AI in the Classroom

Last registered on September 25, 2026
View Trial History

Pre-Trial

Trial Information

General Information

Title
AI in the Classroom
RCT ID
AEARCTR-0019738
Initial registration date
September 16, 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
September 25, 2026, 9:14 AM EDT

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

Locations

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Primary Investigator

Name
Gregory Sun
Affiliation
Washington University in St Louis
Contact Primary Investigator

Other Primary Investigator(s)

PI Name
Kiara Kim
PI Affiliation
Washington University in St Louis
Contact Investigator
PI Name
Nathan Mester
PI Affiliation
Washington University in St Louis
Contact Investigator

Additional Trial Information

Status
In development
Start date
2026-09-17
End date
2026-12-21
Keywords
Behavior, Education
Additional Keywords
JEL code(s)
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
Abstract
This experiment studies how premium AI account use affects school performance and additionally, to what extent students are sophisticated about these effects.
External Link(s)

Registration Citation

Citation
Kim, Kiara, Nathan Mester and Gregory Sun. 2026. "AI in the Classroom." AEA RCT Registry. September 25. https://doi.org/10.1257/rct.19738-1.0
Sponsors & Partners
Experimental Details

Interventions

Intervention(s)
Willingness to pay to upgrade to a premium AI account will be elicited via a BDM mechanism, which will also randomize AI access.
Intervention Start Date
2026-09-18
Intervention End Date
2026-12-20

Primary Outcomes

Primary Outcomes (end points)
Academic Performance (GPA)
Predicted academic performance
Prediction errors (derived from the above two)
WTP for AI at midline
Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)
Stated beliefs about AI
AI use patterns
Secondary Outcomes (explanation)

Experimental Design

Experimental Design
AI will be randomized twice, once at baseline, and once at midline. In both cases, randomization is facilitated through a Becker-Degroot-Marschak mechanism: students will be asked to state their willingness-to-pay for AI, and will receive a subsidy if and only if we draw a subsidy that covers their stated WTP. This thus randomizes AI access as well.
Experimental Design Details
Not available
Randomization Method
Computer
Randomization Unit
Each student gets randomized once, in each of the baseline and midline survey waves.
Was the treatment clustered?
No

Experiment Characteristics

Sample size: planned number of clusters
300-320
Sample size: planned number of observations
300-320
Sample size (or number of clusters) by treatment arms
Roughly 150-160 students each per arm within a survey wave, but with variability due to the BDM-mechanism.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
Supporting Documents and Materials
IRB

Institutional Review Boards (IRBs)

IRB Name
Washington University in St Louis IRB
IRB Approval Date
2026-09-11
IRB Approval Number
202607237
Analysis Plan

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information