Applied Statistics and Probability for

Engineers

Sixth Edition
Douglas C. Montgomery George C. Runger

Chapter 13
Design and Analysis of Single-Factor
Experiments: The Analysis of Variance
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
 13
Design and Analysis of Single-
Factor Experiments: The Analysis
of Variance
CHAPTER OUTLINE
13-1 Designing Engineering 13-3.2 ANOVA & Variance
Experiments Components
13-2 Completely Randomized 13-4 Randomized Complete Block
Single-Factor Experiment Design
13-2.1 Example: Tensile Strength 13-4.1 Design & Statistical Analysis
13-2.2 Analysis of Variance 13-4.2 Multiple Comparisons
13-2.3 Multiple Comparisons 13-4.3 Residual Analysis & Model
Following the ANOVA Checking
13-2.4 Residual Analysis & Model
Checking
13-3 The Random-Effects Model
13-3.1 Fixed Versus Random
Factors
Chapter 13 Title and Outline 2
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
 Learning Objectives for Chapter 13
After careful study of this chapter, you should be able to do the following:
1. Design and conduct engineering experiments involving a single factor with
an arbitrary number of levels.
2. Understand how the analysis of variance is used to analyze the data from
these experiments.
3. Assess model adequacy with residual plots.
4. Use multiple comparison procedures to identify specific differences
between means.
5. Make decisions about sample size in single-factor experiments.
6. Understand the difference between fixed and random factors.
7. Estimate variance components in an experiment involving random factors.
8. Understand the blocking principle and how it is used to isolate the effect of
nuisance factors.
9. Design and conduct experiments involving the randomized complete block
design.

Chapter 13 Learning Objectives 3

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
13-1: Designing Engineering Experiments
Every experiment involves a sequence of activities:
1. Conjecture – the original hypothesis that motivates
the experiment.
2. Experiment – the test performed to investigate the
conjecture.
3. Analysis – the statistical analysis of the data from
the experiment.
4. Conclusion – what has been learned about the
original conjecture from the experiment. Often the
experiment will lead to a revised conjecture, and a
new experiment, and so forth.

Sec 13-1 Designing Engineering Experiments 4

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
13-2: The Completely Randomized Single-Factor Experiment
13-2.1 An Example
A manufacturer of paper used for making grocery bags is
interested in improving the tensile strength of the product. Product
engineering thinks that tensile strength is a function of the
hardwood concentration in the pulp and that the range of
hardwood concentrations of practical interest is between 5 and
20%. A team of engineers responsible for the study decides to
investigate four levels of hardwood concentration: 5%, 10%, 15%,
and 20%. They decide to make up six test specimens at each
concentration level, using a pilot plant. All 24 specimens arc
tested on a laboratory tensile tester, in random order. The data
from this experiment are shown in Table 13-1.

Sec 13-2 Completely Randomized Single-Factor Experiment 5

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
13-2: The Completely Randomized Single-Factor Experiment
13-2.1 An Example

Table 13-1 Tensile Strength of Paper (psi)

Hardwood Observations
Concentration 1 2 3 4 5 6 Totals Averages
()
5 7 8 15 11 9 10 60 10.00
10 12 17 13 18 19 15 94 15.67
15 14 18 19 17 16 18 102 17.00
20 19 25 22 23 18 20 127 21.17
383 15.96

Sec 13-2 Completely Randomized Single-Factor Experiment 6

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
13-2: The Completely Randomized Single-Factor Experiment
13-2.1 An Example

• The levels of the factor are sometimes called

treatments.
• Each treatment has six observations or
replicates.
• The runs are run in random order.

Sec 13-2 Completely Randomized Single-Factor Experiment 7

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
13-2: The Completely Randomized Single-Factor Experiment
13-2.1 An Example

Figure 13-1 (a) Box plots of hardwood concentration data. (b) Display of the model in
Equation 13-1 for the completely randomized single-factor experiment.
Sec 13-2 Completely Randomized Single-Factor Experiment 8
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
13-2 The Completely Randomized Single-Factor
Experiment
13-2.2 The Analysis of Variance

Suppose there are a different levels of a single factor

that we wish to compare. The levels are sometimes
called treatments.

