The goal of the project is to establish, using a Randomized Control Trial (RCT) and a well-defined metric based on Principal Stratification Analysis, whether teachers can give "better" recommendations with the assistance of algorithmic information based on standardised test scores, specifically designed to help them in performing this task.
The RCTs will consist of randomly extracting from the population of schools in the Italian provinces of Firenze, Livorno, Grosseto, and Trento, two sets of treated and control schools. In treated schools, teachers at the beginning of the 8th grade will be offered information about students, including their national and local ranking, based on Invalsi standardised test scores obtained during primary education.
The implications for policy of the results obtained through this project are potentially very relevant. Since tracking is the established way to organise high school studies in continental Europe, it is surprising that no experimental evidence exists to guide educational policymakers in the design of the best system of track advising.
In reacting to receiving the grant, Professors Andrea Ichino and Fabrizia Mealli said: "We are delighted that our project has been selected by Unicredit, and we hope that our results will provide useful information to policymakers. We started thinking about this project long ago when we observed what was happening in our children's school. We would also like to thank the EUI Research Council for having supported in 2022 the preparatory work that led to this project."
Learn more about the project.