Psychology by BogdanovaThe Encyclopedia of Psychology for the Artificial Era

RESEARCH DOSSIER

Social and Emotional Learning in Schools

Abstract. Social and emotional learning (SEL) in schools comprises structured instruction and educational environments that support emotional understanding, self-regulation, interpersonal competence, and responsible action. This fundamental research monograph reconstructs the developmental, preventive, and educational origins of SEL; distinguishes its competency models, instructional mechanisms, measurement strategies, and whole-school systems; and critically synthesizes experimental evaluations and meta-analytic evidence for psychosocial, behavioral, and academic outcomes. Major research programs include PATHS, RULER, and international studies of universal school-based interventions. The evidence demonstrates favorable average effects in several outcome domains, alongside program heterogeneity, implementation-dependent results, limited causal knowledge of active components, and important negative trials. Separate analyses investigate follow-up persistence, international transportability, cultural responsiveness, accessibility, student privacy, and digitally mediated instruction in the Artificial Era. The synthesis identifies research priorities for causal identification, construct validation, independent replication, and equitable implementation. The method is a critical narrative review of identifiable scientific literature; the work reports no original participant data or independently conducted systematic review or meta-analysis.

Keywords: social and emotional learning; social-emotional competence; SEL; school-based universal prevention; emotional literacy; self-regulation; relationship skills; prosocial behavior; academic achievement; school climate; PATHS curriculum; RULER approach; cluster-randomized trials; meta-analysis; implementation fidelity; developmental psychology; inclusive education; cultural adaptation; digital SEL; Artificial Era

The Scientific Object of School-Based Social and Emotional Learning

Social and emotional learning (SEL) in schools is the organized development of knowledge, capacities, dispositions, and context-sensitive practices involved in understanding emotions, regulating conduct, appreciating other perspectives, sustaining relationships, and making decisions with consequences for oneself and others. In educational research, the object comprises both psychological competencies and the instructional and institutional systems through which those competencies are learned, practiced, and assessed. A classroom lesson about identifying frustration is one instance; coordinated teacher practices, school routines, opportunities for student participation, and family partnerships are additional components of the same educational ecology. The CASEL (n.d.) framework characterizes SEL as a process spanning young people and adults, and Greenberg et al. (2003) argues for coordinating social, emotional, and academic development. The central empirical question is how particular educational exposures change particular student outcomes in particular settings, including their durability, costs, and unequal distribution.

School-based SEL is an intervention domain rather than a single psychological faculty. Emotion recognition, self-regulation, empathy, behavioral inhibition, interpersonal negotiation, responsible judgment, perceived belonging, and academic engagement can vary independently. Programs use different combinations of lessons, teacher development, supportive classroom structures, and whole-school practices. A change in a survey score therefore establishes only a change on that particular measure; it cannot automatically demonstrate transferable emotional competence or improved learning. The scientific unit of analysis must specify the learner, the competency, the environment, the mechanism, the intervention, the comparison condition, and the observation interval. Cipriano et al. (2023) demonstrates the field's range by synthesizing hundreds of studies and more than two hundred distinct interventions, while finding meaningful variation associated with content, context, and implementation.

This monograph analyzes universal school-based SEL primarily across kindergarten through the end of secondary education and draws on preschool research where it illuminates developmental foundations. It examines targeted and digital interventions for comparison but keeps universal curricular and whole-school provision at the center. Its method is a critical narrative synthesis of traceable original studies, randomized evaluations, measurement research, and published systematic reviews; no new experiment, preregistered review, or independently computed meta-analysis is claimed. Research questions concern conceptual validity, pathways of change, academic and psychosocial outcomes, implementation, cultural transportability, and causal inference. Related fields—clinical treatment, school-wide discipline, teacher–student relationships, and general achievement motivation—remain distinct subjects whose interactions with SEL require explicit explanation rather than automatic equivalence.

Definitions, Boundaries, and Competing Educational Aims

SEL is variously framed as competency education, positive youth development, preventive mental health, character education, or classroom climate improvement. These traditions partly intersect but supply different evaluative criteria. Competency approaches assess acquired skills and their expression; preventive approaches emphasize changes in risk and protective factors; relational approaches examine the quality of interactions and institutional trust. The influential five-domain structure in CASEL (n.d.) organizes self-awareness, self-management, social awareness, relationship skills, and responsible decision-making. These are broad curricular categories, not five directly observed unitary traits. A measurement model must establish what counts as evidence for each category and whether its meaning is stable across ages, languages, disability profiles, and school cultures.

A learner may describe emotions accurately yet struggle to regulate behavior under pressure, may cooperate within a familiar group while failing to interpret unfamiliar social norms, or may report empathy while taking no prosocial action. Such dissociations make the distinction between verbal knowledge, enacted skill, motivation, and situational opportunity scientifically indispensable. School-based SEL also differs from psychotherapy: universal instruction teaches general competencies within an educational institution; treatment evaluates a specified clinical condition and intervention for an identified need. Universal SEL can complement school mental-health services, while evidence for improved classroom self-regulation does not establish efficacy for a psychiatric disorder. Jones & Bouffard (2012) emphasizes strategies embedded in ordinary school activity, which need evaluation separate from branded curricula.

Normative disagreements arise because emotional expression, appropriate assertiveness, conformity, independence, interpersonal respect, and collective responsibility can be defined differently by students, families, and institutions. Assessment must investigate whose preferences shape a competency rubric and whether an intervention cultivates agency and supportive relationships or simply increases compliance with adult expectations. An equitable SEL design should provide room for disagreement, culturally variable expressions of emotion, and students' interpretations of situations. A school might reduce recorded disciplinary referrals by changing adult responses rather than student competence, making policy change an alternative explanation. Conceptual clarity thus precedes effectiveness claims: investigators must distinguish the educational outcome intended, the measure actually used, and the social standard imposed.

Historical Formation of an Educational Research Field

The contemporary SEL movement emerged from several longer research traditions: developmental studies of emotion and social competence, school prevention science, social-cognitive learning, educational psychology, and efforts to integrate student well-being with academic instruction. Its vocabulary became prominent during the late twentieth century as prevention researchers increasingly examined both children's skills and the social environments supporting them. The Greenberg et al. (2003) synthesis represents an influential integration of prevention and coordinated social, emotional, and academic learning. Jones & Bouffard (2012) later described a movement from stand-alone programs toward repeatable strategies within everyday schooling. These publications document historical shifts in research emphasis, although the psychological processes under examination precede the label SEL.