9
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
13-2 The Completely Randomized Single-Factor
Experiment
13-2.2 The Analysis of Variance
We may describe the observations in Table 13-2 by the
linear statistical model:

The model could be written as

10
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
13-2 The Completely Randomized Single-Factor
Experiment
13-2.2 The Analysis of Variance
Fixed-effects Model
The treatment effects are usually defined as deviations
from the overall mean so that:

Also,

11
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
13-2 The Completely Randomized Single-Factor
Experiment
13-2.2 The Analysis of Variance

We wish to test the hypotheses:

12
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
13-2: The Completely Randomized Single-Factor Experiment

13-2.2 The Analysis of Variance

Definition
The sum of squares identity is

   
a n a a n
  ij
y  y .. 2
 n 
 iy .  y ..2
   ij i
y  y . 2 (13-1)
i 1 j 1 i 1 i 1 j 1

or symbolically

SST = SSTreatments  SSE (13-2)

Sec 13-2 Completely Randomized Single-Factor Experiment 13

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
13-2: The Completely Randomized Single-Factor Experiment

13-2.2 The Analysis of Variance

The expected value of the treatment sum of squares

is
a
E ( SS Treatments )  (a  1) 2  n  i2
i 1

and the expected value of the error sum of squares is

E(SSE) = a(n – 1)2

The ratio MSTreatments = SSTreatments/(a – 1) is called the mean

square for treatments.
Sec 13-2 Completely Randomized Single-Factor Experiment 14
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
13-2: The Completely Randomized Single-Factor Experiment

13-2.2 The Analysis of Variance

The appropriate test statistic is

SS Treatments /( a  1) MS Treatments
F0   (13-3)
SS E /[ a (n  1)] MS E

We would reject H0 if f0 > f,a-1,a(n-1)

Sec 13-2 Completely Randomized Single-Factor Experiment 15

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
13-2: The Completely Randomized Single-Factor Experiment

13-2.2 The Analysis of Variance

Definition
The sums of squares computing formulas for the ANOVA with
equal sample sizes in each treatment are
a n 2
y ..
SST    yij 
2 (13-4)
i 1 j 1 N
and
a
yi2 . y 2..
SS Treatments    (13-5)
i 1 n N
The error sum of squares is obtained by subtraction as

SSE = SST – SSTreatments (13-6)

Sec 13-2 Completely Randomized Single-Factor Experiment 16
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
13-2: The Completely Randomized Single-Factor Experiment

13-2.2 The Analysis of Variance

Analysis of Variance Table

Table 13-3 The Analysis of Variance for a Single-Factor Experiment, Fixed-Effects

Model

Source of Degrees of
Variation Sum of Squares Freedom Mean Square F0
Treatments SSTreatments a1 MSTreatments MS Treatments
MS E
Error SSE a(n  1) MSE
Total SST an  1

Sec 13-2 Completely Randomized Single-Factor Experiment 17

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
13-2: The Completely Randomized Single-Factor Experiment

EXAMPLE 13-1 Tensile Strength ANOVA Consider the

paper tensile strength experiment described in Section
13-2.1. This experiment is a CRD. We can use the
analysis of variance to test the hypothesis that different
hardwood concentrations do not affect the mean tensile
strength of the paper.

The hypotheses are

H0:  1 =  2 =  3 =  4 = 0
H1: i  0 for at least one i

Sec 13-2 Completely Randomized Single-Factor Experiment 18

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
13-2: The Completely Randomized Single-Factor Experiment
Example 13-1
We will use  = 0.01. The sums of squares for the analysis of variance are
computed from Equations 13-4, 13-5, and 13-6 as follows:
4 6
y 2..
SST    yij2 
i 1 j 1 N

(383) 2
 (7)  (8)    (20) 
2 2
 512.96
2
24
4
yi2 . y 2..
SS Treatments   
i 1 n N
(60) 2  (94) 2  (102) 2  (127) 2 (383) 2
 
6 24
 382.79
SS E  SST  SS Treatments
 512.96  382.79  130.17
Sec 13-2 Completely Randomized Single-Factor Experiment 19
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
13-2: The Completely Randomized Single-Factor Experiment
Example 13-1
The ANOVA is summarized in Table 13-4. Since f0.01,3,20 = 4.94, we reject H0
and conclude that hardwood concentration in the pulp significantly affects the
mean strength of the paper. We can also find a P-value for this test statistic as
follows:

P  P( F3,20  19.60) ~ 3.59  10  6

Since P ~ 3.59  10  6 is considerably smaller than  = 0.01, we have strong

evidence to conclude that H0 is not true.

