Sei sulla pagina 1di 8

Exercises

Note to students: The exercises involving data in this and subsequent sections were designed to
be solved using Excel.

Methods
1. In a regression analysis involving 30 observations, the following estimated regression
equation was obtained

ŷ = 17.6 + 3.8x1 – 2.3x2 + 7.6x3 + 2.7x4


and thus interpreted as:
a. b1= 3.8 is the estimate expected value when x2, x3, and x4 are held constant;
b2= -2.3 is the estimate expected value when x1, x3, and x4 are held constant;
b3= 7.6 is the estimate expected value when x1, x2, and x4 are held constant
and;
b4= 2.7 is the estimate expected value when x1, x2, and x3 are held constant.
b. The prediction of the value of ŷ when x1=10, x2=5, x3=1 and x4=2 is 57.1.

2. a. The estimated regression equation relating y to x1 is y = 1.9436x1 + 45.059. The


prediction of the value of ŷ when x1=45 is 132.521.

Relationship of y to x1
y Linear (y)

250

200

150
y

100

50
y = 1.9436x1 + 45.059
R² = 0.66

0
0 10 20 30 40 50 60 70 80

X1
SUMMARY OUTPUT

Regression Statistics
Multiple R 0.778693888
R Square 0.606364172
Adjusted R Square 0.550130482
Standard Error 26.85150605
Observations 9

ANOVA
df SS MS F Significance F
Regression 1 7774.531915 7774.531915 10.78293412 0.013417768
Residual 7 5047.023641 721.0033772
Total 8 12821.55556

Coefficients Standard Error t Stat P-value Lower 95%


Intercept 50.54669031 30.1889829 1.674342275 0.137975207 -20.83891079
x1 1.856382979 0.565326185 3.283737827 0.013417768 0.519598971

Upper 95% Lower 99.0% Upper 99.0%


121.93229 -55.0991511 156.1925317
3.193167 -0.12196656 3.834732521

b. The estimated regression equation relating y to x2 is y = 4.3215x2 + 85.217. The


prediction of the value of ŷ when x2=15 is 150.0395.

Relationship of y to x2
y Linear (y)

250

200

150
y

100

50 y = 4.3215x2 + 85.217
R² = 0.2215
0
0 2 4 6 8 10 12 14 16 18 20
X2
SUMMARY OUTPUT

Regression Statistics
Multiple R 0.489321366
R Square 0.239435399
Adjusted R Square 0.130783313
Standard Error 37.32410421
Observations 9

ANOVA
df SS MS F Significance F
Regression 1 3069.934268 3069.934268 2.20368893 0.181256643
Residual 7 9751.621287 1393.088755
Total 8 12821.55556

Coefficients Standard Error t Stat P-value Lower 95%


Intercept 92.38737624 37.70327215 2.450380854 0.044082638 3.233304538
x2 4.13490099 2.785415384 1.484482715 0.181256643 -2.451559777

Upper 95% Lower 99.0% Upper 99.0%


181.54145 -39.5545949 224.3293474
10.721362 -5.61261362 13.8824156

3. The estimated regression equation for a model involving two independent variables and
10 observations follows.
ŷ = 29.1270 + 0.5906x1 + 0.4980x2
and thus interpreted as:
a. b1= 0.5906 is the estimate expected value when x2 was held constant and;
b2= 0.4980 is the estimate expected value when x1 was held constant.
b. The prediction of the value of ŷ when x1=180 and x2=310 is 289.815.

Applications
4. A shoe store developed the following estimated regression equation relating sales to
inventory investments and advertising expenditures.
ŷ = 25 + 10x1 + 8x2
where
x1 = inventory investments ($1000s)
x2 = advertising expenditures ($1000s)
y = sales ($1000s)
and thus interpreted as:
a. The prediction of the value of sales ŷ when x1=$15000 (15) and x2=$10000 (10) is
255 or $255000 on the shoe store’s sales.
b. b1= 10 is the estimate of the expected value increase in the number of sales of
a shoe store in corresponding to the increase of the shoe store’s inventory
investments when the shoe store’s advertising expenditures are held constant
and;
b2= 8 is the estimate of the expected value increase in the number of sales of a
shoe store in corresponding to the increase of the shore store’s advertising
expenditures when the shoe store’s inventory investments are held constant.

5. a. Showtime Movie Theaters, Inc. estimated regression equation relating weekly gross
revenue to television advertising expenditures.
y = 1.6039x1 + 88.638

Showtime Movie Theaters, Inc.


Television Advertising Expenditures
97
Weekly Gross Revenue ($1000)

96

95

94
Weekly
93 Gross Revenue
($1000s)
92
Linear (Weekly
91
y = 1.6039x1 + 88.638 Gross Revenue
R² = 0.6526 ($1000s))
90

89
0 1 2 3 4 5 6
Television Advertising ($1000)

SUMMARY
OUTPUT

Regression Statistics
Multiple R 0.781166567
R Square 0.610221205
Adjusted R Square 0.532265446
Standard Error 1.23969436
Observations 7
ANOVA
df SS MS F Significance F
Regression 1 12.03007519 12.03007519 7.827789 0.03809156
Residual 5 7.684210526 1.536842105
Total 6 19.71428571

Coefficients Standard Error t Stat P-value Lower 95%


Intercept 87.26315789 2.252914394 38.73345481 2.16E-07 81.47185708
X1 2.105263158 0.752465937 2.797818552 0.038092 0.170987889

