Lecture 13

F-Test and Decomposition of Data

Test of Two Variances
Travel Expenses

Sales ($) Audit ($)

H0 : = 0, H1 : >0 131 130

audit sales audit sales
135 102
Significance Level = 10%
2 146 129
Saudit
Test Statistic: F = 2
F(6,5) , 165 143
S sales
136 149
Fobs = 1.66 , P(F > 1.66) = 29.7% > 142 120
139
The null hypothesis is not rejected. The Variance 149.9 249.2
two variances have no significant dof 5 6

difference. F-Value 1.66

Probability 29.7%

Lecture 13. F-Test and Decomposition of Data 2

Pooling of Errors
Travel Expenses
130 130.3 -0.3
131 142.5 -11.5 102 130.3 -28.3
135 142.5 -7.5 129 130.3 -1.3
146 142.5 3.5 143 = 130.3 + 12.7
165 = 142.5 + 22.5 149 130.3 18.7
136 142.5 -6.5 120 130.3 -10.3
142 142.5 -0.5 139 130.3 8.7
SS = 122587 121837.5 749.5 SS = 120316 118820.6 1495.4
dof = 6 1 5 dof = 7 1 6

749.5 1495.4
2
Ssales = = 149.9 2
Saudit = = 249.2
5 6

749.5 + 1495.4 149.9 × 5 + 249.2 × 6

2
Spooled = = = 204.1
11 5+6

Lecture 13. F-Test and Decomposition of Data 3

Test of Two Variances Route A
52
Route B
59
Two Routes 67 60
56 61

A limousine service offers two routes to the 45 51

70 56
airport. The CEO of a company wants to
54 63
study the variation of the time. He collected
64 57
the following data. Are the variances 65
significantly different, using = 5%. Mean 58.3 59.0
Variance 80.9 19.1
dof 6 7
H0 : 2
A
2
B
= 0, H1 : 2
A
2
B
>0
F-Value 4.23
Significance Level = 5% Probability 4.04%
2
S
Test Statistic: F = A
2
F , Fobs = 4.23 , P(F > 4.23) = 4.04% <
SB

The null hypothesis is rejected. The variance of Route B is significantly greater than
Route A.
Lecture 13. F-Test and Decomposition of Data 4
F-Ditribution

If two sets of data are sampled

from the same population, their
ratio of variances follows F- dof = 31, 31 (m = 1, n = 32)

distribution. dof = 15, 31 (m = 2, n = 32)

dof = 7, 31 (m = 4, n = 32)
Let two variances be Sx2 and
dof = 3, 31 (m = 8, n = 32)
Sy2 , with degrees of freedom dof = 1, 31 (m = 16, n = 32)

dofx and dofy , then

Sx2
2
F(dofx , dofy )
S y

Lecture 13. F-Test and Decomposition of Data 5

How to Generate an F-Ditribution?
Generate 32 random values,
using NORMSINV(RAND()).
Calculate the variance Sy2 , dof = 31, 31 (m = 1, n = 32)

using VAR. dof = 15, 31 (m = 2, n = 32)

Let m = 2; calculate 16 dof = 7, 31 (m = 4, n = 32)

dof = 3, 31 (m = 8, n = 32)
averages; calculate the dof = 1, 31 (m = 16, n = 32)

variance S . Let S = mS .
2
z
2
x
2
z

Calculate F = Sx2 Sy2 . Record

the F value.
Repeat the foregoing steps as
many times as possible and
draw an F-distribution curve for
m = 2.
Lecture 13. F-Test and Decomposition of Data 6
Significance of a Factor
A1 75 78 65 65 Ave Stdev t-Test F-Test
79 77 75 69 72.1 5.5 Pooled variance = 24.0 Var(y) = 24.0
74 79 66 67 Difference = 3.69 mxVar(z) = 108.8
64 68 76 76 Std Dev of Diff = 1.73 F-Value = 4.53
A2 79 77 75 70 Ave Stdev t-Value = 2.13 dof(y) = 30
78 68 71 77 75.8 4.2 dof = 30 dof(z) = 1
79 74 73 78 Probability = 4.16% Probability = 4.16%
82 74 83 74

Two ways of estimating variance: Null hypothesis: factor A is NOT significant

2 2
If A is not significant, then m z y
follows
5.5 × 15 + 4.2 × 15
2 2
2
y
= = 24.0 F-distribution.
30
Observed F = 4.53.
m 2
z
= 16 × Var(72.1, 75.8) = 108.8 The probability that F > 4.53 is 4.16%.

m 2
108.8 If a significant level of 5% is used, the null
F= z
= = 4.53
2
y
24.0 hypothesis must be rejected, i.e., factor A

p = FDIST(4.5, 1, 30) = 4.16% is significant.

