The most alarming effects of the COVID-19 pandemic have nothing to do either with health or infections. It’s that schools have been closed around the globe, affecting the education of approximately 1.6 billion children.

Schools were closed in the majority of countries for more than 4.5 months after the pandemic. This was two years ago. According to the United Nations cultural organisation UNESCO, one in ten countries had shut down schools for more that nine months. Millions of children across the globe had not returned. Although data are still being collected, they are beginning to confirm what everyone was afraid: That the poorest and most disadvantaged children face the greatest setbacks in their learning. It is well known that learning loss can leave lasting scars. This will likely lead to lower incomes and lost opportunities for many years to come.

While many countries want to reduce these losses, it would be foolish to return schools to their normal business operations. They should instead use this opportunity to improve education and teaching systems, as informed by research. Today’s Feature reveals that researchers have accumulated large amounts of evidence, including randomized controlled trial results, which point to cost-effective ways of improving learning and attendance in schools, both in low- and high income countries. These strategies include information for parents and children on the long-term benefits and how to help children understand what they are reading, including how to help them comprehend what they are reading, involve parents in education, give feedback to children and help students plan and evaluate their learning.

This research is often ignored by educators, and even worse by policymakers who think they know best. Children can recover from the educational trauma caused by the pandemic if evidence-based insights are put to use in classrooms all over the globe. It could also help to strengthen whole education systems that were failing children long before COVID-19. Conflict, poverty, and politics are all factors that prevent many children from obtaining education, such as the wars in Ukraine or the Taliban’s exclusion of girls from school in Afghanistan.

Evidence-informed education advocates must be open to the limitations of research. The huge differences in schools and classrooms across countries is a major challenge. If the tutoring program is not effective in one school, it might not work in another. Educational research is a guide, but it does not guarantee that something will work in a particular classroom or for every child. Teachers should be involved in the research process and in the application of results.

The problem with education research is its separation from practice. Most educational researchers don’t teach, and most teachers are not exposed to or involved in research. This is similar to medicine where doctors learn about research during training and then consult evidence-based guidelines as they practice. This isn’t the case everywhere. Teachers’ professional development in Japan and China includes the study of how effective lessons are. This approach should be replicated in other countries.

There is growing evidence that shows how to best bring research insights into the classroom. It is not common for educators to be forced into new ways of thinking by top-down approaches. Rukmini Banerji is the leader of Pratham, an educational nongovernmental organization in New Delhi. She believes that a better approach is to encourage students and teachers to explore evidence-based approaches for their own learning.

Some cases saw the disruption of COVID-19 as a catalyst for new ways of thinking about education and how to work in it. Schools created digital delivery methods for lessons, teachers were more involved in the children’s emotional and social health, and parents were more engaged in the learning of their children at home. These innovations have had a rapid impact on education, but it has not been studied well. Schools and researchers should make use of all data available and gather more information, in order to preserve innovations that have helped, both for children’s education and overall.

To show the long-term effects of missing schooling and other consequences of the pandemic, it is important to track the children in a particular cohort. Innovations and catch-up programs should be evaluated with rigorous research so that data can be retrieved when learning is being reintroduced.

Some people suggest that children may be able to quickly bounce back from COVID-related school closings by engaging in a learning spurt. The reality is that children who are better off will recover faster, while those in need will be more likely to succeed. To help children today and to build education systems for tomorrow, we must first focus on those most marginalized or disadvantaged.