TL;DR
Get the latest gadgets delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
Google says studies in the UK and Italy found that Gemini-supported tutoring performed similarly to human tutoring on some classroom measures, while teachers using Gemini in Italy reported more collaborative lessons. The company says the UK trial did not show gains on STAR standardized tests during its 10-week period, and the findings come from research Google conducted with education partners.
Google has reported findings from two studies of Gemini in classrooms: a 10-week UK trial involving more than 1,100 students found AI math tutoring produced results similar to human tutoring on platform tasks and intermediate quizzes, while a yearlong study across 31 Italian schools linked teacher use of Gemini with more collaborative lessons. The results offer evidence about classroom use, but the UK trial found no improvement on STAR standardized test scores during the study period.
The UK randomized controlled trial involved students aged 12 to 15 using the Eedi school platform. When students answered a math question incorrectly, they received either a static hint, help from a human tutor, or interactive tutoring combining Eedi’s teaching resources with Gemini. Teachers monitored progress and supported students in class. Google said human experts reviewed each AI response before it was shown to a student.
According to Google, students who received AI tutoring performed similarly to students paired with human tutors on in-platform tasks and intermediate quizzes. Expert reviewers approved 95% of AI responses, compared with 76% in a previous study, which Google described as an almost 25% improvement in pedagogical quality. The company also said Gemini guided students rather than giving them answers in nearly 99% of sessions, and students worked through problems themselves in more than 85% of chats.
The Italian study covered 31 schools and included 1,600 teachers working with more than 24,500 students aged 13 to 18. Google said teachers reported feeling more creative in planning and teaching when they co-designed lessons with Gemini: 79% reported this with Gemini, compared with 30% without it. The company also reported that traditional front-of-class lectures fell by 38%, while student-led collaborative learning nearly tripled. These are findings reported by Google from its study, not independently verified conclusions in the material provided.
Classroom Gains and Test-Score Limits
The studies address a practical question for schools: whether generative AI can support learning without replacing teachers or reducing students’ own effort. The UK findings suggest Gemini-based tutoring can provide step-by-step help that compares with human tutoring on certain classroom measures, while the Italian results point to possible changes in how teachers organize lessons. Both studies describe AI as a tool used alongside educator involvement, rather than as a substitute for classroom teaching.
The limits matter as much as the reported benefits. Students using AI tutoring solved about 14 problems per session, compared with 19 among students given static hints. Google said the tutoring took longer per question, so students covered less material over the 10-week trial. Their gains on daily tasks and intermediate quizzes did not translate into higher STAR test scores during that period. The findings therefore do not establish that Gemini improves broad standardized outcomes, or that the same results would apply across subjects, age groups, or school systems.
Popular right now
Shark Navigator Robot Vacuum with 30-Day Self-Emptying Base, White (AV2110S)
Vitamix Propel Series 510 Professional-Grade Blender, 3 Programs, Black
Ninja CREAMi 7-in-1 Ice Cream Maker, NC301
As an affiliate, we earn on qualifying purchases.
Two Studies, Different Classroom Measures
Google said it spent the past year working with educators and researchers in the UK and Italy to study AI in everyday educational settings. The UK trial followed an earlier exploratory study with Eedi, which examined an AI tutor designed to identify a student’s mistake and guide them toward correcting it. The new trial compared that approach with static hints and human tutoring, using assignments, quizzes, and independent STAR tests to track progress.
In Italy, Google worked with the Salesians of Don Bosco school network and the Pontifical Salesian University in Rome. The yearlong study gathered classroom activities and teacher self-reports about lesson planning with and without Gemini for Education. The supplied report describes teacher-reported experiences and changes in classroom activity; it does not provide further detail here about comparison methods or independent assessment of student learning outcomes.
“Technology best serves education when it creates space for inspiring teaching and deeper learning.”
— Google, in its report
Limits of the Reported Evidence
The report does not show whether the UK tutoring results would persist over a full academic year or lead to improved standardized test performance. Google said the 10-week period may have been too short for students to cover enough curriculum for gains to appear on the broader STAR test, because AI tutoring took more time per question. That explanation is the company’s interpretation; the trial did not establish the cause of the test-score result.
The supplied material also does not specify whether the Italian classroom changes were independently measured, how schools were selected, or whether the reported shifts led to lasting improvements in student learning. The study’s reported figures should be read in light of its methods and source: Google and its partners conducted and published the research.
Longer Trials and Further Evaluation
Google said it is exploring follow-up UK trials lasting a full academic year and plans to expand its measures of learning to better understand what contributes to meaningful outcomes over time. Such research could test whether slower, guided problem-solving leads to broader curriculum coverage and whether classroom quiz gains eventually appear on standardized assessments.
The company’s report does not give a timetable for those trials or specify when additional results will be released. For schools considering Gemini, the current findings point to potential uses in tutoring and lesson planning, while leaving questions about long-term achievement, implementation across different settings, and the amount of teacher oversight required.
Key Questions
What did the UK study find about Gemini math tutoring?
Google said students using Gemini-based tutoring performed similarly to students working with human tutors on platform tasks and intermediate quizzes. The 10-week trial did not find gains on STAR standardized test scores during the study period.
Did Gemini tutoring improve standardized test scores?
Google reported no improvement in STAR test scores over the 10-week trial. The company suggested students may have covered too little curriculum because guided tutoring took longer per question, but that explanation has not been established as the cause.
How was AI used in the UK classroom trial?
Students received a static hint, human tutoring, or interactive AI tutoring using Gemini and Eedi’s educational resources after answering a math question incorrectly. Teachers monitored students, and Google said experts reviewed AI responses before students received them.
What did the Italian study report?
Across 31 schools, teachers using Gemini for lesson co-design reported more creativity in planning and teaching. Google also reported fewer traditional front-of-class lectures and nearly three times as much student-led collaborative learning.
Are the findings proof that Gemini improves learning in every school?
No. The findings concern specific studies in the UK and Italy and were reported by Google. The material does not establish that the results apply to all schools, subjects, or students, and questions remain about long-term outcomes.
Source: blog.google
