Lesson 10 - Regression Analysis

LESSON 10
SIMPLE LINEAR
REGRESSION
CHAPTER 14 SECTIONS 14.1 TO 14.3

7 SECTIONS
1.  SIMPLE LINEAR REGRESSION MODEL 14.2
2.  LEAST SQUARE METHOD 14.2
3.  COEFFICIENT OF DETERMINATION 14.2
4.  MODEL ASSUMPTIONS 14.2
5.  TESTING FOR SIGNIFICANCE 14.3
6.  COVARIANCE & COEFFICIENT OF
CORRELATION 14.1
7.  USING THE ESTIMATED REGRESSION 11.7
EQUATION FOR ESTIMATION AND PREDICTION

1 . SIMPLE LINEAR
REGRESSION MODEL
REGRESSION MODEL
REGRESSION EQUATION
ESTIMATED REGRESSION EQUATION
Created by Samie L.S. Ly, must not be used for profitable
purposes without permission.
3

THE OBJECTIVE OF
SLR
Let’s say, I would like to create the ultimate equation to figure
out my Final COMM 215 grade.
What influences my Final COMM 215 grade?
Think of a few examples.
4

FINAL COMM 215 Grade (y) =
Study Time (x1) +
Work Time (x2) +
# of hours spent on Facebook (x3) +
Family Time (x4) +
anything else you can think of (x…) …
5

THE OBJECTIVE OF
SLR
Let’s say,
I study 5 hours a week,
work part-time for 25 hours a week,
I spend 3 hours on facebook,
And I have 2 kids.
Here is 1 scenario.
6

THE OBJECTIVE OF
SLR
What if it was someone else?
With a different profile?
Do I have to make my calculations all over again?
If I set up a regression line, I just need to plug in values and
get an estimation of my Final COMM 215 grade.
Voila!
Then you might ask… how?
7

THE OBJECTIVE OF
SLR
How do I create this regression line?
Answer: By gathering data from history.
I am going to take a sample of individuals who took COMM
215 before and write down their profile.
Hours per
week
Study Time Work Time Family
Time
Facebook
Time
Bob 10 25 5 0
Sally 12 0 15 15
Eric 3 40 10 0
8

THE OBJECTIVE OF
SLR
Since we are only considering 1 independent variable, let’s
take just 1, Study Time as the main indicator of your Final
COMM 215 grade.
Hours
per
week
Study
Time
(x)
Final
Grade
(y)
Bob 10 89
Sally 12 67
Eric 3 45
9

THE OBJECTIVE OF
SLR
We’ve generated this equation!
Now, if Michelle asks, if I study 8 hours a week, what would
be my estimated Final COMM 215 grade?
y = 3.4478x + 38.269
10

SIMPLE LINEAR REGRESSION
1 VARIABLE
FINAL COMM 215 GRADE (y) = STUDY TIME (x)
MULTIPLE LINEAR REGRESSION
MORE THAN 1 VARIABLE
FINAL COMM 215 GRADE (y) =
STUDY TIME (x1) + WORK TIME(x2) + ….
11

As a way of predicting sales
Managerial decisions are often made based
on the relationship between two or more
variables.
The statistical process is called
regression analysis used to develop an
equation showing
how the variables are related.
12

The variable
being predicted is
called the dependent variable.
The variable or variables
being used to predict the value of the
dependence variable are
called independent variables.
13

FINAL COMM 215 Grade (y) =
Study Time (x1) +
Work Time (x2) +
# of hours spent on Facebook (x3) +
Family Time (x4) +
anything else you can think of (x…) …
14

EQUATION
  Positive Linear Relationship
E(y)
x
Slope β1
is positive
Regression line
Intercept
β0
n The relationship between the two variables is
approximated by a straight line.
15
PAGE 572
GROEBNER et Al. (2014)

  Negative Linear Relationship
E(y)
x
Slope β1
is negative
Regression line
Intercept
β0
EQUATION
When would be a case of a negative relationship? If I spend
all my time watching movies instead of studying, would it
increase my grade? Or decrease my grade?
16

  No Relationship
E(y)
x
Slope β1
is 0
Regression lineIntercept
β0
EQUATION
Hmm… what would be an example of No relationship?
If my friend eats 5 ice creams everyday, does it have any
relationship with my grade? Not really right?
17

SIMPLE LINEAR
REGRESSION MODEL
βo and β1 – parameters of the model
ε is a random variable referred to as the error term.
ε – variability in y that cannot be explained
y = β0 + β1x +εE(y)
x
Slope β1
is positive
Regression line
Intercept
β0
18

CHARACTERISTICS OF
THE ERROR TERM
y = β0 + β1x +ε
The more variables you add
The small ε will become
because now you know more!
y=β0+β1x1+ β2x2+ β3x3+ β4x4
+ β5x5+ β6x6+ β7x7+ β8x8…+ βixi
19
PAGE 573

CHARACTERISTICS OF
THE ERROR TERM
y = β0 + β1x +ε
The more variables you add
The small ε will become
because now you know more!
y=β0+β1x1+ β2x2+ β3x3+ β4x4
+ β5x5+ β6x6+ β7x7+ β8x8…+ βixi
Cannot be
explained!
Cannot be
calculated!
Cannot.. Just don’t
ask sigh…
20

REGRESSION EQUATION
Describes how the expected value of y, E(y) is related to x
E(y) = β0 + β1x
•  E(y) is the expected value of y for a given x value.
•  β1 is the slope of the regression line.
•  β0 is the y intercept of the regression line.
•  Graph of the regression equation is a straight line.
21
PAGE 573

ESTIMATED SIMPLE LINEAR
REGRESSION EQUATION
Sample statistics are
E(y) = β0 + β1x
0 1
ˆy b b x= +
•  is the estimated value of y for a given x value.ˆy
•  b1 is the slope of the line.
•  b0 is the y intercept of the line.
•  The graph is called the estimated regression line.
22

PROBLEM #10.1
In the linear regression equation, y = b0 + b1x1, why is the
term at the left given as ŷ instead of simply y?
because it is an estimated value for the dependent
variable given a value of x.
23

2. LEAST SQUARE METHOD
Is a procedure for using sample data to find
the estimated regression equation.
The goal to using the least square method
0 1
ˆy b b x= +
24

Good fit
with the
line
25

LEAST SQUARES METHOD
Least Squares Criterion
min (y yi i−∑  )2
where:
yi = observed value of the dependent variable
for the ith observation
^
yi = estimated value of the dependent variable
for the ith observation
26
PAGE 572

PROBLEM # 10.2
A scatter diagram includes the data points (x=2,y=10), (x=3,
y=12), (x=4, y=20), and (x=5, y=16). Two regression lines are
proposed: (1) y = 10 + x, and (2) y = 8 + 2x. Using the least-
squares criterion, which of these regression lines is the better fit to
the data? Why?
ŷ = 10 + x
x y ŷ (y-ŷ) (y-ŷ)2x y ŷ (y-ŷ) (y-ŷ)2
2 10
3 12
4 20
5 16
ŷ = 10 + = 10 + 2 = 12
27

