Shawn Sun-Lin
ASUSTek Computer Inc.
The Problem of Bias in Higher Education Even in the Age of Data Mining
On Demand
Session Type:
Breakout Session
Delivery Format:
On-Demand
EDM is a powerful tool for academic interventions since it renders data-driven advice immune to human biases. However, biases still exist in the interpretation of data-mined advice. This could make investments in EDM counterproductive. Major biases detected through experiments and corresponding debiasing techniques will be presented.
Outcomes: Know how human cognitive biases undermine assumed benefits of investments in EDM and lead to hidden costs in higher education * Understand what impacts of biases on EDM applications will have if school policy makers fail to face and act on the issue * Comprehend how to embed debiasing techniques into EDM applications and make them effective
Outcomes: Know how human cognitive biases undermine assumed benefits of investments in EDM and lead to hidden costs in higher education * Understand what impacts of biases on EDM applications will have if school policy makers fail to face and act on the issue * Comprehend how to embed debiasing techniques into EDM applications and make them effective