Teacher-Mediated Generative AI use and Student Learning: A Field Experiment on Adaptive Revision in Government Secondary Schools in Rajasthan, India

Last registered on June 23, 2026

Pre-Trial

General Information

Title

RCT ID

AEARCTR-0018900

Initial registration date

June 22, 2026

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published

June 23, 2026, 8:36 AM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

Country

India

Region

Jaipur, Rajasthan

Primary Investigator

Name

Sajal Bhardwaj

Affiliation

IIM Calcutta

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Aditi Bhutoria

PI Affiliation

Associate Professor, IIM Calcutta

Contact Investigator

Additional Trial Information

Status

Completed

Start date

2025-12-01

End date

2026-03-10

Keywords

Behavior, Education

Additional Keywords

Generative AI, Learning Outcomes, Educational Technology, Secondary Education, Personalised Learning, Human-AI Collaboration, Technology-Assisted Learning

JEL code(s)

C93, I21, I24, O33

Secondary IDs

Prior work

This trial does not extend or rely on any prior RCTs.

Abstract

This study reports a school-randomised field experiment on whether teacher-mediated, generative-AI-supported revision improves student learning in resource-constrained government schools. From an eligible pool of schools provided by the Ministry of Education, Government of Rajasthan, we randomised 43 schools to treatment (22) and control (21). In treatment schools, teachers used a curriculum-aligned custom GPT system on classroom smartboards during pre-examination revision in mathematics and science for Classes 7 and 9. At the beginning of each session, the system offered three modules (concept revision, practice questions, and doubt resolution) and generated responses live at three difficulty levels (basic, intermediate, and advanced), enabling within-classroom differentiation that conventional revision strategies cannot easily provide at scale. The usage was teacher-mediated, at an intended dosage of two to three 30-minute sessions per week. Control schools conducted equivalent teacher-led revision without technology.

We conceptualised the intervention as AI-enabled adaptive instructional scaffolding and differentiated instruction. The system configured explanations and practice to students' learning levels and could increase behavioural and emotional engagement, reduce disaffection, and thereby improve academic achievement. The primary outcome is academic achievement, measured by a curriculum-aligned test. Secondary outcomes are academic procrastination (González-Brignardello & Paniagua, 2023) and four engagement/disaffection subscales (Skinner et al., 2008), measured at baseline and endline. We also conduct an exploratory heterogeneity analysis to test whether treatment impacts baseline present bias. The logic being, students who experience studying as more effortful may benefit disproportionately when adaptive support lowers the immediate cost of revision.

After randomisation but before baseline, seven treatment schools withdrew because teachers were assigned to Booth Level Officer duties for cyclical electoral roll revision. This exogenous shock was unrelated to treatment assignment or school characteristics. The analysis sample is 15 treatment and 21 control schools (1,126 students at endline). We assess the administrative balance between retained and withdrawn schools and document the incidence of electoral duty across both arms. Estimation uses ANCOVA, with robustness checks via difference-in-differences, Lee bounds, and the Romanov-Wolf test. This study was registered after data collection; its hypotheses are exploratory rather than prespecified.

External Link(s)

Registration Citation

Citation

Bhardwaj, Sajal and Aditi Bhutoria. 2026. "Teacher-Mediated Generative AI use and Student Learning: A Field Experiment on Adaptive Revision in Government Secondary Schools in Rajasthan, India." AEA RCT Registry. June 23. https://doi.org/10.1257/rct.18900-1.0

Sponsors & Partners

Experimental Details

Interventions

Intervention(s)

The treatment schoolteachers used a custom GPT aligned with the curriculum requirements of classes 7th and 9th. They used the tool on their smartboards for revising the mathematics and science syllabus before the annual examinations. At the beginning of every session, the tool presented three modules: concept revision, practice questions, and doubt resolution. Upon selecting a module, it asked teachers to specify the unit/topic they wanted to revise or the practice questions they wanted to provide. Upon specification, the tool generated live responses at three levels of difficulty- basic, intermediate, and advanced; this enabled a within-classroom differentiation of explanations and practice. The students did not interact with the system directly; teachers mediated the sessions. The intended dosage was 3 times a week, with 30-minute sessions each. Control schools conducted equivalent teacher-led revision without any technological support.

