Teaching and Learning in Quechua: Evidence from Peru for Latin America and the Caribbean

Bilingual intercultural education coexists with two realities: communities where the indigenous language is still alive, and communities where it has nearly been lost. New evidence shows that schools can improve learning outcomes and boost the use of an indigenous language like Quechua in places where it has weakened. The findings come from two experimental evaluations in schools in Peru.

In early childhood, the evaluation of the Mate Raymi program, designed by the Inter-American Development Bank (IDB), Innovations for Poverty Action (IPA), and Peru's Ministry of Education (MINEDU), provides new evidence on the potential of early education programs to strengthen both math learning and cultural identity in indigenous contexts. Together with IPA, the IDB also contributed to the analysis and evaluation of the program through studies on its impact on learning. The project was supported financially by the Old Dart Foundation and the Japan Special Fund Poverty Reduction Program.

In primary and secondary schools, the Quechua Collao app brought the indigenous language back into everyday life, and the most revealing finding was where it reappeared: in public spaces, outside the home. The IDB supported the development of the app and contributed to measuring the initiative's impact on Quechua learning and use.

An interesting aspect of both initiatives is that they combine technological solutions that are feasible in rural settings while also bringing the traditional knowledge of indigenous peoples into the classroom.

On the International Day of the World's Indigenous Peoples, commemorated every August 9, we highlight initiatives that are making a difference in bilingual intercultural education.

The Starting Point: Bilingual Intercultural Education in the Region

In Peru, bilingual intercultural education is a nationwide policy. The Ministry of Education's official records count 26,422 schools of this type, serving 1,203,697 children and adolescents, with 48 indigenous languages recognized as languages of instruction. Quechua is the most widespread; approximately 14% of the population over age five has Quechua as their mother tongue, according to Peru's National Institute of Statistics and Informatics (INEI).

Behind that figure lie very different situations. The Ministry of Education's own model distinguishes two approaches based on the vitality of the language in each community:

Strengthening: This approach is designed for students whose mother tongue is indigenous. In these contexts, the goal is to ensure that students learn school content through a language they already master, while also developing reading and writing skills in it.

Language revitalization: This approach targets students who, despite belonging to an indigenous people, no longer actively speak the language. In these cases, the goal is to reintroduce the heritage language and promote its relearning in contexts that recognize its symbolic and cultural value.

This dilemma extends beyond Peru's borders. Quechua is also spoken in Bolivia, Ecuador (where it is written as Kichwa), Argentina, Colombia, and Chile. Far from being a single language, Quechua comprises a family of closely related linguistic varieties spread across the Andes. Altogether, an estimated 12 million people speak one of these varieties, making Quechua the most widely spoken indigenous language in the Americas. As a result, several Latin American countries implement bilingual intercultural education policies (or "intercultural bilingual education," as it is called in Guatemala) that face a similar question: how to ensure students learn in a language they speak well where indigenous languages remain alive, and how to recover those languages where they are being lost.

Until recently, there was little causal evidence on which type of intervention works for each case. The two evaluations below begin to fill that gap.

Where the Language is Still Alive: Mate Raymi Boosts Math Learning

In communities in Peru such as Apurímac, Ayacucho, and Huancavelica, the Mate Raymi program brought mathematics to the early childhood level through audio programs in Chanka Quechua, accompanied by contextualized materials and teacher training and follow-up.

The program, developed by the IDB and MINEDU and evaluated by IPA, uses interactive radio with 60 bilingual, intercultural radio lessons that serve as guides for teachers and students. It also includes a package of teaching materials, such as guides, number cards, and intercultural illustrations, designed to reflect the Chanka Quechua worldview.

The intervention also includes monthly classroom visits and weekly phone calls to support teachers as they implement the initiative. Finally, the program carried out a systematization of traditional ethnomathematics knowledge and, based on this, produced a curriculum for early childhood education.

The evaluation — detailed in our publication Intercultural Bilingual Education and Early Childhood Mathematics Learning: RCT Evidence from the Mate Raymi Program in Quechua Chanka Schools in Peru — was a randomized controlled trial with 2,891 children from 350 schools that compared two versions of the curriculum: a bilingual one (teaching math in Spanish and Quechua) and an intercultural bilingual one, which additionally incorporates mathematical concepts from the Andean worldview, such as counting with the yupana.

