admin
Thu, 03/28/2024 - 19:07
Edited Text
STEM vs. Non-STEM Course Performance During COVID
Remote Learning in Higher Education
Kelly Smith and Amanda Goodrick
Slippery Rock University

INTRODUCTION
Objectives
• Investigate how student performance was affected during the
stages of COVID remote learning.
• Determine if the effects were similar for STEM courses and
Non-STEM courses at the 300 level
Data
• Fall 2019 – Spring 2021
• 300 level classes from various departments (see methodology
for full list)
• 4000 rows of data
• 531 Different Courses
• Variables
o Semester
o Course ID
o Course Subject
o Course Number
o Final Grade
o Student Count
o Instruction Method
o Instructional Method Description
Assumptions
• Courses unchanged over time except for instructional mode
• Grades BLANK, F, I, IN, NC, WL, X.
• means unsuccessful completion of the course
• Outcomes were unaffected by
o Course instructor
o Class size

METHODOLOGY

Political Science

Homeland Security

Psychology

Computer Science
Cognitive Science & Leadership

Communication

Interdisciplinary
Programs

Safety Management

Cybersecurity

Criminology &
Criminal Justice

Japanese

Secondary
Education/Foundations of
Education

Environmental Geoscience

Corporate Security

Management

Spanish

Dance

Marketing

Special Education

Economics

Military Science

Sport Management

English

Music

Theatre

Finance

Philanthropy,
Nonprofit,
Leadership & Public
Administration

Engineering
Exercise Science & Rehabilitation
Science
Geography/Environmental Science
Health Care Administration &
Management

Managemen
t
Information
Systems
Nursing
Parks &
Conservatio
n
Physics
Petroleum/N
atural Gas
Engineering
Recreationa
l Therapy
Social Work

0.60
0.50
0.40
0.30
0.20
0.10
0.00

Percent

Chemistry

Mathematic
s

Percent

Biology

A

B

P

C

D

P

and 𝑛1 + 𝑛2 + 1

C
D
F

Mean
Std.Dev.

0.60
0.50
0.40
0.30
0.20
0.10
0.00

F

A
B
P
C
A

B

Letter Grade

P

C

D

F

Letter Grade

D
F

Spring 2020 STEM

Spring 2020 Non-STEM

Statistics

0.60
0.50
0.40
0.30
0.20
0.10
0.00

Percent

Business

History

Philosophy

B

Fall 2019 STEM

A

B

Health Science/Public Health

P

C

D

Mean
Std. Dev.

0.60
0.50
0.40
0.30
0.20
0.10
0.00
B

P

C

D

F

A

Letter Grade

Letter Grade

B

• Data cleaning
o Grades coded as 0 due to unsuccessful course
completion: BLANK, F, I, IN, NC, WL, X.
o Coded grades
• P = 2.1 to be coded uniquely since passing grade is at
least a C. There was not enough information to say this
grade was anything higher than a C.

Fall 2020 STEM

Fall 2020 Non-STEM
0.60
0.50
0.40
0.30
0.20
0.10
0.00
A

B

P

C

D

F

C
D
F
Mean
Std.Dev.
A

4

B

3

P

2.1

C

2

D

1

F

0

• Removed variables after classification into STEM and NonSTEM:
o Course ID
o Course Subject
o Course Number
o Instruction Method
o Instructional Method Description

• Restructured data
o Separated data by semester
o Created a count variable which was the total number of
each letter grade over each semester broken into STEM
and Non-STEM
o Computed relative frequency variable for each semester for
STEM and Non-STEM
o New data set contained letter grades, counts for each
semester broken into STEM and Non-STEM, and Relative
Frequencies for each grade for each semester broken into
STEM and Non-STEM.

Increase

Increase

B

Decrease

Decrease

Decrease

P

Increase

Decrease

Decrease

C

Decrease

Decrease

Equal

D

Decrease

Equal

Equal

F

Increase

Increase

Increase

Spring 2020

Fall 2020

Spring 2021

A

Equal

Equal

Increase

B

Decrease

Equal

Decrease

P

Increase

N/A

N/A

C

Decrease

Equal

Decrease

Fall 2019
Fall 2019
Non-STEM
STEM
Reject?
p-value
n=6430
n=2454
52.29%
55.09%
Yes
0.0088
3362
1352
28.26%
27.18%
No
0.3114
1817
667
2.21%
0%
Yes
0
142
0
10.68%
9.90%
No
0.2816
687
243
2.53%
2.61%
No
0.8454
163
64
4.03%
5.22%
Yes
0.0071
259
128
Two Sample t-test equal variances 𝑯𝟎 : 𝝁𝟏 = 𝝁𝟐
3.22
3.24
No
0.4511
1.03
1.08