Intervention (Hidden)

We conceptualised the study as AI-enabled adaptive instructional scaffolding and differentiated instruction rather than as direct student-AI interaction. The customised GPT tool, built on the mathematics and science curriculum and previous-year examination papers for Classes 7 and 9, generated responses live at three difficulty levels (basic, intermediate, advanced), tailoring content to students' differing learning needs rather than adopting a one-size-fits-all revision approach in resource-constrained government school settings. Use was teacher-mediated, on classroom smartboards; students did not interact with the tool directly. The comparison group received teacher-led revision sessions and practice on the same concepts and questions, without technological support. The contrast, therefore, is designed to isolate AI-enabled adaptive differentiation rather than additional instruction time or practice volume.

Intervention Start Date

2025-12-20

Intervention End Date

2026-02-20

Primary Outcomes

Primary Outcomes (end points)

Academic achievement in mathematics and science, measured by a curriculum-aligned assessment administered at baseline and endline.

Primary Outcomes (explanation)

We measure academic achievement as the student's score on a curriculum-aligned test covering the mathematics and science topics (percentage correct/standardised to control-group baseline mean 0, SD 1). The estimated effect is the treatment–control difference in endline achievement conditional on baseline achievement (ANCOVA).

Secondary Outcomes

Secondary Outcomes (end points)

Academic Procrastination, Behavioural engagement, Emotional engagement, Behavioural disaffection, and Emotional disaffection. All these outcomes were measured both at baseline and endline.

Secondary Outcomes (explanation)

The study utilised five scales to measure these outcomes-
1. We measured academic procrastination using MAPS-15 by González-Brignardello & Paniagua (2023). We validated this scale in our context before finalising it as an outcome measure.
2. The behavioural and emotional outcomes were based on the scales by Skinner et al. (2008). Each scale was treated as a separate construct rather than aggregated, since the theoretical interest is in whether the intervention affects emotional and behavioural components differently.
Each scale is scored as the mean of its items (1-5 Likert scale), standardised for analysis. We interpret them as mechanism variables on the pathway from adaptive scaffolding to achievement, not as independent endpoints.

Experimental Design

A two-arm, parallel, cluster-randomised field experiment. Schools were drawn from an eligible pool provided by the Ministry of Education, Government of Rajasthan, and were randomly assigned to the treatment or control group. Treatment schools delivered teacher-mediated, GPT-supported adaptive revision while control schools delivered teacher-led revision without technology. We collected outcomes at the student level at baseline and endline.

Experimental Design Details

Randomisation assigned 43 schools to treatment (22) and control (21). However, before baseline data collection, 7 treatment schools withdrew because their teachers were assigned Booth Level Officer duties for the cyclical electoral roll revision, a government order exogenous to treatment assignment and school characteristics. Because the intervention required active teacher delivery, the exogenous shock could force withdrawal only in the treatment arm. Further, the cyclical scheduling of the duty meant control-school teachers were not drawn during the study window. Hence, the final analysis sample is 15 treatment and 21 control schools.
We assess the balance between retained and withdrawn treatment schools on available administrative characteristics, and document electoral-duty incidence across both arms. We estimate the treatment effect using ANCOVA with baseline controls as covariates. Difference-in-differences and Lee bounds serve as robustness checks for estimation and attrition. Additionally, we use a wild-cluster bootstrap to accommodate the small number of clusters, and a familywise-error-rate correction (Romano–Wolf stepdown) across the outcome family. Lastly, we examine baseline present bias as an exploratory moderator of treatment responsiveness. This study was registered after data collection was completed; all hypotheses are exploratory rather than pre-specified.

Randomization Method

Randomised using a computer

Randomization Unit

Cluster randomised at the school level.

Was the treatment clustered?

Yes

Experiment Characteristics

Sample size: planned number of clusters

43 schools (assigned: 22 treatment, 21 control).
Realised analysis sample following exogenous attrition: 36 schools.

Sample size: planned number of observations

Baseline: 1395 students Endline: 1126 students

Sample size (or number of clusters) by treatment arms

Randomised 22 schools into treatment and 21 into control groups (total 43 schools).

Analysis sample: 15 treatment, 21 control (36 schools), following exogenous attrition of 7 treatment schools to Booth Level Officer electoral duties.

Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

Supporting Documents and Materials

IRB