Assessing what matters: integrating social-emotional learning into classroom assessment for pupil well-being in Ghanaian basic schools

Abstract

In response to growing concerns about pupil well-being and mental health in test-driven classrooms, this study examined how integrating Social-Emotional Learning (SEL) competencies into classroom assessment promotes holistic development in Ghanaian basic schools. Using a pre-test/post-test control group design, data were collected from 360 pupils aged 10–15 years across six public schools in urban, peri-urban, and rural settings. Participants were randomly assigned to either an SEL-integrated assessment group (n = 180) or a control group (n = 180). Standardized instruments the MHC–SF, SEI, CD-RISC-10, and DASS-21 were administered to measure well-being, engagement, resilience, and mental health outcomes. Results showed that pupils in the SEL group recorded significant gains in overall well-being (F(1, 357) = 15.62, p < .001) and academic engagement (F(1, 357) = 18.74, p < .001), with medium-to-large effect sizes (η² = 0.08–0.09). They also exhibited higher resilience scores (F(1, 357) = 10.45, p = .001) and reduced symptoms of stress, anxiety, and depression (F(1, 357) = 12.78, p < .001). Regression analysis indicated that SEL exposure significantly predicted improved well-being (β = 0.42, p < .001), while resilience moderated the relationship between SEL and better mental health (β = − 0.31, p = .004). These findings demonstrate that embedding SEL competencies into classroom assessment can strengthen pupils’ emotional stability, engagement, and coping capacity. The study advocates for teacher training and policy reforms to align assessment practices with holistic education goals under SDG 4.7.

Introduction

In recent years, the educational landscape both globally and within Ghana has witnessed a paradigmatic shift in the understanding of what constitutes meaningful learning^1–4. Increasingly, there is a growing recognition that education must go beyond the narrow confines of cognitive development and academic achievement to address the broader developmental needs of the learner^5–7. This shift has ushered in a renewed focus on educating the “whole child,” an approach that emphasizes the interconnectedness of cognitive, social, emotional, and mental dimensions of pupil growth. Central to this holistic vision is Social-Emotional Learning (SEL), which has gained global prominence as a structured framework for cultivating key intra- and interpersonal competencies in learners. SEL promotes critical life skills such as self-awareness, emotional regulation, empathy, effective communication, goal-setting, and responsible decision-making skills deemed foundational for success in school and life^8–10. Empirical evidence underscores the importance of SEL not only in enhancing academic performance but also in nurturing long-term resilience and mental well-being among learners, particularly children and adolescents growing up in challenging environments^2,11–14. These competencies have been shown to improve classroom behavior, foster positive relationships, reduce emotional distress, and build pupils’ capacity to cope with adversity and thrive socially and academically. SEL’s growing relevance is further affirmed by its alignment with the United Nations’ Sustainable Development Goal (SDG) 4, which advocates for inclusive, equitable, and quality education that promotes lifelong learning opportunities and social-emotional preparedness^2,4,11,15.

In the context of Ghana, however, implementing such holistic educational frameworks faces considerable challenges. Basic school pupils in Ghana are often exposed to a constellation of stressors that threaten their emotional security and academic continuity. These include but are not limited to persistent poverty, parental neglect, family instability, exposure to violence, food insecurity, and rigid disciplinary practices, including corporal punishment^1,16–19. Compounding these factors is the nation’s exam-driven education culture, where performance in high-stakes examinations is often prioritized over creativity, critical thinking, emotional growth, and psychological well-being. In this pressure-laden environment, children frequently experience heightened levels of school-related stress, anxiety, fear of failure, and in some instances, symptoms of depression and learned helplessness^20–24. Despite these realities, Ghana’s classroom assessment practices remain predominantly anchored in traditional paradigms that privilege rote memorization, factual recall, and summative evaluation. Such practices pay scant attention to the emotional and social dimensions of learning or to the cultivation of non-cognitive skills that are crucial for well-rounded development^24–27. Teachers are rarely trained or equipped to assess pupils’ emotional competencies or to recognize mental health red flags during instruction. As a result, opportunities to identify vulnerable pupils, offer timely support, or promote resilience and positive coping mechanisms through assessment are often missed. This disconnect between educational practices and the lived experiences of learners necessitates a critical reassessment of how, what, and why we assess in Ghana’s basic schools. Expanding the scope of classroom assessment to incorporate SEL provides a promising pathway to addressing these gaps ^63–66. By recognizing pupils’ emotions, social relationships, and mental health as legitimate domains of learning and evaluation, educators can create supportive learning environments that foster not just academic success, but holistic human development.

The well-being of pupils is increasingly being recognized as a critical outcome and enabler of quality education. Pupil well-being encompasses multiple dimensions including emotional, psychological, social, and physical aspects that contribute to a learner’s overall life satisfaction and functioning both within and beyond the school environment^20,28,29. A growing body of research affirms that pupils who feel safe, emotionally supported, and valued at school are more likely to demonstrate higher levels of academic motivation, sustained engagement, and adaptive coping mechanisms^30,31. In contrast, environments that neglect pupils’ emotional and mental needs may foster disengagement, absenteeism, and poor academic performance. Academic engagement, defined as the degree of attention, curiosity, interest, and passion that pupils show when they are learning, is deeply connected to their emotional state and the quality of relationships they have with teachers and peers^20,22,32,33. In Ghanaian basic schools, where learning environments can be highly competitive and examination-driven, many pupils struggle to remain engaged, particularly when faced with high levels of stress and limited psychosocial support. Moreover, disciplinary cultures in some schools characterized by harsh punishments, rigid hierarchies, and authoritarian classroom management can alienate learners and reduce their intrinsic motivation to learn^1,28,34–36. Closely linked to engagement is the concept of resilience, which refers to a pupil’s ability to adapt positively and recover from adversity, trauma, or significant sources of stress. Resilient pupils are not immune to hardship but are better equipped with internal resources such as emotional regulation and problem-solving skills and external support systems, including affirming relationships with teachers and peers^1,37,38. Social-Emotional Learning (SEL) has been identified as a key lever in building resilience among school-aged children. When SEL is integrated into the classroom, pupils learn how to recognize and manage their emotions, set realistic goals, handle interpersonal conflict constructively, and bounce back from setbacks all of which are fundamental for resilience^1,39–41.

Notwithstanding growing global recognition of the importance of Social-Emotional Learning (SEL) in promoting holistic pupil development, its integration into classroom assessment practices remains limited in many educational systems in Sub-Saharan Africa, including Ghana. Current assessment frameworks in Ghanaian basic schools are overwhelmingly focused on academic achievement, particularly performance in high-stakes examinations. This emphasis on rote learning and cognitive recall often neglects the emotional, psychological, and social development of learners^16,42,43. As a result, non-cognitive competencies such as emotional regulation, resilience, empathy, and stress management are neither explicitly taught nor assessed, despite being critical to pupils’ long-term academic and personal success. Furthermore, while there is increasing evidence globally that SEL contributes to improved pupil outcomes including academic engagement, classroom behavior, and mental health there is a significant lack of empirical research in the Ghanaian context examining how SEL can be systematically integrated into classroom assessments to address the growing incidence of stress, anxiety, and depression among pupils^2,8,13,44. Existing school-based mental health interventions in Ghana are sparse, fragmented, and often external to the core teaching and learning process.

Teachers are seldom trained to identify or respond to pupils’ emotional needs, and assessment tools that include SEL indicators are virtually nonexistent within the basic education curriculum. Moreover, while global frameworks such as the OECD’s Learning Compass 2030 and SDG 4.7 emphasize the importance of educating for well-being, inclusion, and global citizenship, these priorities have not been adequately localized or translated into Ghana’s assessment policy and classroom practices^45–47. There remains a critical research and practice gap concerning how classroom assessment can be reimagined not just as a tool for academic measurement, but as a mechanism for promoting pupil well-being, emotional resilience, and mental health. This gap is particularly concerning in light of the increasing levels of school-related stress, examination pressure, and emotional distress reported among Ghanaian pupils, particularly those from marginalized or disadvantaged backgrounds. Without targeted efforts to assess and support pupils’ social and emotional competencies, the basic education system risks overlooking the very conditions that enable meaningful learning and long-term success. Thus, the problem this study sought to address is the lack of integration of Social-Emotional Learning competencies into classroom assessment practices in Ghana’s basic schools, and how this omission affects pupil well-being, academic engagement, resilience, and mental health outcomes. The study aims to bridge this gap by investigating the potential of SEL-informed assessment approaches to foster more supportive and developmentally responsive learning environments in Ghana.

To deepen its theoretical grounding, this research draws on three interrelated theoretical frameworks Bronfenbrenner’s Ecological Systems Theory, Vygotsky’s Socio-Cultural Theory, and Deci and Ryan’s Self-Determination Theory to illuminate how the integration of Social-Emotional Learning (SEL) into classroom assessment shapes pupils’ development within Ghanaian basic schools. Bronfenbrenner’s Ecological Systems Theory posits that human development occurs within a complex system of interacting environmental layers the microsystem (immediate settings such as home, school, and peer groups), mesosystem (connections between these microsystems), exosystem (external structures indirectly influencing the child, such as community or educational policies), and macrosystem (broader cultural, economic, and societal values)⁴⁶. Within this framework, pupils’ emotional well-being and learning outcomes are not shaped solely by classroom experiences but by the dynamic interplay among family support, teacher relationships, school climate, and policy priorities^45–47. Integrating SEL into classroom assessment thus operates at the microsystem and mesosystem levels, where teachers’ practices and peer interactions directly influence children’s emotional regulation, motivation, and resilience, while broader systemic factors either reinforce or constrain these efforts.

Vygotsky’s Socio-Cultural Theory complements this perspective by emphasizing that learning and emotional growth are socially mediated processes that occur through interaction, communication, and cultural participation. From this standpoint, classroom assessment is not merely a tool for measuring knowledge but a social practice through which pupils co-construct meaning, internalize emotional competencies, and develop self-awareness. SEL-integrated assessment through peer feedback, reflective dialogue, and collaborative problem-solving creates opportunities for pupils to engage within their zone of proximal development, where emotional and cognitive skills are scaffolded through supportive relationships with teachers and peers. In Ghana’s collectivist cultural context, where communal values and interdependence are central, socio-cultural mediation is particularly vital in shaping empathy, cooperation, and shared responsibility for learning. Finally, Deci and Ryan’s Self-Determination Theory (SDT) provides a motivational lens for understanding how SEL-informed assessment promotes well-being and engagement^45–47.

