Rapid Evaluation of the Impact of Gemini Guided Learning on Students’ Science Scores in Junior Schools in India

How to cite

Aspillaga, Natalia Valdes; Khawar, Usman; Munoz, Ignacio Roman. 2026. "Rapid Evaluation of the Impact of Gemini Guided Learning on Students’ Science Scores in Junior Schools in India." AEA RCT Registry. AEARCTR-0019276. https://doi.org/10.1257/rct.19276-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Natalia Valdes Aspillaga* , Usman Khawar , Ignacio Roman Munoz

* Primary investigator

Registry ID
AEARCTR-0019276
DOI
https://doi.org/10.1257/rct.19276-1.0
First registered
Last updated
Status
On going
Trial dates
2026-04-13 – 2027-02-01
Countries
Private
JEL codes
I24
Registry record
http://www.socialscienceregistry.org/trials/19276

Abstract

This study evaluates the impact of Gemini Guided Learning, an AI-powered instructional experience within Gemini that provides personalised learning through guided questions, step-by-step scaffolding, and multimodal resources rather than supplying direct answers, on physic and chemistry lessons in grades 6,7 and 8 in India (ages 10–14). This evaluation is part of a multisite evaluation. The trial randomly assigns classrooms either to continue with regular teaching or to use Gemini in approximately a third of their science lessons over a 13 week period. The same teachers are responsible for both treatment and control classrooms; all will receive training on how to use the tool, though it will only be applied in treatment classes.
Science performance of students will be assessed at the beginning and end of the intervention. The study will provide rigorous evidence on whether AI supported guided learning can strengthen science learning in junior secondary education.

Intervention

Gemini Guided Learning is an AI-powered instructional tool within Gemini, designed to support students’ science learning through guided questions, step-by-step scaffolding, adaptive feedback, and multimodal resources rather than providing direct answers. Infused with Google’s LearnLM research, Gemini encourages students to reflect on their own thinking, deepening metacognition while fostering curiosity and active engagement.The intervention will be implemented in grades 6, 7 and 8 (ages 10–14) over a 13-week period, during which students in treatment classrooms will use Gemini in one 35-minute chemistry module and one 35-minute physics module per week. This is equivalent to one third of the science lessons. Teachers will receive training on how to use the tool, although the tool will only be applied in treatment classrooms.

Experimental design

This study employs a cluster randomised controlled trial with classrooms as the unit of randomisation. Randomisation is stratified by school, grade level (6, 7 and 8), using computer-generated random assignment with a pre-specified seed. This stratification approach ensures balanced allocation across key prognostic factors and prevents potential confounding between teacher effects and treatment effects.
Classrooms are randomly assigned to either: (1) Treatment arm receiving access to an AI powered science tutoring platform in addition to standard instruction, or (2) Control arm receiving standard science instruction only.

Design details

Randomization method
Done in office by a computer using stata
Randomization unit
Classroom
Clusters
240 classrooms across 16 schools (G6:71; G7:83; G8:86)
Observations
9,000 pupils
Arms
128 control and 112 treated classrooms
Primary outcomes
The primary outcome of interest is student science achievement, measured using standardised assessments aligned with the grade 6,7 and 8 curriculum and administered at the endline.

Cited by

No citation data available for this registration yet.

Keywords