Business Statistics, 4e: by Ken Black

Business Statistics, 4e
by Ken Black
Chapter 11
Discrete Distributions
Analysis of
Variance
& Design of
Experiments
Business Statistics, 4e, by Ken Black. © 2003 John Wiley & Sons. 11-1
Learning Objectives
• Understand the differences between various
experimental designs and when to use them.
• Compute and interpret the results of a one-way
ANOVA.
• Compute and interpret the results of a random
block design.
• Compute and interpret the results of a two-way
ANOVA.
• Understand and interpret interaction.
• Know when and how to use multiple comparison
techniques.
Introduction to Design
of Experiments, #1
Experimental Design
- a plan and a structure to test hypotheses in
which the researcher controls or manipulates
one or more variables.
Introduction to Design of Experiments, #2
Independent Variable
• Treatment variable is one that the experimenter
controls or modifies in the experiment.
• Classification variable is a characteristic of the
experimental subjects that was present prior to the
experiment, and is not a result of the
experimenter’s manipulations or control.
• Levels or Classifications are the subcategories of
the independent variable used by the researcher in
the experimental design.
Introduction to Design
of Experiments, #3
Dependent Variable
- the response to the different levels of the
independent variables.
Three Types
of Experimental Designs
• Completely Randomized Design
• Randomized Block Design
• Factorial Experiments
Completely Randomized Design
Machine Operator
1 2 3 4
Valve Opening
Measurements
. . . .
. . . .
. . . .
Valve Openings by Operator
1 2 3 4
6.33 6.26 6.44 6.29
6.26 6.36 6.38 6.23
6.31 6.23 6.58 6.19
6.29 6.27 6.54 6.21
6.4 6.19 6.56
6.5 6.34
6.19 6.58
6.22
Analysis of Variance:
Variance: Assumptions
Assumptions
• Observations are drawn from normally
distributed populations.
• Observations represent random samples
from the populations.
• Variances of the populations are equal.
One-Way
One-Way ANOVA:
ANOVA: Procedural
Procedural
Overview
Overview
H :        
o
1 2 3 k
Ha: At least one of the means is different from the others
MSC
F
MSE
If F > F , reject H .
c
o
If F  F , do not reject H .
c
o
One-Way ANOVA:
Sums of Squares Definitions
total sum of squares = error sum of squares + between sum of squares
SST = SSC + SSE
   Xij X 
nj
i =1
C
j=1
2 C
 n j
j 1
X j X  2 nj
 
i 1
C
j 1
 X ij  X j  2
where : i  particular member of a treatment level

j = a treatment level
C = number of treatment levels
n j
 number of observations in a given treatment level
X = grand mean
X j
= mean of a treatment group or level
X ij
 individual value
Partitioning Total Sum
of Squares of Variation
SST
(Total Sum of Squares)
SSC SSE
(Treatment Sum of Squares) (Error Sum of Squares)
One-Way
One-Way ANOVA:
ANOVA:
Computational
Computational Formulas
Formulas
SSC   n j
C
X
j 1
j X  2
df C
 C 1
  X 
nj C
ij  X
2
SSE  j df E
 N C
i 1 j 1
  X 
nj C 2
SST 
j 1 i 1
ij  X df T
 N 1
SSC where: i = a particular member of a treatment level

MSC 
df C
j = a treatment level
SSE C = number of treatment levels
MSE 
df n= j
number of observations in a given treatment level
E
X = grand mean
MSC
F
MSE X j
column mean
X =ij
individual value
One-Way
One-Way ANOVA:
ANOVA:
Preliminary
Preliminary Calculations
Calculations
1 2 3 4
6.33 6.26 6.44 6.29
6.26 6.36 6.38 6.23
6.31 6.23 6.58 6.19
6.29 6.27 6.54 6.21
6.4 6.19 6.56
6.5 6.34
6.19 6.58
6.22
Tj T1 = 31.59 T2 = 50.22 T3 = 45.42 T4 = 24.92 T = 152.15
nj n1 = 5 n2 = 8 n3 = 7 n4 = 4 N = 24
Mean 6.318000 6.277500 6.488571 6.230000 6.339583
One-Way
One-Way ANOVA:
ANOVA:
Sum
Sum of
of Squares
Squares Calculations
Calculations
SSC   n j
C
j 1
X j X  2
 [5 (6.318 6.339583)  8 (6.27756.339583)