PROBLEM # 10.2
the data? Why?
ŷ = 10 + x
x y ŷ (y-ŷ) (y-ŷ)2x y ŷ (y-ŷ) (y-ŷ)2
2 10
3 12
4 20
5 16
ŷ = 10 + = 10 + 2 = 12
x y ŷ (y-ŷ) (y-ŷ)2
2 10 12
3 12
4 20
5 16
x y ŷ (y-ŷ) (y-ŷ)2
2 10 12
3 12 13
4 20 14
5 16 15
x y ŷ (y-ŷ) (y-ŷ)2
2 10 12 -2
3 12 13 -1
4 20 14 6
5 16 15 1
x y ŷ (y-ŷ) (y-ŷ)2
2 10 12 -2 4
3 12 13 -1 1
4 20 14 6 36
5 16 15 1 1
x y ŷ (y-ŷ) (y-ŷ)2
2 10 12 -2 4
3 12 13 -1 1
4 20 14 6 36
5 16 15 1 1
42
28

PROBLEM # 10.2
the data? Why?
ŷ = 8 + 2x
x y ŷ (y-ŷ) (y-ŷ)2
2 10
3 12
4 20
5 16
x y ŷ (y-ŷ) (y-ŷ)2
2 10 12
3 12 14
4 20 16
5 16 18
x y ŷ (y-ŷ) (y-ŷ)2
2 10 12 -2
3 12 14 -2
4 20 16 4
5 16 18 -2
x y ŷ (y-ŷ) (y-ŷ)2
2 10 12 -2 4
3 12 14 -2 4
4 20 16 4 16
5 16 18 -2 4
x y ŷ (y-ŷ) (y-ŷ)2
2 10 12 -2 4
3 12 14 -2 4
4 20 16 4 16
5 16 18 -2 4
28
29

PROBLEM # 10.2
the data? Why?
For ŷ = 10 + x, the least square criterion = 42
For ŷ = 8 + 2x, the least square criterion = 28
30

Slope for the Estimated Regression Equation
1 2
( )( )
( )
i i
i
x x y y
b
x x
− −
=
−
∑
∑
where:
xi = value of independent variable for ith
observation
_
y = mean value for dependent variable
_
x = mean value for independent variable
yi = value of dependent variable for ith
observation
31
PAGE 573

  y-Intercept for the Estimated Regression Equation
0 1b y b x= −
32

PROBLEM # 10.3
For a sample of 8 employees, a personnel
director has collected the following data on
ownership of company stock versus years with
the firm.
a. Determine the least-squares regression line
and interpret its slope.
b. For an employee who has been with the firm
10 years, what is the predicted number of shares
of stock owned?
X=
years
Y
=
shares

6
300

12
408

14
560

6
252

9
288

13
650

15
630

9
522

33

PROBLEM # 10.3
For a sample of 8 employees, a personnel director has
collected the following data on ownership of company stock
versus years with the firm.
a.  Determine the least-squares regression line and interpret
its slope.
X=
years
Y
=
shares

6
300

12
408

14
560

6
252

9
288

13
650

15
630

9
522

0
100
200
300
400
500
600
700
0 2 4 6 8 10 12 14 16
shares
years
Ownership of company stock
vs years with the firm
0
100
200
300
400
500
600
700
0 2 4 6 8 10 12 14 16
shares
years
0 1
ˆy b b x= +
34

PROBLEM # 10.3
its slope.
X=
years
Y
=
shares

6
300

12
408

14
560

6
252

9
288

13
650

15
630

9
522

0 1
ˆy b b x= +
35

x y xy x2
6.00 300.00 1800.00 36.00
12.00 408.00 4896.00 144.00
14.00 560.00 7840.00 196.00
6.00 252.00 1512.00 36.00
9.00 288.00 2592.00 81.00
13.00 650.00 8450.00 169.00
15.00 630.00 9450.00 225.00
9.00 522.00 4698.00 81.00
Σ x = 84.00 Σ y = 3610.00 Σ xy = 41,238.00 Σ x2 = 968.00
x y xy x2
6.00 300.00 1800.00 36.00
12.00 408.00 4896.00
14.00 560.00
6.00 252.00 1512.00 36.00
9.00 288.00 81.00
13.00 650.00 8450.00
15.00 630.00 9450.00 225.00
9.00 522.00 81.00
Σ x = Σ y = Σ xy = Σ x2 =
PROBLEM # 10.3
For a sample of 8 employees, a personnel director has collected
the following data on ownership of company stock versus years
with the firm.
a.  Determine the least-squares regression line and interpret its
slope.
X=
years
Y
=
shares

6
300

12
408

14
560

6
252

9
288

13
650

15
630

9
522

x y xy x2
6.00 300.00 1800.00 36.00
12.00 408.00 4896.00
14.00 560.00 7840.00
6.00 252.00 1512.00 36.00
9.00 288.00 2592.00 81.00
13.00 650.00 8450.00
15.00 630.00 9450.00 225.00
9.00 522.00 4698.00 81.00
Σ x = Σ y = Σ xy = Σ x2 =
x y xy x2
6.00 300.00 1800.00 36.00
12.00 408.00 4896.00 144.00
14.00 560.00 7840.00 196.00
6.00 252.00 1512.00 36.00
9.00 288.00 2592.00 81.00
13.00 650.00 8450.00 169.00
15.00 630.00 9450.00 225.00
9.00 522.00 4698.00 81.00
Σ x = Σ y = Σ xy = Σ x2 =
x y xy x2
6.00 300.00 1800.00 36.00
12.00 408.00 4896.00 144.00
14.00 560.00 7840.00 196.00
6.00 252.00 1512.00 36.00
9.00 288.00 2592.00 81.00
13.00 650.00 8450.00 169.00
15.00 630.00 9450.00 225.00
9.00 522.00 4698.00 81.00
Σ x = 84.00 Σ y = 3610.00 Σ xy = 41,238.00 Σ x2 = 968.00
36

PROBLEM # 10.3
its slope.
0 1
ˆy b b x= +
Σ x = 84.00 Σ y = 3610.00 Σ xy = 41,238.00 Σ x2 = 968.00
Do your calculations! Find bo and b1!
37

PROBLEM # 10.3
its slope.
0 1
ˆy b b x= +
Σ x = 84.00 Σ y = 3610.00 Σ xy = 41,238.00 Σ x2 = 968.00
b1=
Σxy –[Σx Σy / n]
Σx2 – [(Σx)2 / n]
41,238 – [(84) (3610)/8]
(968) – [(84)2 / 8]
=
3333
86
= = 38.756
38