Upper 95% Lower 99.0% Upper 99.0%


93.0544587 78.17908493 96.34723086
4.03953843 -0.92878709 5.139313406

b. Thus, the estimated regression equation for both television and newspaper advertising
for Showtime Movie Theaters, Inc.
ŷ = 81.9401 + 2.7687x1 + 1.2929x2

SUMMARY
OUTPUT

Regression Statistics
Multiple R 0.974965335
R Square 0.950557404
Adjusted R Square 0.925836107
Standard Error 0.493640926
Observations 7

ANOVA
df SS MS F Significance F
Regression 2 18.73956026 9.36978013 38.45095 0.00244457
Residual 4 0.974725455 0.243681364
Total 6 19.71428571

Coefficients Standard Error t Stat P-value Lower 95%


Intercept 81.94014096 1.3542042 60.50796544 4.47E-07 78.18026734
X1 2.768726438 0.325214302 8.513544524 0.001044 1.865786782
X2 1.292902803 0.246395312 5.247270311 0.006309 0.608799746
Upper 95% Lower 99.0% Upper 99.0%
85.7000146 75.70525635 88.17502557
3.67166609 1.271408939 4.266043937
1.97700586 0.158475413 2.427330193

c. The estimated regression equation coefficient for television advertising expenditures is


not the same because on part a, we only relate the television advertising expenditures to
Showtime Movie Theaters. Inc.’s Weekly Gross Revenue and we used only one
independent variable. While in part b, we relate the two independent variables (television
and newspaper advertising expenditures) to the dependent variable (weekly gross
revenue) that is why we come up with a new regression equation having the different
estimated regression equation coefficient for x1.
d. The prediction of the value of weekly gross revenue ŷ when x1=$3500 (3.5) and
x2=$1800 (1.8) is 93.95777 or $93,957.77 or an estimated $94,000 weekly gross
revenue for Showtime Movie Theaters, Inc.

6. a. National Basketball Association (NBA) estimated regression equation in predicting the


proportion of games won (PCT) to the proportion of field goals made by the team (FG%).
y = 3.9576x1 - 1.2207

NBA's PCT relating to FG%


PCT Linear (PCT)
0.8

0.7

0.6
PCT Proportion

0.5

0.4

0.3 y = 3.9576x1 - 1.2207


0.2 R² = 0.2008

0.1

0
0.4 0.41 0.42 0.43 0.44 0.45 0.46 0.47 0.48
FG% Proportion
SUMMARY OUTPUT

Regression Statistics
Multiple R 0.473165156
R Square 0.223885264
Adjusted R Square 0.194034698
Standard Error 0.120157331
Observations 28

ANOVA
df SS MS F Significance F
Regression 1 0.10828629 0.10828629 7.500201465 0.010989505
Residual 26 0.375382389 0.014437784
Total 27 0.483668679

Coefficients Standard Error t Stat P-value Lower 95%


Intercept -1.208070105 0.627896264 -1.923996325 0.06536701 -2.498729361
FG% 3.947896533 1.441548629 2.73864957 0.010989505 0.984750889

Upper 95% Lower 99.0% Upper 99.0%


0.082589151 -2.95281458 0.53667437
6.911042178 -0.05775559 7.95354866

b. The slope of the estimated regression equation developed in part a is linear.


c. Thus, the National Basketball Association (NBA) estimated regression equation in
predicting the proportion of games won (PCT) to the proportion of field goals made by
the team (FG%), the proportion of three-point shots made by the team’s opponent (Opp 3
Pt%), and the number of turnovers committed by the team’s opponent (Opp TO).
ŷ = -1.1190 + 4.7292x1 -2.5423x2 + 0.02866x3

SUMMARY OUTPUT

Regression Statistics
Multiple R 0.752469081
R Square 0.566209718
Adjusted R Square 0.511985933
Standard Error 0.093499281
Observations 28
ANOVA
df SS MS F Significance F
Regression 3 0.273857906 0.091285969 10.44209133 0.000138482
Residual 24 0.209810772 0.008742116
Total 27 0.483668679

Coefficients Standard Error t Stat P-value Lower 95%


Intercept -1.119030374 0.581001672 -1.92603641 0.066025026 -2.318158889
FG% 4.729241232 1.138723593 4.153107271 0.00035761 2.379031248
Opp 3 Pt% -2.542341204 0.677607378 -3.751938491 0.000983751 -3.940854096
Opp TO 0.028661443 0.012497038 2.293458993 0.030880608 0.002868825

Upper 95% Lower 99.0% Upper 99.0%


0.08009814 -2.7440569 0.505996154
7.079451217 1.544300231 7.914182234
-1.14382831 -4.43756805 -0.64711436
0.054454061 -0.00629201 0.063614902

d. After getting the regression data analysis and the summary output of the given data sets above,
we can create an estimated regression equation by referring to the last table where we can find the
respective coefficient value for x1 (FG%), x2 (Opp 3 Pt%), and x3 (Opp TO). Also, the table
shows the intercept that may complete the desired estimated regression equation. Looking up the
values and we can get the estimated regression equation ŷ = -1.1190 + 4.7292x1 -2.5423x2 +
0.02866x3

e. The prediction of the value for the proportion of games won for a team ŷ when x1=0.4,
x2=0.34, and x3=17 is 0.395518.

Potrebbero piacerti anche