BEA674 DATA AND BUSINESS DECISION MAKING

Sample Solutions: Chapter 14 Introduction to Multiple Regression Analysis

14.1 (a) Holding constant the effect of X2, for each increase of one unit in X1, the dependent
variable Y is estimated to increase, on average, of 5 units.

Holding constant the effect of X1, for each increase of one unit in X2, the dependent
variable Y is estimated to increase, on average, of 3 units.

(b) The Y-intercept 10 is the estimate of the mean value of Y if X1 and X2 are both 0.

14.3 (a) LTIMP = −0.027 + 0.791 (FOREIMP) + 0.605 (MIDSOLE)

(b) Holding constant the effect of the change in impact properties (MIDSOLE) over time,
each increase of one unit in forefoot impact absorbing capability (FOREIMP) is
estimated to result in, on average, an increase in the long-term ability to absorb shock
(LTIMP) of 0.791 units.

Holding constant the effect of forefoot impact absorbing capability (FOREIMP), each
increase of one unit in measurement of the change in impact properties (MIDSOLE)
over time is estimated to result in, on average, an increase in the long-term ability to
absorb shock (LTIMP) of 0.605 units.

(c ) According to the results of (b), FOREIMP is more effective as each unit

increase in FOREIMP will result in a higher mean increase in LTIMP than the same
amount of increase in MIDSOLE.

14.6 (Question extracted from 7th edition Levine et al.)

(a) The summary output produced using Excel is presented below:

Regression Statistics
Multiple R 0.899273236
R Square 0.808692352
Adjusted R
Square 0.788554705
Standard Error 158.9041256
Observations 22
ANOVA
Significance
df SS MS F F
Regression 2 2028032.69 1014016.345 40.15823435 1.50126E-07
Residual 19 479759.9014 25250.52112
Total 21 2507792.591

Standard
Coefficients Error t Stat P-value Lower 95% Upper 95%
-
Intercept 156.4304345 126.7578563 1.234088672 0.232217275 108.8768071 421.7376762
Radio AD. 13.08068096 1.759373685 7.434850861 4.88861E-07 9.398269521 16.76309239
Newspaper
AD. 16.79527808 2.963377915 5.667612623 1.83069E-05 10.59285683 22.99769932

Page 1 of 5
14.6 (continued)

(a) The multiple regression equation:

Sales = 156.430 + 13.081 (Radio AD.) + 16.795 (Newspaper AD.)

(b) Holding constant the effect of newspaper advertising, each increase of $1,000 in radio
advertising is estimated to result in, on average, an increase in sales of $13,081.

Holding constant the effect of radio advertising, each increase of $1,000 in newspaper
advertising is estimated to result in, on average, an increase in sales of $16,795.

(c) When there is no money spent on radio advertising and newspaper advertising, the
estimated mean amount of sales is $156,430.

(d) According to the results of (b), newspaper advertising is more effective as each
increase of $1,000 in newspaper advertising will result in a higher mean increase in
sales than the same amount of increase in radio advertising.

14.9 (a) MSR = SSR / k = 60 / 2 = 30

MSE = SSE /(n − k − 1) = 120 /18 = 6.67

(b) FSTAT = MSR / MSE = 30 / 6.67 = 4.5

(c) (** Use-F-test for testing the overall significance of a model.)

H0: 𝛽1 = 𝛽2 = 0 (There is no linear relationship between the dependent variable and

the independent variables.)

H1: At least one 𝛽𝑗 ≠ 0 , 𝑗 = 1, 2. (There is a linear relationship between the

dependent variable and at least one of the
independent variables.)

FSTAT = 4.5  F critical value = 3.55 (with 2 and 18 degrees of

freedom for  = 0.05).

Decision: Reject H0. There is evidence of a significant linear relationship between Y

and the two independent variables.

SSR 60
(d) r2 = = = 0.3333
SST 180

 n −1 
(e)
2
radj (

)
= 1 −  1 − rY2.12
n − k − 1 
= 0.2592

Page 2 of 5
14.25 (a) 95% confidence interval on 1 = b1  t ( S b1 )
2
(t-critical value of 2.1788 (with 12 degree of freedoms, α/2=0.025))

0.79116 ± 2.1788 (0.06295)

0.79116 ± 0.13716

0.65400   1  0.92832

(b) LTIMP = −0.027 + 0.791 (FOREIMP) + 0.605 (MIDSOLE)

(** Use t-test for testing the significance of ‘individual’ variable)

For FOREIMP:
H0 : β1=0 (There is no linear relationship between FOREIMP and LTIMP.)
H1 : β1≠ 0 (There is a linear relationship between FOREIMP and LITMP.)

