Towards Equalised Odds as Fairness Metric in Academic Performance Prediction
This work addresses fairness in educational AI systems, but it is incremental as it applies existing guidelines to a specific domain without introducing new methods.
The paper tackled the problem of selecting an appropriate fairness metric for academic performance prediction (APP) systems, finding that equalised odds is the most suitable based on guidelines from recent literature and APP's characteristics.
The literature for fairness-aware machine learning knows a plethora of different fairness notions. It is however wellknown, that it is impossible to satisfy all of them, as certain notions contradict each other. In this paper, we take a closer look at academic performance prediction (APP) systems and try to distil which fairness notions suit this task most. For this, we scan recent literature proposing guidelines as to which fairness notion to use and apply these guidelines onto APP. Our findings suggest equalised odds as most suitable notion for APP, based on APP's WYSIWYG worldview as well as potential long-term improvements for the population.