The impact of a social and emotional learning programme to improve pupils' educational inclusion in vocational education and training

The study addresses the growing educational, social, and policy interest in Social and Emotional Learning (SEL) as a protective factor against pupils’ social, emotional, and behavioral problems and as a driver of performance and well-being. SEL is defined as the process through which pupils acquire and apply knowledge, skills, and attitudes to develop healthy identities, manage emotions, set and achieve goals, show empathy, build supportive relationships, handle interpersonal situations constructively, and make responsible decisions. The research aims to evaluate the effectiveness of a SEL programme for Basic VET pupils from disadvantaged contexts—an educational stage in Spain with high dropout and failure rates—targeting five competencies: self-awareness, social awareness, self-control, relationship skills, and responsible decision-making. The hypotheses are: (1) experimental group pupils will show statistically significant post-test improvements versus pre-test across SEL competencies, whereas the control group will not; and (2) experimental group pupils will outperform control group pupils on post-test SEL scores.

Prior systematic and meta-analytic reviews indicate that SEL interventions improve students’ social-emotional skills, self-perceptions, attitudes, school engagement, prosocial behavior, academic performance, and reduce emotional/behavioral difficulties and substance use across diverse populations and contexts (e.g., Durlak et al., 2010, 2011; Taylor et al., 2017; Corcoran et al., 2018; Wigelsworth et al., 2016; Murano et al., 2020; Sabey et al., 2017; Yang et al., 2019). However, most evidence comes from Anglo-Saxon contexts. In Ibero-American settings, a systematic review (Fernández et al., 2021) found relatively low-quality evidence due to: limited adherence to key implementation indicators, lack of robust evaluation (often only participation/satisfaction data), predominance of qualitative or weak designs (pre-experimental, ex post facto), small samples, and poor intergroup comparability. The present study seeks to strengthen the evidence base in this context by applying a rigorous SEL logic model and quasi-experimental design with monitoring and statistical controls.

Design: Quasi-experimental design with a non-equivalent control group enhanced with statistical control techniques. Classrooms within each of three Basic VET programmes were randomly assigned to experimental or control conditions, respecting natural classroom groupings. Ethical approval: University of Granada Ethics Committee (1736/CEIH/2020); aligned with the Declaration of Helsinki. Informed consent obtained from legal guardians. Sample: N=110 first-year Basic VET pupils from one secondary school in Madrid, Spain. Experimental group: n=55 (14 women, 41 men), mean age 15.98 (SD=0.83); programme distribution: 19 electricity/electronics, 17 computer/communications, 19 administrative services. Control group: n=55 (14 women, 41 men), mean age 15.87 (SD=0.84); programme distribution: 20 electricity/electronics, 15 computer/communications, 20 administrative services. Initial enrollment was 126; attrition led to N=110; comparisons indicated no significant differences between dropouts and final sample. Group equivalence at baseline was confirmed across socio-demographic and educational control variables (e.g., age, nationality, immigration background, household education, funding source, VET programme, previous retakes, prior dropout, and baseline SEL scores). Instruments: (1) Social and Emotional Learning Scale (Fernández et al., 2022): 30 Likert items (1–4) across five subscales (self-awareness, social awareness, self-control, relationship skills, responsible decision-making); reliability: α=0.70–0.84, ω=0.71–0.84; strong construct and predictive validity. (2) Participation Questionnaire (ad hoc, 28 items) capturing socio-demographic and educational control variables. Intervention: Yearlong SEL-integrated programme (Sep 2021–Jun 2022), preceded by teacher/educator training and piloting (2020–2021). The programme delivered explicit instruction of SEL using a sequenced, active, focused, explicit approach (SAFE), integrated into the Basic VET curriculum via six transversal, project-based modules employing active methodologies (project-based learning with service-learning approach, cooperative learning), and reflective tools (portfolios, learning diaries, rubrics, self-/co-assessment). Activities occurred in a specialized flexible classroom and included experiential components: ‘disruptive learning pills’ (artistic-musical, environmental, sports, technological contexts), educational outings (e.g., workplace visits), technical sessions by company professionals, and field exploration/research. Complementary components: peer mentoring, individualized tutoring (risk detection, socio-educational support to pupils/families, personalized learning itineraries), vocational orientation, personalized professional internships (matching company opportunities with pupils’ technical and SEL development), and professional orientation (labor market access, offers, up/reskilling, events). Fidelity and monitoring: Group follow-up sessions with staff and educators at the end of each trimester and with pupils at the end of first and second trimesters to identify challenges, implement solutions, and assess participation. Pre-test (Sep 2021) and post-test (Jun 2022) measurements collected. Data analysis: Power analysis assumed effect size=0.50, α=0.05, power=0.80. Preliminary checks: linearity, outliers, missing/influential cases (Mahalanobis), distribution (Kolmogorov–Smirnov; Mardia’s skewness/kurtosis) indicated non-normality. Descriptive statistics computed. Group equivalence: Mann–Whitney U (age, SEL) and Pearson’s chi-squared (other controls). Hypothesis tests: Wilcoxon z (pre vs post within groups) and Mann–Whitney U (between groups), with Cohen’s d effect sizes. Family-wise error controlled using Bonferroni correction (α=0.01 per test). Software: SPSS 28.0 and STATA 17.