2 2
 7 (6.488571 6.339583)  4 (6.236.339583)

2 2
 0.23658
nj C
SSE   X ij  X j
i 1 j 1
  2
 (6.33 6.318)  (6.266.318)  (6.31 6.318)  (6.29 6.318)

2 2 2 2
 (6.4 6.318)  (6.266.2775)  (6.36 6.2775)

2 2 2
 (6.226.230)  (6.19 6.230)

2 2
 0.15492
One-Way
One-Way ANOVA:
ANOVA:
Sum
Sum of
of Squares
Calculations
X 
nj C 2
SST  
i 1 j 1
ij  X
 (6.336.339583)  (6.266.339583)
2 2
 (6.31 6.339583)    (6.22 6.339583)

2 2
 (6.19 6.339583)
2
 0.39150
One-Way
One-Way
ANOVA:
ANOVA: MeanMean df C
 C 1  4 1  3
Square
Square df E
 N  C  24  4  20
and Calculations df
and FF Calculations T
 N  1  24  1  23
SSC .23658
MSC    .078860
df C
3
SSE .15492
MSE    .007746
df E
20
MSC .078860
F   10.18
MSE .007746
Analysis of Variance
for Valve Openings
Source of Variance df SS MS F
Between 3 0.23658 0.078860 10.18

Error 20 0.15492 0.007746
Total 23 0.39150
A
A Portion
Portion of
of the
the FF Table for 
Table for  == 0.05
0.05
F .05, 3, 20
df1
1 2 3 4 5 6 7 8 9
df 2 1 161.45 199.50 215.71 224.58 230.16 233.99 236.77 238.88 240.54

… … … … … … … … … …
18 4.41 3.55 3.16 2.93 2.77 2.66 2.58 2.51 2.46
19 4.38 3.52 3.13 2.90 2.74 2.63 2.54 2.48 2.42
20 4.35 3.49 3.10 2.87 2.71 2.60 2.51 2.45 2.39
21 4.32 3.47 3.07 2.84 2.68 2.57 2.49 2.42 2.37
One-Way
One-Way ANOVA:
ANOVA:
Procedural
Procedural Summary
Summary
Ho :       
1 2 3 4 Rejection Region
Ha : At least one of the means  1
3
is different from the others
 2
 20
If F > Fc  3.10, reject Ho.

If F  Fc  3.10, do reject Ho. Non rejection 
Region
Since F = 10.18 > Fc  3.10, reject Ho. F .05, 9 ,11
 3.10
Critical Value
Excel
Excel Output
Output
for
for the
the Valve
Valve Opening
Opening Example
Example
Anova: Single Factor
SUMMARY
Groups Count Sum Average Variance
Operator 1 5 31.59 6.318 0.00277
Operator 2 8 50.22 6.2775 0.0110786
Operator 3 7 45.42 6.488571429 0.0101143
Operator 4 4 24.92 6.23 0.0018667
ANOVA
Source of Variation SS df MS F P-value F crit
Between Groups 0.236580119 3 0.07886004 10.181025 0.00028 3.09839
Within Groups 0.154915714 20 0.007745786
Total 0.391495833 23
Multiple Comparison Tests
An analysis of variance (ANOVA) test is an
overall test of differences among groups.
Multiple Comparison techniques are used to
identify which pairs of means are
significantly different given that the
ANOVA test reveals overall significance.
• Tukey’s honestly significant difference
(HSD) test requires equal sample sizes
• Tukey-Kramer Procedure is used when
sample sizes are unequal.
Tukey’s Honestly Significant
Difference (HSD) Test
MSE
HSD  q ,C,N-C
n
where: MSE = mean square error
n = sample size
q ,C,N-C
= critical value of the studentized range distribution from Table A.10
Data
Data from
from Demonstration
Demonstration Problem
Problem 11.1
11.1
PLANT (Employee Age)
1 2 3
29 32 25
27 33 24
30 31 24
27 34 25
28 30 26
Group Means 28.2 32.0 24.8

