Newenkantun: a Play-based Program to Promote Socioemotional Learning in Early Childhood Education in Chile.

STUDY RATIONALE Socioemotional learning (SEL) competencies developed in early childhood underpin later social, behavioral, and academic adjustment, and children in low-income settings are at greater risk of delays in socioemotional and self-regulatory development. Active and sociodramatic play are central vehicles for SEL in the preschool years, and healthy 24-hour movement behaviors (physical activity, limited screen time, and adequate sleep) are associated with better self-regulation and prosocial behavior. In the Araucanía Region of Chile, Mapuche and non-Mapuche educational traditions coexist, and Mapuche concepts such as che (person), newen (force), yamuwün (respect), and aukantun (play) align closely with SEL principles. Rigorously evaluated SEL programs for early childhood education in Latin America rarely integrate play and movement, genuine intercultural co-creation, and multiple levels of influence. Newenkantun, named after a Mapuche hand-strength game, was developed to address this gap and is aligned with the Chilean Ministry of Education's Socioemotional Learning Plan and the national early childhood curriculum. SEL is conceptualized using the five CASEL competencies: self-awareness, social awareness, self-management, relationship skills, and responsible decision-making.

STUDY PHASES Phase 1 comprises program development, participatory cultural adaptation, and a pilot study in two schools, one per municipality, with approximately 40 children and their caregivers. The pilot is used to optimize the intervention, assess feasibility, and refine the program logic model with community participation, using the same instruments and procedures as the main trial. Phase 2 is the stepped-wedge cluster randomized trial in 16 pre-kindergarten and kindergarten classes from 10 schools in Temuco and Padre Las Casas. Phase 3 comprises process, outcome, and impact evaluation and dissemination.

DESIGN RATIONALE The unit of randomization is the class. A stepped-wedge design was chosen because all participating classes eventually receive the intervention, which addresses community and educational concerns about withholding a promising program; rollout can be staggered to match implementation capacity and allow progressive refinement of locally co-created content; repeated measurements within each class increase statistical efficiency compared with a parallel cluster design with the same number of clusters; and the design includes a maintenance period to assess sustained effects.

SAMPLING, RANDOMIZATION, AND BLINDING Randomization is conducted in two stages. In the first stage, 10 schools, five per municipality, are selected by stratified random sampling, with strata defined by municipality, urban or rural location, and socioeconomic level based on the school vulnerability index. Sixteen classes are then identified in the selected schools. In the second stage, classes are allocated to four sequences of four classes each, which determine when each class crosses from the control to the intervention condition. Allocation uses restricted randomization to balance urban and rural classes across sequences, and classes from the same school are allocated to the same sequence to minimize contamination. The allocation is computer-generated by an independent statistician and remains concealed until the baseline assessment has been completed in all classes. Because of the educational nature of the intervention, children, caregivers, and educators cannot be blinded; assessors of directly measured child outcomes are blinded to allocation.

RECRUITMENT AND CONSENT PROCEDURES School principals sign an institutional authorization before any study activity begins. Caregivers are invited through informational meetings scheduled to suit families and supported by accessible graphic materials, and give written informed consent on paper or electronically. Educators and school staff give written informed consent. Before each assessment or program activity, children give developmentally appropriate assent supported by graphic materials, and any verbal or behavioral sign of refusal stops the activity without consequences. Recruitment is conducted by the research team independently of school leadership, and participants are told that declining or withdrawing has no academic, employment, or service-access consequences. Expected participant burden is 20 to 30 minutes per wave for the child assessment and for caregiver questionnaires, 30 to 60 minutes per educator training session or caregiver workshop, 60 to 90 minutes per focus group, and 45 to 60 minutes per individual interview.

