Independent randomised controlled trial across 20 schools and 3,000 students validates Eedi’s measurable impact on learning outcomes

‍

London, UK (November 7, 2025) — A large-scale independent randomised controlled trial (RCT) reveals that secondary school students using Eedi, an evidence-based provider of AI-powered learning tools, gained two to four months of additional progress in maths during the 2023–2024 academic year.

The study, conducted by WhatWorked Education under the direction of Dr. Wayne Harrison and Professor Steve Higgins, evaluated Eedi’s impact on 2,901 students aged 11–12 across 20 UK schools. Students using Eedi achieved the equivalent of two additional months of maths progress compared to peers using other leading platforms. Those who used Eedi for just 10–15 minutes per week saw gains double to four months of progress, equating to 113% of typical annual progress from Y7 to Y8.

“These results prove that evidence-based, teacher and tutor-guided technology can deliver learning gains that rival personalized tutoring, once the gold standard, but never scalable,” said Dr. Bibi Groot, Chief Impact Officer at Eedi. “Just ten minutes a week of diagnostic practice and misconception correction can dramatically accelerate learning, especially for students who start behind.”

‍

Precision Diagnostics that Target Misconceptions

Unlike most math tools that focus on right or wrong answers, Eedi’s approach zeroes in on why a student answers the way they do. Its adaptive diagnostic quizzes, spanning more than 60,000 questions, map each learner’s understanding and misconceptions across mathematical concepts. This insight enables teachers to target the root causes of misunderstanding rather than surface-level errors.

“I’ve been teaching for over 20 years, and Eedi is one of the best maths resources I’ve ever used,” said Julie Kelly, a teacher at Limavady High School. “Our students love it and ask every day to do more.”

Suzanne Walsh, Head of Numeracy at All Hallows Catholic High School, added, "Eedi helps to really quickly identify misconceptions and address them straight away. With the old school methods of marking and assessment, you might see the kids for another three days after they've completed the work you set. By then it's been and gone!"

‍

Independent Evidence Validates Impact

Eedi’s RCT used an active control design, benchmarking performance against students using other digital maths programs, including Sparx and MathsWatch—a higher research standard than most EdTech evaluations.

Eedi achieved statistically significant improvements across two important subgroups:

• For all students starting the year with below-median maths attainment, Eedi delivered an average gain of three additional months of progress.
• Among Pupil Premium students (a proxy for low-income learners), consistent Eedi use (1 quiz per week, just 10-15 minutes of practice) also yielded three months of additional progress, demonstrating its potential to narrow the learning gap.

"This study validates our core belief that the key to accelerating learning is not just more content, but more intelligence," said Ben Caulfield, Co-Founder and CEO of Eedi. "We empower teachers by pinpointing the root cause of student misconceptions, and this report proves that approach is highly effective, especially for students who start the year behind. We are committed to building tools that are not just used, but are evidenced to support teachers and deliver measurable results in real classrooms."

‍

Building the Future of Human-Centered AI Tutoring

Eedi’s growing evidence base is also fueling research into human-in-the-loop AI tutoring in collaboration with Google DeepMind. This work explores how large language models (LLMs) can safely deliver pedagogically sound tutoring informed by real-world teaching data.

‍