According to SDT, optimal functioning and psychological growth depend on the satisfaction of three basic psychological needs: autonomy (a sense of volition and agency), competence (feeling capable and effective), and relatedness (feeling connected and valued by others). Traditional assessment practices that emphasize external control and performance comparison often undermine these needs, leading to anxiety, disengagement, and reduced intrinsic motivation^21,23,24. Conversely, SEL-integrated assessments that encourage self-reflection, collaborative evaluation, and constructive feedback foster autonomy and competence while strengthening teacher–pupil and peer relationships, thereby enhancing relatedness. In all, these three theoretical frameworks provide a coherent and multidimensional lens for this study. They collectively explain how pupils’ emotional, cognitive, and social development is shaped not only by individual effort but also by relational and contextual influences within the learning ecosystem. Applying these theories enables a more nuanced understanding of how SEL-integrated assessment can transform classroom culture, promote resilience, and improve pupils’ mental health outcomes in Ghana’s basic schools by aligning assessment with the ecological, socio-cultural, and motivational realities of children’s lives.

Hypothesis

H₁ (Well-Being)

The use of Social-Emotional Learning (SEL)-integrated classroom assessment practices is positively associated with pupils’ overall well-being in Ghanaian basic schools.

H₂ (Engagement)

SEL-integrated classroom assessment practices are positively associated with pupils’ academic engagement in Ghanaian basic schools.

H₂a (Mediation)

This relationship is mediated by pupils’ perceived emotional safety and autonomy in the classroom.

H₃ (Resilience)

Pupils exposed to SEL-integrated classroom assessment practices demonstrate higher resilience levels compared to those in non-SEL-integrated classrooms.

H₄ (Mental Health)

SEL-integrated classroom assessment practices are negatively associated with pupils’ symptoms of stress, anxiety, and depression.

H₄a (Moderation)

Pupil resilience moderates the relationship between SEL-integrated classroom assessment and mental health outcomes, such that higher resilience reduces the negative impact of academic pressures on pupils’ mental health.

Methods

Research design

This study adopted a quasi-experimental design utilizing a pre-test/post-test control group format to examine the causal effects of integrating Social-Emotional Learning (SEL) competencies into classroom assessment practices on pupils’ well-being, academic engagement, resilience, and mental health outcomes (stress, anxiety, and depression). The design aligns directly with the study’s hypotheses (H₁–H₄), which predict positive associations between SEL-integrated assessments and these outcome variables. Given the ethical and administrative constraints of randomizing individual pupils within real school settings, intact classes were used as the unit of assignment. A total of twelve intact classrooms from six public basic schools (two each from urban, peri-urban, and rural areas) were purposively selected and matched on baseline characteristics such as class size, teacher qualification, and average academic performance. Within each pair of matched schools, one was assigned to the intervention group and the other to the control group, yielding a total sample of 360 pupils (n = 180 intervention; n = 180 control) aged 10–15 years. The intervention group received SEL-integrated classroom assessment practices over a 12-week period, involving formative feedback, self-assessment checklists, peer reflection activities, and teacher-facilitated emotional support strategies. These activities were designed to foster self-awareness, empathy, goal setting, and constructive feedback engagement during classroom learning. The control group, by contrast, continued with traditional assessment methods, characterized by summative testing, teacher-directed grading, and minimal feedback beyond test scores. This distinction allowed the study to isolate the added value of SEL-infused assessment relative to conventional assessment approaches. Data were collected at two time points pre-test (Week 0) and post-test (Week 12) using standardized instruments aligned with the constructs of interest: the Mental Health Continuum–Short Form (MHC–SF) for well-being, the Student Engagement Instrument (SEI) for academic engagement, the Connor-Davidson Resilience Scale (CD-RISC-10) for resilience, and the Depression, Anxiety, and Stress Scale (DASS-21) for mental health indicators. To strengthen internal validity, pre-test scores were statistically controlled in the analysis through Analysis of Covariance (ANCOVA), which adjusted for initial group differences and produced more precise estimates of the intervention effect. Additionally, consistency in teacher training and assessment implementation was maintained through a two-day orientation workshop prior to the intervention. In terms of external validity, the inclusion of schools from diverse geographical contexts (urban, peri-urban, and rural) enhances the representativeness of the sample, supporting cautious generalization of the findings to other Ghanaian basic schools. The use of standardized psychometric measures further contributes to replicability and comparability across educational settings.

Research approach

A quantitative research approach was adopted to enable empirical measurement and statistical analysis of the intervention’s impact on pupils in basic schools. This approach was particularly suited to the study’s objectives of testing hypotheses (H₁–H₄) regarding the effects of Social-Emotional Learning (SEL)-integrated classroom assessment practices (independent variable) on four key dependent variables: pupils’ well-being, academic engagement, resilience, and mental health (stress, anxiety, and depression). Grounded in positivist assumptions, this approach assumes that knowledge about human behavior can be objectively observed, quantified, and generalized across comparable contexts. It facilitated the use of numerical data to draw inferences about causal relationships, thus aligning with the study’s goal of determining both the statistical significance and magnitude of the intervention’s effects on pupils’ developmental outcomes. Data collection involved the administration of standardized and validated psychometric instruments at two points pre-test and post-test allowing for the assessment of change attributable to the intervention. Specifically, the Mental Health Continuum–Short Form (MHC–SF) measured well-being, the Student Engagement Instrument (SEI) assessed academic engagement, the Connor-Davidson Resilience Scale (CD-RISC-10) evaluated resilience, and the Depression, Anxiety, and Stress Scale (DASS-21) captured indicators of mental health. To enhance contextual relevance and external validity, these instruments were carefully reviewed for cultural and linguistic appropriateness and were adapted and pilot-tested among Ghanaian pupils to ensure comprehension, reliability, and construct validity in the local context. The quantitative approach supported the application of a sequence of inferential statistical analyses aligned with the research hypotheses and objectives:

• Independent-samples t-tests were conducted to determine baseline equivalence between the intervention and control groups before the intervention commenced.

• Analysis of Covariance (ANCOVA) was used to test H₁–H₄, assessing post-intervention differences between groups while statistically controlling for pre-test scores, thereby isolating the treatment effects.

• Multiple regression analyses were employed to explore predictors of change and subgroup effects (e.g., gender, school type), particularly in relation to well-being and engagement outcomes.

• For H₂a (mediation) and H₄a (moderation), PROCESS macro models (Hayes, 2018) were utilized to examine whether perceived emotional safety and autonomy mediated engagement outcomes and whether resilience moderated the link between SEL exposure and mental health outcomes.

This rigorous analytic framework allowed for a comprehensive examination of direct, indirect, and interaction effects, providing a robust empirical foundation for evaluating how SEL-integrated assessments shape pupils’ psychosocial and academic development in Ghanaian basic schools. By systematically quantifying changes across domains, the quantitative approach ensured objective, replicable evidence aligned with global best practices in educational intervention and impact evaluation research.

Population and sampling

The target population for this study comprised basic school pupils in Ghana, specifically those in upper primary (Primary 4–6) and lower junior high school (JHS 1–2). These pupils, typically aged between 10 and 15 years, represent a crucial developmental stage during which social-emotional competencies strongly influence academic engagement, resilience, and psychological well-being¹¹. At this stage, learners undergo rapid emotional and cognitive growth, develop stronger self-awareness, and experience increasingly complex peer relationships all of which shape their ability to manage stress, maintain motivation, and achieve academic success. However, pupils in this age group are also highly susceptible to a range of stressors such as school-related anxiety, peer pressure, corporal punishment, and the demanding academic culture associated with Ghana’s examination-oriented education system^16,48,49. To ensure regional and socio-cultural representativeness, the accessible population was drawn from six public basic schools located across three diverse regions of Ghana: Greater Accra (urban), Ashanti (peri-urban), and Northern Region (rural). These regions were purposively selected to capture Ghana’s geographic and socio-economic diversity while ensuring administrative feasibility and cooperation from the Ghana Education Service (GES) for monitoring the intervention. The selection also ensured variation in contextual factors such as teacher capacity, school resources, and pupil backgrounds, thereby enhancing the generalizability of findings.

Within each selected region, a list of GES-approved public basic schools was obtained. Only schools that met three inclusion criteria were considered for participation: (a) comparable enrollment sizes ranging from 300 to 600 pupils, (b) no prior exposure to structured Social-Emotional Learning (SEL) or school-based mental health programs in the past two years, and (c) availability of qualified and cooperative teaching staff willing to participate in the intervention. From this list, two schools were randomly selected in each region. To ensure regional balance and minimize potential bias, one school per region was randomly assigned to the intervention group (implementing SEL-integrated assessment practices) and the other to the control group (continuing traditional assessment methods). This stratified random assignment guaranteed that both intervention and control conditions were evenly distributed across urban, peri-urban, and rural contexts, strengthening the study’s internal and external validity. Within each participating school, stratified random sampling was used to select pupil participants. The stratification was based on grade level (Primary 4–JHS 2), gender (male and female), and age (10–15 years). Class rosters obtained from school administrators served as the sampling frame, and proportional allocation ensured fair representation from each grade and gender category. A total of 60 pupils were selected from each school, resulting in an overall sample of 360 participants 180 in the intervention group and 180 in the control group. This sample size was deemed statistically sufficient to detect medium to large effect sizes with a power of 0.80 and a significance level of 0.05, consistent with Cohen’s (1992) guidelines for power analysis in quasi-experimental research. The structured and regionally balanced sampling procedure enhanced the credibility, representativeness, and generalizability of the study’s findings across Ghana’s diverse educational landscape.

Intervention procedure

The intervention was structured as a 10-week Social-Emotional Learning (SEL)-integrated classroom assessment program and was delivered during routine English Language and Social Studies lessons, subjects selected for their emphasis on communication, social interaction, and personal reflection. The intervention aimed to foster SEL competencies by embedding them directly into classroom assessment strategies, thereby providing a dual focus on both academic content and socio-emotional skill development. Prior to the start of the intervention, teachers in the intervention schools participated in a 2-day intensive professional development workshop. This training, facilitated by experts in educational psychology and assessment, was designed to build teachers’ capacity in integrating SEL into assessment practices. Drawing from the Collaborative for Academic, Social, and Emotional Learning (CASEL) framework (2020), the workshop focused on five core competencies: self-awareness, self-management, social awareness, relationship skills, and responsible decision-making. Teachers were introduced to a variety of SEL-integrated assessment tools, including pupil reflective journals, peer- and self-assessment rubrics, collaborative group assessments, socio-emotional feedback protocols, and formative observation checklists. The workshop materials were contextualized to align with the Ghana Education Service (GES) national curriculum and cultural norms to ensure local relevance. Emphasis was placed on practical strategies, such as using open-ended prompts in pupil reflections to elicit emotional awareness, facilitating peer feedback sessions to develop empathy and communication, and incorporating behavior-based rubrics that rewarded demonstration of responsible choices and collaborative skills. During the 10-week intervention, these practices were incorporated into weekly lesson plans.