INTERVENTION DELIVERY The 16-week program has two components delivered simultaneously and is grounded in sociocultural theory, self-determination theory, and social cognitive theory. The educator component begins with a participatory needs assessment in each school, followed by blended professional development that combines synchronous and asynchronous online modules with three in-person workshops per school. Workshops support the co-creation of intercultural sociodramatic narratives with educators, children, and families, drawing on the Mapuche worldview and the cultural heritage of the region; narrative characters emphasize social and physical abilities linked to self-care and care for the environment. Each school receives a play materials kit for active play during class time, and school-based leaders are trained to sustain the program after the research team withdraws, following a capacity-building, self-sustaining infrastructure model. The family component delivers eight topics every two weeks, about 30 minutes each, through a Moodle-based platform, a mobile app, social media, and printed materials for families with limited connectivity. Topics cover SEL, movement behaviors, family play and physical activity, recognizing changes in socioemotional competencies, goal setting, and sustaining change. Synchronous sessions focus on realistic home strategies, and content is reinforced through moderated peer-support groups, relapse-prevention strategies for behaviors such as excessive screen use, and specific strategies to engage fathers. In rural schools within Mapuche communities, sessions are led by intercultural facilitators in community settings agreed with families. During the control period, classes continue usual educational practice.

DATA COLLECTION SCHEDULE Assessments are conducted in all classes at baseline, before the allocation is disclosed, and at each subsequent step of the stepped-wedge rollout through the end of follow-up. At any given wave, classes may be in the control, intervention, or maintenance condition, and the same procedures are applied to all of them. The primary outcome and process measures are collected at every wave; most secondary measures are collected at baseline, after each class's intervention period, and 12 months later, to limit participant burden. Child assessments take place at school, caregiver questionnaires are completed at home on paper or electronically, and educator and classroom assessments are conducted during scheduled site visits. Standardized evaluation teams are trained for each municipality, and procedures are adapted to connectivity limitations and cultural protocols in rural and Mapuche communities. Device-based measures are collected in a subsample of 120 children selected for logistical feasibility. Instruments used with children and families have been translated and culturally adapted for Chilean preschool populations.

SAMPLE SIZE The number of clusters was fixed by feasibility at 16 classes, with an expected 15 participating children per class. Power was calculated for a closed-cohort stepped-wedge design with five measurement periods, using a multilevel model with fixed period effects solved by generalized least squares and a t-distribution with 14 degrees of freedom. Assuming a within-period intraclass correlation of 0.05, a cluster autocorrelation of 0.8, and an individual autocorrelation of 0.5, the design has 96% power to detect a standardized effect of 0.32, equivalent to an approximately 10% difference in SEL scores, at a two-sided alpha of 0.05; the minimum detectable effect is 0.24. Power remains at least 80% with 10 evaluable children per class, with intraclass correlations up to 0.20, or with up to 20% within-school contamination.

STATISTICAL ANALYSIS Intervention effects will be estimated with linear mixed-effects models, or generalized linear mixed models for non-normally distributed outcomes, including fixed effects for period and intervention condition and random effects for class and child, to account for clustering, repeated measures, and secular trends inherent to the stepped-wedge design. Models will be adjusted for the stratification variables, and a small-sample correction for degrees of freedom will be applied. The primary analysis follows the intention-to-treat principle and includes all randomized participants regardless of adherence; per-protocol and adherence-based sensitivity analyses will also be conducted. Effects will be reported as standardized mean differences. Moderation and mediation analyses will be adjusted for child age and sex. Missing data will be handled by multiple imputation under a missing-at-random assumption. Qualitative data from interviews and focus groups will be analyzed thematically, and quantitative and qualitative findings will be integrated using a convergent parallel mixed-methods design.

PROCESS EVALUATION Process evaluation is embedded throughout the trial and structured by the RE-AIM framework (reach, effectiveness, adoption, implementation, and maintenance), with specific attention to cultural fidelity and to differences between urban and rural delivery. Interviews and focus groups with educators, caregivers, and community stakeholders capture barriers, facilitators, and contextual factors and are also used to monitor within-school contamination.

DATA MANAGEMENT, SAFETY, AND DISSEMINATION Data are managed on secure, password-protected platforms and stored on institutional servers with restricted access; paper questionnaires are coded and stored securely. Data quality is monitored through range and logic checks and regular audits, under a data management plan that meets national requirements for publicly funded research. The study is classified as minimal risk, and procedures for identifying, managing, and reporting adverse events are specified in the approved protocol. Results will be shared with participating schools, families, and communities, with regional and national education authorities, and through peer-reviewed publications.