Joint Learning by Doing

How to cite

Khandelwal, Vatsal; Singh, Juni. 2026. "Joint Learning by Doing." AEA RCT Registry. AEARCTR-0017621. https://doi.org/10.1257/rct.17621-1.0

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

Vatsal Khandelwal* , Juni Singh

* Primary investigator

Registry ID
AEARCTR-0017621
DOI
https://doi.org/10.1257/rct.17621-1.0
First registered
Last updated
Status
On going
Trial dates
2025-05-26 – 2026-12-28
Countries
Nepal ()
Registry record
http://www.socialscienceregistry.org/trials/17621

Abstract

This study examines whether joint learning by doing, experiencing a new technology together with a peer in a socially intimate setting, increases adoption and use relative to individual use. We implement a randomized controlled trial in rural Nepal. The intervention combines temporary access to an electric cooking technology with structured peer meetings centered on meal sharing between randomly matched women. The design creates variation in individual use, social interaction without use, and joint use with a peer, allowing us to isolate the added value of social interaction for technology adoption and the extent to which technology use itself stimulates social interaction. In addition, some participants receive real-time feedback on air quality improvements to study how direct evidence affects belief updating.

Intervention

Villages are first randomly assigned to pure control or treatment.

Within treatment villages, households were cross-randomized to receive a temporary 3-month electric cookstove trial, to participate in structured peer meetings, both, or neither. Peer meetings involved repeated meal-sharing sessions with another randomly assigned household in the village. Among cookstove recipients, half also received real-time air quality feedback.

Experimental design

Villages are first randomly assigned to pure control villages or treatment villages. Within treatment villages, households are cross-randomized to receive a temporary 3-month electric cookstove trial (with a private demonstration) and/or to participate in peer meetings, generating four household-level arms: no cookstove and no meetings, cookstove only, meetings only, and both.

Households assigned to meetings are randomly paired within village and asked to meet every two weeks for three months (six meetings), with the first meeting facilitated by enumerators and subsequent meetings incentivized. Among households receiving cookstoves, half are additionally randomized to receive real-time air quality feedback.

Design details

Randomization method
Randomization done by a computer (using Stata).
Randomization unit
Villages are randomized into pure control and treatment; households in treated villages are randomized into various arms.
Clusters
40
Observations
1555
Arms
Eight villages comprising 339 households form the pure control group. The remaining 32 villages comprise 1,216 households assigned to treatment villages. Within these villages, households were cross-randomized: 350 households were assigned to receive a cookstove and 610 households were assigned to participate in peer meetings, implying that some households received both interventions, some received one of the two, and some received neither. Among households receiving cookstoves, 175 were additionally assigned to receive air quality information.
Primary outcomes
Primary outcomes capture adoption and usage of the technology, willingness to pay for new and used cookstoves, beliefs about benefits and air quality, knowledge around usage, respiratory health outcomes, and social network interactions.
Secondary outcomes
Secondary outcomes capture wellbeing, including loneliness and stress, investment in household electrical capacity, and cooking time and fuel use. Additionally, we measure mechanisms-related outcomes including: (1) Peer interactions — meeting attendance, duration, and peer interactions outside meetings; (2) Learning — confidence in cookstove operation and enumerator-verified knowledge; (3) Attention — awareness of various parameters during cookstove usage; (4) Self-reported benefits of peer meetings and altruistic donations to matched peer; (5) Social norms — perceptions of peer usage, discussions with other members of the community, and incidences of sharing cookstoves.

Cited by

No citation data available for this registration yet.

Keywords