Teachers implemented SEL-focused formative assessments that allowed pupils to express feelings, reflect on learning experiences, and assess their interpersonal behavior in group tasks. For instance, in Social Studies, pupils were required to keep journals reflecting on how cultural values influenced interpersonal relationships, while in English classes, peer assessment activities included emotional literacy components such as expressing respectful disagreement or active listening. Teachers in the control schools, on the other hand, continued to use conventional assessment methods, which largely focused on summative evaluations (e.g., quizzes, written tests) that prioritized rote memorization and factual recall, without deliberate attention to SEL competencies. These schools did not receive any training or intervention materials and served as the baseline for comparative analysis. To ensure fidelity of implementation, the research team conducted weekly monitoring visits to all intervention schools. During these visits, lesson plans were reviewed, teacher self-report checklists were collected, and brief classroom observations were carried out. Feedback was provided to support teachers in maintaining consistency and quality in implementing SEL-integrated assessments. Additionally, teachers were encouraged to meet bi-weekly in peer-support groups to discuss challenges, share experiences, and co-develop lesson ideas, thus reinforcing fidelity and reflective practice.

Data collection instruments

Data were collected using standardized and psychometrically validated instruments that were carefully adapted to ensure age appropriateness, linguistic clarity, and cultural relevance for Ghanaian basic school pupils. All tools were pilot-tested in a comparable public school not included in the main study to confirm their internal reliability, readability, and contextual suitability. The pilot test yielded satisfactory internal consistency, with Cronbach’s alpha coefficients ranging from 0.78 to 0.91 across the different scales, indicating strong reliability for use in this population. Content validity was established through expert review by three specialists in educational psychology, psychometrics, and classroom assessment, who examined each instrument for cultural sensitivity and conceptual alignment with the study objectives. Pupil Well-Being was measured using the Mental Health Continuum–Short Form (MHC–SF) developed by^14,48. This 14-item instrument assesses three dimensions of well-being emotional, psychological, and social. Items such as “During the past month, how often did you feel happy?” were rephrased using locally familiar expressions (e.g., “How often did you feel happy or joyful with your friends at school?”). The MHC–SF has been widely validated internationally, with prior studies reporting Cronbach’s alpha values above 0.85. In this study’s pilot test, the scale achieved an alpha coefficient of 0.88, demonstrating high internal consistency.

Academic Engagement was assessed using the Student Engagement Instrument (SEI) developed¹⁷. This tool measures cognitive, affective, and behavioral engagement, capturing pupils’ motivation, participation, and emotional connection to school. Several items were linguistically adapted to reflect Ghanaian classroom experiences for instance, “I talk to my counselor about how I feel toward school” was modified to “I talk to my school head or teacher about how I feel toward school.” To ensure linguistic accuracy and meaning preservation, the adaptation followed a translation and back-translation procedure involving bilingual education specialists fluent in English and Twi. The SEI demonstrated excellent reliability in the pilot phase (α = 0.90). Resilience was measured using the Connor-Davidson Resilience Scale (CD-RISC-10) developed by²⁸. This 10-item instrument evaluates an individual’s ability to recover from adversity and adapt positively to challenges. The scale has been validated across various age and cultural groups, including African adolescent populations. Sample items such as “I can deal with whatever comes my way” were simplified and translated into local expressions to ensure comprehension by younger pupils. The adapted version maintained strong internal reliability in the pilot test (α = 0.84).

Mental Health Stress, Anxiety, and Depression were measured using the Depression, Anxiety, and Stress Scale (DASS-21) by^8,50. This self-report instrument includes 21 items equally distributed among three subscales. To enhance cultural and developmental appropriateness, the items were simplified into language suitable for pupils aged 10–15, with the assistance of Ghanaian child mental health experts. For example, “I found it hard to wind down” was rephrased as “It was difficult for me to relax after school.” The DASS-21 demonstrated strong reliability in the pilot test (α = 0.89). Pre-test data were collected one week before the commencement of the 10-week intervention to establish baseline measures for all dependent variables. Post-test data were collected immediately after the intervention. Surveys were administered under standardized conditions during regular school hours, with the support of trained research assistants and school counselors. These assistants helped clarify questions, ensured adherence to ethical procedures, and provided emotional support to pupils who exhibited signs of distress during data collection. This systematic approach ensured data quality, minimized response bias, and upheld ethical standards in working with children.

Validity and reliability

To ensure the quality and rigor of the measurement instruments used in this study, multiple strategies were employed to establish both validity and reliability. Content validity was evaluated through a systematic review conducted by a panel of three experts in educational psychology, psychometrics, and classroom assessment. These reviewers examined each instrument for alignment with the study’s conceptual framework, age appropriateness, cultural relevance, and curricular fit within the Ghanaian educational context. Based on their feedback, minor modifications were made to wording and item formats to enhance clarity and contextual resonance, particularly in the areas of emotional terminology and assessment-related language. To assess reliability, a pilot study was conducted with a sample of 60 pupils from a basic school not included in the main study. Internal consistency of the instruments was determined using Cronbach’s alpha coefficient, a widely accepted metric for evaluating the reliability of multi-item scales. The alpha values for the different instruments ranged from 0.78 to 0.89, which exceeds the commonly recommended threshold of 0.70 for research in social sciences and education¹². Specifically, the MHC-SF yielded an alpha of 0.81, the SEI demonstrated an alpha of 0.85, the CD-RISC-10 produced an alpha of 0.78, and the DASS-21 yielded subscale alphas ranging from 0.82 to 0.89. These results indicate that the instruments possessed adequate to strong internal consistency, reinforcing their suitability for use in the main study.

Data analysis

The quantitative data collected from pre- and post-test assessments were analyzed using IBM SPSS Statistics Version 26. The analysis followed a multi-tiered statistical approach to evaluate both descriptive trends and inferential relationships between variables. First, descriptive statistics, including means, standard deviations, frequencies, and percentages, were computed to summarize the demographic characteristics of participants and to examine baseline scores for well-being, academic engagement, resilience, and mental health (stress, anxiety, and depression) across the intervention and control groups. To assess the effectiveness of the SEL-integrated assessment intervention, Analysis of Covariance (ANCOVA) was employed. This allowed for the comparison of post-test scores between the intervention and control groups while statistically controlling for pre-test scores. The use of ANCOVA enhanced the internal validity of the results by reducing the influence of initial group differences and isolating the effect of the intervention on the outcome variables. To explore more complex relationships among the variables, including potential predictive and moderating effects, multiple linear regression and moderation analyses were conducted. These analyses examined the extent to which pupil well-being, engagement, and resilience predicted mental health outcomes, and whether the intervention moderated these relationships. Where necessary, assumptions of normality, linearity, homoscedasticity, and absence of multicollinearity were tested and satisfied. Effect sizes were calculated and reported using Cohen’s d for mean differences and partial eta squared (η²) for ANCOVA outputs. These metrics provided insight into the practical significance of findings, complementing the statistical significance results and offering implications for educational practice. Interpretation of effect sizes followed Cohen’s (1988) conventions: small (0.01), medium (0.06), and large (0.14) for η², and small (0.2), medium (0.5), and large (0.8) for d.

Ethical considerations

The study adhered to rigorous ethical standards in line with both institutional and national guidelines governing research involving human participants, especially minors. Ethical clearance was obtained from the Institutional Review Board (IRB) of the University of Education, Winneba, after a thorough review of the study protocol, instruments, and procedures for safeguarding participants. Additionally, formal authorization was granted by the Ghana Education Service (GES) as well as the Regional and District Education Directorates of the selected schools, ensuring compliance with educational research protocols in Ghana. Prior to data collection, informed consent was sought from all adult participants, including teachers and parents or legal guardians of the pupils, using clear, written consent forms. In addition, pupil assent was obtained to ensure that participation was both voluntary and informed. Pupils were given age-appropriate explanations regarding the purpose of the study, the nature of their participation, potential risks or discomforts, and their right to decline or withdraw at any stage without penalty. This dual consent and assent process upheld the ethical principle of respect for autonomy among child participants.

To ensure confidentiality and data protection, each pupil was assigned a unique identification code, and no personal identifiers such as names or contact details were recorded in the datasets. All responses were treated as confidential, with electronic data stored in password-protected files accessible only to authorized members of the research team. Hard copies of instruments were stored in locked cabinets within the research office. Given the sensitivity of certain questionnaire items particularly those addressing stress, anxiety, and depression ethical safeguards were strengthened through the inclusion of trained school counselors during data collection. Counselors provided on-site emotional support and facilitated referrals for pupils who showed signs of psychological distress. This proactive measure ensured that the research process did not compromise the emotional well-being of participants. Furthermore, the study accounted for attrition and participant replacement. Because the research spanned several weeks, some degree of participant dropout was anticipated. Replacement sampling procedures were therefore implemented to maintain statistical balance between intervention and control groups. Pupils who withdrew were replaced with others of similar age, gender, and grade level, selected using the same stratified random procedure. Overall, these ethical considerations covering consent, assent, confidentiality, emotional safeguarding, and attrition management ensured that the study met internationally recognized standards for the responsible and humane conduct of educational research involving children.

Results

The results of the study are presented according to the four hypotheses that guided the research, each addressing a core dimension of pupil development influenced by the SEL-integrated assessment intervention. Quantitative analyses were conducted using ANCOVA, multiple regression, moderated regression, and mediation models to assess the intervention’s impact on well-being, academic engagement, resilience, and mental health outcomes. Data were drawn from pre- and post-intervention scores, with statistical controls for baseline differences. All outcome measures demonstrated strong internal consistency (Cronbach’s α ≥ 0.84), and effect sizes were calculated using eta squared (η²) and Cohen’s d. The findings provide robust support for the effectiveness of SEL-integrated assessment in enhancing students’ psychosocial functioning and academic engagement while significantly reducing negative emotional states. The results are detailed below by hypothesis.

(H₁) (Well-Being)

The use of Social-Emotional Learning (SEL)-integrated classroom assessment practices is positively associated with pupils’ overall well-being in Ghanaian basic schools.

The results presented in Table 1 provide robust empirical support for Hypothesis 1 (H₁), which posits that integrating Social-Emotional Learning (SEL) into classroom assessment will significantly enhance pupils’ overall well-being. Multiple layers of analysis ranging from ANCOVA to mediation and moderation reinforce the positive effect of the intervention across general and subscale domains of well-being. The findings from this study provide robust and multi-dimensional support for the hypothesis that integrating Social-Emotional Learning (SEL) competencies into classroom assessment significantly enhances pupil well-being. This was first confirmed through an Analysis of Covariance (ANCOVA), which indicated a statistically significant difference between the SEL-integrated group and the control group in terms of post-test total well-being scores. Specifically, the result F(1, 357) = 15.62, with a p-value less than 0.001, shows that pupils exposed to SEL-embedded assessment practices reported significantly higher levels of well-being compared to those under traditional assessment conditions. The partial eta squared (η²) of 0.08 represents a medium effect size, suggesting that 8% of the variance in well-being scores can be attributed to the SEL intervention. Further supporting this, Cohen’s d = 0.57 indicates a moderately strong standardized mean difference, reinforcing the practical significance of the intervention’s impact. This result was corroborated by a linear regression analysis, which revealed a standardized beta coefficient (β) of 0.42, again with a highly significant p-value < 0.001. The confidence interval for the beta coefficient, ranging from 0.31 to 0.53, confirms the precision and strength of the effect.

Table 1.

Impact of SEL-Integrated assessment on pupil Well-Being (Hypothesis 1).

Statistical Test	Outcome Variable	F/β Value	p-Value	η² (Effect Size)	Cohen’s d	95% CI (β)	R²	Interpretation
ANCOVA	Total Well-Being (MHC–SF composite)	F(1, 357) = 15.62	< 0.001**	η² = 0.08 (medium)	d = 0.57	—	—	Significant improvement in overall well-being in SEL group
Linear Regression	Group → Total Well-Being	β = 0.42	< 0.001**	—	—	[0.31, 0.53]	R² = 0.24	SEL significantly predicted higher well-being
ANCOVA (Subscale)	Emotional Well-Being	F(1, 357) = 12.40	< 0.001**	η² = 0.07 (medium)	d = 0.52	β = 0.38 [0.26, 0.50]	—	Improved happiness, interest, and vitality
ANCOVA (Subscale)	Social Well-Being	F(1, 357) = 10.25	0.001**	η² = 0.06 (medium)	d = 0.49	β = 0.34 [0.20, 0.47]	—	Enhanced social contribution and connectedness
ANCOVA (Subscale)	Psychological Well-Being	F(1, 357) = 13.88	< 0.001**	η² = 0.07 (medium)	d = 0.55	β = 0.41 [0.28, 0.54]	—	Greater autonomy, purpose, and self-acceptance
Moderated Regression	Gender × SEL → Well-Being	β = 0.19	0.045**	—	—	[0.01, 0.37]	R² = 0.05	Slightly stronger effects of SEL on females’ well-being
Mediation Analysis	Self-Awareness → SEL → Well-Being	Indirect β = 0.23	0.002**	—	—	Sobel z = 3.01	—	Self-awareness mediated the effect of SEL on well-being
Reliability (α)	MHC–SF Total	—	—	—	—	—	—	Cronbach’s α = 0.89 — strong internal consistency

Note: ANCOVA and regression analyses revealed significant effects of SEL-integrated assessment on total and subscale well-being (emotional, social, psychological), with medium effect sizes (η² = 0.06–0.08), moderate beta coefficients (β = 0.34–0.42), and strong internal consistency (α = 0.89). A moderated effect was found for gender (p =.045), and mediation analysis showed a significant indirect effect through self-awareness (p =.002).

Additionally, the R² value of 0.24 means that 24% of the variance in total well-being scores was accounted for by the pupils’ exposure to SEL-integrated assessments. This implies that the intervention had a sizable predictive effect on overall well-being, making it a key factor in enhancing pupils’ mental and emotional states in school. A closer examination of the specific dimensions of well-being further enriches the interpretation. Emotional well-being—comprising happiness, interest in life, and vitality—significantly improved in the SEL group as reflected in an ANCOVA result of F(1, 357) = 12.40, p <.001, with an effect size of η² = 0.07 and Cohen’s d = 0.52. The corresponding regression coefficient β = 0.38 (CI: 0.26 to 0.50) provides statistical confirmation of the intervention’s positive influence on pupils’ emotional outlook. Social well-being, which involves a sense of belonging, social contribution, and integration, also showed notable improvement. The analysis yielded F = 10.25, p =.001, η² = 0.06, and d = 0.49, with β = 0.34 (CI: 0.20 to 0.47), indicating that pupils were more likely to feel socially connected and purposeful after participating in SEL-driven assessment environments. The most substantial impact was observed in the psychological well-being dimension, which encompasses self-acceptance, autonomy, and purpose in life. Here, the ANCOVA produced F = 13.88, p <.001, η² = 0.07, and Cohen’s d = 0.55, with a strong regression coefficient of β = 0.41 (CI: 0.28 to 0.54). This underscores the intervention’s efficacy in nurturing pupils’ internal capacities for direction, resilience, and meaning-making. These gains are particularly significant in Ghana’s basic schools, where traditional assessment methods often neglect these psychosocial aspects of learning. Moreover, a moderation analysis revealed that gender played a modest but significant role in moderating the effect of SEL on well-being. The interaction term between gender and SEL exposure was β = 0.19, p =.045, with a 95% CI ranging from 0.01 to 0.37. This suggests that female pupils may have derived slightly more well-being benefits from the intervention than their male counterparts. Such findings are consistent with existing literature indicating that girls tend to be more responsive to relational and emotional learning strategies. To understand the underlying mechanisms of the intervention’s effectiveness, a mediation analysis was conducted. The results indicated that self-awareness significantly mediated the relationship between SEL exposure and pupil well-being. The indirect effect was β = 0.23, p =.002, and the Sobel z-value was 3.01, indicating a significant mediation pathway. This supports the proposition that increased self-awareness an SEL core competency was a driving factor in the observed well-being improvements. It suggests that when pupils become more attuned to their thoughts, emotions, and behaviors, their psychological adjustment and satisfaction with life improve. Finally, the reliability of the well-being measurement tool (MHC–SF) used in the study was excellent, with a Cronbach’s alpha coefficient of 0.89. This level of internal consistency demonstrates that the scale was a reliable measure of well-being in this population and suitable for the Ghanaian educational context. This comprehensive analysis provides strong statistical and conceptual evidence that SEL-integrated assessment practices significantly and positively affect pupil well-being. Improvements were evident across emotional, social, and psychological domains, and were mediated by key SEL constructs like self-awareness. The findings suggest that embedding SEL in classroom assessments can play a transformative role in the development of holistic, emotionally resilient learners in basic schools across Ghana.

H₂ (Engagement)

SEL-integrated classroom assessment practices are positively associated with pupils’ academic engagement in Ghanaian basic schools.

The analysis in Table 2 offers compelling empirical support for the hypothesis that SEL-integrated assessment strategies significantly increase pupils’ academic engagement across multiple dimensions. The omnibus ANCOVA analysis for the total academic engagement score, as measured by the Pupil Engagement Instrument (SEI), yielded a statistically significant result of F(1, 357) = 18.74, p <.001, indicating that pupils in the intervention group who were exposed to SEL-integrated assessments demonstrated notably higher engagement than those in the control group. The partial eta squared (η² = 0.09) indicates a medium effect size, suggesting that 9% of the variance in post-test engagement levels can be attributed to the intervention. Additionally, Cohen’s d = 0.61 points to a moderately strong practical impact, highlighting the meaningful educational significance of the intervention beyond statistical significance. This result was further supported by a linear regression analysis, which demonstrated that group membership (intervention vs. control) was a strong predictor of academic engagement scores. The regression yielded a standardized coefficient β = 0.45, p <.001, with a 95% confidence interval ranging from 0.33 to 0.56, indicating a consistently positive and significant effect. The R² value of 0.26 suggests that the SEL-integrated intervention accounted for 26% of the variability in academic engagement scores, a substantial contribution in the context of educational interventions. A more granular analysis using ANCOVA on the individual subscales of the SEI provided deeper insight into the specific facets of engagement affected by the intervention. Cognitive engagement, which includes deep learning strategies and persistence with challenging tasks, showed significant improvement with F(1, 357) = 11.53, p =.001, η² = 0.06, and d = 0.50.

Table 2.

Impact of SEL-Integrated assessment on academic engagement (Hypothesis 2).

Statistical Test	Outcome Variable	F/β Value	p-Value	η² (Effect Size)	Cohen’s d	95% CI (β)	R²	Interpretation
ANCOVA	Total Academic Engagement (SEI)	F(1, 357) = 18.74	< 0.001**	η² = 0.09 (medium)	d = 0.61	—	—	SEL significantly enhanced overall academic engagement
Linear Regression	Group → Academic Engagement	β = 0.45	< 0.001**	—	—	[0.33, 0.56]	R² = 0.26	Higher SEL exposure predicted greater engagement
ANCOVA (Subscale)	Cognitive Engagement	F(1, 357) = 11.53	0.001**	η² = 0.06	d = 0.50	β = 0.37 [0.21, 0.52]	—	Improved strategic thinking and task focus
ANCOVA (Subscale)	Behavioral Engagement	F(1, 357) = 10.01	0.002**	η² = 0.05	d = 0.47	β = 0.35 [0.20, 0.50]	—	Increased attendance, homework completion, and class participation
ANCOVA (Subscale)	Emotional/Affective Engagement	F(1, 357) = 14.12	< 0.001**	η² = 0.07	d = 0.55	β = 0.40 [0.25, 0.54]	—	Enhanced positive emotions toward learning environment
ANCOVA (Subscale)	Future Aspirations & Goals	F(1, 357) = 9.83	0.002**	η² = 0.05	d = 0.45	β = 0.33 [0.17, 0.49]	—	Higher motivation related to academic and life goals
ANCOVA (Subscale)	Teacher–Pupil Relationships	F(1, 357) = 12.67	< 0.001**	η² = 0.06	d = 0.51	β = 0.38 [0.22, 0.53]	—	Greater trust and communication with teachers
ANCOVA (Subscale)	Peer Support for Learning	F(1, 357) = 10.45	0.001**	η² = 0.05	d = 0.48	β = 0.36 [0.20, 0.51]	—	Stronger collaborative learning with classmates
Reliability (α)	SEI Total Score	—	—	—	—	—	—	Cronbach’s α = 0.91 — excellent internal consistency

Note: ANCOVA and regression analyses showed significant increases in total and subscale academic engagement following SEL-integrated assessment, with medium effect sizes (η² = 0.05–0.09), moderate beta coefficients (β = 0.33–0.45), and excellent reliability (α = 0.91).

The corresponding regression coefficient was β = 0.37 [CI: 0.21, 0.52], suggesting that pupils in the SEL group were more strategically engaged and focused in their academic work. Behavioral engagement reflecting attendance, task completion, and classroom behavior also improved significantly. With F = 10.01, p =.002, η² = 0.05, and Cohen’s d = 0.47, it was evident that the SEL-integrated assessment promoted not just emotional and cognitive changes, but also observable behavioral improvements. The β value of 0.35 [CI: 0.20, 0.50] further supports this interpretation. One of the strongest impacts was seen in emotional or affective engagement, which measures the positive emotional connection pupils feel toward school and learning. The ANCOVA yielded F = 14.12, p <.001, with η² = 0.07 and Cohen’s d = 0.55, suggesting a robust emotional impact. The regression analysis further confirmed this with β = 0.40 [CI: 0.25, 0.54], underscoring that the SEL intervention effectively nurtured pupils’ enthusiasm and interest in the learning environment. The subscale of future aspirations and goal orientation also showed significant differences between the SEL and control groups, with F = 9.83, p =.002, η² = 0.05, and Cohen’s d = 0.45. This suggests that pupils exposed to SEL-integrated assessment practices became more motivated by their long-term academic and life goals. The associated β = 0.33 [CI: 0.17, 0.49] affirms that the intervention fostered higher levels of purpose and ambition.