Early developmental evidence helped establish the plausibility of teaching emotional knowledge and regulation. In a study of 143 preschool children, Denham et al. (2003) examined emotional expressiveness, regulation, and knowledge in relation to concurrent and later social competence. Such prospective associations supported inquiry into developmental pathways while leaving open whether instructional manipulation would reproduce the same relations. Prevention trials then evaluated curricular packages with more direct causal designs. The Conduct Problems Prevention Research Group (1999) research on the universal classroom component of Fast Track, and Domitrovich et al. (2007) on preschool PATHS, exemplify attempts to change children's observable behaviors through structured activities. These projects contributed experimentally tractable outcomes, comparisons, and implementation challenges.

Subsequent syntheses expanded the field's scale and evidentiary ambition. Durlak et al. (2011) integrated 213 evaluations, Sklad et al. (2012) compared universal social, emotional, and behavioral programs, and Taylor et al. (2017) analyzed outcomes measured after program completion. Cipriano et al. (2023) later assembled a much larger international evidence corpus, while Ha et al. (2026) isolated academic achievement and Hung et al. (2026) examined prosocial behavior as a distinct outcome. The chronology matters because modern estimates arise from partly overlapping bodies of trials, shifting definitions, and changing standards of measurement. The sequence of increasingly extensive meta-analyses does not imply that every program improved or that the effects measured in different decades have identical meanings.

Competency Frameworks and the Organization of SEL

Self-awareness encompasses recognition and interpretation of emotional experiences and beliefs about oneself. Self-management concerns goal-directed modulation of attention, impulses, stress responses, and action. Social awareness involves representing another person's perspective and responding to situational and cultural circumstances. Relationship skills concern communication, cooperation, conflict negotiation, and repair. Responsible decision-making involves considering consequences, duties, personal goals, and the welfare of others. These categories structure the CASEL (n.d.) framework and supply useful curricular language, provided investigators identify more specific operations within each. A student's ability to identify anger from a vignette, for example, samples semantic emotion knowledge under test conditions; it is not by itself evidence of successful regulation during an actual disagreement.

Competency models can be organized at different levels. At the individual level are emotion concepts, attentional processes, interpretation of social cues, and behavioral repertoires. At the interaction level are turn taking, responding to feedback, conflict resolution, mutual regulation, and communication norms. At the classroom and school levels are routines, perceived safety, relationship networks, discipline procedures, and opportunities for participation. Jennings & Greenberg (2009) proposed a model in which teachers' own social and emotional competence contributes to classroom relationships and program implementation. That model provides a causal hypothesis connecting levels; whether a particular teacher-training component produces student change remains a separate empirical proposition.

The strongest framework is one that allows disconfirming observations. If emotion vocabulary improves without gains in peer cooperation, investigators should report that pattern rather than classify both outcomes as general SEL success. If classroom atmosphere improves while standardized academic scores remain stable, the intervention may have achieved an important social goal without demonstrating academic transfer. If effects depend on teacher support or school climate, the mechanism may lie partly outside individual students. Frameworks therefore support scientific comparison when linked to clear endpoints, developmental theory, and measures sensitive to actual behavior. The broad five-domain structure is best understood as a practical map of research targets whose boundaries require continued construct validation.

Developmental Foundations and Life-Course Variation

Early social-emotional development is characterized by changes in emotion language, inhibition, perspective taking, relational expectations, and the ability to coordinate action with others. Young children often rely on adult co-regulation while acquiring more independent strategies; later learning increasingly involves peers, social identity, evaluation, and autonomy. Denham et al. (2003) associated preschool emotional competence with kindergarten social functioning, providing longitudinal evidence that early competencies may be developmentally meaningful. Domitrovich et al. (2007) evaluated a preschool curriculum in Head Start settings, connecting developmental theory to an experimentally tested instructional design. Generalizing from preschool instruction to adolescents, however, requires new measures and attention to different social challenges.

Across the elementary years, social comparison, cooperative learning, classroom rules, and increased academic evaluation create repeated occasions for emotional appraisal and social practice. During adolescence, peer recognition, belonging, identity, independent choice, and negotiation of institutional authority become salient. A lesson designed around labeling facial emotions in early childhood cannot be presumed suitable for an older student's experience of discrimination, rejection, group conflict, or contradictory loyalties. Taylor et al. (2017) synthesized delayed outcomes from kindergarten through secondary school interventions, and Ha et al. (2026) examined achievement across school grades. Such coverage expands the question beyond immediate skill acquisition but also reveals the need to analyze developmental stage as a moderator rather than a nuisance variable.

Developmental interpretation is complicated by differences in baseline opportunity and challenge. Students enter school with varied linguistic experience, socioeconomic resources, caregiving histories, disability-related access needs, and prior experiences of school discipline or exclusion. Those differences affect both the behaviors educators observe and the instructional supports required to learn new skills. Age-related measurement invariance must be tested before comparing children's survey means across grades. Repeated assessments should distinguish maturation from intervention effects and account for changes in teacher, peer group, and educational environment. Long-term claims should identify what persists—a learned skill, a school relationship, a favorable trajectory, or simply a difference in reporting—and explain the evidence linking the initial intervention to that later endpoint.

Ecological and Relational Mechanisms

SEL occurs within an ecology of repeated interactions rather than through isolated information transfer. A student rehearses conflict resolution in response to an actual peer disagreement; a teacher can model emotion naming and invite reflection rather than escalate a disruptive encounter; classroom routines may provide predictable opportunities for cooperative action. The coordinated prevention position articulated in Greenberg et al. (2003) and the prosocial classroom model in Jennings & Greenberg (2009) suggest that skills, relationships, and institutional climate influence one another. These proposals imply multiple causal pathways: a lesson may expand a behavioral repertoire, while a supportive adult may increase the likelihood that a student can deploy it under stress.

Relational pathways require evidence beyond simple co-occurrence. A school with warm teacher–student relationships may have better adjustment because of selective enrollment, leadership, resources, or established culture. Randomized trials of interventions that explicitly change classroom processes offer stronger tests, as illustrated by Rivers et al. (2013) and Hagelskamp et al. (2013) in research on the RULER approach. Yet a package-level trial identifies effects of the combined intervention, not automatically of every proposed component. Mediation analysis can ask whether improved interaction quality precedes student outcomes, but credible causal mediation depends on temporal ordering, adequate measurement, and assumptions about unobserved confounding.

An institutional mechanism is visible when the opportunity to practice determines whether competence is expressed. A student who has learned to explain a disagreement may remain silent where dissent is punished; a student with good emotion knowledge may become dysregulated in an unpredictable environment. This leads to a dual assessment requirement: measure student capacities and the conditions enabling their use. Context includes schedules, staff development, student voice, disciplinary procedures, family communication, and peer network structure. Such ecological analysis prevents treating a school's outcomes as the aggregate of isolated individual deficits. It also helps separate changes in student conduct from changes in adult perception, enforcement, or the incentives governing what behavior receives attention.