PROBLEM # 10.3
its slope.
0 1
ˆy b b x= +
Σ x = 84.00 Σ y = 3610.00 Σ xy = 41,238.00 Σ x2 = 968.00
b0= ymean – b1 xmean = 451.25 – 38.756 (10.5)
ymean= Σ y / n = 3610.00/8 = 451.25
xmean= Σ x / n = 84 /8 = 10.5
= 44.314
39

PROBLEM # 10.3
its slope.
X=
years
Y
=
shares

6
300

12
408

14
560

6
252

9
288

13
650

15
630

9
522

0
100
200
300
400
500
600
700
0 2 4 6 8 10 12 14 16
shares
years
0
100
200
300
400
500
600
700
0 2 4 6 8 10 12 14 16
shares
years
0 1
ˆy b b x= +
y = 38.756x + 44.314
R² = 0.72009
0
100
200
300
400
500
600
700
0 2 4 6 8 10 12 14 16
shares
years
b1= 38.756
bo= 44.314
40

PROBLEM # 10.3
X=
years
Y
=
shares

6
300

12
408

14
560

6
252

9
288

13
650

15
630

9
522

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
C D E F G H I
SUMMARY OUTPUT
Regression Statistics
Multiple R 0.849
R Square 0.720
Adjusted R Square 0.673
Standard Error 91.479
Observations 8
ANOVA
df SS MS F Significance F
Regression 1 129173.13 129173.13 15.436 0.008
Residual 6 50210.37 8368.40
Total 7 179383.50
Coefficients Standard Error t Stat P-value Lower 95% Upper 95%
Intercept 44.3140 108.51 0.408 0.697 -221.197 309.825
Years 38.7558 9.86 3.929 0.008 14.618 62.893
0 1
ˆy b b x= +
41

PROBLEM # 10.3
b. For an employee who has been with the firm
10 years,
what is the predicted number of shares of stock owned?
X=
years
Y
=
shares

6
300

12
408

14
560

6
252

9
288

13
650

15
630

9
522

ŷ = 44.314 + 38.756 x
ŷ = 44.314 + 38.756 (10)
ŷ = 431.9
the predicted value of Shares will be 431.9
42

3. COEFFICIENT OF DETERMINATION
CORRELATION COEFFICIENT
y = 38.756x + 44.314
R² = 0.72009
0
100
200
300
400
500
600
700
0 2 4 6 8 10 12 14 16
shares
years
43

COEFFICIENT OF
DETERMINATION
0
100
200
300
400
500
600
700
0 2 4 6 8 10 12 14 16 18
SHARES
YEARS
x
y

6
300

9
408

14
560

6
252

6
288

16
650

15
630

13
522

R² = 0.98688
44

PROVIDES A MEASURE OF THE GOODNESS OF FIT FOR THE
ESTIMATED REGRESSION EQUATION
IN OTHER WORDS
DOES THE INDEPENDENT VARIABLE
EXPLAIN
THE DEPENDENT VARIABLE WELL?
COEFFICIENT OF
DETERMINATION
45

FINAL COMM 215 Grade
(y) =
Study Time (x1) +
Work Time (x2) +
# of hours spent on Facebook
(x3) +
Family Time (x4) +
anything else you can think of
(x…) …
Explain 40% of
the grade
Explain 15% of
the grade
Explain 9% of
the grade
Explain 15% of
the grade
46

FINAL COMM 215 Grade
(y) =
Study Time (x1) +
Work Time (x2) +
# of hours spent on Facebook
(x3) +
Family Time (x4) +
anything else you can think of
(x…) …
Explain 40% of
the grade
Explain 15% of
the grade
Explain 9% of
the grade
Explain 15% of
the grade
These 4 Variables
explain
79% of your
grade!
47

ymean=451.25
Sum of Squares due to Error (SSE)
Difference between Observed y and Expected ŷ
y - ŷΣ( )2
48
PAGE 575

ymean=451.25
Sum of Squares due to Regression (SSR)
Difference between Mean of y and Expected ŷ
ymean - ŷΣ( )2
49
PAGE 581

ymean=451.25
Total Sum of Squares (SST)
Difference between Observed y and Mean of y
y- ymeanΣ( )2
50
PAGE 580

COEFFICIENT OF DETERMINATION
Relationship Among SST, SSR, SSE
where:
SST = total sum of squares
SSR = sum of squares due to regression
SSE = sum of squares due to error
SST = SSR + SSE
2
( )iy y−∑ 2
ˆ( )iy y= −∑ 2
ˆ( )i iy y+ −∑
51

  The coefficient of determination is:
where:
SSR = sum of squares due to regression
= explained variation
SST = total sum of squares
= total variation
r2 = SSR/SST
COEFFICIENT OF DETERMINATION
PAGE 392
ADCOCK ET Al. (2011)
52

COEFFICIENT OF
DETERMINATION (R2)
Expresses proportion of
the variation in the
dependent variable (y)
that is explained by the
regression line:
ŷ = b0+b1x1
COEFFICIENT OF
CORRELATION (R)
Describes both the
direction and the
strength of the linear
relationship between
two variables
r = (sign of b1) √ r2
53

PROBLEM #10.4
For a set of data, the total variation or sum of squares for y is
SST = 143.0, and error sum of squares is SSE = 24.0. What
proportion of the variation in y is explained by the regression
equation?
54

PROBLEM #10.4
equation?
= SSR/SST
We are asked to find the
coefficient of determination :
R2
We also know SST = SSR + SSE
55

PROBLEM #10.4
equation?
= SSR/SST
We are asked to find the
coefficient of determination :
R2
We also know SST = SSR + SSE
SST – SSE = SSR
143.0 – 24.0 = 119
119 /143.0 = 0.832 = 83.2%
Interpretation:
83.2% of the variation in y is explained by x
56

4. MODEL
ASSUMPTIONS
Before conducting regression analysis…
What determines an appropriate model for
the relationship between the dependent and
the independent variable(s).
57

Even if the data fits well , the estimated
regression equation should not be used
until further analysis of
how appropriate the assumed model is.
One way to determine is to test for
significance.
58

ASSUMPTIONS-
ERROR TERM
1.  The error term ε is a random variable with
an expected value of zero. In estimating an
element that is unpredictable, it is best to
assume it to be zero.
2.  The variance of ε, denoted by σ2 is the
same for all values of x.
3.  The values of error are independent.
4.  The error term is normally distributed.
PAGE 370
59

Do you see the normal curve?
60

Do you see the normal curve?
Right here is the population
mean, which is also your point
on the line.
61

Assumption # 2: The variance/ standard
deviation for each value x is the same. All
of these normal distributions have the
same standard deviation.
Assumption # 2: The variance/ standard
deviation for each value x is the same. All
of these normal distributions have the
same standard deviation.
62

Assumptions # 3 & 4: the error terms are independent
and are normally distributed.
If the point is right on the line, then you have en error
of 0. If your point is away from the line, the further it is
away, the larger the error term is.
63

Assumptions # 3 & 4: the error terms are independent
and are normally distributed.
If the point is right on the line, then you have en error
of 0. If your point is away from the line, the further it is
away, the larger the error term is.
Away from the line, larger
error = larger standard error.
Right on the line, error = 0,
standard error = 0. It is exactly
where we expect the point to
be.
64

65

5. TESTING FOR
SIGNIFICANCE
ESTIMATE σ2
TESTING FOR SIGNIFICANCE
CONFIDENCE INTERVAL FOR B1
66

An Estimate of σ 2
∑∑ −−=−= 2
10
2
)()ˆ(SSE iiii xbbyyy
where:
s2 = MSE = SSE/(n - 2)
The mean square error (MSE) provides the estimate
of σ 2, and the notation s2 is also used.
67
SSE = ! yi
2
" b0 yi " b1! xi yi!