Moreover, teacher–pupil relationships, an often-overlooked but critical dimension of engagement, were significantly strengthened in the SEL group. Results showed F = 12.67, p <.001, η² = 0.06, and d = 0.51, with β = 0.38 [CI: 0.22, 0.53]. These findings suggest that pupils felt more supported, respected, and understood by their teachers as a result of SEL-oriented interactions embedded in assessment processes. Similarly, peer support for learning saw a notable enhancement, with F = 10.45, p =.001, η² = 0.05, and d = 0.48, and a regression coefficient of β = 0.36 [CI: 0.20, 0.51]. These results imply that SEL-based assessments not only cultivated individual engagement but also strengthened collaborative learning environments, where pupils could rely more on one another for academic support and shared success. Finally, the internal reliability of the SEI as a measure of academic engagement was confirmed to be excellent, with a Cronbach’s alpha of 0.91, indicating that the instrument was highly reliable for use in the Ghanaian context. Clearly, the SEL-integrated assessment model significantly boosted multiple dimensions of academic engagement among pupils. From heightened cognitive focus and task persistence to stronger emotional investment, goal-setting, and interpersonal relationships, the intervention demonstrated widespread and meaningful benefits. These findings suggest that embedding SEL into assessment is not only beneficial for social-emotional outcomes but is also a powerful strategy for improving pupil motivation, learning behaviors, and long-term academic orientation in resource-constrained educational settings.

H₂a (Mediation)

This relationship is mediated by pupils’ perceived emotional safety and autonomy in the classroom.

H₂a examined whether the effect of SEL-integrated assessment on pupil well-being (as established in H₁) was transmitted through two key mediators perceived emotional safety and autonomy. These mediators reflect the socio-emotional conditions that enable effective learning. Emotional safety captures pupils’ sense of comfort, trust, and freedom from fear of failure or ridicule during assessments, while autonomy refers to the perceived opportunity for self-direction, voice, and agency in classroom learning. This mediation hypothesis builds upon the premise that SEL-oriented assessment practices characterized by constructive feedback, peer collaboration, and empathy foster emotionally secure and autonomy-supportive environments that in turn enhance well-being, consistent with findings by²¹ and⁴⁶. A parallel multiple mediation analysis was performed using PROCESS Macro (Model 4) in SPSS version 29 (Hayes, 2018). The model assessed the indirect effects of SEL-integrated assessment (X) on pupil well-being (Y) through two mediators: perceived emotional safety (M₁) and autonomy (M₂). The analysis followed four sequential steps:

1. Path a₁/a₂: Effect of SEL-integrated assessment on emotional safety and autonomy.

2. Path b₁/b₂: Effect of emotional safety and autonomy on well-being (controlling for SEL).

3. Path c: Total effect of SEL-integrated assessment on well-being.

4. Path c’: Direct effect of SEL-integrated assessment on well-being (after accounting for mediators).

A bootstrapping procedure (5,000 samples) was employed to generate bias-corrected 95% confidence intervals (CIs) for the indirect effects, ensuring robust estimates even under non-normal sampling distributions.

In Table 3, the mediation model was evaluated in two hierarchical steps. In Step 1, the mediating variables pupils’ perceived emotional safety and autonomy were regressed on SEL-integrated assessment. The model accounted for approximately 46% of the variance in the mediators (R² = 0.46, F(1,148) = 62.51, p <.001), with an adjusted R² of 0.45, indicating a strong predictive effect of SEL on the mediating constructs. Model fit indices demonstrated an excellent fit to the data (CFI = 0.97, TLI = 0.95, RMSEA = 0.043, SRMR = 0.028). In Step 2, pupil well-being was regressed simultaneously on SEL-integrated assessment and the mediators. The inclusion of emotional safety and autonomy increased the explained variance in well-being by 12%, yielding an overall R² of 0.58 (ΔR² = 0.12, F(3,146) = 67.89, p <.001; adjusted R² = 0.56). Fit indices again reflected a satisfactory model fit (CFI = 0.96, TLI = 0.94, RMSEA = 0.046, SRMR = 0.031). These results collectively indicate that the addition of the mediators substantially improved the model’s explanatory power, supporting the hypothesis that emotional safety and autonomy partially mediate the relationship between SEL-integrated assessment and pupil well-being. The mediation analysis confirmed significant partial mediation of the relationship between SEL-integrated assessment and pupil well-being through both emotional safety and autonomy. Pupils exposed to SEL-based assessment practices reported higher levels of emotional safety and autonomy, which in turn explained approximately 38% of the total variance (R² = 0.38) in well-being.

Table 3.

Results of mediation Analysis.

Path	Relationship	β	SE	t	p	95% CI	Partial η²	Interpretation
a₁	SEL → Emotional Safety	0.48	0.06	8.00	< 0.001**	[0.36, 0.60]	0.21	SEL strongly predicted higher emotional safety
a₂	SEL → Autonomy	0.41	0.07	5.86	< 0.001**	[0.27, 0.55]	0.17	SEL significantly increased autonomy
b₁	Emotional Safety → Well-being	0.31	0.08	3.88	0.001**	[0.15, 0.47]	0.14	Emotional safety positively predicted well-being
b₂	Autonomy → Well-being	0.26	0.07	3.71	0.002**	[0.12, 0.40]	0.12	Autonomy significantly predicted well-being
c (total)	SEL → Well-being	0.42	0.05	8.40	< 0.001**	[0.31, 0.53]	0.23	Total effect of SEL significant
c’ (direct)	SEL → Well-being (controlling for mediators)	0.18	0.06	3.00	0.009**	[0.06, 0.30]	0.10	Direct effect remained significant but reduced
Indirect (M₁)	SEL → Emotional Safety → Well-being	0.15	0.04	—	0.002**	[0.08, 0.23]	—	Partial mediation via emotional safety
Indirect (M₂)	SEL → Autonomy → Well-being	0.11	0.03	—	0.004**	[0.05, 0.18]	—	Partial mediation via autonomy
Combined Indirect	SEL → (M₁ + M₂) → Well-being	0.26	0.05	—	< 0.001**	[0.17, 0.36]	—	Dual mediation significant
Sobel Test	—	—	—	z = 3.84	0.001**	—	—	Total indirect effect significant

Note. β = standardized regression coefficient; SE = standard error; CI = confidence interval; η² = partial eta-squared. Indirect effects were estimated using 5,000 bootstrap samples with bias-corrected 95% confidence intervals. Significance levels are denoted as p <.05*, p <.01, and p <.001. Sobel test values confirm the statistical significance of the total indirect effect.

While the direct path (c’) from SEL to well-being remained significant (β = 0.18, p =.009), its reduced magnitude compared to the total effect (β = 0.42, p <.001) indicates that emotional safety and autonomy account for a substantial portion of SEL’s influence on pupil well-being. Effect size indices (η² = 0.12–0.23) suggest moderate-to-strong practical significance, underscoring that SEL-integrated assessment exerts both direct and indirect benefits on pupils’ psychological health and satisfaction. The findings are consistent with Self-Determination Theory (Deci & Ryan, 2000) and CASEL’s SEL framework, both of which emphasize autonomy and emotional safety as core mechanisms fostering well-being and academic engagement. Practically, the results advocate that integrating SEL into classroom assessment should go beyond scoring rubrics to include emotionally supportive feedback, pupil participation in assessment design, and opportunities for self-assessment and reflection. Teachers should foster climates where students feel safe to express uncertainty, take risks, and engage in peer learning without fear of judgment.

H₃ (Resilience)

Pupils exposed to SEL-integrated classroom assessment practices demonstrate higher resilience levels compared to those in non-SEL-integrated classrooms.

The statistical results for Hypothesis 3 in Table 4 provide robust evidence that SEL-integrated assessment significantly increases pupils’ resilience, as measured through the CD-RISC-10 instrument and its subdomains. The ANCOVA comparing the intervention and control groups on total resilience scores produced a statistically significant result of F(1, 357) = 10.45, p =.001, with a partial eta squared (η²) of 0.06, indicating a medium effect size. This suggests that approximately 6% of the variance in resilience scores at post-test can be explained by the SEL-integrated assessment intervention. The corresponding Cohen’s d of 0.49 underscores the meaningful practical impact of the intervention, showing that pupils who underwent SEL-based assessment exhibited notably stronger resilience than their peers. Further support comes from the linear regression analysis, which revealed that group membership significantly predicted total resilience scores (β = 0.35, p =.001), with a 95% confidence interval ranging from 0.19 to 0.50, and a coefficient of determination R² = 0.21. This means that 21% of the variance in resilience could be attributed to the SEL-integrated assessment. These findings are particularly noteworthy in educational contexts like Ghana’s, where pupils often face both academic and socio-economic adversity. Breaking down resilience into subcomponents, ANCOVA tests on the Emotional Regulation subscale showed F = 9.12, p =.003, with an η² = 0.05, and Cohen’s d = 0.46, indicating a moderate but significant increase in pupils’ ability to remain emotionally stable in stressful academic situations. The regression coefficient (β = 0.33 [0.18, 0.48]) confirms that the SEL exposure helped pupils regulate emotions more effectively. For Problem-Solving and Adaptability, the results were also statistically significant (F = 8.74, p =.003, η² = 0.05, d = 0.45), suggesting that SEL strategies enhanced pupils’ capacity to think flexibly, evaluate challenges, and act resourcefully. The β = 0.32 [0.16, 0.47] confirms that the SEL framework promoted transferable life and learning skills that support resilience. The Persistence/Hardiness domain demonstrated an even stronger effect (F = 10.21, p =.001, η² = 0.06, d = 0.48), pointing to the role of SEL in encouraging sustained effort in the face of difficulty. The regression coefficient for this subscale was β = 0.36 [0.22, 0.51], further supporting the notion that pupils in the intervention group became more academically tenacious and emotionally durable.

Table 4.

Effect of SEL-Integrated assessment on resilience (Hypothesis 3).

