New Study Links AI Learning Tools to Higher Math and English Scores

A four-year study shows students using CENTURY AI are up to 58% more likely to exceed expected math progress. Learn how to use AI tools effectively.

Saturday, October 24, 2026
Related Apps:

Key Takeaways

  • Students who used the CENTURY platform the most were 58% more likely to exceed expected progress in math and 29% more likely in English than peers with low engagement.
  • A meta-analysis of 47 randomized controlled trials shows that the academic benefits of AI tutoring peak at 22.8 hours of usage and then level off.
  • Using personalized learning software consistently throughout the year increases progress by 11 percentage points in math and 7 percentage points in English.
  • AI educational technology is most effective when paired with active human instruction, rather than as a standalone tool.

A new multi-year study of international school students shows that regular use of AI-powered personalized learning software can significantly boost academic progress. The research tracked thousands of student outcomes over four years. It shows that digital tools complement classroom instruction when used consistently. For parents and educators choosing educational technology, these findings offer a clear look at what actually drives student growth.

What Happened

According to an impact report published by CENTURY, researchers analyzed more than 3,300 student outcomes across eight international schools within the Inspired Education Group. The study compared students' regular engagement with the CENTURY platform against their performance on independent, external assessments over four academic years.

The findings revealed a strong connection between active platform use and academic growth. Students with the highest levels of engagement were 58% more likely to exceed expected progress in mathematics compared to their least-engaged peers. In absolute terms, 37.7% of highly engaged students exceeded expectations in math, compared to just 23.7% of low-engagement students. In English, the highly engaged group was 29% more likely to exceed expected progress, translating to 36.9% of highly engaged students versus 28.5% of low-engagement students.

Consistency was also important. Students who used the platform regularly throughout the school year saw an 11 percentage point increase in math progress and a 7 percentage point increase in English progress compared to irregular users. According to Mike Lambert, Global Education Director at the Inspired Education Group, students using the platform were roughly twice as likely to make above-expected progress than below-expected progress, with 31% exceeding progress norms against a standard baseline expectation of 25%.

The Bigger Picture

While these results are promising, they raise an important question: can educational software replace traditional teaching? As we previously reported in our analysis of whether software can replace subject-specialist teachers, digital platforms are most effective as supplements rather than replacements.

This perspective is supported by broader educational research. A comprehensive 2026 meta-analysis published in Educational Psychology Review examined 47 randomized controlled trials of AI interventions in education. While the meta-analysis confirmed that AI tools have a statistically significant positive impact on academic outcomes, it also revealed that these benefits depend heavily on human guidance. AI tools do not work in isolation. They require teachers and parents to actively support and guide the student.

The meta-analysis also identified a clear "dose-response" threshold. Student academic benefits rose rapidly and peaked at approximately 22.8 hours of total usage. After this point, progress stabilized and even showed a slight decline. This suggests that more time on educational software does not automatically equal more learning, and that moderation is key.

It is also worth noting that the CENTURY study was conducted within high-resource private schools, such as those in the Inspired Education network which includes schools like Reddam House. These schools already possess significant technological infrastructure and teacher training. For parents and educators in lower-resource public schools, replicating these gains may require more careful planning around technology access and teacher support.

What This Means for Families

For parents, these findings should shift the focus away from simply "screen time" toward purposeful engagement. Giving a child access to an AI learning platform will not automatically boost their grades. Instead, the software must be integrated into a balanced study routine where human feedback is readily available.

For educators, the data shows the importance of embedding these tools into the daily curriculum rather than treating them as an occasional homework filler. When teachers monitor real-time insights from platforms like CENTURY, they can spot gaps in student understanding early and intervene with targeted instruction.

What You Can Do

To get the most out of these tools, focus on consistency over cramming. Encourage your child to use personalized learning tools for short, regular sessions rather than long, occasional study marathons. Keep human support central. Do not treat software as a digital tutor that can operate entirely on its own. Ensure teachers or parents are available to explain complex concepts when a student gets stuck. Finally, monitor the time spent on platforms. Since research shows that academic benefits peak around 23 hours of use, focus on high-quality, targeted sessions rather than endless repetition.

Share: