residuals allows us to define a pattern around the horizontal axis and
make a conclusion about heteroscedasticity.
ID: Bs2Jaggi 15-48
49) C
Explanation: Serial correlation or autocorrelation means that the error term is uncorrelated across
observations.
ID: Bs2Jaggi 15-49
50) C
Explanation: The competing hypotheses when testing the individual significance of explanatory variables
usually are H0:βj = βj0; HA:βj ≠ βj0; where βj0; = 0.
ID: Bs2Jaggi 15-50

51) A
ID: Bs2Jaggi 15-51
52) C
Explanation: The competing hypotheses when testing individual significance of explanatory variables, in
general, are H0:βj = βj0; HA:βj ≠ βj0.
ID: Bs2Jaggi 15-52

Answer Key
53) B
distribution with df = n – k – 1 and its value is computed as

ID: Bs2Jaggi 15-53
54) D
statistic F(df1, df2) is greater than Fα,(df1, df2); otherwise, do not reject the null hypothesis.
ID: Bs2Jaggi 15-54
55) B
ID: Bs2Jaggi 15-55
56) A
Explanation: The decision rule is to reject the null hypothesis if the confidence interval for the slope
coefficient does not contain the value zero and conclude that the explanatory variable
associated with the regression coefficient is significant; do not reject the null hypothesis if
the confidence interval for the slope coefficient contains the value zero and conclude that
the explanatory variable associated with the regression coefficient is not significant.
ID: Bs2Jaggi 15-56
57) A
Explanation: The competing hypotheses for a test of joint significance are specified as
H0:β1 = β2 = ... = βk = 0;
HA At least one βj ≠ 0.
ID: Bs2Jaggi 15-57
58) A
ID: Bs2Jaggi 15-58
59) A
distribution with df = n – k – 1 and its value is computed as If testing

whether the slope coefficient differs from 0, bj0 = 0.
ID: Bs2Jaggi 15-59

Answer Key
60) D
ID: Bs2Jaggi 15-60
61) A
ID: Bs2Jaggi 15-61
62) D

ID: Bs2Jaggi 15-62
63) A
ID: Bs2Jaggi 15-63
64) A
Explanation: The competing hypotheses for a test of linear restrictions are specified as:
H0: β1 = β2 = ... = βk = 0;
HA: At least one βj ≠ 0.

ID: Bs2Jaggi 15-64
65) C
distribution with df1 equal to the number of linear restrictions and df2 = n – k – 1. The value
of the test statistic for the partial F test is computed as

ID: Bs2Jaggi 15-65
66) A
Explanation: The value under the heading Significance F is the p-value associated with F(df1,df2). Using
the p-value approach, the decision rule is to reject the null hypothesis if the p-value is less
than the significance level α; do not reject the null hypothesis if the p-value is greater than
the significance level α.
ID: Bs2Jaggi 15-66

Answer Key
67) B
Explanation: The predicted value is the point estimate for
ID: Bs2Jaggi 15-67

68) C
Explanation: For a specific values a 100(1 – α)% confidence interval of the expected
value of y is computed as where
is the standard error of and df = n – k

– 1.
ID: Bs2Jaggi 15-68
69) A
Explanation: For a specific values a 100(1 – α)% prediction interval for an individual
value of y is computed as where df = n – k – 1, se

is the standard error of and se is the standard error of the estimate.
ID: Bs2Jaggi 15-69
70) A
reject the null hypothesis if the p-value is less than the significance level α, do not reject the
ID: Bs2Jaggi 15-70
71) C
than the significance level α. p-values for slope coefficients are attached to explanatory
variables in the given table.
ID: Bs2Jaggi 15-71
72) A
Explanation: To test if two explanatory variables, xk and xl, have the same influence on y, we should run
a partial F test. The null hypothesis implies that the restrictions are valid: H0: βk = βl.
ID: Bs2Jaggi 15-72

Answer Key
73) D
Explanation: The competing hypotheses for a test of joint significance are specified as:
H0:β1 = β2 = ... = βk = 0;
HA: At least one βj ≠ 0

