Mindfulness in Schools: A Pilot Cluster Randomised Trial of a Novel Blended Intervention for Early Adolescents
Irena Miksic, Ema Muza, Ivana Marijanovic, Sanja Kisicek, Ivana Buric
Mindfulness July 1, 2026 DOI: 10.1007/s12671-026-02876-w (opens in new tab)
Study at a glance
AI-extracted from the abstract| Characteristics | Pilot cluster randomised controlled trial Randomized Preregistered Peer reviewed |
|---|---|
| Sample size | 178 |
| Population | Early adolescents (ages 10-14) from 12 classrooms in a Croatian school |
| Duration | 8-week intervention |
| Topics | Meditation |
| Key findings | The blended mindfulness intervention showed high feasibility but low home practice and moderate acceptability. Preliminary effect sizes were negligible for behavioral difficulties (β=-0.67, η²=0.00), prosocial behavior (β=0.22, η²=0.00), global assertiveness (β=-0.17, η²=0.03), and trait mindfulness (β=0.07, η²=0.01), with variations by grade. |
Abstract
Abstract Objectives Mindfulness-based interventions (MBIs) are increasingly implemented in schools to support students. However, the effects remain modest and inconsistent, particularly across developmental stages, delivery formats, and school contexts.
Method: In response to these limitations, we conducted a pilot cluster randomised controlled trial (cRCT) of a novel blended MBI in a socioeconomically diverse Croatian school. A total of 178 early adolescents (ages 10–14) from 12 classrooms were randomised to either the intervention group ( n = 95) or a teaching-as-usual (TAU) control group ( n = 83). The 8-week intervention combined pre-recorded mindfulness content designed by an external expert with in-person facilitation by an internal school staff member.
Results: Feasibility was high in terms of recruitment (83.6% of eligible students enrolled), retention (74.7% completed both assessments), and session attendance (96.8% attended at least five of eight sessions). However, home practice was low (8.3% practised mindfulness regularly outside school) and challenges related to data matching, absences, and outlier exclusion reduced the final analytic sample to 103 (57 intervention, 46 control). Acceptability was moderate overall, with younger students generally reporting higher engagement and perceived usefulness, whereas one 7th-grade classroom exhibited peer contagion effects that undermined feasibility and intervention fidelity. Preliminary effect sizes were small, with negligible improvements observed in behavioural difficulties ( β = −0.67, η 2 = 0.00), prosocial behaviour ( β = 0.22, η 2 = 0.00), global assertiveness ( β = −0.17, η 2 = 0.03), and trait mindfulness ( β = 0.07, η 2 = 0.01), but subgroup analysis revealed variations by grade.
Conclusions: These findings provide evidence for the feasibility and acceptability of a blended MBI and align with broader evidence showing modest effects of universal school-based approaches. The study highlights the importance of contextual and individual factors in MBI research and provides recommendations for future studies examining factors that may shape engagement and outcomes. Preregistration The study was pre-registered at the Open Science Framework (identifier: y465x).