nj 5 5 5
C=3
dfE = N - C = 12 MSE = 1.63
q Values for  = .01
q
Number of Populations
Degrees of .01, 3,12
 504
.
2 3 4 5 ...
Freedom
1 90 135 164 186
2 14 19 22.3 24.7
3 8.26 10.6 12.2 13.3
4 6.51 8.12 9.17 9.96
.
.
11 4.39 5.14 5.62 5.97
12 4.32 5.04 5.50 5.84
Tukey’s
Tukey’s HSD
HSD Test
Test
for
for the
the Employee
Employee Age
Age Data
Data
MSE 163
.
HSD  q ,C , N  C  5.04  2.88
n 5
XX 1 2
 28.2  32.0  38
.
XX 1 3
 28.2  24.8  3.4
X X 2 3
 32.0  24.8  7.2
Tukey-Kramer Procedure:
Procedure:
The Case of Unequal Sample Sizes
Sizes
MSE 1 1
HSD  q ,C,N-C (  )
2 nr ns
where: MSE = mean square error
th
n r
= sample size for r sample
th
n s
= sample size for s sample
q  ,C,N-C
= critical value of the studentized range distribution from Table A.10
Freighter
Freighter Example:
Example: Means
Means and
and
Sample
Sample Sizes
Sizes for
for the
the Four
Four Operators
Operators
Operator Sample Size Mean

1 5 6.3180
2 8 6.2775
3 7 6.4886
4 4 6.2300
Tukey-Kramer
Tukey-Kramer Results
Results
for
for the
the Four
Four Operators
Operators
Critical |Actual
Pair Difference Differences|
1 and 2 .1405 .0405
1 and 3 .1443 .1706*
1 and 4 .1653 .0880
2 and 3 .1275 .2111*
2 and 4 .1509 .0475
3 and 4 .1545 .2586*
*denotes significant at .05
Partitioning
Partitioning the
the Total
Total Sum
Sum of
of Squares
Squares
in
in the
the Randomized
Randomized Block
Block Design
Design
SST
(Total Sum of Squares)
SSE
(Error Sum of Squares)
SSC SSR SSE’

(Treatment (Sum of Squares (Sum of Squares
Sum of Squares) Blocks) Error)
A Randomized Block Design
Single Independent Variable
Individual
.
observations
Blocking
Variable .
. . . .
. . . . .
. . . .
.
Randomized
Randomized Block
Block Design
Design Treatment
Treatment
Effects:
Effects: Procedural
Procedural Overview
Overview
Ho :         
1 2 3 k
Ha : At least one of the means is different from the others
MSC
F
MSE
If F > F , reject H .
c
o
If F  F , do not reject H .
c
o
Randomized
Randomized Block
Block Design:
Design:
Computational
Computational Formulas
Formulas
C 2
SSC  n  ( X j  X ) df C
 C 1
j 1
n 2
SSR  C  ( X
i X ) df R
 n 1
i 1
n n 2
SSE    ( X ij  X i  X i  X ) df E
  C  1  n  1  N  n  C  1
j 1 i  1
n n 2
SST    ( X ij  X ) df E
 N 1
j 1 i  1
where: i = block group (row)
SSC
MSC  j = a treatment level (column)
C 1
C = number of treatment levels (columns)
SSR
MSR  n = number of observations in each treatment level (number of blocks - rows)
n 1
SSE X  individual observation
ij SSC  sum of squares columns (treatment)
MSE 
N  n  C 1 X  treatment (column) mean SSR = sum of squares rows (blocking)
j
MSC SSE = sum of squares error
F treatments  MSE X  block (row) mean
i SST = sum of squares total
MSR X = grand mean
F blocks MSE

