PISA 2025: AI and Screen Decline May Be a Measurement Error 2026

The latest PISA 2025 report suggests a historic decline in student performance, with AI and screens often blamed. But a closer look reveals that the drop may be a measurement error rather than a real decline. The OECD's 2025 results, published on September 8, show the worst average scores ever in reading, math, and science across member countries. In Spain, scores fell by 23 points in reading and 16 in math. However, the way PISA measured socioeconomic status changed between 2022 and 2025, potentially skewing the results.

What Did PISA 2025 Actually Report?

The OECD reported the lowest average scores in PISA history for reading, mathematics, and science. This decline was observed not only in Spain but across the entire OECD. Among the many analyses, one finding stood out: where the socioeconomic gap narrowed, it was because advantaged students had fallen, not because disadvantaged students had improved. In Spain, the top socioeconomic quartile lost 33 points in reading and 24 in math—about double the loss of the lowest quartile.

This led to headlines that rich students were the ones who had fallen the most. But is that accurate? The change in how PISA measures socioeconomic status may be the key.

How PISA Measured Socioeconomic Status in 2022 vs. 2025

In 2022, PISA asked students if they had classics, science books, or art books at home to determine if they came from an 'accommodated' background. In 2025, the questions changed. In Spain, at least, students were asked about having a pool, an alarm system, or guest rooms. This shift in indicators could explain why the socioeconomic gap appeared to narrow—and why rich students seemed to drop more.

The new questions may not accurately capture the same socioeconomic groups. For example, having a pool or guest room might not correlate as strongly with academic advantage as having books. This measurement change could artificially inflate the decline among advantaged students.

Why This Matters for Interpreting the Decline

If the socioeconomic gap didn't actually narrow, then the narrative that rich students fell the most is flawed. And if that's not true, then attributing the decline to AI or screens is also problematic. The decline might be partly an artifact of changed measurement, not a real drop in performance.

This is not to say that AI and screens have no effect, but the evidence from PISA 2025 is weaker than headlines suggest. We need to be cautious about drawing strong conclusions from a single assessment, especially when the methodology has changed. Context on how this played out before sits in PISA 2025: AI and Screen Decline May Be a Measurement Error.

Key Takeaways from the PISA 2025 Analysis

• PISA 2025 reported the lowest average scores in reading, math, and science across the OECD.
• The socioeconomic gap appeared to narrow, but this may be due to changed survey questions.
• In Spain, the top quartile lost about twice as many points as the bottom quartile.
• The shift from asking about books to asking about pools and guest rooms could skew results.
• Blaming AI and screens for the decline may be premature without accounting for measurement changes.

Comparing PISA 2022 and 2025 Socioeconomic Indicators

This table shows how the indicators shifted from cultural to material wealth, which may not be equivalent in predicting academic success. The background to this story is set out in Australia's Children's TV Decline: Drama Behind Paywalls. A related thread runs through Science: latest articles.

What This Means for the AI and Screen Debate

The debate about AI and screens affecting student performance is important, but we must ensure the data we use is reliable. If the measurement of socioeconomic status changed, then comparisons to previous years are less valid. The decline in scores might be partly due to this change, not just external factors like technology. Earlier coverage in SpaceX Starship Reaches Orbit for the First Time Ever shows how the argument developed. A closer look at the same subject appears in Miura 5 Rocket Enters System-Level Testing Ahead of 2027 Debut.

Before we blame AI or screens, we should examine other potential causes, such as curriculum changes, teacher shortages, or pandemic aftermath. The PISA 2025 results are a wake-up call, but they require careful interpretation. Anyone following the wider picture can start with Centennial Light: The 125-Year-Old Bulb That Defied Obsolescence. The same theme surfaces in Dreame Prime Day Deals: Save on Vacuums, Hair Tools and Fans, which takes a different angle.

FAQ