An Estimate of σ
2
SSE
MSE
−
==
n
s
•  To estimate σ we take the square root of σ 2.
•  The resulting s is called the standard error of
the estimate.
Why is this n-2?, it actually is n-k-1.
In a simple linear regression, since you always
have 1 independent variable (k=1), therefore
automatically it becomes n-1-1.
68

14-69
Large Standard Error Small Standard Error
STANDARD ERROR
OF THE ESTIMATE

To test for a significant regression relationship, we
must conduct a hypothesis test to determine whether
the value of β1 is zero.
Two tests are commonly used:
t Test and F Test
Both the t test and F test require an estimate of σ 2,
the variance of ε in the regression model.
70

Hypotheses
Test Statistic
TESTING FOR SIGNIFICANCE: T
TEST
0 1: 0H β =
1: 0aH β ≠
1
1
b
b
t
s
= where
1 2
( )
b
i
s
s
x x
=
Σ −
71
PAGE 584

1. Set up Hypotheses.
2. What is the appropriate test statistic to use?.
3. Calculate the test statistic value.
α = .054. Find the critical value for the test statistic.
0 1: 0H β = 1: 0aH β ≠
1
1
b
b
t
s
=
5. Define the decision rule
6. Make your decision
7. Interpret the conclusion in context
TESTING FOR SIGNIFICANCE: T TEST
72

The form of a confidence interval for β1 is:
CONFIDENCE INTERVAL FOR Β1
11 /2 bb t sα±
where is the t value providing an area
of α/2 in the upper tail of a t distribution
with n - 2 degrees of freedom
2/αt
b1 is the
point
estimator
is the
margin
of error
1/2 bt sα
73
PAGE 594

n  H0 is rejected if the hypothesized value of β1 is not
included in the confidence interval for β1.
n  We can use a 95% confidence interval for β1 to test
the hypotheses just used in the t test.
74

Reject H0 if 0 is not included in
the confidence interval for β1.
0 is not included in the confidence interval.
Reject H0
= 5.0 ± 2.048(2.25) = 5.0 ± 4.60812/1 bstb α±
or 0.392 to 9.608
  Rejection Rule
  95% Confidence Interval for β1
  Conclusion
75

SOME CAUTIONS ABOUT THE
INTERPRETATION OF
SIGNIFICANCE TESTS
n  Just because we are able to reject H0: β1 = 0 and
demonstrate statistical significance does not enable
us to conclude that there is a linear relationship
between x and y.
n  Rejecting H0: β1 = 0 and concluding that the
relationship between x and y is significant does
not enable us to conclude that a cause-and-effect
relationship is present between x and y.
76

What is a standard error of the estimate?
What is a standard error of the slope?
2
SSE
MSE
−
==
n
s 1 2
( )
b
i
s
s
x x
=
Σ −
77

PROBLEM # 10.5
At 5% level significance. The manager of Colonial Furniture has
been reviewing weekly advertising expenditures. During the past 6
months, all advertisements for the store have appeared in the local
newspaper. The number of ads per week has varied from one to
seven. The store’s sales staff has been tracking the number of
customers who enter the store each week. The number of ads and
the number of customers per week for the past 26 weeks were
recorded.
a.  Determine the sample regression line
b.  Interpret the coefficients.
c.  Can the manager infer that the larger the number of ads, the larger
the number of customers?
d. Find and interpret the coefficient of determination.
e. In your opinion, is it worthwhile exercise to use the regression
equation to predict the number of customers who will enter the store,
given that Colonial intends to advertise five times in the newspaper? If
so, find the 95% prediction interval. If not, explain why not.
78

PROBLEM # 10.5
At 5% level significance. The manager of Colonial Furniture has been
reviewing weekly advertising expenditures. During the past 6 months, all
advertisements for the store have appeared in the local newspaper. The
number of ads per week has varied from one to seven. The store’s sales
staff has been tracking the number of customers who enter the store each
week. The number of ads and the number of customers per week for the
past 26 weeks were recorded.
a.  Determine the sample regression line
b.  Interpret the coefficients.
Ads Customer
5 353
6 319
3 440
2 332
4 172
2 331
4 344
2 483
4 329
2 532
7 496
5 393
4 376
7 372
2 512
5 254
5 459
2 153
1 426
6 566
6 596
5 395
6 676
3 194
2 135
7 367
!
ŷ = 296.92 + 21.356x
On average each additional ad generates 21.36 customers.
79

PROBLEM # 10.5
c. Can the manager infer that the larger the number of ads, the
larger the number of customers?
1. Set up the hypotheses:
Ho: β1 = 0 ; Ha: β1 > 0
2. What is the appropriate test statistics to use?
One tail t-test, α=0.05
80

1. Set up Hypotheses.
2. What is the appropriate test statistic to use?.
α = .054. Find the critical value for the test statistic.
Testing for Significance: t Test
1
1
b
b
t
s
=
7. Interpret the conclusion in context
81

1
1
b
b
t
s
=
2n
SSE
s
!
=" !
1
3
2
4
5
SSE
Standard Error of Estimate (sε)
= = SSxx
1 2
( )
b
i
s
s
x x
=
Σ −
n-k-1
82

PROBLEM # 10.5
At 5% level significance. The manager of
Colonial Furniture has been reviewing
weekly advertising expenditures. During
the past 6 months, all advertisements for
the store have appeared in the local
newspaper. The number of ads per week
has varied from one to seven. The store’s
sales staff has been tracking the number of
customers who enter the store each week.
The number of ads and the number of
customers per week for the past 26 weeks
were recorded.
c.  Can the manager infer that the larger
the number of ads, the larger the
number of customers?
Ads x Customer y ŷ
=
296.92
+
21.356x
y-‐ŷ
(y-‐ŷ)2

