How to use PISA Insights to Guide Science Curriculum

When the newest Programme for International Student Assessment (PISA) scores were released in September 2026, much of the attention naturally went to the numbers: Which countries improved? Which declined? Where did students rank? But the most important question may be what we do with those numbers next.

Conducted by the Organisation for Economic Co-operation and Development (OECD) with the intent to provide a global snapshot of how well prepared today’s learners are for the future, PISA 2025 focused on assessing science performance of 15-year-olds across countries and economies worldwide, with additional assessment of mathematics, reading, and for the first time, foreign language and “Learning in the Digital World (LDW)”.

No assessment can capture the full complexity of an education system or the many factors that shape student performance. But PISA gives us a valuable global lens into what students know and, importantly, what they can do with that knowledge.

The latest results reveal areas of progress as well as persistent gaps between students. For education leaders, those findings should be more than a scorecard. They should be a prompt for action.

We focus here on science and computational problem-solving as the highlighted and innovative areas in the PISA 2025 assessment, while recognizing that disciplines do not exist in isolation, either in the classroom or in the real world. We approach the findings from our respective roles as Head of Product and Head of Educational Impact at LEGO Education, where we think every day about what high-quality curricular materials should make possible for students and teachers.

As science and computer science educators, instructional leaders, and other decision-makers consider curriculum adoption, PISA 2025 gives us an opportunity to ask not only what students are learning, but what they are being asked to do with what they learn.

• Do the materials we choose give students opportunities to evaluate scientific evidence critically, assess the credibility of information, and apply scientific knowledge to decision-making?
   
• Do the digital learning experiences students participate in engage them in an iterative process of knowledge building and problem solving using computational tools?

These capabilities matter not simply because they appear on an international assessment, but because they are fundamental to what it means to use science and computational thinking practices in an increasingly digital world.


Here are the five takeaways for science leaders to consider:

Look Beyond Coverage to What Students Do

Various movements in science education have rightly shifted away from viewing science and science standards solely through a content lens. This means moving away from teaching science as a series of facts to be memorized and toward experiences that more closely reflect the true nature of science and how science is practiced (Osborne, 2014). This includes explaining phenomena, carrying out investigations, analyzing and interpreting data, using evidence to support claims, and applying knowledge to solve real-world problems.

The questions on the PISA assessment reflect this approach. Rather than simply asking students to recall scientific facts, they ask students to do something arguably more challenging and meaningful: apply what they know.

This is core to our approach when designing science curricular materials: looking not only at what students will learn, but what they will do with that learning. Student agency grows when learners move beyond replicating a standard demonstration of a scientific phenomenon and are instead asked to apply scientific knowledge to authentic problems, test ideas, make decisions, and justify their thinking.

Science leaders should examine curricular materials through this same lens, looking beyond standards alignment and content coverage to consider how students are being asked to use their knowledge.

Consider the released PISA 2025 science unit “Offshore Wind Turbine Farm.” Students draw on scientific knowledge to consider both the benefits of renewable energy and the potential environmental impact of installing wind turbines in a marine environment. Ultimately, they must weigh financial and environmental considerations to determine where turbines should be located and justify their decision.

This same approach can be reflected in curricular design. In the LEGO® Education Science lesson “Blades and Barnacles,” for example, students model, discuss, and propose changes to a wind turbine design within a set of criteria and constraints, with the goal of supporting marine biodiversity.

In both examples, the science content matters, but knowing the content is not the end point. Students are being asked to use that knowledge to investigate a problem, consider competing factors, develop ideas, and make and defend decisions.

Give Students Multiple Ways to Show What They Know

When considering curricular materials that move students beyond recalling facts, education leaders should also consider how students are able to demonstrate what they know. Curricular resources that offer tangible and diverse ways for students to express their understanding not only support the authentic application of science and computational thinking, but also create opportunities for differentiated instruction.

Differentiation is often thought of in terms of how teachers communicate information to students, or the input. But it should also include the different ways students can demonstrate their knowledge and understanding, or the output (Saphier, 1997).

This idea is also reflected in the Universal Design for Learning (UDL) framework, which calls for multiple means of action and expression so learners can demonstrate what they know in different ways.

Don’t Confuse Digital with Innovative

Science is an ever-changing field, shaped by constant discovery and innovation. But giving students access to the latest technology should not be mistaken for innovation itself. While there is much to unpack in the PISA findings around digital learning, two areas stand out for science leaders considering curricular materials: intentionality and engagement.

PISA 2025 reinforces that a binary approach to technology does not suffice. Digital technology does not fall neatly into “good” or “bad” categories.

The more important question, then, is what the technology is doing and how its use contributes to the learning goal.