D

Decrease

Equal

Equal

Spring 2020
Spring 2020
Non-STEM
STEM
Reject?
p-value
n=6104
n=2651
58.85%
56.92%
No
0.7764
3529
1509
22.80%
22.41%
No
0.4087
1392
594
5.55%
3.62%
Yes
0
339
96
7.40%
6.26%
No
0.0550
452
166
0.39%
0.34%
No
0.7064
24
9
6.03%
10.45%
Yes
0
368
277
Two Sample t-test equal variances 𝑯𝟎 : 𝝁𝟏 = 𝝁𝟐
3.27
3.15
Yes
0
1.09
1.26

Conclusion
NS < S
Equal
NS > S
Equal
Equal
NS < S
Equal
Average

Equal proportions of B,
C, and D. STEM had
more A and F while
Non-STEM had more
P. STEM and NonSTEM have Equal
average grades in Fall
2019.

F

Increase

Equal

Increase

In Spring 2020, STEM students earned more P and F, fewer
B, C, and D, and equal portion of A than in Fall 2019.
In Fall 2020, STEM students earned equal portions of A, B,
C, D, and F as in Fall 2019.
In Spring 2021, STEM students earned more A and F, fewer
B and C than Fall 2019. Students earned equal portion of D.
No P grades were earned in Stem courses in the Spring
2021 academic semester.
All statistical conclusions were significant.

LIMITATIONS

Conclusion
Equal
Equal
NS > S
Equal
Equal
NS < S

Equal proportions of A,
B, C, and D. STEM
had more F and NonSTEM had more P.
STEM and Non-STEM
have unequal average
grades in Spring 2020.

• Data was letter grades – more analysis could have been completed with
quantitative percent grades, which are not accessible.
• Unable to incorporate other confounding factors that may have
influenced course grades such as student health and student
circumstances.
• Number of semesters available in data

Unequal
Average

CONCLUSION

B

P

C

D

Fall 2020
Fall 2020
Non-STEM
STEM
Reject?
p-value
n=6464
n=2285
55.45%
54.40%
No
0.3869
3584
1243
25.05%
26.61%
No
0.1407
1619
608
1.71%
0%
Yes
0
111
0
9.53%
9.45%
No
0.9144
616
216
2.48%
3.15%
Yes
0.04199
160
72
5.79%
6.39%
No
0.2942
374
146
Two Sample t-test equal variances 𝑯𝟎 : 𝝁𝟏 = 𝝁𝟐
3.22
3.19
No
0.3423
1.11
1.14

Conclusion
Equal
Equal
NS > S
Equal
NS < S
Equal

Equal proportions of A,
B, C, and F. STEM has
more D. Non-STEM
has more P. STEM and
Non-STEM have Equal
average grades in Fall
2020.

average grades of 300-level STEM and Non-STEM courses are equal for
Fall 2019, Fall 2020, and Spring 2021. This indicates that student grades in

STEM courses were equivalent to student grades in Non-STEM courses in
300-level courses each of those three semesters, regardless of the
modality of instruction. Additionally, the average grade was higher in NonSTEM courses than STEM courses during the Spring 2020 semester,

Equal Average

indicating that STEM students were impacted more than Non-STEM

F

Letter Grade

Spring 2021 STEM

Spring 2021 Non-STEM
0.60
0.50
0.40
0.30
0.20
0.10
0.00

Percent

A

Increase

students during the semester that instruction switched from in-person to

Percent

Numerical Grade

P

0.60
0.50
0.40
0.30
0.20
0.10
0.00

Letter Grade
Letter Grade

A

The results of the two sample mean t-test for equal variances show that the
A

F

Percent

Art

Hospitality and
Tourism
Management

𝑠𝑝 1ൗ𝑛1 +1ൗ𝑛2

where 𝑠𝑝 =

Spring 2021

STEM with Fall 2019 as control

A

𝑛1−1 𝑆12 + 𝑛2−1 𝑠22
𝑛1 +𝑛2 +1

Fall 2019 Non-STEM

Percent

Arabic

STEM

Physical & Health
Education

French

o Test statistic 𝑡 =

Fall 2020

𝑛1 +𝑛2

+ 𝑛
𝑛1
2

• Two Sample t-test for means with equal variances assumed
o 𝐻0: 𝜇1 = 𝜇2 with 𝛼 = 0.05
𝐻𝑎: 𝜇1 ≠ 𝜇2
𝑥ҧ 1−𝑥ҧ 2

