In: Statistics and Probability
Oxnard Petro, Ltd., has three interdisciplinary project
development teams that function on an ongoing basis. Team members
rotate from time to time. Every 4 months (three times a year) each
department head rates the performance of each project team (using a
0 to 100 scale, where 100 is the best rating). Are the main effects
significant? Is there an interaction?
Year | Marketing | Engineering | Finance |
2007 | 82 | 65 | 95 |
84 | 83 | 93 | |
86 | 68 | 91 | |
2009 | 88 | 74 | 85 |
87 | 66 | 80 | |
80 | 84 | 95 | |
2011 | 83 | 70 | 95 |
90 | 71 | 85 | |
81 | 73 | 78 | |
(a-1) Choose the correct row-effect hypotheses.
a. | H0: A1 ≠ A2 ≠ A3 ≠ 0 | ⇐⇐ year means differ |
H1: All the Aj are equal to zero | ⇐⇐ year means are the same | |
b. | H0: A1 = A2 = A3 = 0 | ⇐⇐ year means are the same |
H1: Not all the Aj are equal to zero | ⇐⇐ year means differ |
a
b
(a-2) Choose the correct column-effect
hypotheses.
a. | H0: B1 ≠ B2 ≠ B3 ≠ 0 | ⇐⇐ department means differ |
H1: All the Bj are equal to zero | ⇐⇐ department type means are the same | |
b. | H0: B1 = B2 = B3 = 0 | ⇐⇐ department means are the same |
H1: Not all the Bj are equal to zero | ⇐⇐ department type means differ |
a
b
(a-3) Choose the correct interaction-effect
hypotheses.
a. | H0: Not all the ABjk are equal to zero | ⇐⇐ there is an interaction effect |
H1: All the ABjk are equal to zero | ⇐⇐ there is no interaction effect | |
b. | H0: All the ABjk are equal to zero | ⇐⇐ there is no interaction effect |
H1: Not all the ABjk are equal to zero | ⇐⇐ there is an interaction effect |
a
b
(b) Fill in the missing data. (Round your Table of
Means values to 1 decimal place, SS and F values
to 2 decimal places, MS values to 3 decimal places, and
p-values to 4 decimal places.)
Table of Means | ||||
Factor 2 (Department) | ||||
Factor 1 (Year) | Marketing | Engineering | Finance | Average |
2007 | ||||
2009 | ||||
2011 | ||||
Total | ||||
Source | SS | df | MS | F | p-value |
Factor 1 (Year) | |||||
Factor 2 (Department) | |||||
Interaction | |||||
Error | |||||
Total | |||||
(c) Using α = 0.05, choose the correct
statement.
The main effects of department and year are significant, but there is not a significant interaction effect.
The main effect of department is significant; however, there is no significant effect from year or interaction between department and year.
The main effect of year is significant; however, there is no significant effect from department or interaction between department and year.
(d) Interpret the p-values carefully.
The p-values range from highly significant (Department) to
insignificant (Year). The interaction effect, if any,
is (Click to
select) weak strong since
about (Click to
select) 71 90 39 samples
in 100 would show an F statistic this large in
the (Click to
select) presence absence of
interaction.
(a-1) Choose the correct row-effect hypotheses.
a. | H0: A1 ≠ A2 ≠ A3 ≠ 0 | ⇐⇐ year means differ |
H1: All the Aj are equal to zero | ⇐⇐ year means are the same | |
b. | H0: A1 = A2 = A3 = 0 | ⇐⇐ year means are the same |
H1: Not all the Aj are equal to zero | ⇐⇐ year means differ |
a
b
(a-2) Choose the correct column-effect hypotheses.
a. | H0: B1 ≠ B2 ≠ B3 ≠ 0 | ⇐⇐ department means differ |
H1: All the Bj are equal to zero | ⇐⇐ department type means are the same | |
b. | H0: B1 = B2 = B3 = 0 | ⇐⇐ department means are the same |
H1: Not all the Bj are equal to zero | ⇐⇐ department type means differ |
a
b
(a-3) Choose the correct interaction-effect hypotheses.
a. | H0: Not all the ABjk are equal to zero | ⇐⇐ there is an interaction effect |
H1: All the ABjk are equal to zero | ⇐⇐ there is no interaction effect | |
b. | H0: All the ABjk are equal to zero | ⇐⇐ there is no interaction effect |
H1: Not all the ABjk are equal to zero | ⇐⇐ there is an interaction effect |
a
b
(b) Fill in the missing data. (Round your Table of Means values to
1 decimal place, SS and F values to 2 decimal
places, MS values to 3 decimal places, and
p-values to 4 decimal places.)
Factor 2 | |||||
Means: | |||||
Marketing | Engineering | Finance | |||
2007 | 84.0 | 72.0 | 93.0 | 83.0 | |
Factor 1 | 2009 | 85.0 | 74.7 | 86.7 | 82.1 |
2011 | 84.7 | 71.3 | 86.0 | 80.7 | |
84.6 | 72.7 | 88.6 | 81.9 | ||
ANOVA table | |||||
Source | SS | df | MS | F | p-value |
Factor 1 | 24.96 | 2 | 12.481 | 0.31 | .7342 |
Factor 2 | 1,229.41 | 2 | 614.704 | 15.48 | .0001 |
Interaction | 84.81 | 4 | 21.204 | 0.53 | .7125 |
Error | 714.67 | 18 | 39.704 | ||
Total | 2,053.85 | 26 |
(c) Using α = 0.05, choose the correct statement.
The main effects of department and year are significant, but there is not a significant interaction effect.
The main effect of department is significant; however, there is no significant effect from year or interaction between department and year.
The main effect of year is significant; however, there is no significant effect from department or interaction between department and year.
(d) Interpret the p-values carefully.
The p-values range from highly significant (Department) to
insignificant (Year). The interaction effect, if any,
is (Click to select) weak since
about (Click to select) 71
samples in 100 would show an F statistic this large in
the (Click to select) presence of
interaction.