- An experiment in 60 schools in Ayacucho, Peru, showed that phone tutoring improved the math skills of secondary school students.
- This first experimental evaluation of this type of tutoring shows that governments can strengthen students’ math skills and confidence using only a basic mobile phone and content aligned with the national curriculum.
- Since 2021, the IDB has supported tutoring programs in 17 countries in Latin America and the Caribbean, now delivered both remotely and in person.
Will robots teach us math? Today there are more than 1,600 artificial intelligence tools aimed at education, and tutoring is one of the fastest-growing categories. But we still know little about their effectiveness in Latin America, a region where only six in 10 schools have internet access.
Math is the subject where Latin America lags furthest behind: in PISA 2025, the region scored an average of 370 points, compared with 463 in the countries of the Organisation for Economic Co-operation and Development (OECD), a gap equivalent to about five years of schooling. Behind those numbers are students who carry learning gaps from primary school and who, over time, start to believe that math is not for them. That belief also holds back learning.
In this context, it's worth keeping in mind what already works with much simpler technology. In Ayacucho, an Andean region of Peru, nine in 10 secondary school students have a cell phone, but more than four in 10 have no internet at home. For them, tutoring came by phone: a person on the other end of the line, once a week, patiently reviewing math operations.
Desafía-T a Aprender, a program led by Peru's Ministry of Education with technical and financial support from the Inter-American Development Bank (IDB) and GRADE, shows that phone tutoring can address both at once. Its evaluation, published in the International Journal of Educational Development, found that the program improved math skills among first- and second-year secondary students and strengthened their growth mindset, defined as the conviction that one can improve through practice and effort.
This is the first time an experiment has assessed the effect of phone tutoring on this skill. For governments, the finding has a practical implication: with a basic phone and content aligned with the national curriculum, a single program can strengthen both basic math skills and students' confidence.
Personalized Attention, Even at a Distance
Tutoring is one-on-one or very small group instruction led by someone trained to support the student. The tutor starts from what each student knows rather than from what their grade level dictates: if a secondary school student hasn't yet mastered place value, that's where they begin. Because sessions are frequent, each one picks up where the last left off.
Session by session, a relationship forms, and when students feel that someone is following their progress, they dare to make mistakes and try again. One of the program's tutors saw this in her students: "The girls say they're remembering what they worked on before, and that it's helping them participate in class with more confidence."
Desafía-T a Aprender: An Experiment in 60 Schools in Ayacucho
To measure the program's effect, the evaluation randomly selected 60 schools in Huamanga, Huanta and La Mar: 30 received tutoring and 30 did not. More than 2,000 students were assessed before and after the program.
Over eight weeks, a team of 54 tutors called each student once a week to work on place value, basic operations and problem solving. The Ministry of Education designed the program together with GRADE and the IDB, aligned with its curriculum and its materials on soft skills.
Students who took part in the tutoring made greater progress in math than their peers in schools that did not receive the program. The gain is modest, but encouraging for an intervention of just eight calls.
The most novel finding concerns growth mindset. The program had two groups: one received math tutoring only, and the other, made up exclusively of girls, received three additional sessions focused on this skill.
Overall, the change in growth mindset was nearly twice the gain in math, and it was concentrated among the girls who received those sessions. This doesn't always happen: a similar tutoring program in Spain improved math scores but did not change students' growth mindset.
Math and Confidence, in the Same Call
In math, believing you can improve carries a lot of weight. A mistake can be read as proof that you're "just not good at this," while those who believe they can learn persist longer with a difficult problem.
A 12-year-old first-year secondary student put it this way to her tutor: "I feel like I'm starting to understand. I had no idea about some things in multiplication. I'm getting more confident in myself." Existing tutoring programs could add this component at low cost.
From Pandemic Response to a Regional Tutoring Agenda
Tutoring is part of an agenda the IDB has been advancing since 2021. It began during COVID-19 school closures, with phone tutoring pilots in reading and math evaluated through experiments.
Those pilots showed learning gains, with similar benefits for girls and boys and for students from different socioeconomic backgrounds. Since then, the programs have reached more than 17 countries in the region, with more than 2,500 tutors trained and more than 50,000 students. Today they are also delivered in person, are part of IDB loans to countries, and are beginning to use artificial intelligence to give feedback to tutors.
In Ayacucho, the technology was a cell phone without internet. The results came from a person who called every week, followed each student's progress, and showed them they could learn more than they thought.
Today, together with ministries of education, we are designing and implementing similar experiments that use artificial intelligence in Honduras, Uruguay, Costa Rica, Ecuador and Bolivia. As we bring artificial intelligence into tutoring, that human connection remains the starting point.