Lecture 13. F-Test and Decomposition of Data 7
Decompoition of Data
Exp. A B C D E F G H y1 y2 y3 y4 y5 y6 y7 y S
1 1 1 1 1 1 1 1 1 10.18 10.18 10.12 10.06 10.02 9.98 10.20 10.106 0.0870
2 1 1 2 2 2 2 2 2 10.03 10.01 9.98 9.96 9.91 9.89 10.12 9.986 0.0776
3 1 1 3 3 3 3 3 3 9.81 9.78 9.74 9.74 9.71 9.68 9.87 9.761 0.0641
4 1 2 1 1 2 2 3 3 10.09 10.08 10.07 9.99 9.92 9.88 10.14 10.024 0.0964
5 1 2 2 2 3 3 1 1 10.06 10.05 10.05 9.89 9.85 9.78 10.12 9.971 0.1293
6 1 2 3 3 1 1 2 2 10.20 10.19 10.18 10.17 10.14 10.13 10.22 10.176 0.0321
7 1 3 1 2 1 3 2 3 9.91 9.88 9.88 9.84 9.82 9.80 9.93 9.866 0.0476
8 1 3 2 3 2 1 3 1 10.32 10.28 10.25 10.20 10.18 10.18 10.36 10.253 0.0709
9 1 3 3 1 3 2 1 2 10.04 10.02 10.01 9.98 9.95 9.89 10.11 10.000 0.0698
10 2 1 1 3 3 2 2 1 10.00 9.98 9.93 9.80 9.77 9.70 10.15 9.904 0.1563
11 2 1 2 1 1 3 3 2 9.97 9.97 9.91 9.88 9.87 9.85 10.05 9.929 0.0713
12 2 1 3 2 2 1 1 3 10.06 9.94 9.90 9.88 9.80 9.72 10.12 9.917 0.1392
13 2 2 1 2 3 1 3 2 10.15 10.08 10.04 9.98 9.91 9.90 10.22 10.040 0.1199
14 2 2 2 3 1 2 1 3 9.91 9.87 9.86 9.87 9.85 9.80 10.02 9.883 0.0687
15 2 2 3 1 2 3 2 1 10.02 10.00 9.95 9.92 9.78 9.71 10.06 9.920 0.1295
16 2 3 1 3 2 3 1 2 10.08 10.00 9.99 9.95 9.92 9.85 10.14 9.990 0.0973
17 2 3 2 1 3 1 2 3 10.07 10.02 9.89 9.89 9.85 9.76 10.19 9.953 0.1473
18 2 3 3 2 1 2 3 1 10.10 10.08 10.05 9.99 9.97 9.95 10.12 10.037 0.0673
Ave = 9.984
Level A B C D E F G H
1 10.016 9.934 9.988 9.989 9.999 10.074 9.978 10.032
2 9.953 10.002 9.996 9.970 10.015 9.972 9.967 10.020
3 　 10.016 9.969 9.995 9.938 9.906 10.007 9.901
Lecture 13. F-Test and Decomposition of Data 8
Decompoition of Data
No Control Factors

{ 10.18 10.18 10.12 10.06 10.02 9.98 10.20 }

- { 10.106 10.106 10.106 10.106 10.106 10.106 10.106 }
= { 0.074 0.074 0.014 -0.046 -0.086 -0.126 0.094 }

{ y1 y2 y3 y4 y5 y6 y7 }
- { y y y y y y y }
= { y1 y y2 y y3 y y4 y y5 y y6 y y7 y }

{y } {y } = {y
i i
y }
DOFy DOFy = DOFy y SSy SSy = SSy
i i
i i y

7 1= 6 714.9236 714.8782 = 0.0454

Lecture 13. F-Test and Decomposition of Data 9
Decompoition of Data
One Control Factor

