Mathematics and incentives in the slums

TL;DR

This paper describes a novel incentive-based math learning program in Kenyan slums using cryptocurrency rewards, which significantly increased participation and improved students' mathematical problem-solving skills.

Contribution

It introduces a cryptocurrency reward system integrated with math drills that enhances engagement and learning outcomes in low-income, resource-limited settings.

Findings

High participation: 1301 students across 11 libraries in one year.

Substantial learning gains: students outperform rote and mechanical learning.

Effective incentive: cryptocurrency rewards motivate students to complete extensive math drills.

Abstract

In response to COVID-19, a new project was started to allow students to solve computerised math drills outside of school. In 9 months it has gone from zero to one thousand students, in co-operation with ten community libraries in various slums and low-income regions in Kenya. The program uses the tutor-web as a study environment and access is provided by donating tablet computers to participating community libraries. Students are rewarded using the SmileyCoin cryptocurrency as they progress through the system and the libraries are free to sell for SMLY small food items, sanitary pads and the tablets themselves. The reward system is designed to put an emphasis on secondary school mathematics. Completion of the corresponding collection of drills gives SmileyCoin awards sufficient to purchase a tablet. Conclusions based on the first year indicate that the resulting effect on participation and performance is unprecedented: Eleven libraries with 1301 students opted for voluntary participation in 2021 causing the program to run at full financial capacity. In that year, 450 students earned enough SMLY to purchase the tablets, which involves completing a large collection of drills to a level of excellence. Status exam questions are dispersed within the drill collection. These independent measurements show learning which surpasses both rote learning and mechanical learning and demonstrate an increase in the general ability to address new mathematical problems not seen before.