ID: Bs2Jaggi 15-73
74) D
Explanation: The CAPM expresses the risk-adjusted return of an asset R – Rf, as a function of a
risk-adjusted market return RM – Rf. For empirical estimation, the CAPM is specified as R
– Rf = α + β (RM – Rf) + ε. This is essentially a simple linear regression model. The slope
coefficient β measures how sensitive the stock's return is to changes in the level of the
overall market. The competing hypotheses are H0:β ≤ 1; HA:β > 1.
ID: Bs2Jaggi 15-74
75) C
Explanation: The test statistic for a test of individual significance of slope coefficient is assumed to
follow the tdf distribution with df = n – k – 1 and its value is computed as

ID: Bs2Jaggi 15-75
76) A
Explanation: Using the critical value approach, the decision rule for the right-tailed test is to reject the
null hypothesis H0:β ≤ 1 if the test statistic tdf is greater than tα,df in favor on HA:β > 1.
ID: Bs2Jaggi 15-76
77) A
– Rf = α + β(RM – Rf) + ε. This is essentially a simple linear regression model. The CAPM
theory predicts α to be zero, and thus a nonzero estimate indicates abnormal returns.
Abnormal returns are positive when α > 0, and negative when α > 0. The competing
hypotheses are H0:α = 0; HA:α ≠ 0.
ID: Bs2Jaggi 15-77
78) B
Explanation: The test statistic for a test of individual significance of intercept coefficient is assumed to
follow the tdf distribution with df = n – k – 1 and its value is computed as

ID: Bs2Jaggi 15-78

Answer Key
79) D
than the significance level α. p-value for the intercept can be found in the given table.
ID: Bs2Jaggi 15-79
80) A
Explanation: The competing hypotheses are H0:β0 = 0; HA:β0 ≠ 0.
ID: Bs2Jaggi 15-80
81) B
Explanation: The competing hypotheses are H0:β1 = 0; HA: β1 ≠ 0
ID: Bs2Jaggi 15-81
82) C
Explanation: The decision rule is to reject the null hypothesis if the confidence interval does not contain
the value zero; do not reject the null hypothesis if the confidence interval for the slope
coefficient contains the value zero.
ID: Bs2Jaggi 15-82
83) A
Explanation: The decision rule is to reject the null hypothesis if the confidence interval does not contain
the value zero; do not reject the null hypothesis if the confidence interval for the slope
coefficient contains the value zero.
ID: Bs2Jaggi 15-83
84) B
H0: β1 = β2 = ... = βk = 0;

ID: Bs2Jaggi 15-84
85) D
distribution with df1 = k,
df2 = n – k – 1 and its value is computed as

ID: Bs2Jaggi 15-85
86) A
ID: Bs2Jaggi 15-86

Answer Key
87) A
ID: Bs2Jaggi 15-87
88) D
ID: Bs2Jaggi 15-88
89) C
H0: β1 = β2 = ... = βk = 0;

ID: Bs2Jaggi 15-89
90) C
ID: Bs2Jaggi 15-90
91) B
Explanation: The predicted value is the point estimate for
ID: Bs2Jaggi 15-91

92) C
Explanation: For a specific values a 100(1 – α)% confidence interval of the expected
value of y is computed as where
is the standard error of and df = n – k –

1.
ID: Bs2Jaggi 15-92

Answer Key
93) B
Explanation: For a specific values a 100(1 – α)% prediction interval for an individual

ID: Bs2Jaggi 15-93
94) A
Explanation: The estimate of is computed using the estimated equation with given
values of the slope and intercept coefficients and x value.
ID: Bs2Jaggi 15-94
95) B
Explanation: When testing whether xi significantly influences y, in general, the null hypothesis is: H0:βj
= βj0, where βi0 = 0.
ID: Bs2Jaggi 15-95
96) C
statistic tdf is greater than tα/2,df or less than otherwise, do not reject the null
hypothesis.
ID: Bs2Jaggi 15-96
97) D
Explanation: A 100(1 - α)% confidence interval of the regression coefficient βj is computed as βj ± t
α/2,df , where is the standard error of and βj and df = n – k – 1.
ID: Bs2Jaggi 15-97
98) A
Explanation: The individual significance of all explanatory variables is evaluated first. If the slope
coefficient equals zero, there is no linear relationship between xi and y. And conversely, if
βi does not equal to zero then xi influences y. When testing whether xi significantly
influences y, in general, the null hypothesis is: H0:βj = βj0, where βi0 = 0. Using the critical
value approach, the decision rule is to reject the null hypothesis if the test statistic tdf is
greater than tα/2,df or less than –tα/2,df; otherwise, do not reject the null hypothesis.
ID: Bs2Jaggi 15-98
99) B
H0: β1 = β2 = ... = βk = 0;