Instructional Modalities and Whole-School Designs

Universal SEL is delivered through distinguishable instructional arrangements. A manualized curriculum specifies lessons, sequences, activities, and materials, often with teachers as implementers. Embedded strategies place brief routines within ordinary instruction, such as explicit discussion of perspectives in literature or structured peer problem solving. Whole-school approaches coordinate classroom teaching with adult learning, school norms, leadership practices, and family partnerships. Jones & Bouffard (2012) argued that treating SEL as repeated instructional practice changes the unit of intervention from a bounded program to a pattern of educational interaction. Programs can combine these arrangements, so investigators need descriptions of actual exposure rather than a binary label recording only whether a school adopted SEL.

Effective design begins with an explicit theory of change. If instruction aims to improve recognizing emotions, activities should require discrimination among feelings and contexts and should be tested with unfamiliar examples. If the intended outcome is conflict resolution, students need opportunities to negotiate when interests differ, receive feedback, and practice repair after mistakes. If the endpoint is academic functioning, the proposed chain from regulation or relationships to attention, participation, and learning must be measured rather than assumed. The SAFE features discussed in Durlak et al. (2011)—sequenced, active, focused, and explicit instruction—offer a historically influential characterization of skill-building designs. Their association with outcomes in program synthesis warrants testing specific mechanisms in controlled comparisons.

Whole-school designs raise a problem of causal complexity. Teachers may change discipline responses, school leaders may revise norms, students may encounter different peer interactions, and families may receive new communications during the same intervention. Such coordination is educationally coherent, but an evaluation of the full package cannot identify which element is necessary or sufficient. Factorial trials, planned implementation contrasts, and adequately powered multilevel studies may help isolate elements where feasible. The key practical implication is to specify what the school is doing differently, why that practice should matter, and which outcomes will detect the intended change. A branded curriculum alone provides less information than a transparent record of content, dosage, context, and delivery.

PATHS, RULER, and Experimental Program Traditions

The Promoting Alternative Thinking Strategies curriculum, usually called PATHS, is a prominent prevention-oriented example of explicit teaching about emotion, problem solving, and social behavior. Research on the classroom component of Fast Track Conduct Problems Prevention Research Group (1999) and a randomized evaluation of preschool PATHS Domitrovich et al. (2007) demonstrate how classroom delivery can be studied with comparison groups and external assessments. In the preschool evaluation, twenty classrooms were assigned across intervention and comparison conditions, with classroom lessons delivered across a school year. The importance of such research lies in its defined treatment, measured student outcomes, and attention to developmental stage; findings from a specific preschool setting should not be uncritically transferred to every school system.

RULER is an approach to emotional literacy and classroom climate that integrates professional development, tools, and student instruction. Rivers et al. (2013) tested its effect on the social and emotional climate of classrooms in a cluster-randomized trial, while Hagelskamp et al. (2013) examined proximal and distal classroom outcomes. These studies direct attention to the quality of social interaction as a possible treatment target. The intervention's name should not substitute for specifying the instruments, assignment procedure, implementation fidelity, and outcomes measured. A favorable climate rating is valuable evidence about the setting; whether it mediates academic or mental-health change must be established independently.

Program history also includes rigorously evaluated null results. The large English cluster-randomized evaluation reported by Humphrey et al. (2018) found tentative evidence of very small immediate social-skills improvement from PATHS and no lasting improvement at later follow-up. Related analysis of academic attainment Hennessey & Humphrey (2020) did not support recommending the program as an effective means of improving attainment in the studied English primary-school context. These findings coexist with positive results from earlier evaluations because interventions operate within distinct systems, samples, and implementation conditions. Scientific appraisal must preserve this variation. An intervention label identifies a family of educational practices, while causal efficacy attaches to a tested version delivered under specified conditions and assessed at defined times.

Teachers as Implementers and Participants

Teachers influence SEL through direct instruction, the norms they model, the language they use to interpret behavior, and their management of conflict, error, and participation. The Jennings & Greenberg (2009) prosocial classroom model proposes that teacher social and emotional competence contributes to supportive relationships, classroom management, and student learning. The model places teacher well-being inside the intervention ecology rather than treating staff as neutral conduits. It remains necessary to distinguish this theoretical pathway from empirical confirmation that any particular teacher-training course changes pupil outcomes. Teacher self-reports alone may reflect satisfaction with a program more readily than changes in enacted practice.

Teacher-led delivery is important because it operates within the ordinary school day. Durlak et al. (2011) reported that school teaching staff could implement programs associated with favorable aggregate outcomes. In the U.S.-focused synthesis Cipriano et al. (2024), teacher-delivered programs showed favorable patterns, although such moderator comparisons aggregate programs and contexts that differ in many additional respects. Teacher workload, preparation time, supervision, perceived relevance, and administrative support can influence the quantity and quality of delivery. An implementation study should document both scheduled lessons and how staff respond when students express intense emotion, question the curriculum, or require accommodations.

The educator's social authority complicates measurement. If a teacher delivers the intervention and then rates its success, common-source bias, expectancy, and awareness of treatment status can inflate apparent improvements. Independent observations, student reports, family reports, blinded assessors where practicable, and behavior recorded under ordinary conditions offer complementary evidence. These methods can disagree for substantive reasons: a student may behave differently in class and at home. Rather than averaging away disagreement, investigators should examine setting-specific change and whose behavior actually shifted. A high-quality school intervention therefore includes professional support, clear intervention specifications, and an evaluation plan that does not depend entirely on its implementers' judgments.

Operationalization and Psychometric Measurement

Social and emotional competence is a family of constructs requiring multiple measurement strategies. Self-report questionnaires elicit perceived abilities, attitudes, or experiences; teacher and parent ratings assess observed behavior across different settings; performance tasks sample choices and skill demonstration in structured situations; peer nominations or sociometric methods capture social position; observations record interactions and classroom organization. None is an unqualified gold standard. In Denham et al. (2003), emotional competence was operationalized through emotional expressiveness, emotion regulation, and emotion knowledge, then related to social competence through latent-variable analysis. Such work illustrates the benefit of analyzing components and measurement structure rather than treating one total score as an exhaustive construct.

Behavioral screening instruments also require careful interpretation. The Strengths and Difficulties Questionnaire introduced by Goodman (1997) combines domains of difficulty with prosocial behavior and was developed as a screening measure. It is not a direct measure of the entire CASEL competency model. A reduction in a difficulty scale may reflect an important behavioral change while leaving unexplored emotion understanding, perspective taking, or decision making. Measures must be selected for the exact outcome they operationalize; reliability established in one population does not certify validity in another language or age group. An intervention aimed at interpersonal negotiation needs more direct evidence of negotiation than a generic emotional-adjustment score can provide.

