Predicting Student Success in College Algebra Classes Using Machine Learning
Open accessAbstract
She has been teaching mathematics at Queensborough Community College, CUNY since 2011.She is always looking for innovative ways to teach her classes to inspire her students and enhance their learning experience.Over the years
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
- Findingown data
Hispanic students had a College Algebra passing rate of 0.43.
“Among these groups, the passing rates are 0.43 and 0.42, respectively.”
Quote in the paperNot a substantial claim- Rests on an assumption: Receiving a grade of C or above is a valid indicator of successful course completion.
- Findingown data
Black students had a College Algebra passing rate of 0.42.
“Among these groups, the passing rates are 0.43 and 0.42, respectively.”
Quote in the paperNot a substantial claim- Rests on an assumption: Receiving a grade of C or above is a valid indicator of successful course completion.
- Findingown data
Differences in passing rates among ethnic groups were statistically significant according to the Chi-Square test.
“The difference between passing rates of different ethnic groups are statistically significant based on the Chi-Square Test.”
Quote in the paperNot a substantial claim- Rests on an assumption: Receiving a grade of C or above is a valid indicator of successful course completion.
- Rests on an assumption: The assumptions of the Chi-Square test of independence are met for the analyzed contingency tables.
- Findingown data
Asian or Pacific Islander students had a College Algebra passing rate of 0.60.
“On the other hand, among the students who have identified themselves belonging to the Asian or Pacific Islander ethnicity groups, the passing rate is 0.60.”
Quote in the paperNot a substantial claim- Rests on an assumption: Receiving a grade of C or above is a valid indicator of successful course completion.
- Findingown data
The dependence of passing rate on semester of enrollment was statistically significant by Chi-Square test.
“This dependence is also found to be statistically significant by Chi-Square Test.”
Quote in the paperNot a substantial claim- Rests on an assumption: Receiving a grade of C or above is a valid indicator of successful course completion.
- Rests on an assumption: The assumptions of the Chi-Square test of independence are met for the analyzed contingency tables.
- Findingown data
First-generation college status did not show a significant dependency with course grades.
“Another student feature which did not show significant dependency with course grades was first generation college status of the students.”
Quote in the paperNot a substantial claim- Rests on an assumption: Receiving a grade of C or above is a valid indicator of successful course completion.
- Findingown data
Students categorized as 'not tested' in math, reading, and writing had the highest average College Algebra passing rate, around 0.60 in each subject.
“The average College Algebra passing rate for students who were categorized as 'not tested' was the highest at around 0.60 in each of the three subjects.”
Quote in the paperNot a substantial claim- Rests on an assumption: Receiving a grade of C or above is a valid indicator of successful course completion.
- Findingown data
In-person sections had a College Algebra passing rate of 0.45.
“The passing rates for in-person sections was 0.45, while it was 0.48 for hybrid and 0.57 for online sections.”
Quote in the paperNot a substantial claim- Rests on an assumption: Receiving a grade of C or above is a valid indicator of successful course completion.
- The authors state this as an assumption: Classifying all Spring 2020 sections as hybrid does not bias the modality analysis.
- Findingown data
There was a positive correlation between students' GPA and College Algebra grades.
“A positive correlation is observed between students' GPA and College Algebra grades as shown in Figure 4b via box plot.”
Quote in the paperNot a substantial claim - Findingown data
A quarter of students withdrew from the College Algebra course.
“We observe that a quarter of students have withdrawn from the course.”
Quote in the paperNot a substantial claim - Findingown data
Students performed slightly better during Spring semesters.
“Lastly, students performed slightly better during Spring semesters.”
Quote in the paperNot a substantial claim - Findingown data
Students who passed the initial math exam had a higher College Algebra passing rate (0.51) than those who failed (0.42).
“Students who 'passed' the initial math exam had a higher passing rate (0.51) in College Algebra compared to students who 'failed' this exam (0.42).”
Quote in the paperNot a substantial claim- Rests on an assumption: Receiving a grade of C or above is a valid indicator of successful course completion.
- Findingown data
Hybrid sections had a College Algebra passing rate of 0.48.
“The passing rates for in-person sections was 0.45, while it was 0.48 for hybrid and 0.57 for online sections.”
Quote in the paperNot a substantial claim- Rests on an assumption: Receiving a grade of C or above is a valid indicator of successful course completion.
- The authors state this as an assumption: Classifying all Spring 2020 sections as hybrid does not bias the modality analysis.
- Findingown data
Online sections had a College Algebra passing rate of 0.57.
“The passing rates for in-person sections was 0.45, while it was 0.48 for hybrid and 0.57 for online sections.”
Quote in the paperNot a substantial claim- Rests on an assumption: Receiving a grade of C or above is a valid indicator of successful course completion.
- The authors state this as an assumption: Classifying all Spring 2020 sections as hybrid does not bias the modality analysis.
- Interpretationreasoning
The results can be used by college administrations to predict student success in College Algebra courses based on student characteristics.
“The results of this study can be used by college administrations to make predictions about student success in College Algebra courses based on student characteristics.”
