An economic evaluation of educational interventions in the LOMLOE: Proposals for improvement with Artificial Intelligence.

Abstract

This research aims to demonstrate the need for an economic evaluation of the Organic Law 3/2020, of 29 December, which amends Organic Law 2/2006, of 3 May, on Education (LOMLOE), especially after the investment of EU Next Generation funds that open new opportunities that were lacking in the initial drafting of the law. The challenge for Public Administrations is to use this additional investment efficiently. Our analysis shows that artificial intelligence models can predict whether educational support programmes will help increase the likelihood that students who lag behind will pass the 4th grade of ESO. In this way, we can calculate the social return of these programmes and contribute to their ex-ante design to achieve higher success rates for students. To complement the models already used by public administrations, we use robust Machine Learning (ML) models such as CHAID decision trees and artificial neural networks to analyse the characteristics of the groups of students and the intervention they have been part of. The conclusions allow us to improve educational reinforcement programmes in the coming years to support students with lower chances of academic success.