Evidence quality depends on internal consistency where appropriate, temporal stability when the construct is expected to persist, observer agreement, factorial structure, sensitivity to meaningful change, and validity across groups. Cross-cultural assessment requires attention to translation, item interpretation, and measurement invariance. Students' perceived competence may rise because vocabulary improves or because the intervention changes standards of self-evaluation; either process can alter a survey score. Repeated use of the same task can produce practice effects. Reports should therefore identify instrument version, respondent, scoring, baseline distribution, preregistered primary outcome, and meaningful threshold of change. Multiple methods are particularly valuable when results concern a broad construct whose visible expression depends on context.

Causal Identification and School Evaluation Designs

Schools are clustered environments: students share teachers, classrooms, peers, schedules, and institutional rules. Randomizing individual students within a classroom can allow treatment contamination when instruction changes how teachers treat everyone. Cluster-randomized trials allocate classrooms or schools to conditions and can reduce contamination, although their statistical analysis must account for clustering. Rivers et al. (2013) and Humphrey et al. (2018) illustrate this design logic. Evaluators should report the unit randomized, number of clusters, intraclass correlation where available, attrition by condition, baseline comparability, deviations from assignment, and whether outcomes were registered before analysis. A study with many pupils but few schools may have substantially less independent information than its pupil count suggests.

Quasi-experimental studies can contribute valuable evidence when random allocation is impractical, particularly for policy changes or system-wide programs. They require transparent comparison construction, baseline trends, plausible treatment timing, and attention to selection bias. Cross-sectional associations between reported SEL competence and grades cannot establish that instruction caused achievement. Longitudinal correlations improve temporal description but may still be explained by family resources, teacher ratings, or other unmeasured determinants. Zhao & Sang (2025) reported that quasi-experimental studies tended to produce larger academic effects than randomized trials within its reviewed evidence, a pattern that strengthens the case for investigating selection and design features. No single design determines quality automatically; fit between the causal question and identification strategy is central.

Outcome selection is equally decisive. An intervention may have a proximal effect on emotion vocabulary, a more distant effect on peer behavior, and a still more distant effect on standardized achievement. Each step increases the number of additional influences that can dilute or modify the signal. Investigators should specify primary and secondary outcomes, test timing, missing-data handling, subgroup hypotheses, and multiplicity corrections. The Ha et al. (2026) achievement synthesis distinguished test scores from grades, while Hung et al. (2026) focused specifically on prosocial behavior. Separating outcomes provides more interpretable evidence than an undifferentiated claim that SEL works. The most informative trials report null and unfavorable results alongside favorable outcomes, with uncertainty intervals and complete denominators.

Implementation Fidelity, Dosage, Adaptation, and Reach

Implementation describes what participants actually receive rather than what a manual intends. Fidelity can refer to adherence to prescribed elements, dosage to the amount delivered, quality to the interactional manner of teaching, and responsiveness to students' engagement. Reach captures participation across eligible populations, while sustainability concerns maintenance after external support is reduced. Durlak et al. (2011) found implementation-related moderation, and Cipriano et al. (2023) identified implementation quality among the sources of variation across programs. These patterns make implementation a scientific explanatory variable as well as an operational concern. A trial with weak average outcomes may reflect an ineffective mechanism, inadequate delivery, poor contextual fit, or some combination.

The mechanism connecting dosage to outcomes is rarely a simple linear increase. Students need sufficient exposure to acquire concepts, but repeated low-quality lessons can cause disengagement. Shi & Cheung (2024) reported a pooled effect across a selected set of higher-quality programs and found smaller effects in large-scale studies and associations involving teacher development and program dosage. Such component comparisons remain observational at the level of programs unless components were experimentally manipulated. The relation between adherence and outcome can also be biased because more motivated schools may implement more effectively. Reliable implementation measurement should include observations and records collected independently of the outcome whenever feasible.

Adaptation is unavoidable when curricula move across languages, cultures, ages, and disability contexts. The scientific problem is to distinguish core functions from particular surface forms and then test whether adapted materials preserve the intended mechanism. Li et al. (2025) reviewed eleven U.S. cultural adaptation studies involving 5,173 students and reported mixed effectiveness results alongside frequent acceptability or feasibility. An adaptation can increase accessibility and relevance without reproducing every original effect estimate. Conversely, strict adherence can preserve an unsuitable example or interaction norm. Trials should document adaptations before and during delivery, reasons for making them, involvement of students and communities, and the resulting pattern of outcomes. Fidelity and responsiveness should be studied jointly, not treated as automatically competing values.

The 2011 Meta-Analysis and Its Evidentiary Meaning

The influential synthesis by Durlak et al. (2011) combined 213 evaluations of universal school-based SEL involving 270,034 participants from kindergarten through high school. Compared with comparison groups, the combined interventions were associated with improvements in social-emotional skills, attitudes, behavior, and academic achievement. The paper reported an average achievement difference commonly expressed as an eleven-percentile-point gain. This is a group-level summary of effects from the studies reviewed, not a prediction that every student or every program gains eleven percentage points on an examination. The distinction between a percentile-position change and an increase in raw test accuracy is crucial. The synthesis helped establish that psychosocial instruction could be studied together with academic endpoints while supporting attention to implementation and instructional design.

Several inferential limits accompany that contribution. The evidence pooled programs that varied in curriculum, school level, intervention length, implementation support, and measurement. A positive pooled average is compatible with small, null, and unfavorable effects in particular settings. Many measures were reported by adults aware of intervention status, and the availability of appropriate controls differed among evaluations. Historical estimates also depend on the studies accessible at the time, their publication selection, and the methods used to combine outcomes. Program features associated with larger effects in a meta-regression are candidate moderators, not necessarily causal ingredients. Interpreting the study as definitive proof of one standardized school method would discard the heterogeneity that motivated later research.

The meta-analysis has enduring conceptual value because it explicitly linked outcomes to the possibility of teacher-led implementation and explored features of instruction. It also suggests a hierarchy of questions for future trials: whether skills improve, whether behavior changes outside the taught exercise, whether academic effects occur, and whether benefits persist once delivery ends. The relationship among these questions should be investigated as a set of potentially mediated pathways rather than inferred from the common label SEL. Later reviews such as Cipriano et al. (2023) build on this foundation with a larger and more internationally varied evidence base, offering an opportunity to examine changes in the research ecosystem without erasing the earlier contribution.

Contemporary International Evidence and Heterogeneity

The contemporary systematic review and meta-analysis by Cipriano et al. (2023) identified 424 studies from 53 countries, representing 252 distinct universal SEL interventions and 575,361 students. Its search covered research available for the period from 2008 through 2020, and it examined outcomes across skills, attitudes, behavior, climate and safety, peer relationships, school functioning, and academic performance. In aggregate, students in studied intervention conditions showed favorable results compared with controls across those domains. The review also identified significant heterogeneity linked to intervention content, features, context, and implementation quality. This combination—a favorable general tendency alongside substantial variation—is the central contemporary evidence pattern.