N = total number of observations
Randomized
Randomized Block
Block Design:
Design:
Tread-Wear
Tread-Wear Example
Example
Speed
Block
Supplier Slow Medium Fast Means
( X )
i
1 3.7 4.5 3.1 3.77

2 3.4 3.9 2.8 3.37 N = 15
n=5 3 3.5 4.1 3.0 3.53
4 3.2 3.5 2.6 3.10
5 3.9 4.8 3.4 4.03
Treatment
Means( X ) j
3.54 4.16 2.98 3.56
X
C=3
Randomized
Randomized Block
Block Design:
Design:
Sum
Sum of
of Squares
Calculations (Part
(Part 1)
1)
C 2
SSC  n ( X j  X )
j 1
 5[(3.54  356
. )  (4.16 356
. )  (2.98 356
2 2 2
. )
 3484
.
n 2
SSR  C  ( X
i 1
i X )
 3[(3.77  356
. )  (3.37  356
. )  (3.53 356
. )  (3.10 356
. )  (4.03 356
2 2 2 2 2
. )]
 1549
.
Randomized
Randomized Block
Block Design:
Design:
Sum
Sum of
of Squares
Calculations (Part
(Part 2)
2)
n C 2
SSE   ( X ij  X
i 1 j 1
j X i X )
 (3.7 3.543.77  3.56)  (3.4 3.54 3.37  3.56)   

2 2
(2.6 2.983.103.56)  (3.4  2.98 4.03 3.56)

2 2
 0.143
n C 2
SST   ( X ij  X )
i 1 j 1
 (3.7 3.56)  (3.4 3.56)    (2.6 3.56)  (3.4 3.56)

2 2 2 2
 5.176
Randomized Block Design:
Mean Square Calculations
Calculations
SSC 3.484
MSC    1742
.
C 1 2
SSR 1549 .
MSR    0.387
n 1 4
SSE 0143
.
MSE    0.018
N  n  C 1 8
MSC 1742 .
F   96.78
MSE 0.018
for the Tread-Wear Example
Source of Variance SS df MS F
Treatment 3.484 2 1.742 96.78
Block 1.549 4 0.387 21.50
Error 0.143 8 0.018
Total 5.176 14
Randomized
Randomized Block
Block Design
Design Treatment
Effects: Procedural Summary
Ho: 1   2   3
Ha: At least one of the means is different from the others
MSC 1.742
F   96.78
MSE 0.018
F = 96.78 > F .01,2,8

= 8.65, reject Ho.
Randomized Block Design Blocking
Effects: Procedural Overview
Ho: 1   2   3   4   5
Ha: At least one of the blocking means is different from the others
MSR .387
F   21.5
MSE .018
F = 21.5 > F .01, 4 ,8

= 7.01, reject Ho.
Excel
Excel Output
Output for
for Tread-Wear
Tread-Wear
Example:
Example: Randomized
Randomized Block
Block Design
Design
Anova: Two-Factor Without Replication
SUMMARY Count Sum Average Variance

Suplier 1 3 11.3 3.7666667 0.4933333
Suplier 2 3 10.1 3.3666667 0.3033333
Suplier 3 3 10.6 3.5333333 0.3033333
Suplier 4 3 9.3 3.1 0.21
Suplier 5 3 12.1 4.0333333 0.5033333
Slow 5 17.7 3.54 0.073

Medium 5 20.8 4.16 0.258
Fast 5 14.9 2.98 0.092
ANOVA
Rows 1.5493333 4 0.3873333 21.719626 0.0002357 7.0060651
Columns 3.484 2 1.742 97.682243 2.395E-06 8.6490672
Error 0.1426667 8 0.0178333
Total 5.176
Business Statistics, 4e, by Ken Black. © 2003 John Wiley & Sons. 14 11-41
Two-Way Factorial Design
Column Treatment
.
Cells
Row
Treatment .
. . . .
. . . . .
. . . .
.
Two-Way ANOVA: Hypotheses
Row Effects: Ho: Row Means are all equal.