5.00 353.00 403.70
-‐50.70
2570.49

6.00 319.00 425.06
-‐106.06
11247.88

3.00 440.00 360.99
79.01
6242.90

2.00 332.00 339.63
-‐7.63
58.25

4.00 172.00 382.34
-‐210.34
44244.60

2.00 331.00 339.63
-‐8.63
74.51

4.00 344.00 382.34
-‐38.34
1470.26

2.00 483.00 339.63
143.37
20554.38

4.00 329.00 382.34
-‐53.34
2845.58

2.00 532.00 339.63
192.37
37005.45

7.00 496.00 446.41
49.59
2458.97

5.00 393.00 403.70
-‐10.70
114.49

4.00 376.00 382.34
-‐6.34
40.25

7.00 372.00 446.41
-‐74.41
5537.15

2.00 512.00 339.63
172.37
29710.73

5.00 254.00 403.70
-‐149.70
22410.09

5.00 459.00 403.70
55.30
3058.09

2.00 153.00 339.63
-‐186.63
34831.50

1.00 426.00 318.28
107.72
11604.46

6.00 566.00 425.06
140.94
19865.21

6.00 596.00 425.06
170.94
29221.85

5.00 395.00 403.70
-‐8.70
75.69

6.00 676.00 425.06
250.94
62972.89

3.00 194.00 360.99
-‐166.99
27884.99

2.00 135.00 339.63
-‐204.63
41874.26

7.00 367.00 446.41
-‐79.41
6306.27

424281.17

SSE = Σ(y-ŷ)2
1
SSE = Σ(y-ŷ)2 = 424281.17
83

PROBLEM # 10.5
Colonial Furniture has been reviewing weekly
advertising expenditures. During the past 6
months, all advertisements for the store have
appeared in the local newspaper. The number
of ads per week has varied from one to seven.
The store’s sales staff has been tracking the
number of customers who enter the store
each week. The number of ads and the
number of customers per week for the past 26
weeks were recorded.
c.  Can the manager infer that the larger the
number of ads, the larger the number of
customers?
2
2n
SSE
s
!
=" !
SSE = Σ(y-ŷ)2 = 424281.17

2n
SSE
s
!
=" !
n-k-1
424281.17

26-1-1
2n
SSE
s
!
=" !132.96
84

PROBLEM # 10.5
At 5% level significance. The manager of Colonial
Furniture has been reviewing weekly advertising
expenditures. During the past 6 months, all
advertisements for the store have appeared in the
local newspaper. The number of ads per week has
varied from one to seven. The store’s sales staff has
been tracking the number of customers who enter
the store each week. The number of ads and the
number of customers per week for the past 26
weeks were recorded.
customers?
SSxx=
Ads x x-‐xbar
(x-‐xbar)2

5.00 0.88
0.7744

6.00 1.88
3.5344

3.00 -‐1.12
1.2544

2.00 -‐2.12
4.4944

4.00 -‐0.12
0.0144

2.00 -‐2.12
4.4944

4.00 -‐0.12
0.0144

2.00 -‐2.12
4.4944

4.00 -‐0.12
0.0144

2.00 -‐2.12
4.4944

7.00 2.88
8.2944

5.00 0.88
0.7744

4.00 -‐0.12
0.0144

7.00 2.88
8.2944

2.00 -‐2.12
4.4944

5.00 0.88
0.7744

5.00 0.88
0.7744

2.00 -‐2.12
4.4944

1.00 -‐3.12
9.7344

6.00 1.88
3.5344

6.00 1.88
3.5344

5.00 0.88
0.7744

6.00 1.88
3.5344

3.00 -‐1.12
1.2544

2.00 -‐2.12
4.4944

7.00 2.88
8.2944

4.12
86.6544

3
1 2
( )
b
i
s
s
x x
=
Σ −
=132.96
1 2
( )
b
i
s
s
x x
=
Σ −86.6544
132.96
= 14.28
4
85

PROBLEM # 10.5
Colonial Furniture has been reviewing weekly
advertising expenditures. During the past 6
months, all advertisements for the store have
appeared in the local newspaper. The number
of ads per week has varied from one to seven.
The store’s sales staff has been tracking the
number of customers who enter the store each
week. The number of ads and the number of
customers per week for the past 26 weeks
were recorded.
customers?
5
1
1
b
b
t
s
=
1 2
( )
b
i
s
s
x x
=
Σ −
= 14.28
ŷ = 296.92 + 21.356x
= + 21.356
= 14.28
1.4955
86

PROBLEM # 10.5
At 5% level significance. The manager of Colonial Furniture has been reviewing weekly
advertising expenditures. During the past 6 months, all advertisements for the store have
appeared in the local newspaper. The number of ads per week has varied from one to seven.
The store’s sales staff has been tracking the number of customers who enter the store each
week. The number of ads and the number of customers per week for the past 26 weeks were
recorded.
c.  Can the manager infer that the larger the number of ads, the larger the number of
customers?
t =
b1 – β1
sb1
=
21.356 - 0
14.28
= 1.496
87

PROBLEM # 10.5
4. Find the critical value of the test statistics
tα, n-k-1= t0.05,24 = 1.711
Reject Ho, if tobserved > tcritical, otherwise do not reject.
Since tobserved= 1.4955 < tcritical , then we do not reject Ho
7. Interpret in the context
There is not enough evidence to conclude that the larger the number
of ads the larger the number of customers.
88

6. COVARIANCE &
COEFFICIENT OF CORRELATION
Covariance
Interpretation of the covariance
Correlation coefficient
89

MEASURES OF ASSOCIATION
BETWEEN TWO VARIABLES
Thus far we have examined numerical methods used
to summarize the data for one variable at a time.
Often a manager or decision maker is interested in
the relationship between two variables.
Two descriptive measures of the relationship
between two variables are covariance and correlation
coefficient.
90

COVARIANCE
Positive values indicate a positive relationship.
Negative values indicate a negative relationship.
The covariance is a measure of the linear association
between two variables.
92
PAGE 560

The covariance is computed as follows:
for
samples
for
populations
s
x x y y
n
xy
i i=
− −∑
−
( )( )
1
σ
µ µ
xy
i x i yx y
N
=
− −∑( )( )
COVARIANCE
93

Just because two variables are highly correlated, it
does not mean that one variable is the cause of the
other.
Correlation is a measure of linear association and not
necessarily causation.
94

The correlation coefficient is computed as follows:
for
samples
for
populations
r
s
s s
xy
xy
x y
= ρ
σ
σ σxy
xy
x y
=
Pearson Product Moment
Correlation Coefficient.
95

∑
∑
∑ ∑
∑ ∑
∑ ∑ ∑ ∑
r - Sample correlation coefficient
n - Sample size
x - Value of the independent variable
y - Value of the dependent variable
PAGE 561

Values near +1 indicate a strong positive linear
relationship.
Values near -1 indicate a strong negative linear
relationship.
The coefficient can take on values between -1 and +1.
The closer the correlation is to zero, the weaker the
relationship.
97

