Statistics for Business and Economics, 9e, Global Edition (Newbold)

Chapter 12 Multiple Regression

1) The two regressions Y = β0 + β1X1 + β2X2 + β3X3 + β4X4 + ε and Y = β0 + β1X1 + β2X2 + ε were run
using a sample of 30 observations. Assume the SSE for the first regression is 298.4, and is 382.3 for the
second regression. Test H0 : β3 = β4 = 0 at α = 0.05. Interpret your results.

Answer: = SSE / (n - K - 1) = 298.4 / (30 - 4 - 1) = 11.936

F= = = 3.51

Since F = 3.51 is greater than FR,n-K-R-1, α = F2, 23, 0.05 = 3.42, we reject H0 : β3 = β4 = 0 at α = 0.05. We
conclude that at least one of the variables X3 and X4 is statistically significant.

THE NEXT QUESTIONS ARE BASED ON THE FOLLOWING INFORMATION:

Suppose you are interested in determining the factors that influence the time required to prepare a tax
return, and developed the model: Y = β0 + β1X1 + β2X2 + β3X3 + β4X4 + ε, where Y is the amount of time
(in minutes), X1 is the income of the individual (in thousands of dollars), X2 is the age of the individual,
X3 is the number of people living in the household, and X4 is a dummy variable that takes the value 1, if
the individual owns his or her own home. After interviewing 40 accountants, you get the following
results: = 17.2 + 3.8x1 - 1.04x2 + 2.15x3 + 15.1x4, = 5.3, = 0.13, = 0.33, = 1.51, = 4.7, SSR =

164.2, SSE = 200.7, and R2 = 0.45.

2) Test H0 : β1 = 0 against H1 : β1 ≠ 0 and interpret your result.

Answer: t = (j = 1, 2,..., K)

t= = 29.23

Since t = 29.23, we can reject H0 for any value of α. This means that X1 is a statistically significant
predictor variable in explaining the amount of time required to prepare a tax return.

3) Test H0 : β3 = 0 against H1 : β3 ≠ 0 and interpret your result.

Answer: t = (j = 1, 2,..., K)

t= = 1.42

Since t = 1.42, we fail to reject H0 at α ≤ 0.10. This means that X3 is not a statistically significant predictor
variable in explaining the amount of time required to prepare a tax return.

12-1
Copyright © 2020 Pearson Education, Inc.
4) Test for H0 : β1 = β2 = β3 = β4 = 0 and interpret your result.
Answer: SSR = 164.2 and SSE = 200.7
F= = = 7.16

Since F = 7.16, we can reject H0 at α > 0.0005. This means that at least one of the predictor variables, X1,
X2, X3, and X4, is statistically significant in explaining the amount of time required to prepare a tax
return.

THE NEXT QUESTIONS ARE BASED ON THE FOLLOWING INFORMATION:

A professor investigated some of the factors that affect an individual student's final grade in his course.
He proposed the multiple regression model Y = β0 + β1X1 + β2X2 + β3X3 + ε , where Y is the final mark
(out of 100), X1 is the number of lectures skipped, X2 is the number of late assignments, and X3 is the
mid-term test mark (out of 100). The professor recorded the data for 50 randomly selected students. The
computer output is shown below.

The regression equation is = 41.6 - 3.18x1 - 1.17x2 + 0.63x3.

Predictor Coefficient StDev t

Constant 41.6 17.8 2.337
x1 -3.18 1.66 -1.916
x2 -1.17 1.13 -1.035
x2 0.63 0.13 4.846

S = 13.74 R-Sq = 30.0%

ANALYSIS OF VARIANCE

Source of Variation df SS MS F
Regression 3 3,716 1,238.667 6.558
Error 46 8,688 188.870
Total 49 12,404

5) Does the data provide enough evidence to conclude that, at the 5% significance level, the final mark
and the number of skipped lectures are linearly related?
Answer: H0 : β1 = 0 against H1 : β1 ≠ 0
Test statistic: t = -1.916
Reject H0 if |t| > t0.05, 46. Since t0.05, 46 = 2.013, we fail to reject H0. Hence, the data does not provide
enough evidence to conclude that, at the 5% significance level, the final mark and the number of skipped
lectures are linearly related.

6) Interpret the coefficient b3.

Answer: The coefficient b3 = 0.63. This tells us that for each additional mid-term test score (out of 100),
the student's final score on average increases by 0.63 points, assuming that the number of lectures
skipped, and the number of late assignments in the model are held constant.