Quote in the paperSubstantial claim- Rests on an assumption: Historical data from 2017-2021 is representative of future student cohorts for the purpose of predicting success.
- Rests on an assumption: The student characteristics used in the models are available for incoming students at the time predictions are needed.
- Rests on an assumption: The machine learning models' predictions are accurate enough to guide interventions and advisement.
- Interpretationown data
Most of the student and course characteristics analyzed display correlation with College Algebra course outcomes.
“The analysis shows that most of the student and course characteristics analyzed display correlation with the College Algebra course outcomes.”
Quote in the paperSubstantial claim - Interpretationown data
Machine learning algorithms, specifically KNN and decision tree, can be used to predict course outcomes based on students' demographics.
“We conclude that machine learning algorithms, specifically KNN and decision tree algorithms, can be used to predict course outcomes based on the students' demographics.”
Quote in the paperSubstantial claim- Rests on an assumption: The machine learning models' predictions are accurate enough to guide interventions and advisement.
- Interpretationown data
Transfer students tend to perform better than first-time freshmen.
“Transfer students tend to perform better than first-time freshmen.”
Quote in the paperNot a substantial claim
- May combine two claims. 12
- Classified as probably not a substantial claim. 2
May combine two claims.
Findingown dataFemale students had an overall passing rate of 0.51, significantly higher than the 0.44 rate for male students.
“The overall passing rates for female students (0.51) is significantly higher than the passing rate for male students (0.44) which is consistent with the graph.”
Quote in the paperNot a substantial claim- Rests on an assumption: Receiving a grade of C or above is a valid indicator of successful course completion.
May combine two claims.
Findingown dataFirst-time freshmen had a passing rate of 0.48, while transfer students had a passing rate of 0.58.
“Passing rate for first-time freshmen students was 0.48 while it was 0.58 for transfer students.”
Quote in the paperNot a substantial claim- Rests on an assumption: Receiving a grade of C or above is a valid indicator of successful course completion.
May combine two claims.
Findingown dataThe KNN model achieved a test accuracy of 0.842, and the Decision Tree model achieved a test accuracy of 0.849.
“| Accuracy | 0.854 | 0.842 | 0.861 | 0.849 |”
Quote in the paperSubstantial claimMay combine two claims.
Findingown dataBoth KNN and Decision Tree models produced F1 scores of 0.91 and Jaccard indices of 0.83.
“F1 scores were found to be 0.91 while the Jaccard indices were found to be 0.83 for both models.”
Quote in the paperSubstantial claimMay combine two claims.
Findingown dataPassing rates did not differ significantly among students who passed, failed, or were exempt in reading and writing entrance exams, and were around 0.46 for all three categories.
“The passing rates did not differ significantly among students who passed, failed or who were exempt in reading and writing entrance exams, and was around 0.46 for all three categories.”
Quote in the paperNot a substantial claim- Rests on an assumption: Receiving a grade of C or above is a valid indicator of successful course completion.
May combine two claims.
Findingown dataThere was a positive correlation between student age and College Algebra grades, confirmed by Chi-Square test.
“An increase in the average student grades is observed along with the increase in age groups.Chi-Square test confirms this positive correlation.”
Quote in the paperSubstantial claim- Rests on an assumption: The assumptions of the Chi-Square test of independence are met for the analyzed contingency tables.
May combine two claims.
Findingown dataThe average passing rate was 0.46 for Fall semesters and 0.50 for Spring semesters.
“The average passing rate was 0.46 for College Algebra sections taken in Fall semesters while it was 0.50 for sections taken in Spring semesters.”
Quote in the paperNot a substantial claim- Rests on an assumption: Receiving a grade of C or above is a valid indicator of successful course completion.
Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.
Findingown dataMore students were enrolled during Fall semesters, and a slight decline in total enrollment over the years was observed.
“More students have been enrolled during Fall semesters while a slight decline in the number of total enrollment over the years is observed.”
Quote in the paperNot a substantial claimMay combine two claims.
Interpretationown dataThe dependency of course success on student and course features was statistically significant, and their cumulative effect was strong enough for the machine learning algorithms to predict course outcomes with 85 percent accuracy.
“The dependency of course success to student and course features were shown to be statistically significant and their cumulative effect was sufficiently strong for the machine learning algorithms to predict the course outcomes with 85 percent accuracy.”
Quote in the paperSubstantial claim- Rests on an assumption: The machine learning models' predictions are accurate enough to guide interventions and advisement.
Classified as probably not a substantial claim. May combine two claims.
Study design and conductnone givenAll models were evaluated through cross-validation, and the best combination for each model was evaluated against a testing data set for final evaluation.
“All models were evaluated through cross-validation and the combination with the highest accuracy for each model was evaluated against testing data set for final evaluation.”
Quote in the paperSubstantial claimMay combine two claims.
Interpretationown dataFemale students, Asian or Pacific Islander students, older students, and students enrolled in online sections have a higher chance of performing better than their peers.
