TESTA at UNSW, Sean Brawley, TESTA Summit 16 Sept 2013
STAT 3510 Presentation
1. STAT 3510 Final Project
CSUEB Fall Teaching Faculty and Academic Calendars
2. Motivation for our project
Due to the fact that CSUEB is changing from a quarter system to a semester system,
we wanted to get a real world sampling of the teaching faculties thoughts on
which is a better academic calendar. We thought it would be easier to get a true
Simple Random Sample from this population.
We asked them which system they individually prefer and which system they
thought benefitted students more.
They were given three choices: semesters, quarters, and no preference.
3. Methods/Procedure
Our sampling frame was the fall teaching faculty at CSUEB, a total of 804.
We used an online random name generator to sample 300 of them.
We sent out an 8 question survey to their CSUEB e-mail account. Two
separate e-mails were sent out over the course of two weeks.
We received 163 responses back for a response rate of 54.3%, we discarded
two of the responses due to completely unusable data, so our final sample size
was 161 respondents.
4. Descriptive Statistics
What Academic Calendar do Teaching Faculty Prefer
1. Semester System
The 95% confidence interval for teaching faculty who prefer semester system is (0.288, 0.420)
Thus, we are 95 % confident that between 28.8% and 42% teaching faculty prefer semester system.
2. Quarter System
The 95% confidence interval for teaching faculty who prefer quarter system is (0.403, 0.541)
Therefore, we are 95% confident that between 40.3% and 54.1% teaching faculty prefer quarter system.
3. No preference
The 95% confidence interval for teaching faculty who have no preference of systems is (0.121, 0.227)
Therefore, we are 95% confident that between 12.1% and 22.7% teaching faculty have no preference on specific system either
5. Descriptive Statistics
Which Benefits Students : Quarter? Semester? No Preference?
1. Semester System
The 95 % C. I for teaching faculty who said that the semester system benefits students more is (0.325,0.457)
Therefore, we are 95% confident that between 32.5 % and 45.7% teaching faculty say that semester system benefits students
more.
2. Quarter System
The 95 % C. I for teaching faculty who said that the quarter system benefits students more is (0.258,0.386)
Thus, we are 95 % confident that between 25.8% and 38.6% of the teaching faculty say that quarter system benefits students
more.
3. No Preference
The 95 % C. I for teaching faculty who had no preference for which system benefits students more is (0.230,0.340)
Thus, we are 95% confident that between 23% and 34% teaching faculty have no preference which system will benefit students.
6. Hypothesis(Likelihood of response by college)
Ho:
When looked at by college, there is no significant statistical difference in the
probability of getting a response.
Ha:
There is at least one college that is more or less likely to respond to our survey.
7. College Count of potential Response
#s
Count of actual
responses
response
percentage
Business and
Economics
36 18 0.5
Education and Allied
Studies
35 19 0.542857143
Letters, Arts and
Social Sciences
127 78 0.614173228
Science 102 49 0.480392157
Grand Total 300 164 0.546666667
Table of Values
8. Results
Average response rate is about 53% with a sample standard deviation of about 6%.
It is 95% confident that response rates will be within 44% and 62%.
All of the response rates recorded are within this range so don’t reject Ho.
In other words each college is equally likely to respond to the college or, despite
Statistics teachers being counted under Math and thus Science they are no more likely
to help out Statistic students than any other teacher.
9. Hypothesis (College Within CSUEB)
Ho: The college within CSUEB that teaching
faculty works in makes no difference as to
what academic calendar they prefer.
Ha: The college within CSUEB that teaching
faculty works in makes a difference as to
what academic calendar they prefer.
H0: The college within CSUEB that teaching
faculty works in makes no difference as to
what academic calendar they think is more
beneficial to the students.
Ha: The college within CSUEB that teaching
faculty works in makes a difference as to what
academic calendar they think is more
beneficial to the students.