ID: Bs2Jaggi 15-99

Answer Key
100) C
Explanation: The test statistic for a test of joint significance is assumed to follow the F(df1, df2)

ID: Bs2Jaggi 15-100
101) D
Explanation: To test individual significance of explanatory variables, p-value approach and confidence
intervals of the regression coefficients can be used.
ID: Bs2Jaggi 15-101
102) A
distribution with df1 equal to the number of linear restrictions and df2 = n – k – 1. The value
of the test statistic for the partial F test is computed as

ID: Bs2Jaggi 15-102
103) B
Explanation: For a specific value a 100(1 – α)% prediction interval for an individual value
of y is computed as where df = n – k – 1, se is the

standard error of and se is the standard error of the estimate.
ID: Bs2Jaggi 15-103
104) C
Explanation: For a specific value a 100(1 – α)% confidence interval of the expected value
of y is computed as where
1.
ID: Bs2Jaggi 15-104
105) D
Explanation: One of the required assumptions of regression analysis states that there is no exact linear
relationship among the explanatory variables, or in statistical terminology, there is no
perfect multicollinearity.
ID: Bs2Jaggi 15-105
106) a. H0:β1 = 0; HA:β1 ≠ 0.
b. Yes, since the p-value of 0.0106 is less than 0.05, we reject H0: β1 = 0 and conclude that x is significant.
Explanation: The competing hypotheses when testing individual significance of explanatory variables
usually are H0:βj = βj0; HA:βj ≠ βj0; where βj0 = 0. Using the p-value approach, the
decision rule is to reject the null hypothesis if the p-value is less than the significance level
α; do not reject the null hypothesis if the p-value is greater than the significance level α.
ID: Bs2Jaggi 15-106
Answer Key
107) a. H0:β1 = 1; HA:β1 ≠ 1.

b. t38 = –3.71.
c. –t0.025,38 = –2.024 and t0.025,38 = 2.024.
d. Because −3.71 < −2.024, we reject H0 and conclude that the slope differs from one.
usually are H0:βj = βj0; HA:βj ≠ βj0. The test statistic for a test of individual significance is
assumed to follow the tdf distribution with df = n – k – 1 and its value is computed as
Using the critical value approach, the decision rule for the two-tailed test
is to reject the null hypothesis H0:βj = βj0, if the test statistic tdf is less than –tα/2,df or
greater than –tα/2,df; otherwise, do not reject the null hypothesis. Use t table to get critical
values.
ID: Bs2Jaggi 15-107
108) a. H0:β1 = 0; HA:β1 ≠ 0; because the p-value = 0.0264 < 0.05 = α, we reject H0; x1 and y are linearly
related.
b. The confidence interval for β2 is reported as [−0.07, 2.59]; we cannot conclude that x2 is significant in
explaining y because the interval contains 0.
c. H0:β1 = –1; HA:β1 ≠ –1; t27 = 2.89; critical value t0,025,27 = 2.052. Because t27 > t0,025,27 we
reject H0 and conclude that β1 differs from –1.
Explanation: The competing hypotheses when testing individual significance of explanatory variables, in
general, are H0:βj = βj0; HA:βj ≠ βj0. Using the p-value approach, the decision rule is to
null hypothesis if the p-value is greater than the significance level α. If the confidence
interval for the slope coefficient contains the value zero, then the explanatory variable
associated with the regression coefficient is not significant. Conversely, if the confidence
interval does not contain the value zero, then the explanatory variable associated with the
regression coefficient is significant. The test statistic for a test of individual significance is
assumed to follow the tdf distribution with df = n – k – 1 and its value is computed as
Using the critical value approach, the decision rule for the two-tailed test
is to reject the null hypothesis H0:βj = βj0, if the test statistic tdf is less than –tα/2,df or
greater than –tα/2,df; otherwise, do not reject the null hypothesis. Use t table to get critical
values.
ID: Bs2Jaggi 15-108

