Fostering belongingness through recognition: A randomized evaluation of reputational incentives in Indian schools

The intervention is split into two days.
1. Day 1 – Teacher-Level Recognition and Student-Teacher Council:
a) Teacher recognition/award: A research assistant (RA) meets the Class 9 (sample) teacher, presents a certificate, and reads a brief, scripted acknowledgment of classroom achievements.
b) Student-teacher council: After this, the RA facilitates a 20-minute student-teacher council to discuss classroom climate and shared expectations.
2. Day 2 – School-Level Recognition and Principal-Student-Teacher Council:
a) School and Principal recognition/award: The PI presents an award note, certificate and poster to the principal and reads a scripted acknowledgment and posts the poster in a prominent location in the classroom of class 9.
b) Principal–Student-Teacher council: The PI and RAs then convene a 20-minute Principal–Student-Teacher council to discuss the topic of collectiveness and belongingness.

2025-07-22

2025-09-01

The primary outcomes focus on beliefs, perceptions and behaviour. Constructs such as reflection and mindfulness, both measured using 4-item battery of questions in the form of likert scales. We gather information on altruism, fairness, reciprocity, and perception of teacher and student behaviour and that of inequality inside the classroom (Bottiani et al., 2016; Garcia Castro et al., 2020; Diehl et al., 2023; Falk et al., 2016; Falk et al., 2018). Behavioural dimensions such as pro-sociality, collective action and cheerfulness or patience in responding to questions are captured using direct questions or vignettes (Murphy et al., 2020; Alan et al., 2025; Diehl et al., 2023; Falk et al., 2016; Falk et al., 2018).
Alongside, we ask direct questions on school attractiveness and time preferences in the follow-up as a proxy for school climate and student’s psychological states.

Within a set of secondary outcomes, psychological factors assessed using validated instruments. These include depressive symptoms measured by the Patient Health Questionnaire (PHQ-9) (Kroenke et al., 2009), perceived stress assessed through the Perceived Stress Scale (Cohen et al., 1994; White, 2014), and Hope is evaluated using the Adult Hope Scale (AHS) and Children’s Hope Scale (CHS) (Snyder et al., 1997; Snyder et al., 2007). Aspects of children’s mental health are also assessed through children’s self-portraits and drawing features (Baird et al., 2022; Glewwe et al., 2018). Some additional secondary outcomes include students’ beliefs related to gender norms and stereotypes that are based on measures from Nyqvist and Jayachandran (2017) and Bhan and Wen (2021). Moreover, academic performance is evaluated.
Alongside, we try to capture the school climate through questionnaires from teachers and principals and by tracking the choice of opting into a program that is premised to foster student wellbeing in the school premises. We also observe factors that contribute to student wellbeing such as teacher responsiveness or peer networks.

In this study, we aim to examine the impact of a reputational incentive ascribed at an institutional (school) level on school climare and student’s psychological well-being and academic performance. Randomisation happens at school level. Non-financial incentives and reflective exercises are two strategies that motivate our intervention that combines visible, school-level recognition with small-group reflection exercises to ignite intrinsic commitment and strengthen a virtuous cycle of belongingness and high performance. The (2-part) intervention program will be rolled out in July-August 2025. The term starts in April, followed immediately by 2-month summer holidays in May and June. Follow-up data collection will be conducted from August 2025 till April 2026. Data analysis will start in 2025.

Not available

Primary and secondary data will be collected from children and teachers of 72 government schools. These schools are randomly assigned into two homogeneous and comparable groups: one treatment and one control group.

Randomisation happens at school level.

Yes

72 schools

The sample consists of 8th and 9th class students of 72 government schools in Jaipur, Rajasthan during the period 2023-2026. Average class size is 23 students with 6-12 teachers in each school in the sample.

36 schools in each arm (36*2 = 72 schools)

Estimates for detecting 0.1, 0.25 and 0.3 sd effects at 5% significance level within conventional limits predicts a sample of 22-35 schools per treatment on our primary outcomes. We started off with reaching out to 90 schools of which 72 agreed to participate in the study.