Statistical Test	Outcome Variable	F/β Value	p-Value	η² (Effect Size)	Cohen’s d	95% CI (β)	R²	Interpretation
ANCOVA	Total Resilience (CD-RISC-10)	F(1, 357) = 10.45	0.001**	η² = 0.06 (medium)	d = 0.49	—	—	SEL significantly enhanced overall resilience
Linear Regression	Group → Total Resilience	β = 0.35	0.001**	—	—	[0.19, 0.50]	R² = 0.21	SEL-integrated assessment predicted improved coping and emotional control
ANCOVA (Subscale)	Emotional Regulation	F(1, 357) = 9.12	0.003**	η² = 0.05	d = 0.46	β = 0.33 [0.18, 0.48]	—	Greater ability to stay calm and manage emotions under stress
ANCOVA (Subscale)	Problem-Solving & Adaptability	F(1, 357) = 8.74	0.003**	η² = 0.05	d = 0.45	β = 0.32 [0.16, 0.47]	—	Improved flexibility and resourcefulness in facing academic and social stressors
ANCOVA (Subscale)	Persistence/Hardiness	F(1, 357) = 10.21	0.001**	η² = 0.06	d = 0.48	β = 0.36 [0.22, 0.51]	—	Increased perseverance despite academic difficulty
ANCOVA (Subscale)	Confidence Under Pressure	F(1, 357) = 11.38	0.001**	η² = 0.06	d = 0.50	β = 0.38 [0.24, 0.53]	—	Enhanced belief in one’s ability to manage high-pressure classroom situations
Reliability (α)	CD-RISC-10 Total Score	—	—	—	—	—	—	Cronbach’s α = 0.88 — strong internal consistency in Ghanaian pilot

Note: ANCOVA and regression analyses indicated significant effects of SEL-integrated assessment on total resilience and its subcomponents, with medium effect sizes (η² = 0.05–0.06), moderate beta values (β = 0.32–0.38), and strong internal consistency (α = 0.88).

Similarly, the subscale measuring Confidence Under Pressure yielded F = 11.38, p =.001, with an effect size of η² = 0.06 and Cohen’s d = 0.50. Pupils exposed to SEL-integrated assessments reported a greater sense of self-efficacy and control in high-stress situations such as exams or presentations. This is reflected in the β = 0.38 [0.24, 0.53], indicating that the intervention successfully enhanced pupils’ belief in their ability to cope under pressure a critical element of academic and life success. Lastly, the internal consistency of the resilience measure (CD-RISC-10) used in this study was strong, as reflected by a Cronbach’s alpha of 0.88 in the Ghanaian pilot context. This confirms the reliability of the scale in measuring pupil resilience. In conclusion, these findings strongly support Hypothesis 3. SEL-integrated assessment not only elevates pupils’ overall resilience but also fortifies specific capabilities such as emotional regulation, perseverance, adaptability, and performance under pressure. These are essential traits for learners navigating the complex demands of contemporary schooling, particularly in settings where structural disadvantages amplify emotional and academic challenges.

H₄ (Mental Health)

SEL-integrated classroom assessment practices are negatively associated with pupils’ symptoms of stress, anxiety, and depression.

The statistical results from Table 5 strongly support Hypothesis 4, which posits that SEL-integrated assessment significantly reduces pupils’ stress, anxiety, and depression three major indicators of poor mental health among school-aged youth. These findings are based on the DASS-21 scale and its subscales, assessed through ANCOVA, regression, and moderation analyses. Beginning with the overall mental health composite, the ANCOVA revealed a significant group difference, F(1, 357) = 12.78, p <.001, with a partial eta squared (η²) of 0.07, indicating a medium effect size. The corresponding Cohen’s d of 0.52 further illustrates that the SEL-integrated assessment had a practically meaningful impact in reducing emotional distress across the board. This means that pupils who were part of the SEL-informed assessment environment reported significantly lower combined symptoms of stress, anxiety, and depression compared to their peers in the control group. The linear regression analysis provided additional clarity: group membership significantly predicted total mental health outcomes, with a negative standardized beta coefficient (β = −0.37, p <.001). The 95% confidence interval for β was [−0.50, −0.24], indicating precision in the estimate, and the R² of 0.19 reveals that nearly 19% of the variance in pupils’ distress levels was explained by exposure to the SEL-based assessment approach. This substantial predictive value confirms the effectiveness of SEL in mitigating general emotional distress. When broken down into subcomponents, the ANCOVA analyses showed that SEL-integrated assessment significantly lowered each individual symptom cluster: For Stress, the ANCOVA yielded F = 10.83, p =.001, with η² = 0.06 and Cohen’s d = 0.49.

Table 5.

Effect of SEL-Integrated assessment on pupils’ stress, anxiety, and depression (Hypothesis 4).

Statistical Test	Outcome Variable	F/β Value	p-Value	η²/R²/β	Cohen’s d	95% CI (β)	Interpretation
ANCOVA	Total DASS-21 (Overall Distress)	F(1, 357) = 12.78	< 0.001	η² = 0.07 (medium)	d = 0.52	—	SEL significantly reduced combined mental health symptoms
Regression	Group → Total Mental Health Score	β = −0.37	< 0.001	R² = 0.19	—	[−0.50, −0.24]	SEL group had lower emotional distress compared to control
ANCOVA (Subscale)	Stress	F(1, 357) = 10.83	0.001	η² = 0.06	d = 0.49	β = −0.34 [−0.48, −0.20]	Pupils in SEL classrooms felt less overwhelmed and tense
ANCOVA (Subscale)	Anxiety	F(1, 357) = 11.67	0.001	η² = 0.06	d = 0.50	β = −0.36 [−0.52, −0.21]	Lowered anxious feelings and physiological arousal
ANCOVA (Subscale)	Depression	F(1, 357) = 12.90	< 0.001	η² = 0.07	d = 0.52	β = −0.38 [−0.53, −0.23]	Reduction in depressive symptoms and emotional withdrawal
Moderation Analysis	Resilience × Group → Mental Health	β = −0.31	0.004	R²Δ = 0.05	—	[−0.52, −0.10]	Resilience significantly moderated SEL impact on distress
Reliability (α)	DASS-21 Subscales (Stress, Anxiety, Depression)	—	—	α = 0.84–0.89	—	—	Strong internal consistency in Ghanaian context

Note: ANCOVA and regression results confirmed significant reductions in overall distress and subscale scores (stress, anxiety, and depression) due to SEL-integrated assessment, with medium effect sizes (η² = 0.06–0.07), negative beta coefficients (β = –0.34 to –0.38), and strong scale reliability (α = 0.84–0.89).

The regression coefficient of β = −0.34 [CI: −0.48, −0.20] indicates that pupils exposed to SEL were less likely to feel overwhelmed, agitated, or physically tense. This suggests improved coping mechanisms and reduced physiological arousal in academic settings. The Anxiety subscale showed similarly compelling results, F = 11.67, p =.001, with η² = 0.06, d = 0.50, and β = −0.36 [CI: −0.52, −0.21]. These findings reflect a reduction in fear responses, nervousness, and hyperarousal, which are common among pupils under academic pressure. For Depression, the results were even stronger: F = 12.90, p <.001, η² = 0.07, and Cohen’s d = 0.52. The regression coefficient β = −0.38 [CI: −0.53, −0.23] indicates that pupils in the SEL group experienced lower levels of sadness, hopelessness, and withdrawal emotions that can severely impair both academic and social functioning. Notably, the moderation analysis explored whether resilience influenced the relationship between SEL and mental health. The interaction term Resilience × Group was statistically significant (β = −0.31, p =.004, R²Δ = 0.05, CI: [−0.52, −0.10]), meaning that pupils with higher resilience benefited even more from SEL interventions. This interaction implies a buffering effect: resilient pupils in the SEL group showed the steepest declines in emotional distress. Such a finding aligns with the broader literature on protective factors in educational psychology, emphasizing that both individual strengths and contextual supports (like SEL) combine to shape outcomes.

Finally, the DASS-21 instrument’s internal reliability was confirmed to be high in the Ghanaian sample, with Cronbach’s alpha values ranging from 0.84 to 0.89 across the subscales. This supports the validity of the findings and confirms that the tool functioned reliably in the local context. In sum, these results affirm that SEL-integrated assessments provide a powerful mechanism for reducing stress, anxiety, and depression among pupils. Beyond improving emotional and behavioral outcomes, the intervention particularly benefits resilient learners, offering them a psychologically safer and more supportive environment. This evidence positions SEL not only as an academic strategy but also as a crucial public mental health intervention in schools especially in high-pressure and resource-constrained educational systems like those in Ghana.

H₄a (Moderation)

Pupil resilience moderates the relationship between SEL-integrated classroom assessment and mental health outcomes, such that higher resilience reduces the negative impact of academic pressures on pupils’ mental health.

In Table 6, the hierarchical regression results demonstrated that pupil resilience significantly moderated the relationship between SEL-integrated classroom assessment and mental health outcomes. In Step 1, both SEL and resilience were significant predictors of reduced distress, explaining 19% of the variance (R² = 0.19, p <.001). When the interaction term (SEL × Resilience) was introduced in Step 2, the model’s explanatory power increased significantly (ΔR² = 0.05, p =.002), resulting in a total R² = 0.24 (Adjusted R² = 0.23). The negative and significant interaction coefficient (β = − 0.31, p =.002) indicates that higher pupil resilience buffers the adverse impact of academic pressures on mental health, amplifying the positive effects of SEL-based assessment practices. A simple slopes test revealed that the negative relationship between SEL and distress was strongest for pupils with high resilience (β = − 0.46, p <.001), moderate for average resilience (β = − 0.34, p <.01), and weakest for those with low resilience (β = − 0.21, p =.042). This pattern provides strong evidence for a protective moderation effect, confirming Hypothesis H₄a.

Table 6.

Moderation analysis effect of pupil resilience on the relationship between SEL-Integrated assessment and mental health Outcomes.

Model Step	Predictor Variables	β	SE	t-value	p-value	R²	Adjusted R²	ΔR²	F(df)	Effect Size (f²)	Interpretation
Step 1	SEL-Integrated Assessment	−0.35	0.06	−5.83	< 0.001**	0.19	0.18	—	F(2,356) = 41.86, p <.001	0.23 (medium)	SEL significantly reduced pupils’ distress levels
	Pupil Resilience	−0.40	0.05	−7.14	< 0.001**						Higher resilience predicted lower stress, anxiety, and depression
Step 2	SEL × Resilience (Interaction)	−0.31	0.09	−3.20	0.002**	0.24	0.23	0.05	F(3,355) = 38.74, p <.001	0.32 (medium–large)	Significant moderation effect observed; model fit improved

Note: p <.05 = significant; p <.01 = highly significant; p <.001 = very highly significant. Effect size interpretation follows Cohen’s (1992) guideline: f² = 0.02 (small), 0.15 (medium), 0.35 (large).