The broader geographic scope improves the diversity of the evidence corpus but introduces additional interpretive challenges. Schools differ in discipline policy, curriculum, languages of instruction, age of compulsory attendance, expectations for emotional expression, and resources available for staff training. Studies may define achievement differently or use locally developed self-report instruments. Comparisons across these measures require confidence that pooled outcomes refer to sufficiently related constructs. Researchers must also avoid conflating the number of studies with the number of independent populations, because related evaluations, shared programs, and dependent measures can contribute correlated evidence. The study's findings should be read as a synthesis of evaluated interventions under the sampled conditions, with moderator analysis guiding rather than completing explanation.

The review documents the value of cumulative evidence while indicating where claims remain unresolved. Positive effects on school functioning do not specify whether attendance, grades, disciplinary incidents, or perceptions of school engagement changed. Perceived school safety is scientifically meaningful but distinct from independently recorded physical safety. Peer-relationship measures may reflect network position or adult appraisal. Future studies should name the exact endpoint and predefine its role in the causal model. Because the evidence base contains many versions of interventions, the practical translation from an aggregate mean to an individual school's adoption decision requires local assessment of fit, capacity, student needs, and outcome quality.

Durability, Follow-Up, and Developmental Trajectories

Immediate postintervention improvements have a different evidentiary status from sustained changes months or years later. Taylor et al. (2017) synthesized follow-up findings from 82 universal school-based interventions involving 97,406 students. Its included follow-up observations ranged from six months to eighteen years after intervention completion. Compared with controls, intervention participants showed favorable follow-up outcomes for social-emotional skills, attitudes, and indicators of well-being. The review identified postintervention social-emotional skill development as a strong predictor of later well-being. Those findings support investigation of lasting pathways, while variation in follow-up interval, outcome availability, and participant retention complicates a single account of durability.

Follow-up evaluation is particularly vulnerable to selective observation. Students who can still be located may differ from those lost to follow-up; later schooling, family circumstances, community conditions, and additional programs introduce new exposures. If an early SEL program predicts later graduation or social functioning, the claim should specify whether the result is an intention-to-treat intervention contrast, a subgroup association, or an observational mediation. Taylor et al. (2017) reported several longer-term indicators that were measured less frequently; those outcomes need to be distinguished from the more common intermediate skill and attitude outcomes. A theoretically plausible long-term pathway is not a substitute for the actual data required to estimate its magnitude.

Evidence from national or program-specific evaluations can diverge from the international follow-up average. Humphrey et al. (2018) found no sustained later benefit from the tested version of PATHS in its English elementary-school trial. The U.S.-focused review Cipriano et al. (2024) also reported no overall evidence of benefits persisting six months or longer after program completion in the studies it analyzed, even while finding favorable nearer-term outcomes. Different syntheses can therefore reach different persistence conclusions because their samples, intervention sets, endpoints, and analysis choices differ. Responsible interpretation asks which populations and time horizons are represented in each result rather than selecting the most reassuring headline. The scientific objective is a predictive account of when skills persist, generalize, decay, or require continued institutional reinforcement.

Academic Achievement and Educational Transfer

SEL is frequently justified partly by its potential contribution to academic achievement. The theoretical pathway can involve emotion regulation that supports sustained attention, interpersonal cooperation that improves classroom participation, or a supportive climate that reduces disruption. These are mechanisms to be tested, and academic learning remains a distinct domain requiring content instruction and appropriately valid assessments. Durlak et al. (2011) reported positive aggregate academic effects, but a review dedicated to the achievement endpoint allows more precise interpretation. Ha et al. (2026) synthesized forty empirical studies and 76 achievement effect sizes involving 33,737 students across grades one through twelve. The authors reported a positive average effect, including favorable patterns in literacy and mathematics, while explicitly examining assessment type and age.

An additional meta-analytic review Zhao & Sang (2025) analyzed twenty-two studies with 24,510 elementary and middle-school students and reported a modest average academic effect (Hedges' g approximately 0.08). Middle-school evidence was based on only two studies in that review, making age comparisons especially uncertain. The reported difference between quasi-experimental and randomized study results underscores the danger of treating every favorable association as causally equivalent. Even among randomized trials, academic effects can depend on alignment between a program's intended pathway and the timing and difficulty of school assessments. School grades also incorporate teacher judgment and local policy, whereas standardized tests may sample a narrower or more comparable set of academic competencies.

The academically relevant conclusion is therefore quantitative and qualified: the reviewed evidence supports a small favorable average in several syntheses, together with substantial contextual variation and documented null trials. An eleven-percentile-point description from an earlier review, a standardized mean difference from a later analysis, and a report of months of academic progress are different metrics and must not be substituted for one another. Hennessey & Humphrey (2020) found no support for the tested PATHS curriculum as a means of improving academic attainment in its English primary-school context. A school adopting SEL for academic reasons should specify realistic targeted outcomes and collect direct attainment evidence, while also valuing independently established social and emotional benefits.

Prosocial Behavior, Disruption, and Peer Outcomes

Prosocial behavior describes voluntary action intended to benefit other people, including helping, cooperating, sharing, and responding constructively to others' needs. It is an important endpoint but cannot be reduced to obedience or a low count of disciplinary referrals. Hung et al. (2026) examined this specific construct in a meta-analysis of 66 studies involving 52,914 young people and reported a mean standardized effect of approximately g = 0.24, with substantial heterogeneity. The estimated effect concerned prosocial behavior and therefore should not be used as a numerical estimate for academic achievement or psychiatric symptoms. Behavioral outcomes may be measured by teachers, peers, observations, or self-reports, each exposing different aspects of interpersonal conduct.

Trials such as Conduct Problems Prevention Research Group (1999) and Domitrovich et al. (2007) show why it matters to document behavior within concrete classroom and developmental contexts. General social problem-solving knowledge may improve before spontaneous behavior does; genuine peer interaction is shaped by social status, prior conflict, incentives, and the responses of others. A classroom that permits student participation and offers low-stakes opportunities for repair may enable skill generalization differently from one relying primarily on behavior sanctions. When a teacher both delivers SEL and judges prosocial improvement, expectancy effects can complicate the evidence. Peer reports and structured observations may reduce one source of bias but introduce their own vulnerabilities.

Behavioral outcomes should also be interpreted through school policy. Referral counts can change because reporting rules or staff decisions change. A student avoiding punishment may appear better adjusted even if belonging, agency, or genuine relationships have deteriorated. Conversely, a short-term increase in expressed disagreement might accompany healthier self-advocacy. Evaluation therefore benefits from multi-informant reporting and from endpoints that capture constructive action, conflict resolution, mutual respect, and perceived safety. Cipriano et al. (2023) found favorable overall results across behavioral and peer-related outcomes, but translating that average into an explanation of a particular school's social network requires specific evidence about the interactions and procedures altered by the intervention.

