Bias in MOOC Completion Prediction: A Fairness Analysis Across African Countries and Socio-Economic Contexts
Domain:
educationRecord type:
paper
Creator:
JadSébMélFra
Editor:
BotT. SinOga
Publisher:
Int
Host:
Low completion rates are a well-documented challenge in MOOCs, and Machine Learning (ML)-based predictors have been proposed as a way to detect at-risk learners and support them through personalised interventions. Yet, evidence shows that ML models trained on such data can be biased against certain groups of learners, often the historically discriminated ones. In particular, usage patterns can substantially vary across world regions due to contextual factors, which can affect predictions.
In this work, we analyse ML models trained on a MOOC dataset including African and OECD learners to examine whether, and for whom, biases arise.
Our findings show that learners from Middle and Western Africa and from countries with low HDI and low literacy rates are affected by significant biases in ML predictions.
Our findings highlight the need to systematically evaluate ML unfairness in African educational contexts.
Similar
Socio-economic factors associated with intention to quit across the four African countries.Socio-economic factors associated with previous quit attempts across the four African countries.A Spatial Analysis of COVID-19 in African Countries: Evaluating the Effects of Socio-Economic Vulnerabilities and NeighbouringOn a smart digital gender ecosystem for managing the socio-economic development in the african contextsjoynaomi81/Fairness-and-Bias-Evaluation-of-Transformer-Models-for-Yoruba-Sentiment-AnalysisNaomi-NLP/Fairness-and-Bias-Evaluation-of-Transformer-Models-for-Yoruba-Sentiment-Analysis