HERO Teachers - Harbouring Empathy to Raise Opportunities: a study of teachers' burnout and students' outcomes

Our treatment is inspired by the Balint (1957) group from the psychological literature. During Balint groups, medical physicians share cases and challenges faced with patients and receive support and suggestions from peers. The experience aims to improve clinicians’ relationships with patients. In our context, we ask teachers to meet twice a month to discuss the challenges they face with students, a psychologist mediates the discussion and provides guidance on how to best address the teachers’ struggles. We will organize 8 meetings in each school throughout the year.

2023-09-01

2024-06-01

We measure the impact of the intervention on four groups of outcomes: i) teachers’ well-being ii) their sense of community; iii) teachers/student relationships; iv) students’ outcomes.

We use validated psychometric scales to measure students and teachers’ well-being (MBI, BAT). We measure teachers’ sense of community through their participation in coffee gatherings we organize in treated and control schools, and via answers to open-ended survey questions (i.e., “what is your definition of working community?”).
We measure teachers/students cooperation through a behavioural task: we distribute half of lottery tickets to teacher/student pairs. The lottery tickets display a barcode with no digits. Teachers and students need to cooperate to match the missing half in order to participate in the lottery. We compare the share of matched tickets in treated vs control schools. We also track noise dynamics in three random lectures to see whether the intervention affected conflictuality in the classroom, as suggested by smoother noise dynamics.
We rely on students’ grades and their performance on a sample of questions from a standardized test to measure students’ outcomes.

We will collect three types of data: survey, observational and administrative data.
i) Survey Data: We will interview teachers, students, and schools’ principals at baseline (September 2023) and endline (May 2024), through tablet-based questionnaires.
ii) Observational Data: We will: a) collect teachers’ mood via monthly SMS; b) collect teachers’ attendance to social events organized at baseline and follow-up; c) observe students' and teachers’ cooperation in games; d) measure decibels in classrooms via noise detectors; e) use qualitative data psychologists collect during the intervention.
iii) Administrative Data: Students’ grades and attendance, high school choice (8th-graders only).

Participants, sample and selection procedures
We recruit schools in the periphery of Naples, Italy, on a voluntary basis. In Summer 2023, we will organize an event at Ufficio Scolastico Regionale (Regional School Bureau) and contact school principals personally to explain the program. In each school, principals will identify one class for each grade between 6th and 8th with the highest risk of drop-out, and sign it up for the program. Sign-up is conditional on teachers agreeing to participate.

Research design
We will carry out a Pairwise Clustered Randomized Control Trial involving 60 middle schools. We pair schools by similarity in teachers, students and neighbourhood characteristics. Within each pair, we randomly assign one middle school to treatment (mutual support group), and one to control.

Not available

Randomization using a computer.

School.

Yes

60

600 teachers, 3600 students

30 clusters, 300 teachers, 1800 students in T and in C group.