12-2
Copyright © 2020 Pearson Education, Inc.
THE NEXT QUESTIONS ARE BASED ON THE FOLLOWING INFORMATION:
As director of the local tourist board, you are interested in determining the factors that influence the hotel
occupancy rate in your city each month. Hotel occupancy can be measured as the percentage of available
hotel rooms that are occupied by paying customers. You develop the following model: Y = β0 + β1X1 +
β2X2 + β3X3 + β4X4 + ε, where Y is the hotel occupancy rate, X1 is the total number of passengers arriving
at the airport, X2 is a price index of local hotel room rates, X3 is the consumer confidence index, and X4 is
a dummy variable = 1 during the months of June, July, and August. You look at data from the past 36
months and obtain the following results: = 67.1 + 0.02x1 - 0.055x2 + 0.08x3 + 12.3x4, R2 = 0.67, = 58.3,
= 0.008, = 0.01, = 0.06, = 4.7, and SSE = 576.

7) What are your hypotheses regarding this model? Explain. Are there any other variables you would
have liked to include?
Answer: We would expect that as the number of airline passengers increases, demand for rooms would
also increase. We would expect lower occupancy rates with higher room rates. We might expect that
higher consumer confidence might lead to more leisure travel, which might mean a higher occupancy
rate. We might also expect higher occupancy during the summer months when there are typically more
tourists staying in hotels.

8) Interpret the estimated regression coefficient b4.

Answer: With all other independent variables in the model held constant, the occupancy rate in June,
July, and August is about 12.3 percent higher compared to the other months.

THE NEXT QUESTIONS ARE BASED ON THE FOLLOWING INFORMATION:

The model y = β0 + β1X1 + β2X2 + ε was fitted to a sample of 25 families in order to explain household
milk consumption: where y = Milk consumption, in quarts, per week, x1 = Weekly income, in hundreds of
dollars, and x2 = Family size. The least squares estimates of the regression parameters were b0 = -0.03, b1 =
0.05, and b2 = 1.1, with coefficient standard errors = 0.02; = 0.38. The total sum of squares and the
error sum of squares were found to be 165.8 and 66.32 respectively.

9) Find 95% confidence interval for β1.

Answer: The 95% confidence interval for β1 is
b1 ± tn-K-1, α/2 = 0.05 ± (2.074)(0.02) = 0.05 ± 0.0415, or 0.0085 < β1 < 0.0915

10) For a fixed weekly income, test the null hypothesis that milk consumption does not depend linearly
on the size of the family against an appropriate one-sided alternative. Use α = 0.025.
Answer: H0 : β2 = 0; H1 : β2 > 0

The test statistic t = = = 2.89

Reject H0 if t > tn-K-1,α. Since t22,0.025 = 2.074, we reject H0 and conclude that for fixed weekly income,
milk consumption depends linearly on family size.

12-3
Copyright © 2020 Pearson Education, Inc.
THE NEXT QUESTIONS ARE BASED ON THE FOLLOWING INFORMATION:
In a study of foreign holdings in Egyptian banks, the following sample regression results were obtained,
based on 14 annual observations: = -3.25 + - + , and R2 = 0.92,

Where the numbers in parentheses under the coefficient estimates are the estimated coefficient standard
errors, and
y = Year-end share of assets in Egyptian bank subsidiaries held by foreigners, as a percentage of total
assets
x1 = Annual change, in billions of Egyptian pounds, in foreign direct investment in Egypt
x2 = Bank price-earnings ratio
x3 = Index of the exchange value of the Egyptian pounds

11) Find a 95% confidence interval for β1.

Answer: n = 14, b1 = 0.10, = 0.025, t10, 0.025 = 2.228
b1 ± tn-K-1, α/2 = 0.10 ± 2.228(0.025) = 0.10 ± 0.0557
or 0.0443 < β1 < 0.1557

12) Test the null hypothesis that β1 is zero, against the alternative that it is positive at the 5% significance
level, and interpret your result.
Answer: H0 : β1 = 0, H1 : β1 > 0
Reject H0 if t > tn-K-1,α = t10, 0.05 = 1.812

The test statistic is t = = = 4, therefore, reject H0. We conclude that the annual change in foreign

direct investment in Egypt is a statistically significant predictor of year-end share of assets in Egyptian
bank subsidiaries held by foreigners given that we have controlled the effect of bank price-earnings ratio
and the index of the exchange value of the Egyptian pounds.

12-4
Copyright © 2020 Pearson Education, Inc.