Answer Key
109) a. H0: β1 = β2 = β3 = 0; HA: At least one βj ≠ 0.

b. Yes, because the p-value of 0.0003 is less than 0.05, we reject H0 and conclude that the explanatory
variables are jointly significant.
c. H0:β2 = 0; HA:β2 ≠ 0; because the p-value of 0.5972 is greater than 0.05, we do not reject H0 and
conclude that x2 is not significant in explaining y.
d. H0:β3 = 2; HA:β3 ≠ – 2. We calculate the value of the test statistic as t13 = –1.71. At the 5%
significance level, the critical values are –t0,025,13 = –2.160 and t0,025,13 = 2.160 Because −2.160 < −1.71
< 2.160, we do not reject H0 and conclude that β3 is not significantly different from −2.
Explanation: The competing hypotheses for the test whether the explanatory variables are jointly
significant, specified as H0:β1 = β2 = ... = βk = 0;HA: At least one βj ≠ 0. To conduct the
test of joint significance, we employ a one-tailed F test. The value under the heading
Significance F is the p-value. To test the individual significance the competing hypotheses
are H0:βj = βj0; HA:βj ≠ βj0. Using the p-value approach, the decision rule is to reject the
null hypothesis if the p-value is less than the significance level α; do not reject the null
hypothesis if the p-value is greater than the significance level α. The test statistic for a test
of individual significance is assumed to follow the tdf distribution with df = n – k – 1 and its
value is computed as Using the critical value approach, the decision rule
for the two-tailed test is to reject the null hypothesis H0:βj = βj0, if the test statistic tdf is
less than –tα/2,df or greater than –tα/2,df; otherwise, do not reject the null hypothesis. Use t
table to get critical values.
ID: Bs2Jaggi 15-109
110) a. H0:β2 = β3 = 0; HA: At least one βj ≠ 0.
b. Restricted model: y = β0 + β1x + ε; unrestricted model: y = β0 + β1x1 + β2x2 + β3x3 + ε.
c. F(2,26) = 19.87.
d. F0.05,(2,26) = 3.37.
e. We reject H0 because 19.87 is greater than 3.37. At the 5% level, we conclude that x2 and x3 are jointly
significant in explaining y.
Explanation: To test if the influence of explanatory variables is different, the competing hypotheses are
H0:βi = βj; HA:βi ≠ βj. The test statistic is assumed to follow the F(df1,df2) distribution
with df1 equal to the number of linear restrictions and df2 = n – k – 1. The value of the test
statistic for the partial F test is computed as Use F table to get the
critical value Fα(df1,df2). Using the critical value approach, the decision rule is to reject the
null hypothesis if the test statistic F(df1,df2) is greater than the critical value Fα(df1,df2);
otherwise, do not reject the null hypothesis.
ID: Bs2Jaggi 15-110

Answer Key
111) a. H0:β2 = β3; HA:β2 ≠ β3.

b. F(1,26) = 0.91.
c. F0.05(1,26) = 4.23.
d. We do not reject H0 because 0.91 is less than 3.37. At the 5% level, we cannot conclude that the
influence of x2 is different from the influence of x3.
ID: Bs2Jaggi 15-111
112) a. [58.54, 63.86] b. [54.13, 68.27]
Explanation: For specific values a 100(1 – α)% confidence interval of the expected value

– 1. For specific values a 100(1 – α)% prediction interval for an individual

ID: Bs2Jaggi 15-112

Answer Key
113) a. [34.85, 46.35]

b. [28.12, 53.08]
c. The confidence interval is narrower because it assumes that the expected value of the error term is zero, wherea
the prediction interval incorporates the nonzero error term.

1. For specific values a 100(1 – α)% prediction interval for an individual
value of y is computed as where df = n – k – 1, se is

the standard error of and se is the standard error of the estimate. The prediction interval
will be wider than the confidence interval, because it also incorporates the error term.
ID: Bs2Jaggi 15-113
114) a. [373.38, 499.62]
b. [367.73, 505.27]
c. The confidence interval is narrower because it assumes that the expected value of the error term is zero, wherea
the prediction interval incorporates the nonzero error term.