10. No preference Quarter Semester Total
College of Business and
Economics
4
3.019
0.31905
8
8.609
0.04304
6
6.373
0.02179
18
College of Education and
Allied Studies
4
3.019
0.31905
6
8.609
0.79051
8
6.373
0.41556
18
College of Letter, Arts, and
Social Sciences
11
12.745
0.23901
39
36.348
0.19352
26
26.907
0.03056
76
College of Science 8
8.217
0.00575
24
23.435
0.01363
17
17.348
0.00697
49
Total 27 77 57 161
Cell contents: Count
Expected count
Contribution to Chi-Square
Pearson Chi-Square = 3.684. DF = 6, P-Value = 0.719
Chi Square Test: CSUEB College, Benefits the Teacher
11. No preference Quarter Semester Total
College of Business and
Economics
5
5.143
0.00397
6
5.814
0.00597
7
7.304
0.00027
18
College of Education and
Allied Studies
7
5.143
0.67063
4
5.814
0.56580
7
7.043
0.00027
18
College of Letters, Arts, and
Social Science
23
21.714
0.07613
21
24.547
0.51242
32
29.739
0.17188
76
College of Science 11
14.000
0.64286
21
15.826
1.69147
17
19.174
0.24648
39
Total 46 52 63 161
Chi Square Test: CSUEB College, Benefits for the Students
Cell contents: Count
Expected count
Contribution to Chi-square
Pearson Chi-square = 5.711, DF = 6, P-Value = 0.456
12. Results
There doesn’t seem to be a difference
between the college that the
teaching faculty works in and
teachers and students benefits in
academic calendar.
Therefore we fail to reject the null
hypothesis Ho in both situations.
Teaching faculty had about 20 more responses
for quarter in terms of benefits for the
teachers, and 11 more think that quarter
would benefit the students
If we had just semester and quarter, and took
out the no preference answer, maybe the
data would have changed and caused a
preference towards one or the other.
Especially, in the college and benefits students
test because there was 46 total no preference
responses
13. Hypothesis (Teaching Title)
Ho:The title and rank of the teaching
faculty at CSUEB makes no difference
as to what system they prefer
Ha: The title and rank of the teaching
faculty at CSUEB makes a difference in
what system they prefer
Ho: The title and rank of the teaching
faculty at CSUEB makes no difference
in what system they think is beneficial
to the students
Ha: The title and rank of the teaching
faculty at CSUEB makes a difference in
what system they think is beneficial to
the students
14. Teaching Title vs Personal Preference
Chi-Sq=23.046, DF=6 P-Value=0.001, Reject Null Hypothesis
Full professor Associate
professor
Assistant
Professor
Adjunct
Professor
Total
Quarter 17
12.27
1.820
9
8.50
0.030
2
10.86
7.226
48
44.37
0.297
76
Semester 4
9.20
2.943
7
6.37
0.062
17
8.14
9.634
29
33.28
0.550
57
No Preference 5
4.52
0.051
2
3.313
0.408
4
4.00
17
16.35
0.026
28
Total 26 18 23 94 161
15. Teaching Title vs Student Benefit
Chi-Sq= 23.788; DF= 6; P-Value=0.001, Reject Null Hypothesis
Full Professor Associate
Professor
Assistant
Professor
Adjunct
Professor
Total
Quarter 11
8.40
0.807
6
5.81
0.006
2
7.43
3.967
33
30.36
0.230
52
Semester 6
10.17
1.712
10
7.04
1.241
18
9.00
9.000
29
36.78
1.647
63
No Preference 9
7.43
0.332
2
5.14
1.921
3
6.57
1.941
32
26.86
0.985
46
Total 26 18 23 94 161
16. Results
From the data there is a correlation between
job title and the type of system they prefer
Since there is a preference we reject Ho
After running pairwise tests, the data shows
that full professors show a preference for
quarters over semesters, while assistant
professors show a strong preference for
semesters over quarters, the other two group
don’t show a preference.