Table 13-4 ANOVA for the Tensile Strength Data

Source of Sum of Degrees of Mean P-value
Variation Squares Freedom Square f0
Hardwood
concentration 382.79 3 127.60 19.60 3.59 E-6
Error 130.17 20 6.51
Total 512.96 23

Sec 13-2 Completely Randomized Single-Factor Experiment 20

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
13-2: The Completely Randomized Single-Factor Experiment

Confidence Interval on a Treatment Mean

A 100(1  ) percent confidence interval on the mean of the ith

treatment t is

MS E MS E
yi .  t/2,an 1   i  yi .  t/2,a n 1 (13-7)
n n

For 20% hardwood, the resulting confidence interval on the mean

is

19.00 psi  m4  23.34 psi

Sec 13-2 Completely Randomized Single-Factor Experiment 21

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
13-2: The Completely Randomized Single-Factor Experiment

Confidence Interval on a Difference in Treatment Means

A 100(1  ) percent confidence interval on the difference in two

treatment means μi  μj is

2MS E 2MS E
yi .  y j . t/2,an 1  i   j  yi .  y j .  t/2,an 1 (13-8)
n n

For the hardwood concentration example,

1.74  m3  m2  4.40

Sec 13-2 Completely Randomized Single-Factor Experiment 22

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
13-2: The Completely Randomized Single-Factor Experiment

An Unbalanced Experiment
The sums of squares computing formulas for the ANOVA with
unequal sample sizes ni in each treatment are

a ni 2
y ..
SST    yij 
2 (13-9)
i 1 j 1 N
a
yi2 . y 2..
SS Treatments    (13-10)
i 1 ni N
and

SSE = SST  SSTreatments (13-11)

Sec 13-2 Completely Randomized Single-Factor Experiment 23

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
13-2: The Completely Randomized Single-Factor Experiment

13-2.3 Multiple Comparisons Following the ANOVA

The least significant difference (LSD) (Fisher’s Test)

is
2MS E
LSD  t /2,a n 1 (13-12)
n

If the sample sizes are different in each treatment:

1 1 
LSD  t /2,N  a MS E   
 ni n j 
 

Sec 13-2 Completely Randomized Single-Factor Experiment 24

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
13-2: The Completely Randomized Single-Factor Experiment

Example 13-2
We will apply the Fisher LSD method to the hardwood
concentration experiment. There are a = 4 means, n = 6,
MSE = 6.51, and t0.025,20 = 2.086. The treatment means are

y1.  10.00 psi

y2 .  15.67 psi
y3.  17.00 psi
y4 .  21.17 psi

The value of LSD is LSD  t0.025,20 2MSE /n  2.086 26.51/6  3.07 .

Therefore, any pair of treatment averages that differs by more
than 3.07 implies that the corresponding pair of treatment
means are different.

Sec 13-2 Completely Randomized Single-Factor Experiment 25

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
13-2: The Completely Randomized Single-Factor Experiment

Example 13-2
The comparisons among the observed treatment averages are as follows:

4 vs. 1 = 21.17  10.00 = 11.17 > 3.07

4 vs. 2 = 21.17  15.67 = 5.50 > 3.07
4 vs. 3 = 21.17  17.00 = 4.17 > 3.07
3 vs. 1 = 17.00  10.00 = 7.00 > 3.07
3 vs. 2 = 17.00  15.67 = 1.33 < 3.07
2 vs. 1 = 15.67  10.00 = 5.67 > 3.07

Conclusions: From this analysis, we see that there are significant differences
between all pairs of means except 2 and 3. This implies that 10% and 15%
hardwood concentration produce approximately the same tensile strength and
that all other concentration levels tested produce different tensile strengths.

