You are on page 1of 7

ESGC6122 EXPERIMENTAL DESIGN

TUTORIAL 1
EGE 120016 TAN CHUN HEE

Chapter 2: Single Factor ANOVA


Question 4
H
0
: t
1
= t
2
= t
3
= t
4
= t
5
= 0
H
1
: not all t
j
= 0

Test Statistics:
ANOVA
STAFFING_RATIO
Sum of
Squares
df Mean
Square
F Sig.
Between Groups 6.108 4 1.527 6.797 .002
Within Groups 3.370 15 .225

Total 9.478 19

Since p-value = 0.002 < 0.05, there is sufficient evidence to reject H
0.
Hence, we concluded that
there are differences in staffing ratio due to dominant technology.



Question 6
H
0
: t
1
= t
2
= t
3
= t
4
= 0
H
1
: not all t
j
= 0

Test Statistics:
ANOVA
MOTION_ASSESSMENT
Sum of
Squares
df Mean
Square
F Sig.
Between Groups 313.333 3 104.444 4.719 .012
Within Groups 442.667 20 22.133

Total 756.000 23

Since p-value = 0.012 < 0.05, there is sufficient evidence to reject H
0.
Hence, we concluded that the
level of the room on the cruise ship does affect the perception of the degree of motion.
The p-value of the test is 0.012









Question 9
H
0
:
1
=
2
=
3

H
1
: not all
j
are equal


Test Statistics:
ANOVA
PRICE
Sum of
Squares
df Mean
Square
F Sig.
Between Groups .261 2 .130 36.024 .000
Within Groups .043 12 .004

Total .304 14

Since p-value = 0.000 < 0.05, there is sufficient evidence to reject H
0.
Hence, we concluded that
there is difference in price of the brand of bottled water among the three cities.

Question 10
Between Boynton Beach and Boca Raton
H
0
:
1
=
2

H
1
:
1
=
2

Test Statistics:
ANOVA
PRICE
Sum of
Squares
df Mean
Square
F Sig.
Between Groups .246 1 .246 55.205 .000
Within Groups .036 8 .004

Total .282 9

Since p-value = 0.000 < 0.05, there is sufficient evidence to reject H
0.
Hence, we concluded that
there is a difference in price of the brand of bottled water between Boynton Beach and Boca Raton.

Between Delray Beach and Boca Raton
H
0
:
1
=
2

H
1
:
1
=
2

Test Statistics:
ANOVA
PRICE
Sum of
Squares
df Mean
Square
F Sig.
Between Groups .123 1 .123 32.424 .000
Within Groups .030 8 .004

Total .154 9

Since p-value = 0.000 < 0.05, there is sufficient evidence to reject H
0.
Hence, we concluded that
there is a difference in price of the brand of bottled water between Delray Beach and Boca Raton.



Between Boynton Beach and Delray Beach
H
0
:
1
=
2

H
1
:
1
=
2

Test Statistics:
ANOVA
PRICE
Sum of
Squares
df Mean
Square
F Sig.
Between Groups .021 1 .021 8.186 .021
Within Groups .021 8 .003

Total .042 9

Since p-value = 0.021< 0.05, there is sufficient evidence to reject H
0.
Hence, we concluded that
there is a difference in price of the brand of bottled water between Boynton Beach and Delray
Beach.

Question 12

H
0
:
1
=
2
=
3
=
4

H
1
: not all
j
are equal


Test Statistics:
ANOVA
WAITING_TIME
Sum of
Squares
df Mean
Square
F Sig.
Between Groups 1007.394 3 335.798 106.383 .000
Within Groups 555.543 176 3.156

Total 1562.937 179

Since p-value = 0.000 < 0.05, there is sufficient evidence to reject H
0.
Hence, we concluded that
there is a difference in average waiting time for a receptionist among the four MVPC offices.
The p-value of the test is 0.000

















Question 14

H
0
: t
1
= t
2
= t
3
= 0
H
1
: not all t
j
= 0

Test Statistics:
ANOVA
INSURANCE
Sum of
Squares
df Mean
Square
F Sig.
Between Groups 2859.524 2 1429.762 .339 .717
Within Groups 75907.143 18 4217.063

Total 78766.667 20

Since p-value = 0.717 > 0.05, there is insufficient evidence to reject H
0.
Hence, we concluded that
there are no differences in amount of life insurance due to state.