– 1. For specific values a 100(1 – α)% prediction interval for an individual

is the standard error of and se is the standard error of the estimate. The prediction
interval will be wider than the confidence interval, because it also incorporates the error
term.
ID: Bs2Jaggi 15-114
115) a. H0:β1 = 0; HA:β1 ≠ 0.
b. Yes, since the p-value of 0.01 is less than 0.05, we reject H0:β1 = 0 and conclude that Service is significant
in explaining Salary.
usually are H0:βj = βj0; HA:βj ≠ βj0. Using the p-value approach, the decision rule is to
ID: Bs2Jaggi 15-115

Answer Key
116) a. Here the competing hypotheses for a joint significance test of Turnover and Expense take the form:
H0:β1 = β2 = 0 At least one βj ≠ 0Because the p-value associated with F(2,7) = 4.90 is 0.047 and this
value is less than 0.10, we reject H0 and conclude that at least one of the explanatory variables is significant in
explaining Return.
b. The competing hypotheses for an individual significance test of Turnover take the form: H0:β1 = 0; HA
Because the p-value associated with t7 = 2.23 is 0.061 and this value is less than 0.10, we reject H0 and
conclude that Turnover is significant. The competing hypotheses for an individual significance test of
Expense take the form:H0:β2 = 0; HA:β2 ≠ 0. Because the p-value associated with t7 = 0.22 is 0.831 and
this value is greater than 0.10, we do not reject H0 and conclude that Expense is not significant.
Explanation: To conduct the test of joint significance: H0:β1 = β2 = ... = βk = 0; HA: At least one βj ≠ 0,
we employ a one-tailed F test. The value under the heading Significance F is the p-value.
The competing hypotheses when testing individual significance of explanatory variables, in
ID: Bs2Jaggi 15-116
117) a. H0:β ≥ 1; HA:β < 1. The value of the test statistic is t58 = –2.387. The critical values at the 5%
significance level are –t0.025,58 = –1.672 and t0.025,58 = 1.672. Since −2.387 < −1.672, we reject H0:β ≥ 1
in favor of HA:β < 1; implying that the return on Pfizer stock is less risky than the return on the market.
b. H0:α = 0; HA: α ≠ 0. Because the p-value of 0.5364 is greater than the significance level of 0.05, we
cannot reject H0:α = 0 and thus cannot conclude that there are abnormal returns.
– Rf = α + β(RM – Rf) + ε. This is essentially a simple linear regression model. The slope
coefficient β measures how sensitive the stock's return is to changes in the level of the
overall market. The competing hypotheses are H0:β ≤ 1; HA:β > 1. The CAPM theory
predicts α to be zero, and thus a nonzero estimate indicates abnormal returns. Abnormal
returns are positive when α > 0, and negative when α < 0. The competing hypotheses are
H0:α = 0; HA:α ≠ 0.
ID: Bs2Jaggi 15-117

Answer Key
118) a. Here the competing hypotheses for a joint significance test of Housing Starts and Lumber Sales take
the form: H0:β1 = β2 = 0; HA: At least one βj ≠ 0. Because the p-value associated with F(2,45) = 103.3 is
0.000 and this value is less than 0.05, we reject H0 and conclude that at least one of the explanatory variables
is significant in explaining Lumber Sales.
b. The competing hypotheses for an individual significance test of Commercial Construction take the form:
= 0; HA:β2 ≠ 0. Because the p-value associated with t45 = 3.78 is 0.0005 and this value is less than 0.05,
we reject H0 and conclude that Commercial Construction is significant.
c. H0:β1 = 1; HA:β1 ≠ 1. The value of the test statistic is t45 = –2.67. The critical values at the 5%
significance level are –t0.025,45 = –2.014 and t0.025,45 = 2.014. Because −2.67 < −2.014, we reject H0:β1 =
1 and conclude that the slope coefficient attached to Housing Starts differs from 1.
null hypothesis if the p-value is greater than the significance level α. The test statistic for a
test of individual significance is assumed to follow the tdf distribution with df = n – k – 1
and its value is computed as Using the critical value approach, the
decision rule for the two-tailed test is to reject the null hypothesis H0:βj = βj0, if the test
statistic tdf is less than –tα/2,df or greater than –tα/2,df; otherwise, do not reject the null
hypothesis. Use t table to get critical values.
ID: Bs2Jaggi 15-118