Ha: At least one row mean is different from the others.
Columns Effects: Ho: Column Means are all equal.
Ha: At least one column mean is different from the others.
Interaction Effects: Ho: The interaction effects are zero.
Ha: There is an interaction effect.
Formulas
Formulas for
for Computing
Computing
aa Two-Way
Two-Way ANOVA
ANOVA
R 2
SSR  nC  ( X
i X ) df R
 R 1
i 1
C 2
SSC  nR  ( X j  X ) df C
 C 1 where:
j 1 n = number of observations per cell
R C 2
SSI  n  ( X ij  X i  X j  X ) df I
  R  1  C  1 C = number of column treatments
i 1 j 1
R C n R = number of row treatments
SSE     ( X ijk  X ij )
2
df E
 RC n  1 i = row treatment level
i 1 j 1 k 1
C R n 2 j = column treatment level
SST     ( X ijk  X ) df T
 N 1 k = cell member
c 1 r 1 a 1
SSR MSR Xijk = individual observation
MSR 
R 1 FR  MSE X ij
= cell mean
SSC MSC
MSC 
C 1 F C

MSE
X i
= row mean
SSI MSI X j
= column mean
MSI 
 R  1  C  1 F I

MSE X = grand mean
SSE
MSE 
RC n  1
A 22  33 Factorial
Factorial Design
Design
with Interaction
Row effects
Cell
R1
Means
R2
C1 C2 C3
Column
Factorial Design
Design
with Some Interaction
Row effects
Cell
R1
Means
R2
C1 C2 C3
Column
Factorial Design
Design
with No Interaction
Row effects
Cell
R1
Means
R2
C1 C2 C3
Column
A 22  33 Factorial
A Factorial Design:
Design: Data
Data and
and
Measurements
Measurements for for CEO
CEO Dividend
Dividend Example
Example
Location Where Company
Stock is Traded
How Stockholders
are Informed of NYSE AMEX OTC Xi
Dividends
2 2 4
1 3 3
Annual/Quarterly
2 3 4 2.5
Reports
1 2 3
X11=1.5 X12=2.5 X13=3.5 X=2.7083
2 3 4 N = 24
3 3 4 n=4
Presentations to
1 2 3 2.9167
Analysts
2 4 4
X21=2.0 X22=3.0 X23=3.75
Xj 1.75 2.75 3.625
A 2  3 Factorial Design: Calculations
for
for the
the CEO
CEO Dividend
Dividend Example
Example (Part
(Part 1)
1)
R 2
SSR  nC  ( X i  X )
i 1
 ( 4)( 3)[( 2.5  2.7083) 2  ( 2.9167  2.7083) 2 ]

 10418
.
C 2
SSC  nR  ( X j  X )
j 1
 ( 4)( 2)[(1.75  2.7083) 2  ( 2.75  2.7083) 2  (3.625  2.7083) 2 ]

 14.0833
R C 2
SSI  n  ( X ij  X i  X j  X )
i 1 j 1
 4[(15
.  2.5  1.75  2.7083) 2  (2.5  2.5  2.75  2.7083) 2
 ( 3.5  2.5  3.625  2.7083) 2  (2.0  2.9167  1.75  2.7083) 2
 ( 3.0  2.9167  2.75  2.7083) 2  (3.75  2.9167  3.625  2.7083) 2 ]
 0.0833
Factorial Design:
Design: Calculations
Calculations
for
for the
the CEO
CEO Dividend
Dividend Example
Example (Part
(Part 2)
2)
R C n
SSE     ( X ijk  X ij )
2
i 1 j 1 k 1
 (2 15
. )  (115
. )  (3 375
. )  (4  375
2 2 2 2
. )
 7.7500
C R n 2
SST     ( X ijk  X )
c 1 r 1 a 1
 (2  2.7083)  (1 2.7083)  (3 2.7083)  (4  2.7083)

2 2 2 2
 22.9583
Factorial Design:
Design: Calculations
Calculations
for
for the
the CEO
CEO Dividend
Dividend Example
Example (Part
(Part 3)
3)
SSR 10418 . MSR 10418
.
MSR 
R 1