- Preliminary analyses found no atypical/missing/influential cases; SEL scores were non-normally distributed. - Intragroup (pre vs post): Control group showed no statistically significant changes across SEL competencies and trivial effect sizes (d≈0.06–0.15). Experimental group showed statistically significant improvements across all five competencies with medium to large effect sizes: • Self-awareness: 2.90→3.28 (SD 0.35→0.28); Wilcoxon z≈-6.45; p<0.01; d=1.20. • Social awareness: 2.99→3.31 (SD 0.39→0.35); z≈-6.48; p<0.01; d=0.86. • Self-control: 2.43→2.77 (SD 0.39→0.34); z≈-6.51; p<0.01; d=0.93. • Relationship skills: 2.93→3.10 (SD 0.34→0.32); z≈-6.45; p<0.01; d=0.52. • Responsible decision-making: 2.20→2.77 (SD 0.68→0.43); z≈-6.46; p<0.01; d=1.00. - Intergroup (experimental vs control): No significant baseline differences. Post-test differences favored the experimental group across all competencies, with medium to large effect sizes: • Self-awareness: U=608.50; d=1.25 (Exp 3.28±0.28 vs Ctrl 2.89±0.34). • Social awareness: U=726.50; d=0.80 (3.31±0.35 vs 3.03±0.35). • Self-control: U=875.00; d=0.84 (2.77±0.34 vs 2.47±0.37). • Relationship skills: U=1130.00; d=0.50 (3.10±0.32 vs 2.94±0.32). • Responsible decision-making: U=835.50; d=0.86 (2.77±0.43 vs 2.30±0.64). - All hypothesis tests used Bonferroni-adjusted α=0.01 and remained statistically significant in the experimental group outcomes and intergroup post-test comparisons.

Findings support both hypotheses: the SEL programme produced statistically significant and practically meaningful gains in all five SEL competencies for the experimental group, with no comparable changes in the control group. Post-test intergroup differences were medium to large, underscoring robust programme impact even under conservative Bonferroni correction. Interpreted via effect-size benchmarks, improvements are of practical significance and likely observable in everyday settings; illustrative calculations suggest an experimental-group pupil could outperform 69% to over 88% of control peers across various competencies. Results align with international meta-analytic evidence showing SEL’s positive effects on socio-emotional outcomes, engagement, and academic-related behaviors, and they contribute higher-quality causal evidence from an Ibero-American context. Educationally, strengthened SEL competencies in Basic VET pupils may facilitate school progression, enhance prospects for social and labor market integration, and support informed life choices.

The study demonstrates that a yearlong, curriculum-embedded SEL programme using sequenced, active, focused, and explicit instruction—augmented by mentoring, tutoring, vocational guidance, and personalized internships—significantly improves self-awareness, social awareness, self-control, relationship skills, and responsible decision-making among Basic VET pupils from disadvantaged contexts. The programme’s methodological rigor and outcomes support its consideration as an evidence-based intervention. Future research should broaden implementation across more VET programmes and years, employ designs with stronger randomization where feasible, incorporate additional covariates, expand sample sizes, and iteratively refine implementation processes (e.g., coordination among educators) to further validate and enhance impacts.

- Sampling and design constraints: convenience sampling and a quasi-experimental, non-equivalent control group (despite random assignment at classroom level) limit control over potential biases compared with fully randomized designs. - Covariate selection and measurement: future iterations should expand and refine covariates to strengthen assumptions of unconfoundedness. - Sample size and scope: although powered for expected effects, larger and more diverse samples across additional Basic VET programmes and years would improve generalizability. - Implementation/process factors: opportunities exist to improve coordination among teachers and educators based on process evaluation feedback.