Studies in recent years have documented the detrimental effects that low-achieving students can have on the academic performance of their classmates. This is of concern everywhere, but particularly so in Latin America and the Caribbean where test results for reading, math, and science lag considerably behind those of OECD countries and where most students need to improve their learning, rather than have it further handicapped.
To improve the abilities of low achievers, the IDB together with the Secretary of Education of Manizales, a medium-sized city in Colombia and Fundación Luker, provided a remedial literacy program for third graders struggling to read at the appropriate age-level in a sequence of experiments running between 2015 and 2017. Reading is a pillar of all types of learning, and even before the COVID-19 pandemic more than half of 10-year-old children in Latin America and the Caribbean were suffering from serious reading problems, unable to read and understand a small written paragraph. The experiments, involving 2,000 mostly poor children at 90 different schools in Manizales, sought to correct the problem at a local level.
Using small, cost-effective tutorial sessions and a structured phonetics approach, the first time the programs was implemented improved the literacy skills of low achieving participants by 35% compared to the average student in just 12 weeks. In the next two implementations of the program—this time of 16 weeks each and with tweaks to improve instruction—, the gains were three times higher. That was greater progress than made by other popular interventions for literacy improvement, like introducing computers and decreasing class size.
Significant Gains in Learning for Higher Achieving Students
But that was not all. The program not only improved the performance of low-achieving students. An evaluation run as a by-product of the program showed that raising the level of the poorest performers also increased that of their higher achieving peers, who attained around 30% of the considerable learning gains of those benefitting from the tutorials. Indeed, looking only at the progress of those who got remedial instruction, would miss nearly half of the overall benefits of the program, which substantially aided both categories of students.
The reasons for this are not entirely clear. But the evidence suggests that a reduction in classroom disruptions may explain some of the benefits for higher achieving students. When low achieving ones fail to understand something, they may continually raise their hands, force teachers to give them extra help, or act out in fits of boredom. Increasing their literacy level can ameliorate those problems in the interests of all. It can also make peer-to-peer collaboration on projects more fruitful, similarly benefitting those students whose level was higher at the outset.
The Importance of Peer Effects
What is clear is that the results are not due to a smaller class size, a change in teaching strategies, or a change in classroom composition. The experiment was carefully designed so that the number of students taken out of normal classroom instruction for extra tutorial help was small. Moreover, the students were taken out of different classrooms during different subject instruction, but the improvements in the high achievers’ performance occurred only in reading classes, not in math or other areas, where class size was temporarily reduced as well—proof that class size was not a significant actor. Teachers also did not interact with tutors or change their strategies, and classroom composition was not altered.
A Scalable Experiment for the Benefit of Society as a Whole
The key elements instead appear to have been peer effects. Improvement in the performance of lower achievers, while critical in helping them succeed in school and the workplace also is critical for every student—something that both parents and policymakers should keep in mind. With the program in Colombia costing $89 for every student tutored, it is eminently scalable and would be good for society as a whole.