Uruguay's Machine-Learning Early-Warning Model for Primary-School Repetition Risk (Panel-EIT / ANEP)

Uruguay's Panel-EIT study tracked 15,529 children from early-childhood assessment in 2016 through primary grades 1–3 (2017–2019). Researchers from Universidad de la República, working with ANEP-CODICEN administrative data, built supervised machine-learning models to predict which students would repeat a grade.

The research aimed to identify students at risk of grade repetition early enough to inform early-warning and intervention policy, and to test which early-childhood indicators best predict later repetition.

Researchers trained and compared several supervised models (logistic regression, Bayesian networks, classification trees) on the cohort's administrative and early-childhood assessment data, then tested how model performance changed when the early-childhood assessment score was removed.

The best models achieved 80% accuracy and an AUC of 0.81, with sensitivity of 62–64% and specificity of 83%; 16.2% of the cohort repeated at least one grade. Removing the early-childhood assessment score dropped accuracy to 70–72% and sensitivity as low as 42%, confirming it as the strongest single predictor.

The peer-reviewed authors explicitly caution that the models show statistical association rather than causal mechanisms over a narrow three-year window, and the Inter-American Development Bank's own fAIrLAC pilot page leaves this initiative's implementation status blank — so it is best read as a validated research prototype rather than a proven, actively operating national early-warning system.