Mental Health, Well-Being, and Prevention

Schools increasingly connect SEL with psychological well-being, stress management, and prevention of later difficulties. These aims have plausible developmental links: recognizing emotions may support help-seeking, supportive relationships can buffer stress, and self-regulation may help students manage conflict and attention. Yet improvement on a social competence measure is not a clinical diagnosis or a treatment effect. Sklad et al. (2012) reviewed universal social, emotional, and behavioral programs across multiple outcome domains, including mental health, while Taylor et al. (2017) reported favorable follow-up indicators of well-being. These findings support prevention-oriented research but do not license claims that a universal school curriculum treats an identified disorder.

Clinical outcomes should be defined with validated tools and, when appropriate, diagnostic procedures. Teacher ratings may capture behavior that is disruptive in the classroom without accurately characterizing internal distress. A student may become more socially cooperative while still experiencing severe anxiety or depression. Screening instruments such as the Goodman (1997) SDQ can support population-level investigation of difficulties and strengths, but their scores require appropriate population norms and careful interpretation. SEL programs should provide links to qualified school or community support for students who need individualized assessment, rather than representing universal instruction as a substitute for clinical care.

Prevention has a temporal component that must be measured explicitly. Fewer symptoms immediately after lessons may reflect transient reassurance; reduction in the incidence of a later disorder requires longer observation, clinically meaningful outcome criteria, and adequate statistical power. Risks are also worth monitoring: poorly adapted emotion disclosure exercises may expose students to embarrassment, inadvertent stigma, or disclosure beyond their control. School policy should protect privacy and establish voluntary, developmentally appropriate forms of participation where sensitive personal experiences are involved. The scientific aim is to evaluate benefit and potential burden with equal clarity, preserving the distinction among competence education, well-being promotion, indicated prevention, and specialized treatment.

Cultural Responsiveness, Equity, and the Meaning of Competence

School-based social and emotional learning necessarily operates through culturally organized expectations about expression, authority, disagreement, cooperation, and appropriate conduct. Its content should distinguish broad psychological capacities such as recognizing a person's distress from local rules prescribing exactly how that distress ought to be displayed. The framework developed by CASEL (n.d.) explicitly places supportive classrooms, families, and communities around individual competencies. A critical evaluation therefore asks whose social norms are taught, whether multiple ways of expressing competence are recognized, and whether students can contest norms that make their participation less secure. A curriculum can improve knowledge of a dominant institutional code while failing to change exclusionary practices within that institution. Those outcomes must be described separately.

The systematic review by Li et al. (2025) investigated cultural adaptations of SEL programs for students in United States schools, analyzing eleven studies encompassing 5,173 participants. Its findings indicate that adaptation is a real design and implementation practice, but that the reviewed effectiveness results were mixed. Improved acceptability or feasibility does not automatically demonstrate a favorable causal effect on student competence. Such findings support planned comparisons of adaptation processes and outcome measures: the language used, examples selected, adult assumptions about participation, and the relation between intervention content and students' lived contexts. Involvement of students, caregivers, and local educators may improve explanatory and cultural fit, while evidence of changed outcomes still requires prospective assessment.

Research with historically marginalized student groups requires attention to baseline disparities and institutional processes. The United States-focused synthesis Cipriano et al. (2024) examined universal SEL interventions and explicitly considered students marginalized by race, ethnicity, language, gender, and disability. Variations in study reporting limited the precision of subgroup conclusions. Stronger equity evidence would require prespecified hypotheses, representative enrollment, appropriate subgroup sample sizes, and measures validated across groups. Aggregate improvement that conceals deterioration for a subgroup is an incomplete account of impact. Outcomes should also include procedural fairness, access to support, student voice, and consistency of adult discipline, because these may shape the very conditions under which social-emotional competence can be displayed.

Disability, Neurodiversity, and Accessible Participation

Universal SEL includes students with diverse sensory, communication, cognitive, developmental, physical, and emotional characteristics. Accessibility is both an ethical requirement and a condition of valid scientific measurement. An assessment that expects every student to interpret facial expressions, report inner states verbally, or sustain visual attention in a standardized interaction may misrepresent the skills of those who communicate differently. The domains described by CASEL (n.d.) require operational forms appropriate to different learning pathways rather than one rigid behavioral appearance. A student who negotiates by assistive communication or requests a quiet environment may be displaying sophisticated self-management or relationship competence, even when such actions differ from a normative classroom script.

Adaptation should be documented at the level of access and mechanism. Alternative communication supports, flexible response modalities, predictable routines, accessible examples, and opportunities for self-directed regulation change what students can demonstrate. However, a program tested in general education classrooms should not be assumed to have identical outcomes in specialized educational settings, nor should evidence from a small targeted group be generalized across distinct disabilities. Cipriano et al. (2024) identifies gaps in population-specific reporting within the United States SEL literature, and Li et al. (2025) highlights the limited causal knowledge available about program adaptations. The relevant design is one that identifies participants, supports offered, plausible mechanisms, and reliable outcomes with enough clarity to permit replication.

Neurodiversity-related interpretations also require distinction between social adaptation and compelled masking. An intervention may teach contextual awareness and shared problem-solving while respecting a student's communication style and autonomy. Evaluation should measure whether students gain useful options, reciprocal understanding, and access to relationships, rather than only whether they more closely imitate an assumed norm. Student-reported burden, fatigue, and perceived authenticity can complement adult ratings of classroom behavior. Inclusion is substantively measurable through attendance, participation opportunities, accessible materials, peer interaction, and meaningful choice. No single test score exhausts those dimensions.

Family, Community, and Educational Continuity

Children encounter emotional and interpersonal demands across home, school, peer, and community settings. School curricula operate within this broader developmental ecology. The coordinated model proposed by Greenberg et al. (2003) emphasizes linking social, emotional, and academic learning, while CASEL (n.d.) explicitly recognizes family and community partnerships. The logic is that concepts practiced in multiple relationships may become easier to generalize, although causal evidence is needed to determine which forms of coordination improve outcomes and under what conditions. Families may also interpret emotional expression, privacy, authority, and cooperation differently from a school curriculum; involving them can reveal rather than erase those differences.

Family engagement should be described through identifiable activities and reciprocal decisions. A parent information sheet, a collaborative workshop, a shared problem-solving routine, and joint modification of school policies are different exposures with different causal possibilities. Participation can be constrained by work schedules, language, transportation, technological access, and prior experiences of exclusion. If better-resourced families participate more often, an apparently successful intervention could widen inequality even as its participant mean improves. Implementation research should therefore record eligible families, actual reach, barriers, accommodations, and whether outcomes differ by participation status. Outcomes need independent observation because family satisfaction, while useful, measures a different construct from a child's behavioral generalization.

