In: Statistics and Probability
Proficiency of several biology sections is evaluated using a
special departmental assessment. Three different instructors at two
different class times are assessed. The data are below. What can
the department conclude with an α of 0.01?
Instructor A Time 1 |
Instructor B Time 1 |
Instructor C Time 1 |
Instructor A Time 2 |
Instructor B Time 2 |
Instructor C Time 2 |
---|---|---|---|---|---|
100 88 96 98 |
85 79 80 90 |
89 78 89 61 |
94 81 80 85 |
72 63 73 77 |
90 94 96 87 |
a) What is the appropriate test statistic?
---Select--- na One-Way ANOVA Within-Subjects ANOVA Two-Way
ANOVA
b) Compute the appropriate test statistic(s) to
make a decision about H0.
Time: critical value = ; test statistic
=
Decision: ---Select--- Reject H0 Fail to reject H0
Instructor: critical value = ; test
statistic =
Decision: ---Select--- Reject H0 Fail to reject H0
Interaction: critical value = ; test
statistic =
Decision: ---Select--- Reject H0 Fail to reject H0
c) Compute the corresponding effect size(s) and
indicate magnitude(s).
Time: η2
= ; ---Select--- na trivial effect small
effect medium effect large effect
Instructor: η2
= ; ---Select--- na trivial effect small
effect medium effect large effect
Interaction: η2
= ; ---Select--- na trivial effect small
effect medium effect large effect
d) Make an interpretation based on the
results.
There is a time difference in the assessment.
There is no time difference in the assessment.
There is an instructor difference in the assessment.
There is no instructor difference in the assessment.
There is a time by instructor interaction in the assessment.
There is no time by instructor interaction in the assessment.
Based on the given data, we have to test the effect of Instructors and different Class times on proficiency of biology sections. We have to study the effect of two factors (Categorical) and their interaction on the dependent variable ' Proficiency'.
a) The appropriate test statistic would be that
of a
- Two-Way ANOVA
Running a Two way ANOVA with replication (4 times in each group),
We get the output:
From the output obtained,
b) To test H0. Class time has no effect on Proficiency Vs Ha. Class time has a significant effect on Proficiency
The appropriate test statistic can be expressed as:
with critical / rejection region given by F >
FCritical
Time: Critical value = 8.285 ; Test statistic =
1.313 Since F = 1.313 < 8.285 does not lie in the rejection
region, we fail to reject H0.
Instructor: critical value = 6.013 ; test
statistic = 6.357
Since F = 6.357 > 6.013 lie in the rejection region, Reject
H0
Interaction: critical value = 6.013 ; test
statistic = 7.152
Since F = 7.152 > 6.013 lie in the rejection region, Reject
H0.
c) The corresponding effect size(s) for the factors and their interaction can be obtained from partial eta square using the formula:
It gives the a proportion of variance in the dependent variable accounted for by the effect. It quantifies how strong the effects are.
By Cohen's rule of thumb, η2 = 0.01 implies small effect; η2 = 0.06 implies moderate effect and η2 = 0.14 implies large effect.
Using excel,
Time: η2 = 0.068 ; medium effect
Instructor: η2 = 0.414 ; large
effect
Interaction: η2 = 0.443 ; large
effect
d) Based on the results, we may conclude that, There is no time difference in the assessment. (p-value 0.267 > 0.01) There is an instructor difference in the assessment. (p-value 0.008 < 0.01) There is a time by instructor interaction in the assessment. (p-value 0.005 < 0.01)