1
 10418
. F R  MSE  0.4306  2.42
SSC 14.0833 MSC 7.0417
MSC 
C 1

2
 7.0417 F C MSE  0.4306  16.35

SSI 0.0833 MSI 0.0417
MSI 
 R  1  C  1

2
 0.0417 F I  MSE  0.4306  010
.
SSE 7.7500
MSE    0.4306
RC n  1 18
for the CEO Dividend Problem
Source of Variance SS df MS F
Row 1.0418 1 1.0418 2.42
Column 14.0833 2 7.0417 16.35*
Interaction 0.0833 2 0.0417 0.10
Error 7.7500 18 0.4306
Total 22.9583 23
*
Denotes significance at = .01.
Excel
Excel Anova: Two-Factor With Replication
SUMMARY NYSE ASE OTC Total

Output
Output AQReport
Count 4 4 4 12
for
for the
the
Sum
Average
Variance
6
1.5
10
2.5
0.3333 0.3333 0.3333
14
3.5
30
2.5
1
CEO
CEO Presentation
Dividend
Count 4 4 4 12
Dividend Sum
Average
8
2
12
3
15
3.75
35
2.9167
Example
Example Variance 0.6667 0.6667 0.25 0.9924
Total
(Part
(Part 1)
1) Count
Sum
8
14
8
22
8
29
Average 1.75 2.75 3.625
Variance 0.5 0.5 0.2679
Excel Output for the
CEO Dividend Example (Part 2)
ANOVA
Sample 1.0417 1 1.0417 2.4194 0.1373 4.4139
Columns 14.083 2 7.0417 16.355 9E-05 3.5546
Interaction 0.0833 2 0.0417 0.0968 0.9082 3.5546
Within 7.75 18 0.4306
Total 22.958 23

Business Statistics, 4e: by Ken Black

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Business Statistics, 4e: by Ken Black

Uploaded by

Copyright:

Available Formats

Business Statistics, 4e

6.26 6.36 6.38 6.23

6.31 6.23 6.58 6.19

6.29 6.27 6.54 6.21

6.4 6.19 6.56

Ha: At least one of the means is different from the others

where : i  particular member of a treatment level

SSC where: i = a particular member of a treatment level

 [5 (6.318 6.339583)  8 (6.27756.339583)

 7 (6.488571 6.339583)  4 (6.236.339583)

 (6.33 6.318)  (6.266.318)  (6.31 6.318)  (6.29 6.318)

 (6.4 6.318)  (6.266.2775)  (6.36 6.2775)

 (6.226.230)  (6.19 6.230)

 (6.31 6.339583)    (6.22 6.339583)

Between 3 0.23658 0.078860 10.18

df 2 1 161.45 199.50 215.71 224.58 230.16 233.99 236.77 238.88 240.54

If F > Fc  3.10, reject Ho.

Group Means 28.2 32.0 24.8

Operator Sample Size Mean

1 and 3 .1443 .1706*

1 and 4 .1653 .0880

2 and 3 .1275 .2111*

2 and 4 .1509 .0475

3 and 4 .1545 .2586*

*denotes significant at .05

SSC SSR SSE’

Single Independent Variable

Ha : At least one of the means is different from the others

1 3.7 4.5 3.1 3.77

 (3.7 3.543.77  3.56)  (3.4 3.54 3.37  3.56)   

(2.6 2.983.103.56)  (3.4  2.98 4.03 3.56)

 (3.7 3.56)  (3.4 3.56)    (2.6 3.56)  (3.4 3.56)

F = 96.78 > F .01,2,8

F = 21.5 > F .01, 4 ,8

SUMMARY Count Sum Average Variance

Slow 5 17.7 3.54 0.073

Row Effects: Ho: Row Means are all equal.

 ( 4)( 3)[( 2.5  2.7083) 2  ( 2.9167  2.7083) 2 ]

 ( 4)( 2)[(1.75  2.7083) 2  ( 2.75  2.7083) 2  (3.625  2.7083) 2 ]

 (2  2.7083)  (1 2.7083)  (3 2.7083)  (4  2.7083)

SUMMARY NYSE ASE OTC Total

You might also like