Spring 2020

In Spring 2020, Non-STEM students earned more A, P, and
F and fewer B, C, D than Fall 2019.
In Fall 2020, Non-STEM students earned more A and F and
fewer B, P and C than Fall 2019.
In Spring 2021, Non-STEM students earned more A and F,
fewer B and P than Fall 2019. Students earned equal portion
of C and D. All statistical conclusions were significant.

o Compare STEM and Non-STEM each semester
o Compare STEM and compare Non-STEM across all semesters with Fall
2019 as control
o Compare STEM and Non-STEM overall for all four semesters

Percent

Accounting

𝑝ො𝑝 1−𝑝ො𝑝

RESULTS AND DISCUSSION

Industrial &
Systems
Engineering

Non-STEM with Fall 2019 as control

• Inferences for Two Population Proportions 𝐻0: 𝑝Ƹ1 = 𝑝Ƹ 2
o 𝐻𝑎 : 𝑝Ƹ1 ≠ 𝑝Ƹ 2 if 𝐻0 rejected, 𝐻𝑎: 𝑝Ƹ1 < 𝑝Ƹ 2 or 𝐻𝑎: 𝑝Ƹ 1 > 𝑝Ƹ 2 based on direction of
sample proportions
𝑝ො1−𝑝ො2
(𝑥1+𝑥2)
o Test statistic 𝑧 =
,
where
𝑝
Ƹ
=
with 𝛼 = 0.05
𝑝
1
1

df
o Test for equal average grade for all STEM and Non-STEM comparisons

• This project was IRB approved
o Data was organized into STEM and Non-STEM courses
based on Slippery Rock program guide.
Non-STEM

• Descriptive Analysis
o Frequencies
• Initial Frequencies per letter grade for each semester (STEM and NonSTEM) were investigated
o Graphs
• Bar graphs of each semester’s relative frequencies for STEM and NonSTEM
o Cluster Graphs
• Non-STEM and STEM each semester
• Non-STEM courses by semester
• STEM courses by semester.

A

B

P

C

Letter Grade

D

F

0.60
0.50
0.40
0.30
0.20
0.10
0.00

A
B
P
A

B

P

C

Letter Grade

D

F

C
D
F
Mean
Std. Dev.

Spring 2021
Spring 2021
Non-STEM
STEM
Reject?
p-value
n=6135
n=3032
55.40%
57.85%
Yes
0.0132
3399
1754
24.50%
22.69%
No
0.0562
1503
688
1.14%
0%
Yes
0
70
0
10.17%
8.01%
Yes
0
624
243
2.87%
2.51%
No
0.3184
176
76
5.92%
8.94%
Yes
0
363
271
Two Sample t-test equal variances 𝑯𝟎 : 𝝁𝟏 = 𝝁𝟐
3.21
3.18
No
0.2947
1.13
1.24

online. In general, students in both STEM and Non-STEM courses earned
Conclusion
NS < S
Equal
NS > S
NS > S
Equal
NS < S

Equal proportions of B
and D. STEM has
more A and F while
Non-STEM has more P
and C. STEM and
Non-STEM have Equal
average grades for
Spring 2021.

more A and F grades throughout online learning and fewer average grades
than the last in-person semester of Fall 2019. Prior to COVID-remote

learning and during the fully online learning, STEM and Non-STEM student
grades were not significantly impacted in comparison to one another each
semester. Overall, remote learning did not significantly impact the average
grade distribution for STEM and Non-STEM courses except during the

Equal Average

semester of transitioning to online learning, where STEM grades were
Fall 2019-Spring
Fall 2019-Spring 2021
2021
Non-STEM
STEM
Reject?
n=25133
n=10422
A
55.20%
56.21%
No
13874
5858
B
25.19%
24.53%
No
6331
2557
P
2.63%
0.92%
Yes
662
96
C
9.47%
8.33%
Yes
2379
868
D
2.08%
2.12%
No
523
221
F
5.43%
7.89%
Yes
1364
822
Two Sample t-test equal variances 𝑯𝟎 : 𝝁𝟏 = 𝝁𝟐
Mean
3.23
3.19
Yes
Std.Dev.
1.09
1.19

p-value

Conclusio
n

0.0824

Equal

0.1939

Equal

0

NS > S

0

NS > S

0.8124

Equal

0

NS < S

0.0038

Unequal
Average

Overall, from Fall 2019
through Spring 2021,
STEM students and
Non-STEM students
earned equal
proportions of A, B,
and D. Non-STEM
students earned more
P and C, and STEM
students earned more
F. Unequal average
grades.

more negatively impacted.

FUTURE WORKS





Compare pre-COVID online to during and post COVID online
Same study comparing departments
Investigate 100 and 200 level courses
Investigate Graduate level courses