It is often helpful to draw a graph of the treatment means, such as in Fig. 13-2,
with the means that are not different underlined. This graph clearly reveals the
results of the experiment and shows that 20% hardwood produces the maximum
tensile strength.
Sec 13-2 Completely Randomized Single-Factor Experiment 26
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
13-2: The Completely Randomized Single-Factor Experiment

13-2.4 Residual Analysis and Model Checking

Table 13-6 Residuals for the Tensile Strength Experiment

Hardwood
Concentration (%) Residuals
5 3.00 -2.00 5.00 1.00 -1.00 0.00
10 -3.67 1.33 -2.67 2.33 3.33 -0.67
15 -3.00 1.00 2.00 0.00 -1.00 1.00
20 -2.17 3.83 0.83 1.83 -3.17 -1.17

Sec 13-2 Completely Randomized Single-Factor Experiment 27

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
13-2: The Completely Randomized Single-Factor Experiment

13-2.4 Residual Analysis and Model Checking

Figure 13-4 Normal probability plot

of residuals from the hardwood
concentration experiment.

Sec 13-2 Completely Randomized Single-Factor Experiment 28

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
13-2: The Completely Randomized Single-Factor Experiment

13-2.4 Residual Analysis and Model Checking

Figure 13-5 Plot of residuals

versus factor levels (hardwood
concentration).

Sec 13-2 Completely Randomized Single-Factor Experiment 29

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
13-2: The Completely Randomized Single-Factor Experiment

13-2.4 Residual Analysis and Model Checking

Figure 13-6 Plot of residuals

versus yi .

Sec 13-2 Completely Randomized Single-Factor Experiment 30

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
 13-3: The Random-Effects Model
13-3.2 ANOVA and Variance Components

The linear statistical model is

i  1, 2, , a
Yij    i  ij 
 j  1, 2, , n

The variance of the response is V Yij   2  2

Where each term on the right hand side is called
a variance component.

Sec 13-3 The Random-Effects Model 31

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
 13-3: The Random-Effects Model
13-3.2 ANOVA and Variance Components
For a random-effects model, the appropriate
hypotheses to test are:

H 0 :  2  0
H1 :  2  0
The ANOVA decomposition of total variability is
still valid:

SST = SSTreatments + SSE

Sec 13-3 The Random-Effects Model 32

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
 13-3: The Random-Effects Model
13-3.2 ANOVA and Variance Components
The expected values of the mean squares are
In the random-effects model for a single-factor, completely
randomized experiment, the expected mean square for treatments
is
 SSTreatments 
E ( MSTreatments )  E  
 a 1  (13-13)

 2  n2

and the expected mean square for error is

 SS E 
E ( MS E )  E  
 a n  1 
(13-14)

 2
Sec 13-3 The Random-Effects Model 33
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
 13-3: The Random-Effects Model
13-3.2 ANOVA and Variance Components

The estimators of the variance components

are

  MS E
ˆ 2 (13-15)

and
MSTreatments  MS E
ˆ 
  (13-16)
n

Sec 13-3 The Random-Effects Model 34

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
 13-3: The Random-Effects Model
Example 13-4 Textile Manufacturing In Design and Analysis of Experiments, 7th edition
(John Wiley, 2009), D. C. Montgomery describes a single-factor experiment involving the
random-effects model in which a textile manufacturing company weaves a fabric on a
large number of looms. The company is interested in loom-to-loom variability in tensile
strength. To investigate this variability, a manufacturing engineer selects four looms at
random and makes four strength determinations on fabric samples chosen
Table 13-7 Strength Data for Example 13-4

Observations
Loom 1 2 3 4 Total Average
1 98 97 99 96 390 97.5
2 91 90 93 92 366 91.5
3 96 95 97 95 383 95.8
4 95 96 99 98 388 97.0
1527 95.45
Table 13-8 Analysis of Variance for the Strength Data
Source of Sum of Degrees of Mean Square f0 P-value
Variation Squares Freedom
Looms 89.19 3 29.73 15.68 1.88 E-4
Error 22.75 12 1.90
Total 111.94 15

Sec 13-3 The Random-Effects Model 35

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
 13-3: The Random-Effects Model
Example 13-4
From the analysis of variance, we conclude that the looms in the
plant differ significantly in their ability to produce fabric of uniform
strength. The variance components are estimated by ˆ   1.90 and

29.73  1.90
ˆ    6.96
4
Therefore, the variance of strength in the manufacturing process is
estimated by
V (Yij )  ˆ 2  ˆ 2  6.96  1.90  8.86

Conclusions: Most of the variability in strength in the output

product is attributable to differences between looms.
Sec 13-3 The Random-Effects Model 36
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
 13-4: Randomized Complete Block Design
13-4.1 Design and Statistical Analysis
The randomized block design is an extension of
the paired t-test to situations where the factor of
interest has more than two levels.