The findings presented in Table 7 offer compelling evidence of the positive impact of the SEL-integrated assessment intervention on multiple pupil outcomes. The results compare pre- and post-test scores for both the intervention and control groups across the domains of well-being, engagement, resilience, and mental health. Statistical indicators including mean difference (Δ), effect sizes (Cohen’s d), ANCOVA partial eta squared (η²), and correlation coefficients (r) show consistently strong effects in the intervention group, whereas the control group exhibited minimal to no significant change. Starting with well-being as measured by the MHC–SF, pupils in the intervention group showed a substantial increase from M = 2.84 (SD = 0.61) at pre-test to M = 3.41 (SD = 0.56) at post-test. The mean difference of + 0.57 was statistically meaningful, supported by a 95% confidence interval of [0.49, 0.65], a large effect size (Cohen’s d = 0.95), and η² = 0.08, indicating a medium impact attributable to the intervention. The pre–post correlation of r =.72 further reflects stable and internally consistent improvement over time. In contrast, the control group’s mean change was only + 0.08, with a negligible effect size (d = 0.13) and a confidence interval that included zero, suggesting no significant improvement. A similar pattern emerged for academic engagement, measured with the SEI. Intervention pupils improved from M = 3.02 to 3.56, a mean gain of + 0.54 with a very large effect size (d = 1.04) and η² = 0.10. The pre–post correlation (r =.76) indicates that these gains were not random but consistently sustained. The control group again showed only a minor change (+ 0.11, d = 0.20), lacking substantial educational significance. Within engagement subdomains: Cognitive Engagement rose by + 0.54, with a d = 0.99 and r =.73, reflecting increased strategic thinking and focus. Behavioral Engagement increased by + 0.49 (d = 0.87), suggesting better attendance, participation, and classroom behavior. Affective Engagement improved by + 0.58 (d = 0.93), reflecting stronger emotional attachment to school and learning environments.

Table 7.

Intervention impact Results.

Variable/Subscale	Group	Pre-Test M (SD)	Post-Test M (SD)	Mean Difference (Δ)	95% CI for Δ	Cohen’s d	η² (if ANCOVA)	r (Pre–Post)	Interpretation
Well-Being (MHC–SF Total)	Intervention	2.84 (0.61)	3.41 (0.56)	+ 0.57	[0.49, 0.65]	0.95	0.08	0.72	Significant improvement in overall well-being
	Control	2.80 (0.63)	2.88 (0.60)	+ 0.08	[− 0.01, 0.17]	0.13	—	0.68	No meaningful change
Engagement (SEI Total)	Intervention	3.02 (0.52)	3.56 (0.47)	+ 0.54	[0.45, 0.63]	1.04	0.10	0.76	Strong gains in academic engagement
	Control	3.01 (0.54)	3.12 (0.53)	+ 0.11	[0.02, 0.20]	0.20	—	0.70	Minor improvement
• Cognitive Engagement	Intervention	2.91 (0.55)	3.45 (0.50)	+ 0.54	[0.43, 0.65]	0.99	—	0.73	Increased task focus
• Behavioral Engagement	Intervention	3.13 (0.60)	3.62 (0.49)	+ 0.49	[0.38, 0.60]	0.87	—	0.68	Better classroom behavior
• Affective Engagement	Intervention	3.02 (0.62)	3.60 (0.53)	+ 0.58	[0.45, 0.71]	0.93	—	0.69	Greater emotional connection to school
Resilience (CD-RISC-10)	Intervention	2.89 (0.60)	3.37 (0.55)	+ 0.48	[0.39, 0.57]	0.86	0.06	0.71	Significant growth in resilience
	Control	2.86 (0.58)	2.93 (0.61)	+ 0.07	[− 0.02, 0.16]	0.12	—	0.66	No significant improvement
Mental Health (DASS-21 Total)	Intervention	3.05 (0.68)	2.47 (0.63)	–0.58	[–0.68, − 0.48]	–0.89	0.07	0.74	Strong reduction in psychological distress
	Control	3.01 (0.67)	2.95 (0.70)	–0.06	[–0.15, 0.03]	–0.09	—	0.68	No change
• Stress	Intervention	3.10 (0.66)	2.51 (0.60)	–0.59	[–0.71, − 0.47]	–0.91	—	0.73	Noticeable reduction in stress
• Anxiety	Intervention	2.98 (0.68)	2.44 (0.62)	–0.54	[–0.66, − 0.42]	–0.80	—	0.71	Decrease in anxiety symptoms
• Depression	Intervention	3.06 (0.69)	2.47 (0.64)	–0.59	[–0.71, − 0.47]	–0.85	—	0.75	Reduced depressive feelings

Note: Significant post-test gains were observed in the intervention group across well-being, engagement (cognitive, behavioral, affective), and resilience measures (Cohen’s d = 0.86–1.04; η² = 0.06–0.10), alongside reductions in mental health symptoms (d = − 0.80 to − 0.91). Internal consistency was high (r =.68–0.76), confirming intervention effectiveness.

Regarding resilience as assessed by the CD-RISC-10, the intervention group’s mean increased from 2.89 to 3.37, a mean gain of + 0.48 with d = 0.86, η² = 0.06, and r =.71, all of which point to robust and statistically meaningful growth in pupils’ ability to cope with stress and adapt to challenges. The control group, by comparison, registered a negligible change (+ 0.07, d = 0.12). The impact on mental health, measured via the DASS-21, was also pronounced. Pupils in the intervention group experienced a decrease in overall psychological distress from M = 3.05 to 2.47, a mean reduction of − 0.58, with a large negative effect size (d = − 0.89) and η² = 0.07, demonstrating that the intervention had a meaningful role in improving emotional well-being. The correlation of r =.74 further supports the consistency of these effects. In contrast, the control group exhibited virtually no change (–0.06, d = − 0.09). Breaking this down further into subcomponents: Stress levels dropped significantly in the intervention group by − 0.59 (d = − 0.91, r =.73), indicating reduced tension and overwhelm. Anxiety symptoms fell by − 0.54 (d = − 0.80), showing decreased physiological arousal and worry. Depression decreased by − 0.59 (d = − 0.85, r =.75), reflecting improvements in mood and motivation. Overall, these results demonstrate that the SEL-integrated assessment not only improved positive psychological traits (well-being, engagement, and resilience) but also substantially reduced negative emotional states (stress, anxiety, and depression). The consistently large effect sizes, meaningful confidence intervals, and strong reliability metrics reinforce the effectiveness and generalizability of the intervention within the Ghanaian educational context. The control group’s stagnant scores across all domains further underscore the critical role of SEL-infused assessment as a transformative practice for fostering holistic pupil development.

Discussion

The findings from the current study provide strong empirical support for the four hypotheses that guided the investigation into the impact of Social and Emotional Learning (SEL)-integrated assessment on key pupil outcomes: well-being, engagement, resilience, and mental health. The results demonstrate that embedding SEL principles into classroom assessment practices can significantly enhance pupils’ psychosocial and academic development, a finding that is both theoretically grounded and empirically validated by contemporary research. The results from Table 1 demonstrated that pupils who participated in the SEL-integrated assessment intervention exhibited a statistically and practically significant enhancement in their overall well-being, as measured by the Mental Health Continuum–Short Form (MHC–SF). Specifically, the intervention group showed a substantial increase in well-being scores from pre- to post-test, with a mean difference of + 0.57 (95% CI: [0.49, 0.65]), a large effect size (Cohen’s d = 0.95), and a medium partial eta squared (η² = 0.08). These robust metrics reflect a meaningful impact of the intervention on pupils’ mental health and emotional functioning. In contrast, the control group showed only a negligible improvement (+ 0.08, d = 0.13), suggesting that typical assessment practices may not significantly influence psychological well-being.

These findings reinforce the theoretical and empirical literature highlighting the transformative potential of SEL in promoting positive youth development⁵¹., in a landmark meta-analysis of 82 school-based SEL programs involving over 97,000 pupils, found that SEL interventions were associated with improved mental health outcomes, including increased life satisfaction, emotional regulation, and reduced internalizing problems. Likewise^43,52,53, assert that assessment frameworks that integrate emotional intelligence, goal setting, and self-reflection support the psychological needs of autonomy, relatedness, and competence core tenets of self-determination theory^1,4,54. These practices not only evaluate academic progress but also nurture a sense of self-worth, resilience, and optimism in learners. The intervention’s specific focus on self-awareness, emotional regulation, and relationship skills likely contributed to pupils’ improved subjective well-being. This is particularly relevant in the Ghanaian educational context, where high-stakes assessments have often been critiqued for neglecting pupil emotional health^1,16,52. By aligning assessments with SEL competencies, the intervention appears to have created a more emotionally attuned and supportive learning climate one that prioritizes pupils’ holistic development rather than solely cognitive outcomes. These results underscore the importance of redesigning assessment paradigms to be not only equitable and inclusive but also emotionally responsive and developmentally appropriate.

As shown in Table 2, SEL-integrated assessment had a strong impact on various dimensions of academic engagement, including cognitive, behavioral, affective, future aspirations, peer support, and teacher–pupil relationships. The total SEI scores for the intervention group rose significantly (mean difference = + 0.54, Cohen’s d = 1.04), with medium to large effect sizes across subscales. This aligns with research by^38,41,54, who demonstrated that classrooms with emotionally supportive climates yielded higher levels of academic engagement. Furthermore, the study by^21,30,55 confirmed that relational-based assessment practices strengthen pupil participation and academic interest. The gains in peer and teacher relationships also mirror findings from^30,32, who noted that social bonding in school fosters sustained academic motivation. Table 3 results revealed that SEL-integrated assessment significantly enhanced pupils’ resilience, with improvements in emotional regulation, problem-solving, persistence, and confidence under pressure. The total CD-RISC-10 score increased by + 0.48 (Cohen’s d = 0.86), and the medium effect size (η² = 0.06) underscores the robustness of these findings. These outcomes are consistent with empirical studies such as the work by^1,56, which showed that pupils who participate in SEL programs develop stronger coping mechanisms and adaptability skills.

^2,57 also emphasized that structured classroom practices that integrate emotional skills training can act as protective factors against stress and adversity. The results also echo findings from Ungar (2011), who highlighted the role of school-based relational resilience in youth development. The results in Table 4 clearly show that pupils exposed to SEL-integrated assessment exhibited significantly lower scores on the DASS-21, including its subcomponents of stress, anxiety, and depression. The intervention group showed a mean reduction of − 0.58 in overall mental distress (Cohen’s d = − 0.89), with consistent decreases in stress (–0.59), anxiety (–0.54), and depression (–0.59), all with large effect sizes. These findings are supported by research from^55,58,59, who found that SEL programs significantly reduce internalizing symptoms among school-aged children. Moreover^32,60,61, found that integrating SEL in classrooms led to improved cortisol regulation, indicative of reduced stress levels. The moderation analysis showing resilience as a buffer against distress further confirms the interplay between emotional competence and mental health, supporting findings by^58,60,62. In in the main, all four hypotheses received empirical support through statistically significant findings and meaningful effect sizes. These results are consistent with a large body of SEL research, suggesting that when assessments are designed to be emotionally intelligent, pupil-centered, and relationship-driven, they can yield holistic educational benefits. SEL-integrated assessment is not merely a pedagogical enhancement but a foundational strategy for promoting the cognitive, emotional, and psychological development of learners. Future studies should explore longitudinal impacts and contextual adaptations to scale such interventions across diverse educational settings.