The transfer of skills between school and home also depends on environmental affordances. A student can learn a conflict-resolution technique in class but lack the authority or safety to use it in a different setting. An ecological model predicts that supportive opportunities, adult responsiveness, and perceived consequences moderate expression of competence. Jennings & Greenberg (2009) places adult competencies within the classroom system, providing a parallel theoretical explanation for why family and community participation may matter. These hypotheses can be tested through longitudinal measures, cross-setting observations, and interventions varying the degree of coordination. Their plausibility should not be elevated to proof that any family-partnership component necessarily produces stronger outcomes.

Digital Social and Emotional Learning in the Artificial Era

Digital instruction changes how SEL is delivered, observed, and practiced. Interactive stories, educational games, simulated social scenarios, mobile activities, and adaptive systems may offer repeated decision points and immediate feedback. They can also shift control over pacing and expose students to standardized tasks beyond the physical classroom. The systematic review and meta-analysis by Cai et al. (2026) analyzed 37 experimental and quasi-experimental studies involving 4,742 kindergarten-through-grade-twelve learners. It reported favorable average standardized effects on social-emotional skills, affect and attitudes, and prosocial or antisocial behavior, with different magnitudes across these domains. Because the studies varied by intervention design, setting, reporter, and comparison condition, the pooled results indicate a promising research field rather than a universal guarantee of digital effectiveness.

Digital delivery introduces measurement opportunities and problems. Scenario choices, response timing, game progression, and conversation logs can make learning processes observable, but behavior within a simulated interaction may not generalize to peers, families, or teachers. Adaptive software might change task difficulty as a learner progresses, creating a need to track exposure and preserve comparability of outcomes. The mechanism may depend on feedback quality, sense of agency, narrative engagement, adult mediation, or repeated rehearsal. A randomized comparison of digital and conventional delivery should therefore distinguish content from medium, equivalent instructional time from unequal access, and automated logging from actual psychological validity. Digital abundance of data is not evidence of valid competence measurement.

Within the Artificial Era, artificial agents create new possibilities for conversational rehearsal, perspective simulation, and personalized instructional feedback. The potential research object is the effect of a specified artificial instructional system on human social and emotional learning. A system's fluent conversational performance is an observable property of its interaction; claims about its inner emotional life, psychological reciprocity, or independent agency require separate theoretical and empirical treatment. Research must evaluate what students learn, how reliance develops, whether interactions generalize to human relationships, and whether systems reinforce inaccurate social expectations. Privacy safeguards and human oversight remain central when minors' disclosures, inferred emotional states, or behavioral histories are involved. This is a testable extension of SEL science rather than an automatic inference from technological novelty.

International Systems and Scientific Transportability

Cross-national research broadens the evidence available for comparing school-based SEL, but transportability depends on more than finding statistically positive results in many locations. The review by Cipriano et al. (2023) drew studies from 53 countries, making its evidence base international while documenting substantial heterogeneity. School systems differ in curriculum hours, teacher preparation, assessment regimes, language, discipline procedures, family roles, and access to school psychology services. An intervention evaluated in one context becomes a new causal question when the institutions organizing it change. A school system importing a program must therefore examine whether the underlying psychological mechanism, the means of delivery, and the meaning of measured outcomes remain similar enough to support the intended inference.

A geographically specific synthesis can illuminate differences in that question. Cipriano et al. (2024) pooled United States studies and reported favorable aggregate outcomes while finding limited evidence for durable effects at six months or more. Hua et al. (2026) analyzed interventions and practices in primary schools, offering another design-sensitive synthesis within a different evidence selection. The different conclusions of overlapping reviews should be interpreted with attention to eligibility rules, student populations, time windows, outcome classes, and statistical choices. Apparent country differences can emerge from study composition even when program mechanisms are similar. Moderator claims about culture require direct comparisons or carefully specified multilevel analyses rather than intuition drawn from national stereotypes.

Transportability is strengthened by prospective replication using common outcome definitions, transparent adaptation logs, and locally meaningful supplementary measures. Students and educators can participate in identifying which examples or practices are culturally appropriate while researchers retain measurement anchors necessary for cross-site comparison. International trials should report language versions, socioeconomic context, implementation funding, teacher preparation, educational access, and participant selection. Such reporting allows readers to assess whether an average effect is applicable to a new setting and helps reveal cases where a locally effective intervention is more important than a universally branded curriculum. The strongest international evidence permits both carefully justified generalization and principled recognition of contextual differences.

Converging Findings and the Strength of the Evidence

Across major syntheses, the most reproducible broad pattern is an average favorable association between exposure to evaluated universal school-based SEL interventions and several measured social, emotional, behavioral, and school outcomes. Durlak et al. (2011) found positive pooled results in its earlier school-program corpus; Cipriano et al. (2023) found a similarly favorable direction in a larger international evidence set; Taylor et al. (2017) identified some positive outcomes after program completion. Together, these findings justify treating SEL as an established field of educational intervention research. They do not justify attributing the same magnitude of benefit to every intervention, learner, outcome, or duration. Heterogeneity, differences in measurement, and variation in study quality are part of the result, not incidental complications to be set aside.

Domain-specific syntheses sharpen this conclusion. Ha et al. (2026) and Zhao & Sang (2025) both found positive average academic effects but used different eligibility criteria and estimated magnitudes. Hung et al. (2026) focused on prosocial behavior, a construct that differs from both standardized attainment and internal emotional distress. Cai et al. (2026) reported favorable digital intervention effects while relying on a smaller and technologically heterogeneous set of evaluations. These estimates are not directly interchangeable; nor can their participant counts be added to produce a unique total, because the underlying evidence pools may overlap. The most informative synthesis treats each review as an answer to its own research question with its own temporal, geographical, and methodological boundaries.

Evidence confidence varies by claim. Broad short-term changes in measured SEL-related outcomes are supported by numerous studies and reviews, while precise predictions for an individual branded program, cultural subgroup, long-term clinical prevention, or specific active ingredient often face narrower evidence. The presence of randomized studies increases confidence about intervention-package causal effects in their study settings, but construct validity and the quality of blinding remain necessary. When different raters disagree, that disagreement can reveal real contextual differences or bias in measurement. A scientific conclusion should identify which claim is supported, at which level of explanation, and by which designs. Unsupported extensions from program averages to diagnoses, moral character, or universal developmental laws should be excluded.

Negative Trials, Heterogeneity, and Replication

Null findings provide essential evidence about the boundary conditions of programs. The English cluster-randomized evaluation described by Humphrey et al. (2018) assessed a specified PATHS implementation and found small immediate improvements in some outcomes without clear sustained effects at later follow-up. Its related academic analysis Hennessey & Humphrey (2020) reported no meaningful attainment gains under intention-to-treat analysis. Such results indicate that an intervention with a positive evidence history may perform differently under new conditions. Differences can involve educational system, teacher support, baseline competence, content fit, exposure, outcome selection, and ordinary classroom practice in the comparison group. The proper scientific response is to characterize these possible moderators and test them, not to suppress unfavorable trials.