Covariance and Correlation Coefficient
277.6
259.5
269.1
267.0
255.6
272.9
69
71
70
70
71
69
x y
10.65
-7.45
2.15
0.05
-11.35
5.95
-1.0
1.0
0
0
1.0
-1.0
-10.65
-7.45
0
0
-11.35
-5.95
( )ix x− ( )( )i ix x y y− −( )iy y−
Average
Std. Dev.
267.0 70.0 -35.40
8.2192 .8944
Total
98

Ads x Customer y x-‐xmean
(x-‐xmean)2
y-‐ymean
(y-‐ymean)2
(x-‐xmean)(y-‐ymean)

5.00 353.00 0.88
0.77
-‐31.81
1011.88
-‐27.993

6.00 319.00 1.88
3.53
-‐65.81
4330.96
-‐123.723

3.00 440.00 -‐1.12
1.25
55.19
3045.94
-‐61.813

2.00 332.00 -‐2.12
4.49
-‐52.81
2788.90
111.957

4.00 172.00 -‐0.12
0.01
-‐212.81
45288.10
25.537

2.00 331.00 -‐2.12
4.49
-‐53.81
2895.52
114.077

4.00 344.00 -‐0.12
0.01
-‐40.81
1665.46
4.897

2.00 483.00 -‐2.12
4.49
98.19
9641.28
-‐208.163

4.00 329.00 -‐0.12
0.01
-‐55.81
3114.76
6.697

2.00 532.00 -‐2.12
4.49
147.19
21664.90
-‐312.043

7.00 496.00 2.88
8.29
111.19
12363.22
320.227

5.00 393.00 0.88
0.77
8.19
67.08
7.207

4.00 376.00 -‐0.12
0.01
-‐8.81
77.62
1.057

7.00 372.00 2.88
8.29
-‐12.81
164.10
-‐36.893

2.00 512.00 -‐2.12
4.49
127.19
16177.30
-‐269.643

5.00 254.00 0.88
0.77
-‐130.81
17111.26
-‐115.113

5.00 459.00 0.88
0.77
74.19
5504.16
65.287

2.00 153.00 -‐2.12
4.49
-‐231.81
53735.88
491.437

1.00 426.00 -‐3.12
9.73
41.19
1696.62
-‐128.513

6.00 566.00 1.88
3.53
181.19
32829.82
340.637

6.00 596.00 1.88
3.53
211.19
44601.22
397.037

5.00 395.00 0.88
0.77
10.19
103.84
8.967

6.00 676.00 1.88
3.53
291.19
84791.62
547.437

3.00 194.00 -‐1.12
1.25
-‐190.81
36408.46
213.707

2.00 135.00 -‐2.12
4.49
-‐249.81
62405.04
529.597

7.00 367.00 2.88
8.29
-‐17.81
317.20
-‐51.293

4.12
384.81
86.65
463802.04
1850.577

1.86
136.21
74.023

3.466176
18552.081544
99

Ads x Customer y
x-‐xmean
(x-‐xmean)2
y-‐ymean
(y-‐ymean)2

(x-‐xmean)

(y-‐ymean)

5.00 353.00 0.88
0.77
-‐31.81
1011.88
-‐27.993

… …. …
…
…
…
…

6.00 566.00 1.88
3.53
181.19
32829.82
340.637

6.00 596.00 1.88
3.53
211.19
44601.22
397.037

5.00 395.00 0.88
0.77
10.19
103.84
8.967

6.00 676.00 1.88
3.53
291.19
84791.62
547.437

3.00 194.00 -‐1.12
1.25
-‐190.81
36408.46
213.707

2.00 135.00 -‐2.12
4.49
-‐249.81
62405.04
529.597

7.00 367.00 2.88
8.29
-‐17.81
317.20
-‐51.293

4.12
384.81
86.65
463802.04
1850.577

1.86
136.21
74.023

100

Ads x Customer y x-‐xmean
(x-‐xmean)2
y-‐ymean
(y-‐ymean)2
(x-‐xmean)(y-‐ymean)

5.00 353.00 0.88
0.77
-‐31.81
1011.88
-‐27.993

6.00 319.00 1.88
3.53
-‐65.81
4330.96
-‐123.723

3.00 440.00 -‐1.12
1.25
55.19
3045.94
-‐61.813

2.00 332.00 -‐2.12
4.49
-‐52.81
2788.90
111.957

4.00 172.00 -‐0.12
0.01
-‐212.81
45288.10
25.537

2.00 331.00 -‐2.12
4.49
-‐53.81
2895.52
114.077

4.00 344.00 -‐0.12
0.01
-‐40.81
1665.46
4.897

2.00 483.00 -‐2.12
4.49
98.19
9641.28
-‐208.163

4.00 329.00 -‐0.12
0.01
-‐55.81
3114.76
6.697

2.00 532.00 -‐2.12
4.49
147.19
21664.90
-‐312.043

7.00 496.00 2.88
8.29
111.19
12363.22
320.227

5.00 393.00 0.88
0.77
8.19
67.08
7.207

4.00 376.00 -‐0.12
0.01
-‐8.81
77.62
1.057

7.00 372.00 2.88
8.29
-‐12.81
164.10
-‐36.893

2.00 512.00 -‐2.12
4.49
127.19
16177.30
-‐269.643

5.00 254.00 0.88
0.77
-‐130.81
17111.26
-‐115.113

5.00 459.00 0.88
0.77
74.19
5504.16
65.287

2.00 153.00 -‐2.12
4.49
-‐231.81
53735.88
491.437

1.00 426.00 -‐3.12
9.73
41.19
1696.62
-‐128.513

6.00 566.00 1.88
3.53
181.19
32829.82
340.637

6.00 596.00 1.88
3.53
211.19
44601.22
397.037

5.00 395.00 0.88
0.77
10.19
103.84
8.967

6.00 676.00 1.88
3.53
291.19
84791.62
547.437

3.00 194.00 -‐1.12
1.25
-‐190.81
36408.46
213.707

2.00 135.00 -‐2.12
4.49
-‐249.81
62405.04
529.597

7.00 367.00 2.88
8.29
-‐17.81
317.20
-‐51.293

4.12
384.81
86.65
463802.04
1850.577

1.86
136.21
74.023

s
x x y y
n
xy
i i=
− −∑
−
( )( )
1
r
s
s s
xy
xy
x y
=
101

PROBLEM # 10.5
recorded.
e. In your opinion, is it worthwhile exercise to use the regression
equation to predict the number of customers who will enter the store,
given that Colonial intends to advertise five times in the newspaper? If
so, find the 95% prediction interval. If not, explain why not.
102

Ads x Customer y
x-‐xmean
(x-‐xmean)2
y-‐ymean
(y-‐ymean)2

(x-‐xmean)

(y-‐ymean)

