Towards Equity and Algorithmic Fairness in Student Grade Prediction

Equity of educational outcome and fairness of AI with respect to race have been topics of increasing importance in education. In this work, we address both with empirical evaluations of grade prediction in higher education, an important task to improve curriculum design, plan interventions for academic support, and offer course guidance to students. With fairness as the aim, we trial several strategies for both label and instance balancing to attempt to minimize differences in algorithm performance with respect to race. We find that an adversarial learning approach, combined with grade label balancing, achieved by far the fairest results. With equity of educational outcome as the aim, we trial strategies for boosting predictive performance on historically underserved groups and find success in sampling those groups in inverse proportion to their historic outcomes. With AI-infused technology supports increasingly prevalent on campuses, our methodologies fill a need for frameworks to consider performance trade-offs with respect to sensitive student attributes and allow institutions to instrument their AI resources in ways that are attentive to equity and fairness.

[1]  Chris Piech,et al.  Achieving Fairness through Adversarial Learning: an Application to Recidivism Prediction , 2018, ArXiv.

[2]  Zhe Zhao,et al.  Data Decisions and Theoretical Implications when Adversarially Learning Fair Representations , 2017, ArXiv.

[3]  Toniann Pitassi,et al.  Fairness through awareness , 2011, ITCS '12.

[4]  Kevin G. Welner,et al.  Closing the opportunity gap : what America must do to give every child an even chance , 2013 .

[5]  Erez Shmueli,et al.  Algorithmic Fairness , 2020, ArXiv.

[6]  Christopher Brooks,et al.  Evaluating the Fairness of Predictive Student Models Through Slicing Analysis , 2019, LAK.

[7]  Horacio Matos-Díaz,et al.  Do student evaluations of teaching depend on the distribution of expected grade? , 2010 .

[8]  Tom Routen,et al.  Intelligent Tutoring Systems , 1996, Lecture Notes in Computer Science.

[9]  Zachary C. Lipton,et al.  Algorithmic Fairness from a Non-ideal Perspective , 2020, AIES.

[10]  Charlton D. McIlwain.  Computerize the Race Problem?: Why We Must Plan for a Just AI Future , 2020, AIES.

[11]  Nathan Srebro,et al.  Equality of Opportunity in Supervised Learning , 2016, NIPS.

[12]  George Chen,et al.  Measuring financial implications of an early alert system , 2016, LAK.

[13]  C. Weaver,et al.  A more explicit grading scale decreases grade inflation in a clinical clerkship. , 2007, Academic emergency medicine : official journal of the Society for Academic Emergency Medicine.

[14]  Xia Hu,et al.  Fairness in Deep Learning: A Computational Perspective , 2019, IEEE Intelligent Systems.

[15]  Jon M. Kleinberg,et al.  Inherent Trade-Offs in the Fair Determination of Risk Scores , 2016, ITCS.

[16]  Zachary A. Pardos,et al.  Time slice imputation for personalized goal-based recommendation in higher education , 2019, RecSys.

[17]  Allison Woodruff,et al.  Putting Fairness Principles into Practice: Challenges, Metrics, and Improvements , 2019, AIES.

[18]  Christoph Bartneck,et al.  Robots Can Be More Than Black And White: Examining Racial Bias Towards Robots , 2019, AIES.

[19]  Toon Calders,et al.  Building Classifiers with Independency Constraints , 2009, 2009 IEEE International Conference on Data Mining Workshops.

[20]  Iris Howley,et al.  Assessing Post-hoc Explainability of the BKT Algorithm , 2020, AIES.

[21]  Emma Brunskill,et al.  Fairer but Not Fair Enough On the Equitability of Knowledge Tracing , 2019, LAK.

[22]  Eitel J. M. Lauría,et al.  Early Alert of Academically At-Risk Students: An Open Source Analytics Initiative , 2014, J. Learn. Anal..

[23]  Zachary A. Pardos,et al.  Goal-based Course Recommendation , 2018, LAK.

[24]  Shayan Doroudi,et al.  Towards Accurate and Fair Prediction of College Success: Evaluating Different Sources of Student Data , 2020, EDM.

[25]  Toniann Pitassi,et al.  Learning Adversarially Fair and Transferable Representations , 2018, ICML.

[26]  Toon Calders,et al.  Data preprocessing techniques for classification without discrimination , 2011, Knowledge and Information Systems.

[27]  Ramesh Johari,et al.  How a data-driven course planning tool affects college students' GPA: evidence from two field experiments , 2018, L@S.

[28]  Mitchell L. Stevens,et al.  AI and Holistic Review: Informing Human Reading in College Admissions , 2019, AIES.

[29]  Sidney K. D'Mello,et al.  Evaluating Fairness and Generalizability in Models Predicting On-Time Graduation from College Applications , 2019, EDM.

[30]  George Karypis,et al.  Feature Extraction for Next-Term Prediction of Poor Student Performance , 2019, IEEE Transactions on Learning Technologies.

[31]  Saif Mohammad,et al.  Examining Gender and Race Bias in Two Hundred Sentiment Analysis Systems , 2018, *SEMEVAL.

[32]  Xing Xie,et al.  Fairness-aware News Recommendation with Decomposed Adversarial Learning , 2020, AAAI.

[33]  Jieyu Zhao,et al.  Balanced Datasets Are Not Enough: Estimating and Mitigating Gender Bias in Deep Image Representations , 2018, 2019 IEEE/CVF International Conference on Computer Vision (ICCV).

[34]  Kris D. Gutiérrez,et al.  Social Design Experiments: Toward Equity by Design , 2016, Cultural-Historical Activity Theory Approaches to Design-Based Research.

[35]  G. Wyness,et al.  Grade Expectations: How well can we predict future grades based on past performance? , 2020 .

[36]  Blake Lemoine,et al.  Mitigating Unwanted Biases with Adversarial Learning , 2018, AIES.

[37]  Hansol Lee,et al.  Algorithmic Fairness in Education , 2020, ArXiv.

[38]  Matthew D. Pistilli,et al.  Course signals at Purdue: using learning analytics to increase student success , 2012, LAK.

[39]  Zachary A. Pardos,et al.  Connectionist recommendation in the wild: on the utility and scrutability of neural networks for personalized course guidance , 2018, User Modeling and User-Adapted Interaction.

[40]  Franco Turini,et al.  Discrimination-aware data mining , 2008, KDD.

[41]  Krishna P. Gummadi,et al.  Accounting for Model Uncertainty in Algorithmic Discrimination , 2019, AIES.

[42]  Pratik Gajane,et al.  On formalizing fairness in prediction with machine learning , 2017, ArXiv.

[43]  Sergey Karayev,et al.  Grades are not Normal: Improving Exam Score Models Using the Logit-Normal Distribution , 2019, EDM.