Exp. A y1 y2 y3 y4 y5 y6 y7 y S
1 1 10.18 10.18 10.12 10.06 10.02 9.98 10.20 10.106 0.0870
2 2 10.06 10.18 10.05 9.88 9.96 9.93 10.04 10.014 0.0996
Ave = 10.060

　 A
Level 1 10.106
Level 2 10.014
10.060 10.060 10.060 10.060 10.060 10.060 10.060
10.060 10.060 10.060 10.060 10.060 10.060 10.060
} 0.046 0.046 0.046 0.046 0.046 0.046 0.046
-0.046 -0.046 -0.046 -0.046 -0.046 -0.046 -0.046
10.106 10.106 10.106 10.106 10.106 10.106 10.106
10.014 10.014 10.014 10.014 10.014 10.014 10.014
} 0.074
0.046
0.074
0.166
0.014 -0.046 -0.086 -0.126 0.094
0.036 -0.134 -0.054 -0.084 0.026
10.180 10.180 10.120 10.060 10.020 9.980 10.200
10.060 10.180 10.050 9.880 9.960 9.930 10.040

Lecture 13. F-Test and Decomposition of Data 10

Decompoition of Data
One Control Factor

10.180 10.180 10.120 10.060 10.020 9.980 10.200

10.060 10.180 10.050 9.880 9.960 9.930 10.040
} Total variations
10.060 10.060 10.060 10.060 10.060 10.060 10.060
-　
10.060 10.060 10.060 10.060 10.060 10.060 10.060

0.046 0.046 0.046 0.046 0.046 0.046 0.046 Factor effect

=
=　
-0.046 -0.046 -0.046 -0.046 -0.046 -0.046 -0.046

0.074 0.074 0.014 -0.046 -0.086 -0.126 0.094 Error

+
+　
0.046 0.166 0.036 -0.134 -0.054 -0.084 0.026

14 1= 1+ 12
1416.9846 1416.8504 = 0.0293 + 0.1049

Lecture 13. F-Test and Decomposition of Data 11

Decompoition of Data
Mutiple Control Factors

DOFT = DOFA + DOFB + DOFC + + DOFe

{y } {y } = {E } + {E } + {E } +
ij A B C {}
+ e
SST = SSA + SSB + SSC + + SSe

Exp. A B C y1 y2 y3 y4 y5 y S
1 1 1 1 69.7 72.4 55.0 65.3 68.7 66.2 6.77
2 1 2 2 62.1 77.2 72.5 69.8 71.9 70.7 5.52
3 2 1 2 100.2 89.6 95.2 102.5 98.8 97.3 5.03
4 2 2 1 99.1 104.4 100.7 105.8 95.6 101.1 4.10
Ave. = 83.83

　 A B C
Level 1 68.46 81.74 83.67
Level 2 99.19 85.91 83.98

Lecture 13. F-Test and Decomposition of Data 12

Decompoition of Data
Mutiple Control Factors
69.7 72.4 55.0 65.3 68.7 83.83 83.83 83.83 83.83 83.83
62.1 77.2 72.5 69.8 71.9 83.83 83.83 83.83 83.83 83.83
-
100.2 89.6 95.2 102.5 98.8 83.83 83.83 83.83 83.83 83.83
99.1 104.4 100.7 105.8 95.6 83.83 83.83 83.83 83.83 83.83

-15.37 -15.37 -15.37 -15.37 -15.37 -2.08 -2.08 -2.08 -2.08 -2.08
-15.37 -15.37 -15.37 -15.37 -15.37 2.08 2.08 2.08 2.08 2.08
= +
15.37 15.37 15.37 15.37 15.37 -2.08 -2.08 -2.08 -2.08 -2.08
15.37 15.37 15.37 15.37 15.37 2.08 2.08 2.08 2.08 2.08

-0.16 -0.16 -0.16 -0.16 -0.16 3.48 6.18 -11.22 -0.92 2.48
0.16 0.16 0.16 0.16 0.16 -8.60 6.50 1.80 -0.90 1.20
+ +
0.16 0.16 0.16 0.16 0.16 2.94 -7.66 -2.06 5.24 1.54
-0.16 -0.16 -0.16 -0.16 -0.16 -2.02 3.28 -0.42 4.68 -5.52
20 1= 1+ 1+ 1+ 16
145815.37 140532.61= 4721.67 + 86.94 + 0.48 + 473.67
Lecture 13. F-Test and Decomposition of Data 13