Answer Key
119) a. The competing hypotheses for a joint significance test of the three explanatory variables take the form:
H0:β1 = β2 = β3 = 0; HA: At least one βj ≠ 0. Because the p-value associated with F(3,6) = 10 is 0.0095
and this value is less than 0.10, we reject H0 and conclude that at least one of the explanatory variables is
significant in explaining crop yield.
b. The competing hypotheses for an individual significance test of fertilizer take the form: H0:β3 = 0; HA:β
Because the p-value associated with t6 = 2 is 0.0574 and this value is less than 0.10, we reject H0 and
conclude that fertilizer is significant.
c. H0:β1 = 9; HA:β1 ≠ 9. The value of the test statistic is t6 = –0.6. The critical values at the 10%
significance level are –t0.05,6 = –1.943 and t0.05,6 = 1.943. Because −1.943 < −0.6 < 1.943, we do not reject
H0:β1 = 9 and cannot conclude that the slope coefficient attached to rainfall differs from 9.
null hypothesis if the p-value is greater than the significance level α. The test statistic for a
test of individual significance is assumed to follow the tdf distribution with df = n – k – 1
and its value is computed as Using the critical value approach, the
decision rule for the two-tailed test is to reject the null hypothesis H0:βj = βj0, if the test
statistic tdf is less than –tα/2,df or greater than –tα/2,df; otherwise, do not reject the null
hypothesis. Use t table to get critical values.
ID: Bs2Jaggi 15-119
120) a. H0:β2 = β3 = 0; HA: At least one βj ≠ 0
b. F(2,35) = 26.83
c. Using the F table, the critical value approximately is F0.05,(2,35) ≈ 3.32
d. Because 26.83 > 3.32, we reject the null hypothesis and conclude that Median Income and the Mortality Rate ar
jointly significant in explaining the Poverty.
Explanation: To conduct the test of joint significance, the competing hypotheses are H0:β1 = β2 = ... =
βk = 0; HA: At least one βj ≠ 0, When testing linear restrictions, the test statistic is assumed
to follow the F(df1,df2) distribution with df1 equal to the number of linear restrictions and
df2 = n – k – 1. The value of the test statistic for the partial F test is computed as
Use F table to get the critical value Fα(df1,df2). Using the critical
value approach, the decision rule is to reject the null hypothesis if the test statistic
F(df1,df2) is greater than the critical value Fα(df1,df2); otherwise, do not reject the null
hypothesis.
ID: Bs2Jaggi 15-120

Answer Key
121) a. = 86.8 or 86,800 people; in the modified model, the point estimate of the intercept is the predicted value for
Attendance when Price equals $80.
b. In the modified model, we have estimates of the intercept and its standard error. We then find 78.2 ≤ E(y
Converting to the correct units of measurement, we have [78,200, 95,400]. Excel also provides the 95% confidenc
interval directly.
c. For prediction interval we have estimates of the intercept, its standard error, and the standard error of estimate.
We then find 42.9 ≤ (y0) ≤ 130.7. Converting to the correct units of measurement, we have [42,900,
130,700].
Explanation: If the value of the explanatory variable in the modified model after substitution equals zero,
then the point estimate for the intercept becomes the predicted value. For specific values
a 100(1 – α)% confidence interval of the expected value of y is computed as
where is the
standard error of and df = n – k – 1. A 100(1 – α)% prediction interval for an individual

ID: Bs2Jaggi 15-121
122) a. = 86.8; in the modified model, the point estimate of the intercept is when x1, x2, and x3 equal 85%,
$50,000, and 10, respectively.
b. In the modified model, we have estimates of the intercept and its standard error. We then find 14.14 ≤ E(
15.64. Excel also provides the 95% confidence interval directly.
c. For prediction interval we have estimates of the intercept, its standard error, and the standard error of estimate.
We then find 12.15 ≤ y0 ≤ 17.63.