5.00 353.00 0.88
0.77
-‐31.81
1011.88
-‐27.993

… … …
…
…
…
…

6.00 319.00 1.88
3.53
-‐65.81
4330.96
-‐123.723

2.00 153.00 -‐2.12
4.49
-‐231.81
53735.88
491.437

1.00 426.00 -‐3.12
9.73
41.19
1696.62
-‐128.513

6.00 566.00 1.88
3.53
181.19
32829.82
340.637

6.00 596.00 1.88
3.53
211.19
44601.22
397.037

5.00 395.00 0.88
0.77
10.19
103.84
8.967

6.00 676.00 1.88
3.53
291.19
84791.62
547.437

3.00 194.00 -‐1.12
1.25
-‐190.81
36408.46
213.707

2.00 135.00 -‐2.12
4.49
-‐249.81
62405.04
529.597

7.00 367.00 2.88
8.29
-‐17.81
317.20
-‐51.293

4.12
384.81
86.65
463802.04
1850.577

1.86
136.21
74.023

s
x x y y
n
xy
i i=
− −∑
−
( )( )
1
r
s
s s
xy
xy
x y
=
103

PROBLEM # 10.5
recorded.
e. In your opinion, is it worthwhile exercise to use the regression equation to predict the
number of customers who will enter the store, given that Colonial intends to advertise five
times in the newspaper? If so, find the 95% prediction interval. If not, explain why not.
2
y
2
x
2
xy2
ss
s
R = ! 0851.
)552,18)(47.3(
)02.74( 2
= !
There is a weak linear relationship between the number of ads and the
number of customers.
=
104

PROBLEM # 10.5
recorded.
e. In your opinion, is it worthwhile exercise to use the regression equation to predict the
number of customers who will enter the store, given that Colonial intends to advertise five
times in the newspaper? If so, find the 95% prediction interval. If not, explain why not.
The linear relationship is too weak for the model to produce predictions.
105

7. ESTIMATION
POINT ESTIMATION
INTERVAL ESTIMATION
CONFIDENCE INTERVAL FOR THE MEAN VALUE OF Y
PREDICTION INTERVAL FOR AN INDIVIDUAL VALUE OF Y
106

If 3 TV ads are run prior to a sale, we expect
the mean number of cars sold to be:
^
y = 10 + 5(3) = 25 cars
POINT ESTIMATION
107

Using the Estimated Regression
Equation for Estimation and Prediction
 / y t sp yp
± α 2
where:
confidence coefficient is 1 - α and

tα/2 is based on a t distribution
with n - 2 degrees of freedom
/2 indpy t sα±
  Confidence Interval Estimate of E(yp)
  Prediction Interval Estimate of yp
108

2
ˆ 2
( )1
( )p
p
y
i
x x
s s
n x x
−
= +
−∑
  Estimate of the Standard Deviation of ˆpy
Confidence Interval for E(yp)
 / y t sp yp
± α 2
109

CONFIDENCE VS PREDICTION
INTERVAL

PROBLEM #10.6
For n=6 data points, the following quantities have been calculated:
a.  Determine the least-squares regression line.
!!"!#$!!!!!!!!! !!"!%&!!!!!!!! !"!"!#$$!!
!!
"!'#&!!!!! !!
"!((&$!!!!!!!! !! ! !!!
")*+''#!
!
111

PROBLEM #10.6
a.  Determine the least-squares regression line.
To determine the least squares regression line, we must calculate the slope
and y‑intercept.
i i
1 2 22
i
x y nxy 400 6(6.67)(12.67)
b 1.354
346 6(6.67)x nx
! !
= = = !
!!
"
"
!
0 1b y b x 12.67 ( 1.354)(6.67) 21.701= ! = ! ! = !
400 – [(40)(76)/6]
346 – [(40)2/6]
b1 -1.345
!!"!#$!!!!!!!!! !!"!%&!!!!!!!! !"!"!#$$!!
!!
"!'#&!!!!! !!
"!((&$!!!!!!!! !! ! !!!
")*+''#!
!
12.67 – (-1.345)(6.67) = 21.641
The regression equation is ŷ = 21.641 – 1.345 x
112

PROBLEM #10.6
b.  Determine the standard error of estimate.
!!"!#$!!!!!!!!! !!"!%&!!!!!!!! !"!"!#$$!!
!!
"!'#&!!!!! !!
"!((&$!!!!!!!! !! ! !!!
")*+''#!
!
2
i i
y.x
ˆ(y y ) 52.334
s
n 2 6 2
!
= = =
! !
"
113

PROBLEM #10.6
b.  Determine the standard error of estimate.
The standard error of the estimate is
2
i i
y.x
ˆ(y y ) 52.334
s 3.617
n 2 6 2
!
= = =
! !
" !
!!"!#$!!!!!!!!! !!"!%&!!!!!!!! !"!"!#$$!!
!!
"!'#&!!!!! !!
"!((&$!!!!!!!! !! ! !!!
")*+''#!
!
2
i i
y.x
ˆ(y y ) 52.334
s
n 2 6 2
!
= = =
! !
"
114

PROBLEM #10.6
c.  Construct the 95% confidence interval for the mean of y when x=7.0
!!"!#$!!!!!!!!! !!"!%&!!!!!!!! !"!"!#$$!!
!!
"!'#&!!!!! !!
"!((&$!!!!!!!! !! ! !!!
")*+''#!
!
The regression equation is ŷ = 21.641 – 1.345 (7) = 12.226
115

2 2
y.x 2 2
i2
i
(x x) (7 6.67)1 1
ˆy ts 12.223 2.776(3.617)
n 6( x ) 40
(346 )( x )
6n
12.223 4.116 (8.107, 16.339)
! !
± + = ± +
!!
= ± =
"
" !
12.226
2 2
y.x 2 2
i2
i
(x x) (7 6.67)1 1
ˆy ts 12.223 2.776(3.617)
n 6( x ) 40
(346 )( x )
6n
12.223 4.116 (8.107, 16.339)
! !
± + = ± +
!!
= ± =
"
" !
12.226
2 2
y.x 2 2
i2
i
(x x) (7 6.67)1 1
ˆy ts 12.223 2.776(3.617)
n 6( x ) 40
(346 )( x )
6n
12.223 4.116 (8.107, 16.339)
! !
± + = ± +
!!
= ± =
"
" !79.333
0.1089
10.041 0.4099± x
12.226 + 10.041 x 0.4099 = 16.342
12.226 - 10.041 x 0.4099 = 8.11
12.226 10.041
116

