Project Based Learning

How to cite

Malik, Samreen; Bharti, Nitin. 2026. "Project Based Learning." AEA RCT Registry. AEARCTR-0018266. https://doi.org/10.1257/rct.18266-1.0

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

Samreen Malik* , Nitin Bharti

* Primary investigator

Registry ID
AEARCTR-0018266
DOI
https://doi.org/10.1257/rct.18266-1.0
First registered
Last updated
Status
On going
Trial dates
2024-09-01 – 2026-05-31
Countries
India (Bihar)
JEL codes
C93, D83, I21, I24, O15
Registry record
http://www.socialscienceregistry.org/trials/18266

Abstract

Large-scale pedagogical reforms often underperform because policy adoption does not automatically translate into classroom implementation. This paper studies whether scalable implementation supports can improve the adoption of activity-based science pedagogy in government schools. We examine a statewide Project-Based Learning (PBL) reform in Bihar and embed a stratified cluster randomized controlled trial across 926 schools in Vaishali district. Schools were assigned to one of three groups: control, monitoring only, or monitoring plus mentoring. Relative to the control, the monitoring arm includes regular follow-up on project completion, whereas the mentoring arm includes monitoring and recurring hands-on teacher training through a cascade model. We estimate the effects of these supports on teacher implementation of PBL, instructional practices, and downstream student outcomes. The design allows us to identify both the effect of monitoring relative to the common rollout and the additional effect of mentoring beyond monitoring.

Intervention

The study experimentally varies two implementation supports: (i) T1: monitoring of project completion and (ii) T2: mentoring teachers through regular and structured teacher training.

Experimental design

This study employs a cluster randomized controlled trial with the school as the unit of randomization. The study is implemented in the Vaishali district (in Bihar state), which had 960 government middle schools in administrative records. Before randomization, we selected 34 Master Trainers (MTs) to support the teacher-training cascade model. These MTs's schools were kept out of the randomization sample, leaving 926 schools for random assignment.

Randomization was stratified by administrative blocks. There are a total of 16 blocks, with different number of schools. Within each block, all schools were randomly assigned to three groups, Treatment 1 (T1), Treatment 2 (T2), or Control (C). The distribution of schools is done to ensure similar proportion across T1, T2, and C arms within each block. This resulted into yielding a total of 308 schools in T1, 308 schools in T2, and 310 schools in Control.

Treatment arm 1 (T1) – Monitoring: Schools in this group receive regular monitoring of PBL project completion.

Treatment arm 2 (T2) – Monitoring and Mentoring: Schools in this group receive the same monthly monitoring as T1, and science teachers additionally receive structured, hands-on PBL training.

Control group (C) – Schools in this group continued under business-as-usual implementation of the state-wide PBL rollout (including receipt of the PBL handbook and annual PBL training) and do not receive additional mentoring support or regular monitoring as part of the study.

Design details

Randomization method
Randomization was stratified by administrative blocks. There are a total of 16 blocks, with different number of schools. Within each block, all schools were randomly assigned to three groups, Treatment 1 (T1), Treatment 2 (T2), or Control (C). The distribution of schools is done to ensure similar proportion across T1, T2, and C arms within each block. This resulted into yielding a total of 308 schools in T1, 308 schools in T2, and 310 schools in Control. This was done in Stata in office.
Randomization unit
School
Clusters
16 clusters/blocks
Observations
The total number of teachers is roughly 900. In-person surveys administered to roughly 151 teachers, and telephonic surveys for the remaining teachers. In person surveys in randomly selected 151 schools comprising roughly 10000-12000 students
Arms
The distribution of schools is done to ensure similar proportion across T1, T2, and C arms within each block. This resulted into yielding a total of 308 schools in T1, 308 schools in T2, and 310 schools in Control.
Primary outcomes
Teacher Level: PBL implementation, Time spent in experiments, General Pedagogical Index, Science-based Pedagogical Index, Confidence Student Level: Personal-level factors (curiosity, importance of science index, aspiration, barriers to education, confidence); Engagement Outcomes (Science Interest Index, Affective Engagement, Engagement Index)
Secondary outcomes
Teacher Level: Quality of PBL sessions, Teaching Philosophy, Homework Assignment, Teacher's Network, Teacher's Personal Satisfaction Student Level: Cognitive Ability (Test Scores), Raven Matrices, Creativity, Other personal and engagement factors

Cited by

No citation data available for this registration yet.

Keywords