• Can the learning goal be achieved without it?
   
• Does the technology provide students with new opportunities to investigate, create, model, receive feedback, or make their thinking visible?

In other words, education leaders should examine what the technology is doing pedagogically and how intentionally it connects to the intended learning outcomes.

Engagement is another important part of this equation.

Higher levels of reported digital distraction were also associated with lower science performance in more than two-thirds of participating countries and economies with available data, although PISA cautions that the relationship varies across education systems and should not be interpreted as straightforward causation.

From our perspective, this finding also raises a question about the kind of engagement curricular materials promote. Science is a social practice. Students learn science not only by interacting with content, but by asking questions, exchanging ideas, working together, arguing from evidence, and making sense of phenomena with others. When learners share their understanding and reasoning of scientific phenomena and engineering problems with one another, multiple representations of the concept support depth of understanding.

We intentionally chose to design both LEGO® Education Science and LEGO Education Computer Science & AI with a group collaboration model where learners are scaffolded to collaborate in groups of four. The lesson design provides meaningful and equitable roles for all learners within their first group activity, which we’ve observed in classroom testing to significantly increase the participation of all learners in group discussion.

When evaluating curricular materials, whether digital or non-digital, leaders should therefore consider not only whether students are “on task,” but what that task asks them to do with one another. Digital tools should not unnecessarily pull students away from those interactions. When appropriate, they should be intentionally designed to support and strengthen them.

Evaluate the Whole Learning Experience and the Whole Child

PISA 2025 reminds us that student success is about more than academic proficiency. The report examines learner characteristics through both proficiency and engagement, including curiosity, perseverance, and growth.

When evaluating curricular materials, science leaders should consider the whole learning experience and the whole child.

• What opportunities do students have for choice and agency?
   
• Are they motivated to engage and persist?
   
• Do they have space to struggle productively, test an idea, learn from what did not work, and try again?

When designing an instructional task in our products , one of our design principles is to aim for a level of solution diversity alongside demonstration of applied scientific knowledge or computational thinking principles. If all learners come up with the same solution to the task posed when we test a lesson, we redesign the task. Better outcomes result when learners construct knowledge together socially by sharing a broad array of solutions and justifications relating to the same scientific phenomenon.

High-quality curricular materials should not only build science knowledge, or computational thinking practices, but also create experiences that recognize students as curious, capable learners who bring their own ideas, interests, and approaches to a problem.

Look for materials that:

• Create meaningful choices
   
• Invite multiple pathways to a solution
   
• Encourage iteration
   
• Give students ownership over their learning

Consider whether the learning experience has built-in moments for iterative struggle, not only essential to the learner’s growth but also an authentic part of the true scientific process. The question is not only whether students are learning subject matter content and skills, but what they are experiencing as learners while they do it.

Examine the Instructional Support for Educators

A curriculum is only as strong as its implementation, which means curriculum review should include a careful look at how materials support educators in bringing that curriculum to life.

This matters in any subject, but especially in science and computer science. According to the Learning Policy Institute report (2026), 41 of the 50 states reported science teacher shortages during the 2024–25 school year. These shortages can mean that educators are asked to teach content or facilitate practices for which they have had limited professional development.

Similarly, even as the demand for CS and AI education is rapidly expanding, many teachers report limited preparation, confidence, and access to resources for teaching CS and AI concepts (UNESCO, 2023; Twarek et al., 2025)

Evaluating instructional support means looking beyond whether resources simply exist and asking whether they help teachers implement the intended pedagogy. Materials should not only tell educators what to do but provide concrete tools that help them do it well.

In the development of LEGO® Education Science and LEGO® Education Computer Science & AI, we involved hundreds of educators with a wide range of confidence in hands-on science and computer science teaching. Their direction was clear – though the desire and belief in the value of hands-on science and computer science learning experiences was strong, educators were not feeling supported with their currently available resources to confidently deliver frequent, high-quality, hands-on learning in these subjects.

Their asks for ready-to-use classroom materials, straightforward facilitation guidance, and flexible implementation options were critical guides in the product design process.

Leaders should also look beyond one-time professional development and consider how a curriculum supports educators over time. Strong implementation requires opportunities to build confidence and deepen practice as teachers begin using the materials, encounter challenges, learn from their students, and refine their instruction.

Curriculum selection, therefore, should consider not only what students receive, but what teachers receive to make high-quality learning possible.

Summary of Actions

PISA 2025 gives science and computer science leaders a reason to look beyond the scores and take a closer look at the experiences we create for students.

The next step is not simply to raise the scores. It is to create the learning experiences that better prepare students for what those scores are intended to measure and, more importantly, for the world beyond them.