Figure 13-9 A randomized complete block design.

Sec 13-4 Randomized Complete Block Design 37

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
 13-4: Randomized Complete Block Design
13-4.1 Design and Statistical Analysis
General procedure for a randomized complete block design:
Table 13-10 A Randomized Complete Block Design with a Treatments and b Blocks

Blocks

Treatments 1 2  b Totals Averages

1 y11 y12  y1b y1. y1.
2 y21 y22  y2b y2. y2 .

     
a ya1 ya2  yab ya. ya .
Totals y.1 y.2  y.b y..
Averages y.1 y.2  y.b y..

Sec 13-4 Randomized Complete Block Design 38

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
 13-4: Randomized Complete Block Design
13-4.1 Design and Statistical Analysis
The appropriate linear statistical model:
i  1, 2, , a
Yij    i   j  ij 
 j  1, 2, , b

We assume
• treatments and blocks are initially fixed effects
• blocks do not interact
•
i 1  j 1
a b
  0 and i 0 j

Sec 13-4 Randomized Complete Block Design 39

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
 13-4: Randomized Complete Block Design
13-4.1 Design and Statistical Analysis
We are interested in testing:
The sum of squares identity for the randomized complete block
design is
a b a b
 ( yij  y..)2  b  ( yi .  y..)2  a  ( y. j  y..)2
i 1 j 1 i 1 j 1 (13-17)
a b
  ( yij  y. j  yi .  y.. ) 2
i 1 j 1
or symbolically

SST = SSTreatments + SSBlocks + SSE

Sec 13-4 Randomized Complete Block Design 40
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
 13-4: Randomized Complete Block Design
13-4.1 Design and Statistical Analysis
The mean squares are:
SSTreatments
MSTreatments 
a 1
SS Blocks
MSBlocks 
b 1
SS E
MS E 
a  1 b  1

Sec 13-4 Randomized Complete Block Design 41

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
 13-4: Randomized Complete Block Design
13-4.1 Design and Statistical Analysis
The expected values of these mean squares are:
a
b  i2
E ( MSTreatments )   2  i 1
a 1
b
a  2j
j 1
E ( MSBlocks)   2 
b 1
E ( MS E )   2

Sec 13-4 Randomized Complete Block Design 42

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
 13-4: Randomized Complete Block Design
13-4.1 Design and Statistical Analysis
Definition
The computing formulas for the sums of squares in the analysis of variance
for a randomized complete block design are
a b 2
..
y
SST    yij2  (13-18)
i 1 j 1 ab
a 2
1 y..
SSTreatments 
b
 yi2 . 
ab
(13-19)
i 1

b 2
..
1 y
SSBlocks 
a
 y 2. j 
ab (13-20)
j 1
and
SSE = SST - SSTreatments - SSBlocks (13-21)
Sec 13-4 Randomized Complete Block Design 43
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
 13-4: Randomized Complete Block Design
13-4.1 Design and Statistical Analysis
Table 13-11 ANOVA for a Randomized Complete Block Design

Source of Degrees of
Variation Sum of Squares Freedom Mean Square F0
Treatments SSTreatments a-1 SSTreatments MSTreatments
a 1 MS E
SS Blocks
Blocks SSBlocks b-1
b 1
SS E
Error SSE (by subtraction) (a -1)(b-1) a  1b  1
Total SST ab – 1

Sec 13-4 Randomized Complete Block Design 44

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
 13-4: Randomized Complete Block Design
Example 13-5 Fabric Strength An experiment was performed to
determine the effect of four different chemicals on the strength of a
fabric. These chemicals are used as part of the permanent press
finishing process. Five fabric samples were selected, and a RCBD was
run by testing each chemical type once in random order on each fabric
sample. The data are shown in Table 13-12. We will test for differences
in means using an ANOVA with a = 0.01.
Table 13-12 Fabric Strength Data—Randomized Complete Block Design
Fabric Treatment Treatment
Sample Totals Averages
Chemical Type 1 2 3 4 5 y i. y. j

1 1.3 1.6 0.5 1.2 1.1 5.7 1.14

2 2.2 2.4 0.4 2.0 1.8 8.8 1.76
3 1.8 1.7 0.6 1.5 1.3 6.9 1.38
4 3.9 4.4 2.0 4.1 3.4 17.8 3.56
Block totals y.j 9.2 10.1 3.5 8.8 7.6 39.2(y..)

