Algorithmic Fairness, Bias, and Justice

I completed a final paper for my Algorithmic Fairness, Bias, and Justice course titled “Ethical Algorithms in Education: A Critical Analysis of Ofqual’s A-Level Grading Solution.”

The paper examines how Ofqual deployed an algorithm in 2020 to estimate A-level grades when exams were cancelled due to COVID-19. Ofqual’s algorithm has been widely reported as biased and unfair in its treatment of disadvantaged students and schools, and my paper focuses on issues from the problem specification phase.

Key course insights included:

1. Society must evaluate which applications of machine learning to encourage or discourage.

2. Statistical patterns valid for majority groups may not apply to minority populations.

3. Fairness implementation can be computationally demanding — achieving equitable outcomes across different groups might require complex, non-linear decision rules rather than simple linear classifiers, creating algorithmic challenges without efficient solutions currently available.