School, Better Measures for Better Decisions

Public debate on Italian schools often focuses on resources, curricula, and technology. These are crucial issues, but there is a deeper level at which a decisive part of the system's fairness is played out: the way decisions are made. 

School is not only a place of learning, but also an institution that distributes opportunities. Every grade, every piece of career advice, every classroom dynamic contributes to defining educational and professional trajectories. However, many of the forces that influence these choices operate through cognitive mechanisms that are subject to systematic biases. In a system that relies heavily on discretion, these mechanisms can have unintended effects on students' opportunities.

When data reveals isolation

This is why data is an indispensable resource: it does not replace human judgment, but rather makes it more informed. In a study conducted in Italian primary schools, a map of relationships between classmates was created together with teachers, asking children to indicate their friends. The graphic representation revealed a higher percentage of unnamed pupils than teachers had imagined. Making isolation visible changed behaviors: in the following months, both the number of children without friends and antisocial behaviors decreased.

Assessments, stereotypes, and hidden gaps

Large-scale analysis of administrative data shows that students with a migrant background receive lower average grades than their Italian peers, even when they achieve the same results in INVALSI tests. The gap is wider among high-achieving students. When teachers are provided with tailor-made information about their implicit stereotypes, measured using established psychological tools, the gap in assessment narrows. Here too, the data acts as a mirror.

A similar bias emerges in career choices at the end of middle school: even with equally high grades in mathematics, girls are less frequently advised to pursue scientific studies than their male peers, despite showing comparable performance later on. Expectations and representations affect future opportunities.

Algorithms as a support, not a replacement

To address these distortions, the point is not replacing professional judgment with automation, but rather supplementing it with rigorous information tools. Ongoing experiments are evaluating algorithmic indicators based on past results or objective benchmarks, with the aim of making the exercise of discretion more informed.

In a context where technology is making strong headway in education systems, the decisive question is not whether to use it, but how to do so. An algorithm can cement inequalities if it incorporates distorted data, but it can also act as a corrective if designed to compensate for systematic errors or signal aggregate imbalances.

Data as democratic infrastructure

Relying solely on individual sensitivity means ignoring cognitive limitations; blindly relying on technology means neglecting the complexity of educational judgment. The most robust solution integrates professionalism and data, autonomy and responsibility.

A modern school accepts measuring its results, questioning its practices, and introducing tools capable of reducing predictable distortions. Data is not an accessory technical element, but a democratic infrastructure that transforms subjective perceptions into shared evidence. If schools want to remain the main driver of social mobility in the country, they must also invest in the quality of the decisions they make every day.