From the data there is a correlation between
job title and the type of system they believe
to be beneficial to the students
Since there is a preference we reject Ho
The data shows that overall the teaching
faculty believes that semester system is
better for the students
The results are more interesting in this
comparison because the adjunct professors
are almost equally split three ways
17. Hypothesis (system they have previously taught in)
Ho: There is no difference in the proportion
of teaching faculty that have only taught in a
quarter system compared to those who have
taught in both a quarter and semester
system when it comes to what system they
prefer.
Ha: There is a difference in the proportion of
teaching faculty that have only taught in a
quarter system compared to those who have
taught in both a quarter and semester
system when it comes to the system they
prefer
Ho: There is no difference in the proportion
of teaching faculty that have only taught in a
quarter system compared to those who have
taught in both a quarter and semester
system when it comes to what system they
thinks benefits students.
Ha: There is a difference in the proportion of
teaching faculty that have only taught in a
quarter system compared to those who have
taught in both a quarter and semester
system when it comes to the system they
think benefits students.
18. No preference Quarter Semester Total
Taught in only
quarter system
16
10.61
2.7398
30
28.80
0.0504
15
21.60
2.0147
61
Taught in both
quarter and
semester system
12
17.39
1.6713
46
47.20
0.0308
42
35.40
1.2290
100
Total 28 76 57 161
Cell contents: Count
Expected count
Contribution to Chi-Square
Pearson Chi-Square = 7.736 DF = 2, P-Value = 0.021, Reject Null Hypothesis
Chi Square Test: Previous System Taught in, Teaching Faculty Prefers
19. No preference Quarter Semester Total
Taught in only
quarter system
25
17.43
3.2892
20
19.70
0.0045
16
23.87
2.5945
61
Taught in both
quarter and
semester system
21
28.57
2.0064
32
32.30
0.0028
47
39.13
1.5287
100
Total 46 52 63 161
Cell contents: Count
Expected count
Contribution to Chi-Square
Pearson Chi-Square = 9.480 DF = 2, P-Value = 0.009, Reject Null Hypothesis
Chi Square Test: Previous System Taught in, Benefits Students
20. Results
The null hypothesis was rejected in both chi
square tests of what does teaching faculty
prefer and what benefits students when
comparing the three responses of no
preference, semesters and quarters.
After running pairwise tests on the data to
determine what differences actually exist,
the only contrast that was shown was
between the responses “no preference”
and “semesters”
Professors who have taught in both systems
show a higher preference for semesters
than expected, but a lower no preference
response than expected
Professors who have only taught in a quarter
system show a higher no response than
expected and lower preference for
semesters than expected
Even though the null hypothesis was rejected
in both cases, there is no evidence that
shows an actual difference between the
system that is preferred or the one that
benefits students
21. Conclusions
Overall, it appears that CSUEB teaching faculty prefer quarters over semesters, while between 12-
22% have no preference.
Overall, it appears that CSUEB teaching faculty think semesters have a slight edge over quarters
when it comes to what system benefits students, while between 23 - 34% have no preference.
The college within CSUEB that a professor teaches in does not appear to affect their preference or
what they think benefits students.
Assistant professors prefer semesters over quarters and also think semesters benefit students more,
we are unsure why. Full professors prefer quarters over semesters, possibly due to the high amount
of time they have spent at CSUEB, a quarter system.
22. More Conclusions
There is no evidence indicating that the previous system that teaching faculty have
been in, has any effect on what system they prefer for themselves or students.
Teaching faculty who have taught only in quarter system probably chose no preference
over semesters because they don’t have enough information on semesters.
Teaching faculty who have taught in both systems chose semesters over no preference
because they feel strongly about semesters.
The addition of the “no preference” category made some our data more difficult to
interpret.
23. Shortcomings
Our response rate was lower than we wanted, around 53%
Some colleges responded more than others. The College of Letter, Arts, and Social
Sciences, and The College of Science had a majority of the responses.
Depending on the rank of the teaching staff, it affected the response rate
If we took out the no preference, then maybe we could have seen a difference in
preference.
One of the questions was misunderstood by 20% of the respondents, it could have
been worded differently
The survey questions could have been randomized to eliminate bias. The order of