PROBLEM #10.6
c.  Construct the 95% confidence interval for the mean of y when x=7.0
We also need the t‑value with 6 ‑ 2 = 4 degrees of freedom for a 95% interval;
this value is 2.776.
Therefore the 95% confidence interval for the mean value of y when x = 7 is:
2 2
y.x 2 2
i2
i
(x x) (7 6.67)1 1
ˆy ts 12.223 2.776(3.617)
n 6( x ) 40
(346 )( x )
6n
12.223 4.116 (8.107, 16.339)
! !
± + = ± +
!!
= ± =
"
" !
!!"!#$!!!!!!!!! !!"!%&!!!!!!!! !"!"!#$$!!
!!
"!'#&!!!!! !!
"!((&$!!!!!!!! !! ! !!!
")*+''#!
!
2 2
y.x 2 2
i2
i
(x x) (7 6.67)1 1
ˆy ts 12.223 2.776(3.617)
n 6( x ) 40
(346 )( x )
6n
12.223 4.116 (8.107, 16.339)
! !
± + = ± +
!!
= ± =
"
" !
12.226
The confidence interval ranges from 8.110 to 16.342Created by Samie L.S. Ly, must not be used for profitable
117

PROBLEM #10.6
d. Construct the 95% confidence interval for the mean of y when x=9.0
!!"!#$!!!!!!!!! !!"!%&!!!!!!!! !"!"!#$$!!
!!
"!'#&!!!!! !!
"!((&$!!!!!!!! !! ! !!!
")*+''#!
!
118

2 2
y.x 2 2
i2
i
(x x) (7 6.67)1 1
ˆy ts 12.223 2.776(3.617)
n 6( x ) 40
(346 )( x )
6n
12.223 4.116 (8.107, 16.339)
! !
± + = ± +
!!
= ± =
"
" !
9.536
2 2
y.x 2 2
i2
i
(x x) (7 6.67)1 1
ˆy ts 12.223 2.776(3.617)
n 6( x ) 40
(346 )( x )
6n
12.223 4.116 (8.107, 16.339)
! !
± + = ± +
!!
= ± =
"
" !79.333
5.4289
10.041 0.4849± x
9.536 + 10.041 x 0.4849 = 14.4048
9.536 - 10.041 x 0.4849 = 4.668
9.536 10.041
2 2
y.x 2 2
i2
i
(x x) (9 6.67)1 1
ˆy ts 9.515 2.776(3.617)
n 6( x ) 40
(346 )( x )
6n
9.515 4.868 (4.647, 14.383)
! !
± + = ± +
!!
= ± =
"
" !
9.536
119

PROBLEM #10.6
d. Construct the 95% confidence interval for the mean of y when x=9.0
We also need the t‑value with 6 ‑ 2 = 4 degrees of freedom for a 95% interval;
this value is 2.776.
Therefore the 95% confidence interval for the mean value of y when x = 9 is:
2 2
y.x 2 2
i2
i
(x x) (9 6.67)1 1
ˆy ts 9.515 2.776(3.617)
n 6( x ) 40
(346 )( x )
6n
9.515 4.868 (4.647, 14.383)
! !
± + = ± +
!!
= ± =
"
" !
!!"!#$!!!!!!!!! !!"!%&!!!!!!!! !"!"!#$$!!
!!
"!'#&!!!!! !!
"!((&$!!!!!!!! !! ! !!!
")*+''#!
!
2 2
y.x 2 2
i2
i
(x x) (9 6.67)1 1
ˆy ts 9.515 2.776(3.617)
n 6( x ) 40
(346 )( x )
6n
9.515 4.868 (4.647, 14.383)
! !
± + = ± +
!!
= ± =
"
" !
9.536
The confidence interval ranges from 4.668 to 14.404Created by Samie L.S. Ly, must not be used for profitable
120

PROBLEM #10.6
e. Compare the width of the confidence interval obtained in part (c ) with
the obtained in part (d). Which is wider and why?
!!"!#$!!!!!!!!! !!"!%&!!!!!!!! !"!"!#$$!!
!!
"!'#&!!!!! !!
"!((&$!!!!!!!! !! ! !!!
")*+''#!
!
121

PROBLEM #10.6
e. Compare the width of the confidence interval obtained in part (c ) with
the obtained in part (d). Which is wider and why?
The confidence interval in d is wider because 9 is farther
from the mean of x than 7.
!!"!#$!!!!!!!!! !!"!%&!!!!!!!! !"!"!#$$!!
!!
"!'#&!!!!! !!
"!((&$!!!!!!!! !! ! !!!
")*+''#!
!
For x = 7,
8.110 to 16.342
For x = 9,
4.668 to 14.404
122

2
ind 2
( )1
1
( )
p
i
x x
s s
n x x
−
= + +
−∑
  Estimate of the Standard Deviation
of an Individual Value of yp
Prediction Interval for yp
/2 indpy t sα±
123

PROBLEM #10.7
For the summary data provided in Problem #10.18, construct a
95% prediction interval for an individual y value whenever
a. x=2
124

PROBLEM #10.7
a. x=2
We also need the t‑value with 6 ‑ 2 = 4 degrees of freedom for a 95%
interval; this value is 2.776. Therefore, the 95% prediction interval for
an individual y value when x = 2 is:
125

2 2
y.x 2 2
i2
i
(x x) (2 6.67)1 1
ˆy ts 1 18.993 2.776(3.617) 1
n 6( x ) 40
(346 )( x )
6n
18.993 12.056 (6.937, 31.049)
! !
± + + = ± + +
!!
= ± =
"
" !
18.951
79.333
21.8089
10.041 1.2006± x
18.951 + 10.041 x 1.2006 = 31.006
18.951 - 10.041 x 1.2006 = 6.896
18.951 10.041
2 2
y.x 2 2
i2
i
(x x) (2 6.67)1 1
ˆy ts 1 18.993 2.776(3.617) 1
n 6( x ) 40
(346 )( x )
6n
18.993 12.056 (6.937, 31.049)
! !
± + + = ± + +
!!
= ± =
"
" !
18.951
2 2
y.x 2 2
i2
i
(x x) (2 6.67)1 1
ˆy ts 1 18.993 2.776(3.617) 1
n 6( x ) 40
(346 )( x )
6n
18.993 12.056 (6.937, 31.049)
! !
± + + = ± + +
!!
= ± =
"
" !
126

PROBLEM #10.7
a. x=2
We also need the t‑value with 6 ‑ 2 = 4 degrees of freedom for a 95%
interval; this value is 2.776. Therefore, the 95% prediction interval for
an individual y value when x = 2 is:
2 2
y.x 2 2
i2
i
(x x) (2 6.67)1 1
ˆy ts 1 18.993 2.776(3.617) 1
n 6( x ) 40
(346 )( x )
6n
18.993 12.056 (6.937, 31.049)
! !
± + + = ± + +
!!
= ± =
"
" !
18.951
2 2
y.x 2 2
i2
i
(x x) (2 6.67)1 1
ˆy ts 1 18.993 2.776(3.617) 1
n 6( x ) 40
(346 )( x )
6n
18.993 12.056 (6.937, 31.049)
! !
± + + = ± + +
!!
= ± =
"
" !
The prediction interval ranges from 6.895 to 31.007
127

Lesson 10 - Regression Analysis

More Related Content

What's hot (18)

Similar to Lesson 10 - Regression Analysis (20)

Recently uploaded (20)

Lesson 10 - Regression Analysis