Question 15

H
0
: t
1
= t
2
= t
3
= 0
H
1
: not all t
j
= 0

Test Statistics:
ANOVA
DONATION
Sum of
Squares
df Mean Square F Sig.
Between Groups 4905500.000 2 2452750.000 3.651 .039
Within Groups 18137500.000 27 671759.259

Total 23043000.000 29

Since p-value = 0.039 < 0.05, there is sufficient evidence to reject H
0.
Hence, we concluded that
there are differences in amount of charitable donations due to solicitation approach.


















Chapter 3: Assumption of ANOVA, Kruskal-Wallis Test, Power
Question 2

H
0
: Amounts of life insurance does not differed by state
H
1
: Amounts of life insurance differed by state

Test Statistics:

Ranks
STATE N Mean Rank
INSURANCE
1.00 7 12.79
2.00 7 9.36
3.00 7 10.86
Total 21


Test Statistics
a,b

INSURANCE
Chi-
Square
1.079
df 2
Asymp.
Sig.
.583
a. Kruskal Wallis Test
b. Grouping Variable:
STATE
Since p-value = 0.583 > 0.05, there is insufficient evidence to reject H
0.
Hence, we concluded that
the amounts of life insurance does not differed by state

Question 3

H
0
: t
1
= t
2
= t
3
= 0
H
1
: not all t
j
= 0

Test Statistics:
ANOVA
INSURANCE
Sum of
Squares
df Mean
Square
F Sig.
Between Groups 2859.524 2 1429.762 .339 .717
Within Groups 75907.143 18 4217.063

Total 78766.667 20

Since p-value = 0.717 > 0.05, there is insufficient evidence to reject H
0.
Hence, we concluded that
there are no differences in amount of life insurance due to state.

Both conventional F test and Kruskal Wallis Test lead to the same result, even though conventional
F test is more powerful (higher probability of rejecting H
0
when it indeed should be rejected). Hence,
we can strongly conclude that there are no differences in amount of life insurance due to state.
Question 7

C=4
o = 0.01
Power = 0.8
/o = 2

R = 10, 10 number of replicates needed at each level of the factor

Question 9

H
0
: There are no differences in solicitation approach with respect to amount of charitable donations
H
1
: There are differences in solicitation approach with respect to amount of charitable donations

Test Statistics:

Ranks
APPROACH N Mean Rank
DONATIONS
1.00 10 18.50
2.00 10 17.95
3.00 10 10.05
Total 30



Test Statistics
a,b

DONATIONS
Chi-Square 5.811
df 2
Asymp. Sig. .055
a. Kruskal Wallis Test
b. Grouping Variable:
APPROACH
Since p-value = 0.055 > 0.05, there is insufficient evidence to reject H
0.
Hence, we concluded that
the amounts of charitable donations does not differed by solicitation approach

Question 10(a)
ANOVA
DONATION
Sum of
Squares
df Mean Square F Sig.
Between Groups 4905500.000 2 2452750.000 3.651 .039
Within Groups 18137500.000 27 671759.259

Total 23043000.000 29





Approach A

95% confidence interval for the true column mean of the approach A

2205 2.05*(/ ) = 2205 531.3255

Hence, the confidence interval is (1673.674, 2736.326)

Approach B

95% confidence interval for the true column mean of the approach B

2125 2.05*(/ ) = 2125 531.3255

Hence, the confidence interval is (1593.674, 2656.326)

Approach C

95% confidence interval for the true column mean of the approach C

1310 2.05*(/ ) = 1310 531.3255

Hence, the confidence interval is (778.6745, 1841.326)

Question 10(b)

Yes, the values from 1673.674 to 1841.326 are included in all three of these confidence interval

Question 10(c)

Since there are overlapping between these three interval, we can conclude that there are similarity
between these three approaches, and therefore there are no significant differences between these
three approaches.

You might also like