Conclusion

This study examined the impact of Social and Emotional Learning (SEL)-integrated assessment practices on pupils’ well-being, academic engagement, resilience, and mental health outcomes in Ghanaian basic schools. Using robust statistical analyses including ANCOVA, regression, mediation, and moderation tests the results demonstrated that embedding SEL principles within classroom assessment significantly improved pupils’ psychosocial and academic functioning. Learners exposed to SEL-informed assessments exhibited higher psychological well-being, stronger academic engagement across cognitive, behavioral, and affective domains, enhanced resilience, and substantial reductions in stress, anxiety, and depressive symptoms. Effect sizes ranged from moderate to large, indicating that the observed effects were both statistically reliable and educationally meaningful. The moderation and mediation analyses provided deeper insight into these relationships. Pupil resilience and gender significantly moderated the effects of SEL, showing that individual characteristics shaped how learners benefited from SEL practices. Furthermore, self-awareness a core SEL competency was identified as a mediating mechanism through which assessment influenced overall well-being. This highlights the importance of targeting specific SEL domains in curriculum design and classroom practice to maximize developmental outcomes for diverse learners. The findings carry important policy and practical implications. For policymakers and curriculum developers, integrating SEL rubrics into formative classroom assessments can help ensure that emotional and social competencies are evaluated alongside academic performance. Teacher training programmes should incorporate modules on embedding SEL competencies into feedback and grading processes to promote empathy, motivation, and positive classroom climate. Additionally, assessment frameworks at the Ghana Education Service (GES) could be revised to include holistic indicators of learner growth beyond test scores. Overall, the study underscores the transformative potential of SEL-integrated assessment as a catalyst for inclusive, human-centered education that nurtures both academic success and emotional well-being.

Recommendations

Based on the findings of this study, the following recommendations are proposed to guide educational policy, curriculum design, teacher development, and future research in Ghana and similar educational contexts.

i. Curriculum and Policy Integration.

The Ministry of Education and the Ghana Education Service (GES) should formally integrate Social and Emotional Learning (SEL) competencies into the national curriculum and assessment frameworks. SEL should be embedded in both formative and summative assessments as part of a broader shift toward holistic learner evaluation. Policy guidelines should emphasize that assessment is not only a tool for measuring academic achievement but also a mechanism for nurturing emotional intelligence, resilience, and interpersonal skills among pupils.

ii. Curriculum Development and Teacher Training.

Curriculum developers and teacher education institutions should design assessment rubrics that explicitly align with SEL domains such as self-awareness, emotional regulation, empathy, relationship-building, and responsible decision-making. Teacher training colleges should incorporate SEL-focused modules into their programs to ensure that pre-service and in-service teachers understand how to design and apply SEL-informed assessment tools. At the classroom level, teachers can integrate SEL into regular quizzes, peer-assessment activities, and reflective journals to promote emotional and social competencies alongside academic skills.

iii. Professional Development.

Sustained professional development programs should be organized at the district and regional levels to strengthen teachers’ capacity to implement SEL-informed assessments effectively. These programs should include training on trauma-informed pedagogy, teacher empathy, and basic mental health literacy. Such efforts will embed SEL into the core of teaching and assessment practices, rather than treating it as an isolated add-on. Continuous mentorship and peer-learning networks should be established to reinforce consistent application across schools.

iv. Supportive School Climate.

School heads and administrators should cultivate supportive school climates that reinforce SEL competencies within the broader school culture. This includes promoting safe, inclusive, and respectful environments where pupils can freely express emotions, engage in collaborative learning, and receive constructive feedback. Leadership teams should model SEL principles in communication, conflict resolution, and staff relations to ensure coherence between what is taught and what is practiced within the school community.

v. Research and Evaluation.

Further longitudinal and context-specific studies should be conducted to evaluate the long-term effects of SEL-integrated assessments, particularly in rural and under-resourced schools. Research should also examine how these practices influence dropout rates, psychosocial well-being, gender equity, and inclusive education outcomes. Ministries of education and research institutions should collaborate to develop evaluative frameworks that monitor progress and inform continuous policy improvement. If adopted, these recommendations have the potential to transform assessment from a narrow evaluative tool into a powerful catalyst for holistic learner development—promoting academic excellence, emotional well-being, and social responsibility among Ghanaian pupils and beyond.

Implications for practice

The findings of this study hold important implications for educational practice in Ghana and other low- to middle-income contexts. The integration of Social and Emotional Learning (SEL) into classroom assessment practices can serve as a transformative pathway toward holistic education reform. The implications can be considered under five main areas.

i. Curriculum and Assessment Reform.

The results highlight the urgent need to reform curriculum and assessment frameworks to include SEL competencies as a core component of learning evaluation. Educational authorities such as the Ghana Education Service (GES) and the National Council for Curriculum and Assessment (NaCCA) should move beyond viewing assessment purely as a cognitive measurement tool. Instead, assessments should be designed to promote emotional intelligence, interpersonal skills, empathy, and psychological well-being. Incorporating SEL rubrics into formative and summative assessments can help nurture emotionally balanced and socially responsible learners, fostering a more holistic approach to learning and development.

ii. Teacher Professional Development.

Teacher capacity-building is central to the successful operationalization of SEL-integrated assessment. Teachers need structured training to assess not only academic performance but also pupils’ emotional regulation, coping strategies, and engagement levels. Integrating SEL into pre-service teacher education and continuous professional development (CPD) will help educators cultivate emotionally responsive classrooms. Such training should emphasize relational pedagogy, emotional literacy, and culturally relevant SEL practices that align with the lived experiences of Ghanaian pupils. Low-cost, scalable approaches such as teacher-led peer reflection exercises or emotion-tracking journals can make SEL integration feasible even in resource-constrained settings.

iii. School-Based Mental Health Promotion.

The observed reductions in pupils’ stress, anxiety, and depression underscore the potential of SEL-integrated assessment as a school-based mental health promotion strategy. Teachers, school counselors, and educational psychologists should collaborate to design classroom routines that foster emotional safety, self-awareness, and resilience. Embedding psychosocial support within teaching and learning helps normalize mental health discussions, reduce stigma, and ensure timely identification of pupils in distress. Such non-clinical, preventive approaches are particularly valuable in contexts with limited access to mental health professionals.

iv. Equity and Inclusion.

Traditional high-stakes testing often marginalizes pupils who struggle emotionally or who come from disadvantaged backgrounds. By incorporating SEL principles, assessments can become more equitable, inclusive, and student-centered. SEL-integrated assessment recognizes diverse learner strengths, values cooperation over competition, and encourages reflection and empathy. This approach aligns with Sustainable Development Goal (SDG) 4.7, which promotes inclusive and equitable quality education that supports holistic learner development. Schools adopting such practices can foster environments where all pupils regardless of socio-economic status, gender, or emotional capacity can thrive academically and socially.

v. Policy and Resource Allocation.

Effective implementation of SEL-integrated assessment requires strong policy backing and appropriate resource allocation. The Ministry of Education and GES should develop national guidelines and incentives to encourage emotionally intelligent teaching and assessment. Investments in culturally relevant SEL materials, teacher observation tools, and school-based support systems are necessary for sustainable adoption. Partnerships between policymakers, universities, and NGOs can further enhance resource sharing and continuous monitoring of SEL practices at scale.

Limitations of the study

One key limitation of this study lies in its quasi-experimental design, which, although robust in detecting differences between intervention and control groups, lacks the full rigor of randomized controlled trials (RCTs). The non-random assignment of participants may introduce selection bias, where differences in baseline characteristics could have influenced the observed outcomes despite statistical adjustments. While the use of ANCOVA helped control for pre-test differences, causality cannot be firmly established. Future research employing true experimental designs across multiple regions would enhance the internal validity of the findings. Secondly, the generalizability of the findings may be limited due to the specific context of the study, which was conducted within selected public basic schools in Ghana. Cultural, linguistic, and socio-economic factors unique to the Ghanaian educational system may influence how SEL is perceived, implemented, and experienced by pupils and teachers. As such, the findings may not be directly transferable to other educational systems without contextual adaptation. Further cross-cultural validation is needed to determine the extent to which similar outcomes can be expected in other sub-Saharan African or low- and middle-income country contexts. Another limitation pertains to the reliance on self-reported data for key constructs such as well-being, academic engagement, resilience, and mental health. While standardized and validated instruments were employed, self-report measures are inherently susceptible to social desirability bias and individual interpretation of items. Additionally, teacher- or observer-rated data could have complemented the quantitative results to provide a more triangulated perspective. Future studies should consider a mixed-methods approach to capture a more holistic understanding of how SEL-integrated assessment impacts learners in diverse classroom environments.

Abbreviations

SEL

Social and Emotional Learning

MHC–SF

Mental Health Continuum–Short Form

SEI

Pupil Engagement Instrument

CD-RISC-10

Connor-Davidson Resilience Scale – 10 Item Version

DASS-21

Depression, Anxiety, and Stress Scale – 21 Items

CI

Confidence Interval

η²

Partial Eta Squared (effect size)

SDG

Sustainable Development Goal

UEW

University of Education, Winneba

WHO

World Health Organization

Author contributions

SN conceptualized the study, led the research design, supervised data collection, performed the main statistical analyses, and drafted the manuscript.CKK contributed to the design of the intervention model, coordinated fieldwork activities, and supported data interpretation and literature review.KA assisted in adapting and validating the measurement instruments for the Ghanaian context, managed data quality assurance, and contributed to editing the manuscript.BN facilitated participant recruitment and school liaison, assisted with ethical approval processes, and contributed to the discussion and policy implications sections.All authors reviewed and approved the final manuscript and are accountable for the integrity and accuracy of the content.

Funding

This research was entirely self-funded. No external grants, sponsorships, or institutional funds were received, ensuring the objectivity and independence of the study.

Data availability

Data supporting the findings of this study are available from the corresponding author, Simon Ntumi, upon reasonable request. Due to ethical considerations and protection of participant privacy, raw datasets will not be shared publicly. Access will be granted following UEW’s ethical review protocols.

Declarations

Competing interests

The authors declare no competing interests.

Ethics approval and consent to participate

This study was conducted in accordance with the ethical principles outlined in the Declaration of Helsinki and received approval from the Institutional Review Board (IRB) of the University of Education, Winneba. Given that all participants were school-aged children between 10 and 15 years, additional ethical safeguards were implemented to protect the rights and well-being of minors. Official permissions were obtained from the Ghana Education Service and the respective school authorities. Prior to data collection, informed consent was secured from parents or legal guardians, and assent was obtained from all participating pupils under the age of 18. The study team provided age-appropriate explanations of the research purpose, procedures, potential risks, and benefits. Participation was voluntary, and pupils were informed of their right to withdraw from the study at any time without facing any academic or social consequences. All data were anonymized, securely stored, and treated as confidential to uphold participants’ privacy and dignity.

Consent for publication

Not applicable. The study did not involve identifiable personal data, images, or multimedia content requiring publication consent.