1. A 100(1 – α)% prediction interval for an individual value of y is computed as
where df = n – k – 1, se is the standard error of

and se is the standard error of the estimate.
ID: Bs2Jaggi 15-122

Answer Key
123) a. H0:β1 = β2; HA:β1 ≠ β2

b. F(1,22) = 7.33
c. Using the F table, the critical value is F0.05,(1,22) = 4.30
d. Because 7.33 > 4.30, we reject the null hypothesis and conclude that β1 ≠ β2; thus, the influence of the two
variables on test scores is different.
ID: Bs2Jaggi 15-123
124) a.
b. Heteroscedasticity is likely a problem in this application in that the residuals seem to fan out across the horizont
axis. We can correct the biased standard errors using White's correction method.
Explanation: In the presence of heteroscedasticity (when the variance of the error term is not the same for
all observations), the OLS estimators are still unbiased, though they are no longer efficient.
Consequently, we cannot put much faith in the confidence intervals, or the standard t or F
tests, in the presence of heteroscedasticity.
ID: Bs2Jaggi 15-124

Answer Key
125) a.
b. Positive serial correlation is likely a problem in this application because the residuals form a wavelike pattern
over time. We can correct the biased standard errors using the Newey-West procedure.
Explanation: The wavelike movement in the residuals over time creates a pattern around the horizontal
axis and we conclude that positive serial correlation is likely to be a problem.
ID: Bs2Jaggi 15-125

Bus Stats Ch15 PDF

Uploaded by

Copyright:

Available Formats

Bus Stats Ch15 PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Bus Stats Ch15 PDF

Uploaded by

Copyright:

Available Formats

Exam Chapter 15

4) The term BLUE stands for Best Linear Unbiased Estimator. T or F

7) A wavelike movement in residuals suggests positive serial correlation. T or F

Exam Chapter 15 pg-1

Exam Chapter 15 pg-2

Exam Chapter 15 pg-3

Coefficients Standard Error t-stat p-value

Coefficients Standard Error t-stat p-value

Is x significantly related to y at the 5% significance level?

Coefficients Standard Error t-stat p-value

Exam Chapter 15 pg-4

A) Neither of the explanatory variables is related to the response variable.

A) 1.96 B) 3.49 C) 8.04 D) 10

Exam Chapter 15 pg-5

A) Neither of the explanatory variables is significantly related to the response

A) 1.96 B) 2.59 C) 3.49 D) 10

Exam Chapter 15 pg-6

Exam Chapter 15 pg-7

Exam Chapter 15 pg-8

Exam Chapter 15 pg-9

42) Multicollinearity is suspected when ________. 42)

43) One of the assumptions of regression analysis is ________. 43)

Exam Chapter 15 pg-10

Which of the following can be inferred from the residual plot?

Exam Chapter 15 pg-11

Coefficients Standard Error t-stat p-value

Sales = β0 + β1 Advertising + ε. The following table shows a portion of the regression

Coefficients Standard Error t-stat p-value

Exam Chapter 15 pg-12

Coefficients Standard Error t-stat p-value

Sales = β0 + β1 Advertising + ε. The following table shows a portion of the regression

Coefficients Standard Error t-stat p-value

Exam Chapter 15 pg-13

Sales = β0 + β1 Advertising + ε. The following table shows a portion of the regression

Coefficients Standard Error t-stat p-value

Exam Chapter 15 pg-14

Win = β0 + β1 Double Faults + β2Aces + ε, where Win is the proportion of winning,

Exam Chapter 15 pg-15

Exam Chapter 15 pg-16

Win = β0 + β1 Double Faults + β2 Aces + ε, where Win is the proportion of winning,

Exam Chapter 15 pg-17

Win = β0 + β1 Double Faults + β2 Aces + ε, where Win is the proportion of winning,

Exam Chapter 15 pg-18

Exam Chapter 15 pg-19

When testing whether Temperature is significant at the 5% significance level, she

Exam Chapter 15 pg-20

Exam Chapter 15 pg-21

Exam Chapter 15 pg-22

Exam Chapter 15 pg-23

Exam Chapter 15 pg-24

Exam Chapter 15 pg-25

Exam Chapter 15 pg-26

Coefficients Standard t-stat p-value Lower Upper

Exam Chapter 15 pg-27

Attendance = β0 + β1 Price + β2 Rides + ε, where Attendance is the daily attendance (in

Coefficients Standard t-stat p-value Lower Upper

Exam Chapter 15 pg-28

Coefficients Standard t-stat p-value Lower Upper

Exam Chapter 15 pg-29