“Specifically, female students, students of Asian or Pacific Islander origin, older students, students enroll in online sections have a higher chance of performing better compared to their peers.”
Quote in the paperSubstantial claimMay combine two claims.
Interpretationown dataAlthough differences between passing rates in different categories are not very high, the cumulative effect of these features has strong predictive power.
“Although the differences between passing rates in different categories are not very high, the cumulative effect of these features has a strong predictive power.”
Quote in the paperSubstantial claim
The paper relies on this without support.
Historical data from 2017-2021 is representative of future student cohorts for the purpose of predicting success.
“The results of this study can be used by college administrations to make predictions about student success in College Algebra courses based on student characteristics.”
The paper relies on this without support.
Receiving a grade of C or above is a valid indicator of successful course completion.
“A successful completion of the course is considered as receiving a grade of C or above.”
The paper relies on this without support.
The student characteristics used in the models are available for incoming students at the time predictions are needed.
“The results of this study can be used by college administrations to make predictions about student success in College Algebra courses based on student characteristics.”
The paper relies on this without support.
The machine learning models' predictions are accurate enough to guide interventions and advisement.
“We conclude that machine learning algorithms, specifically KNN and decision tree algorithms, can be used to predict course outcomes based on the students' demographics.”
The paper relies on this without support.
The data obtained from the Office of Institutional Research and Assessment is accurate and complete for the variables used.
“Data from students enrolled in the College Algebra courses between the years 2017 and 2021 have been requested from the Office of Institutional Research and Assessment.”
The authors state this as an assumption.
Classifying all Spring 2020 sections as hybrid does not bias the modality analysis.
“All sections that started in-person or hybrid transitioned to online in Spring 2020 semester due to the Covid-19 pandemic, therefore all sections are considered as hybrid in this semester.”
The paper relies on this without support.
The assumptions of the Chi-Square test of independence are met for the analyzed contingency tables.
“We also use Chi-Square Test of Independence to show that these relations are statistically significant.”
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Study period has two different end years
MinorThe introduction says the analyzed records extend through 2022, but the methodology says the requested dataset ends in 2021.
“Data of over twenty thousand students who have taken College Algebra courses Queensborough Community College between the years 2017 and 2022 have been analyzed.”
“Data from students enrolled in the College Algebra courses between the years 2017 and 2021 have been requested from the Office of Institutional Research and Assessment.”
Reported age range conflicts with the data table
MinorThe text describes student ages as ranging from 16 to around 60, whereas Table 1 gives an age range of 13-73.
“The students' ages range from 16 to around 60, shown in bins of two.”
“| Age | 13-73 |”
Cites
Handbook of Distance Education
(2018) · doi:10.4324/9781315296135 · Cited by 795 · see citing works
Predicting Student Performance by Using Data Mining Methods for Classification
Kabakchieva, D. (2013) · Cybernetics and Information Technologies · doi:10.2478/cait-2013-0006 · Cited by 299 · see citing works
Enhancing prediction of student success: Automated machine learning approach
Zeineddine, H., Braendle, U., & Farah, A. (2020) · Computers & Electrical Engineering · doi:10.1016/j.compeleceng.2020.106903 · Cited by 214 · see citing works
Jiao, P., Ouyang, F., Zhang, Q., & Alavi, A. H. (2022) · Artificial Intelligence Review · doi:10.1007/s10462-022-10155-y · Cited by 141 · see citing works
Success in Algebra Among Community College Students
Reyes, C. S. (2010) · Community College Journal of Research and Practice · doi:10.1080/10668920802505538 · Cited by 23 · see citing works
Student Ability Best Predicts Final Grade in a College Algebra Course
O’Connell, K. A., Wostl, E., Crosslin, M., Berry, T. L., & Grover, J. P. (2018) · Journal of Learning Analytics · doi:10.18608/jla.2018.53.11 · Cited by 19 · see citing works
Application of Data mining Techniques for student success and failure prediction
Yehuala, M. A. (2013) · Cited by 5 · see citing works
Flagged first, then by citations
- Authors
- Akcay Ozkan, Z., Yu, Y., & Stelmach, E.
- Affiliations
- City University of New York, Queensborough Community College
- Rutgers, The State University of New Jersey (US)
- ORCID
- Akcay Ozkan, Zeynep
- Venue
- 2023 ASEE Annual Conference & Exposition Proceedings
- Publisher
- ASEE Conferences
- Published
- 2024-02-07
- OpenAlex
- W4391603118
- Cited by
- 0
- References
- 7
- FWCI
- 0.00
- Type
- conference paper
- Language
- en
- Open access
- bronze
- doi.org
- Locations
- ASEE Annual Conference & Exposition (publishedVersion)
- Topics
- Physical Sciences › Computer Science › Computer Science Applications › Online Learning and Analytics
- Physical Sciences › Mathematics › Modeling and Simulation › Mathematics Education and Programs
- Physical Sciences › Mathematics › Statistics and Probability › Statistics Education and Methodologies
- Keywords
- student performance prediction, college algebra, machine learning, mathematics education, community college students, academic performance
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-01.
Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-01.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