Block averages 2.30 2.53 0.88 2.20 1.90 1.96( y..)

yi .
Sec 13-4 Randomized Complete Block Design 45
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
 13-4: Randomized Complete Block Design
Example 13-5
The sums of squares for the analysis of variance are
computed as follows:
4 5
y..2
SST   yij2 
ab
i 1 j 1

(39.2) 2
 (1.3)  (1.6)    (3.4) 
2 2
 25.69
2
20
4
yi2. y..2
SSTreatments  
i 1 b ab
(5.7) 2  (8.8) 2  (6.9) 2  (17.8) 2

5
(39.2) 2
  18.04
20
Sec 13-4 Randomized Complete Block Design 46
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
 13-4: Randomized Complete Block Design
Example 13-5
5 y.2j 2
y ..
SS Blocks   
j 1 a ab

(9.2) 2  (10.1) 2  (3.5) 2  (8.8) 2  (7.6) 2


4
(39.2) 2
  6.69
20
SS E  SST  SS Blocks  SSTreatments
 25.69  6.69  18.04  0.96

The ANOVA is summarized in Table 13-13.

Since f0 = 75.13  f0.01,3,12 = 5.95 (the P-value is 4.79  10-8), we
conclude that there is a significant difference in the chemical types so
far as their effect on strength is concerned.
Sec 13-4 Randomized Complete Block Design 47
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
 13-4: Randomized Complete Block Design
Example 13-5

Table 13-13 Analysis of Variance for the Randomized Complete Block

Experiment
Source of Sum of Degrees of Mean Square f0 P-value
Variation Squares Freedom
Chemical types 18.04 3 6.01 75.13 4.79 E-8
(treatments)
Fabric samples 6.69 4 1.67
(blocks)
Error 0.96 12 0.08
Total 25.69 19

Sec 13-4 Randomized Complete Block Design 48

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
 49
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
 50
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
 13-4: Randomized Complete Block Design

13-4.2 Multiple Comparisons

Fisher’s Least Significant Difference for Example 13-5

Figure 13-10 Results of Fisher’s LSD method.

Sec 13-4 Randomized Complete Block Design 51

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
 13-4: Randomized Complete Block Design

13-4.3 Residual Analysis and Model Checking

Figure 13-11 Normal

probability plot of residuals
from the randomized
complete block design.

Sec 13-4 Randomized Complete Block Design 52

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
 13-4: Randomized Complete Block Design

Figure 13-12 Residuals by treatment from the

randomized complete block design.

Sec 13-4 Randomized Complete Block Design 53

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
 13-4: Randomized Complete Block Design

Figure 13-13 Residuals by block from the

randomized complete block design.
Sec 13-4 Randomized Complete Block Design 54
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
 13-4: Randomized Complete Block Design

Figure 13-14 Residuals versus ŷij from the

randomized complete block design.
Sec 13-4 Randomized Complete Block Design 55
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
13-5 Nonparametric Methods in the Analysis of
Variance

The single-factor analysis of variance model for

comparing a population means is

The hypothesis of interest is

56
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
13-5 Nonparametric Methods in the Analysis of
Variance (Kruskal–Wallis Test )

Since it is a non-parametric method, the Kruskal–Wallis test does not assume

a Normal Distribution of the residuals (Ch.13), unlike the analogous one-way
analysis of variance

The test statistic is

Computational method

57
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
We reject H0 when
58
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Example

59
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
 60
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
 Important Terms & Concepts of Chapter 13
Analysis of variance Multiple comparisons
(ANOVA) Nuisance factors
Blocking Random factor
Completely randomized Randomization
experiment Randomized complete block
Expected mean squares design
Fisher’s least significant Residual analysis & model
difference (LSD) method adequacy checking
Fixed factor Sample size & replication in
Graphical comparison of an experiment
means Treatment effect
Levels of a factor Variance component
Mean square
Chapter 13 Summary 61
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.