A yupana is a mathematical calculation tool of Inca origin, used to perform arithmetic operations of varying complexity. Generally described as an "Inca abacus," it comes in various forms, ranging from trays or boxes of different heights to boards subdivided into grids, similar to a chessboard.

Where the Language Was Nearly Lost: What Happens When an App Helps You Learn Quechua Collao?

In Cusco and Puno, where many students belong to Quechua peoples but no longer speak the language, the Quechua Collao app aimed to boost its learning as a second language among sixth-grade primary and first-year secondary students. Integrated into the school day with pedagogical guides and teacher support, the app was evaluated with 4,464 students in 200 schools. The initiative was carried out in partnership with a range of actors. In addition to the IDB, the Organization of Ibero-American States led its implementation in schools; Innovations for Poverty Action (IPA) supported the project's evaluation; and MINEDU, through its Directorate of Bilingual Intercultural Education, served as a strategic institutional partner, providing technical and institutional support for the initiative's implementation, validating the instruments in Quechua Collao, and coordinating its execution with participating schools. The project received financial support from the IDB's Japan Poverty Reduction Fund.

Students improved their Quechua skills: 0.137 SD on a general test and 0.330 SD on a test aligned with the app's content, with the greatest gains in reading comprehension. The most revealing finding was where the language's use increased. Outside the classroom, Quechua was used more in public spaces (0.187 SD) and with people outside the family (0.176 SD), with no change within the household. The school expanded the spaces where the language is spoken, extending it to the market and the street. The effect was greatest among those who spoke the least Quechua at the start, in line with the goal of revitalization.

What Works to Revitalize Indigenous Languages

Both studies show that technological solutions like apps or radio can support learning, for example by helping teachers close gaps in their training and facilitating the delivery of new content in bilingual settings. However, technology alone is not enough.

In Mate Raymi, the study suggests that the effects stem from improvements in teaching practices enabled by interactive radio and the delivery of content that was not previously covered by the traditional curriculum. An important factor was teacher support, which was valued by 95% of participants.

In Quechua Collao, the importance of support was even more evident: in 91% of classrooms, most sessions were led by specialized trainers from the Organization of Ibero-American States, the project's implementing partner, because many teachers did not speak Quechua fluently. Of the 250 teachers in the treated classrooms, 61 were not even listed in the National Registry of Native Language Teachers.

This is, above all, a policy signal. Educational technology delivers results when paired with a sound pedagogical approach and well-trained teachers. Investing in teacher training for bilingual intercultural education is, in large part, the condition for everything else to work.

What the Evidence on Bilingual Intercultural Education Still Doesn't Answer

The results are encouraging, and at the same time, they define the limits of what these interventions can achieve on their own.

• In Quechua Collao, students' self-reported perceptions of their own language did not change: using Quechua more in public is not, by itself, enough to change how much value people place on their language.
• In Mate Raymi, the intercultural bilingual version did not improve conventional math performance and had a small negative effect, of 0.12 SD, on Spanish oral comprehension, which the authors caution should be interpreted carefully. However, a similar program in Panama showed results in both conventional and intercultural math, with no negative effect on Spanish.

This suggests that further research is needed into design and context factors that could help fine-tune these programs for greater impact.

These findings provide new evidence for bilingual intercultural education in Latin America and the Caribbean and offer concrete lessons on how to strengthen Quechua, revitalize indigenous languages, and improve learning outcomes across diverse contexts.

On the International Day of the World's Indigenous Peoples, these two evaluations offer a concrete lesson for education policy. Schools can improve math learning when they teach in a language students speak well, and they can help recover a heritage language where it has weakened, as long as the intervention is well designed and well supported. For the first time, there is causal evidence to answer a question shared by education systems across Latin America: what works in each context.

Note: I would like to thank Juan Hernández-Agramonte, Senior Director of the Embedded Labs program at IPA (Innovations for Poverty Action), for his valuable contribution to this blog.