Test Statistic:
t STAT = b1 − 0 / sb1 = 0.79116 / 0.06295 = 12.57  t critical value = 2.1788 (with 12
degrees of freedom for /2 = 0.025).

Decision: Reject H0. There is a significant linear relationship between LTIMP

and the variable FOREIMP, taking into account variable MIDSOLE.

For MIDSOLE:
H0 : β2=0 (There is no linear relationship between MIDSOLE and LITMP.)
H1 : β2≠ 0 (There is a linear relationship between MIDSOLE and LITMP.)

Test Statistic: t STAT = b2 − 0 / sb2 = 0.60484 / 0.07174 = 8.43  t critical value = 2.1788
(with 12 degrees of freedom for /2 = 0.025).

Decision: Reject H0. There is a significant linear relationship between LITMP and
variable MIDSOLE, taking into account variable FOREIMIP.

Both variables FOREIMP and MIDSOLE should be included in the model.

Page 3 of 5
14.31 (a) To determine the contribution of an independent variable to the
regression model – Use partial F-test.

(i) The null and alternative hypotheses to test for the contribution of X1 to the
model are:

H0: Variable X1 does not significantly improve the model after variable
X2 has been included.

H1: Variable X1 significantly improves the model after variable X2 has

been included.

For X1: SSR( X 1 X 2 ) = SSR( X 1 and X 2 ) − SSR( X 2 ) = 60 − 25 = 35 .

SSR ( X 1 X 2 ) 35
FSTAT = = = 5.25  F critical value = 3.55 (with 2 (df1)
MSE 120 / 18
and 18 (df2) degrees of freedom and  = 0.05 ).

Decision: Reject H0. The addition of variable X1 significantly improves

the model after variable X2 has been included.
.

(ii) The null and alternative hypotheses to test for the contribution of X2 to the
model are:

H0: Variable X2 does not significantly improve the model after variable
X1 has been included.

H1: Variable X2 significantly improves the model after variable X1 has

been included.

For X2: SSR( X 2 X 1 ) = SSR( X 1 and X 2 ) − SSR( X 1 ) = 60 − 45 = 15

SSR ( X 2 X 1 ) 15
FSTAT = = = 2.25  F critical value = 3.55 (with 2 and
MSE 120 / 18
18 degrees of freedom and  = 0.05 ).

Decision: Do not reject H0. The addition of variable X2 does not significantly
improve the model after X1 has been included.

Since variable X2 does not significantly contribute to the model in the presence of X1,
only variable X1 should be included and a simple linear regression model should be
developed.

Page 4 of 5
14.32 (a)

(i) The null and alternative hypotheses to test for the contribution of X1 to the model are:

H0: Variable X1 does not significantly improve the model after variable
X2 has been included.

H1: Variable X1 significantly improves the model after variable X2 has been
included.

For X1: SSR( X 1 X 2 ) = SSR( X 1 and X 2 ) − SSR( X 2 ) = 30 − 15 = 15 .

SSR ( X 1 X 2 ) 15
FSTAT = = = 1.25  F critical value = 4.10 (with 2 and 10
MSE 120 / 10
degrees of freedom and  = 0.05 ).
[*MSE=SSE/(n-k-1)].

Decision: Do not reject H0. Variable X1 does not significantly improve the model
after variable X2 has been included.

(ii) The null and alternative hypotheses to test for the contribution of X2 to the model are:

H0: Variable X2 does not significantly improve the model after variable
X1 has been included.

H1: Variable X2 significantly improves the model after variable X1 has

been included.

For X2: SSR( X 2 X 1 ) = SSR( X 1 and X 2 ) − SSR( X 1 ) = 30 − 20 = 10 .

SSR ( X 2 X 1 ) 10
FSTAT = = = 0.833  F critical value = 4.10 (with 2 and 10
MSE 120 / 10
degrees of freedom and  = 0.05 ).

Decision: Do not reject H0. Variable X2 does not significantly improve the model
after variable X1 has been included.

Neither variable should be included in the model and consequently other variables
should be investigated.

Page 5 of 5