Teaching that works: Lessons from 3,000 classrooms

We know millions of children aren't learning in school. But why? What teaching practices make a difference? How does teachers' knowledge impact student learning? And crucially, how can education systems better support their teachers?

A new study—What’s at Play? Unpacking the Relationship Between Teaching and Learning—draws on unique data from the World Bank's Global Education Policy Dashboard (GEPD) to offer insights into these questions. For the first time, we have comprehensive, nationally representative data from 13 education systems that connects the dots between teaching practices, student learning, and system support. And this is just the beginning. Thanks to a collaboration with the LEGO Foundation through the Foundational Learning Compact (FLC), the GEPD has reached 18 education systems and has 10 more under implementation, which means future research will be able to cover nearly a quarter of WB client countries.

The study combines:

• Classroom observations of nearly 3,000 teachers;
• Learning assessments of 50,000 students;
• Content knowledge tests of 12,000 teachers; and
• Extensive interviews with teachers, principals, and education officials

This rich dataset allows us to answer three fundamental questions:

1. How do teachers' pedagogical practices and content knowledge affect student learning?
2. What role does student engagement play in learning outcomes?
3. What system-level factors help create better teaching quality?

The timing is critical. As education systems worldwide grapple with learning losses from the pandemic, understanding what works in teaching – and how to support it effectively – has never been more important.  The findings offer clear guidance for policymakers seeking to improve learning outcomes through better teaching.

The current crisis

Our findings reveal a troubling reality: Students in our study averaged just 65% correct answers on literacy assessments and 40% on mathematics. While most can handle basic tasks like recognizing letters, they struggle with grade-appropriate material like understanding simple stories. Even more concerning, teachers themselves often struggle with the content they teach, averaging only 54% on fourth grade-level assessments. Teaching quality, as measured through Teach Primary, also shows substantial room for improvement: In most education systems, the average teacher scored below 3 on a 1-to-5 scale of teaching practices.

Better teaching makes a difference

Our research provides compelling evidence that teaching quality significantly impacts student learning. We measured this through two key dimensions: pedagogical practices and content knowledge. When teachers display stronger pedagogical skills (as measured by Teach)—meaning that they excel at explaining concepts, checking understanding, and providing feedback—their students achieve better learning outcomes. This relationship is most visible in literacy, where a one-point improvement in teaching quality predicts a two-point increase in student scores on average across the 13 countries. While this overall impact is modest, the impact is strong in middle-income countries, where the same improvement in teaching quality predicts substantial increases of up to five points in literacy scores.

Teacher content knowledge also matters. When teachers have a better grasp of the material they teach, their students learn more. This relationship holds true across both literacy and mathematics and is particularly strong in low-income countries. Importantly, these relationships are robust to controlling for numerous other factors like student background, teacher characteristics, and school resources.

The power of student engagement

Our study breaks new ground in understanding how student engagement in class affects learning. Using the innovative Playful Learning Across Years (PLAY) tool  (now known as Engage), we measured specific teaching practices that support student engagement—practices like connecting lessons to students' lives, encouraging exploration, and fostering student agency. The results are striking: Teachers who excel at these engagement practices see their students perform up to 13 percentage points better on literacy assessments, even after controlling for other factors.

This finding holds particular promise because it suggests a concrete way to improve learning outcomes: by helping teachers adopt practices that actively engage students. The data shows that while teachers generally do well at creating supportive emotional environments, they often struggle with practices to promote student agency and exploration, which could be important in improving learning.

The support teachers need

Our findings point to specific ways education systems can help teachers improve. And that should be the focus—not on the teachers themselves, who are often doing the best they can, but within systems that are failing to equip them adequately. Three types of systems support consistently predict better teaching quality in this set of countries:

• First, practical pedagogical support makes a real difference. Teachers who receive mentoring, participate in induction programs, and get feedback on their lesson plans demonstrate better teaching practices. The key here is practicality.  It's not enough to just provide training; that training needs to include hands-on, classroom-based components.
• Second, strong instructional leadership at the school level matters a lot. When school leaders regularly observe classes and provide meaningful feedback, teaching quality improves. However, our data shows a crucial gap: While 75% of teachers have their classes observed, only 54% receive feedback afterward.  This signals a missed opportunity for improvement.
• Third, regular evaluation with consequences helps drive improvement, but only when combined with support. Teachers in schools with robust evaluation systems show better pedagogical practices, as long as evaluation is paired with practical support and feedback.

Yet currently, these support systems are often lacking. Only 39% of teachers reported receiving any in-service training in the previous year, and a striking 85% received no training with an in-classroom component. This support deficit helps explain why teaching quality often remains low despite teachers' best efforts.

Moving forward

These findings point to clear actions for education systems seeking to improve learning outcomes. First, strengthen practical support for teachers through mentoring, feedback, and classroom-based training. Second, invest in instructional leadership at the school level to ensure teachers receive regular, meaningful feedback. Finally, embed strategies for fostering student engagement into teacher training programs and learning materials.

The challenge of improving education in developing countries is complex, but our research suggests that focusing on teaching quality—and supporting teachers effectively—can make a significant difference in student learning. With targeted investments in teacher support and development, education systems can help ensure that every child not only attends school but learns effectively.

To receive weekly articles, sign-up here