Replication should preserve the relevant causal question while varying context systematically. An independent evaluation can use the same curriculum and independent assessors in new schools; a conceptual replication can test a proposed mechanism through a different instructional activity; an implementation replication can investigate whether similar results occur under routine operating conditions rather than intensive research support. Publication bias may arise when favorable outcomes are more likely to be submitted or prominently reported than null results. Selective analysis, multiple unregistered endpoints, and choice of effect-size metric can similarly distort the literature. Cipriano et al. (2023) provides an extensive evidence synthesis, but continued transparency about protocols, data, analytic decisions, and study inclusion remains important.

A further problem concerns mechanism inference. A treatment package may improve classroom behavior, but attributing that change specifically to emotional vocabulary, empathy exercises, teacher warmth, or rule consistency requires additional causal tests. Post hoc mediation can offer evidence compatible with a model, yet treatment may alter the proposed mediator and outcome through shared unmeasured changes. Appropriately powered experimental component comparisons, repeated intermediate measures, and credible causal assumptions are needed. The central goal is to progress from the question of whether a package works on average toward a mechanistic account explaining when, through what pathway, and for which students it works. Null results improve that account by limiting its scope.

Scale, Cost, Implementation Economics, and Institutional Capacity

An intervention shown effective under research conditions may be difficult to sustain when extended across a school system. Scale changes the availability of coaching, teacher time, class materials, supervision, and evaluation capacity. It also changes the composition of schools and students reached. Shi & Cheung (2024) identified smaller effects in large-scale studies in its selected meta-analysis, a result that invites analysis of implementation support and selection without proving that scale itself causes lower efficacy. Expansion should therefore be understood as a new implementation question, not merely the multiplication of an effect size by a larger student population.

Economic appraisal requires distinguishing intervention costs, displaced instructional time, existing school resources, and longer-term benefits that may or may not have been directly observed. Teacher professional development can be a substantive part of program delivery and a material cost. The relevant comparator is the educational practice that would otherwise occur, including any existing advisory activities or mental-health supports. A claim of cost-effectiveness depends on measured outcome units, valid comparison costs, uncertainty, and jurisdiction-specific values. It cannot be inferred from a positive standardized mean difference alone. Trial reports should specify staff hours, materials, technical infrastructure, implementation assistance, and outcome follow-up; transparent cost reporting helps decision-makers distinguish small inexpensive effects from larger but impractical packages.

Sustainability additionally depends on institutional ownership and conditions for staff learning. A school might start a curriculum with external support but struggle to maintain routines after staff turnover or leadership changes. Whole-school approaches may be more resilient when practices are integrated into existing structures, but this is a hypothesis requiring comparative evaluation. Students, teachers, administrators, and families may identify benefits and burdens that aggregate outcome scores overlook. Implementation science should therefore report reach, adaptations, fidelity, costs, routine feasibility, and persistence as distinct endpoints. The practical aim is an intervention that can be delivered, understood, evaluated, and maintained at the scale for which a school system is responsible.

A Research Agenda for Precise and Generalizable SEL Science

The next stage of SEL research should specify interventions in terms of mechanisms rather than only proprietary program labels. A trial could compare emotional labeling alone with labeling plus practiced self-regulation, or conflict-resolution instruction with and without structured peer rehearsal. Predefined proximal and distal outcomes would clarify whether a component changes the intended psychological operation and whether that change transfers. Tests must be developmentally appropriate and ideally include independent, blinded assessments when practicable. Prespecifying analyses and publishing complete outcomes would make the literature more informative about effects that fail to materialize as well as effects that do.

Mechanistic research should examine interactions between competence and opportunity. An intervention may equip a child to negotiate disagreement, but the outcome depends on whether peers and adults respond constructively. Multilevel and repeated-measure designs can test whether teacher practices, student skills, and classroom climate change in the expected temporal order. Such studies should measure both institutional behavior and student functioning, while accounting for clustered assignment and differential attrition. Longitudinal follow-up can distinguish persistence of explicit knowledge from stable changes in behavior, relationships, or academic performance. Taylor et al. (2017) offers a foundation for investigating these later pathways; Humphrey et al. (2018) demonstrates why independent replication and longer-term observation are necessary.

Equity research should test whether outcomes are reliable across languages, cultures, disabilities, and socioeconomic settings, rather than assuming that a scale developed for one population has identical meaning everywhere. Li et al. (2025) shows that cultural adaptation studies remain a relatively small and mixed evidence base. Cipriano et al. (2024) highlights limits in subgroup reporting. Future trials can improve that foundation by including meaningful participation in program design, planned subgroup analyses with adequate power, and explicit assessment of institutional fairness. Digital and artificial instructional systems should be compared against active controls with matched time and instructional content; their educational data practices must be independently assessed. Cai et al. (2026) provides a starting evidence synthesis, not a final account of agent-mediated social learning.

Educational Decisions and Evidence-Informed Practice

Evidence-informed adoption begins with a clearly specified educational problem. A school seeking more constructive peer disagreement requires different intervention mechanisms and outcomes from a school seeking better emotion vocabulary or academic persistence. Leaders should examine the original trials' ages, languages, implementation requirements, comparison conditions, and measures before selecting or adapting a program. Cipriano et al. (2023) supports the general potential of universal SEL, but its reported heterogeneity requires local judgment grounded in specific evidence. A school should identify what counts as success in student experience, classroom relations, and any intended academic transfer, then distinguish those endpoints in its monitoring plan.

Implementation should allow teachers to learn and demonstrate practices while permitting students to exercise agency and participate through accessible formats. Adults can model constructive disagreement and repair, offer structured practice in ordinary classroom activities, and make explicit the difference between recognizing emotion and acting responsibly under pressure. Implementation records should capture dosage and reach without turning student emotional disclosure into a compulsory performance. Transparent feedback from students and families can reveal whether the program feels relevant, respectful, and safe. Observations and independent assessments complement satisfaction reports. Schools should also maintain distinct pathways for students requiring individual psychological or clinical assistance.

Local evaluation is essential after adoption. Tracking only implementation completion cannot show an outcome effect, while comparing raw grades before and after a rollout cannot rule out changes in curriculum, staffing, or assessment. More informative designs use planned baselines, suitable comparison groups where feasible, repeated direct measures, and attention to missing data and unintended outcomes. Positive averages should be interpreted alongside uncertainty and the possible distribution of effects across groups. If a program improves relational climate but shows no attainment gain, a school can recognize the verified outcome without embellishment. If a program creates burden or unequal participation, adaptation and reassessment are substantive educational responsibilities. Sound implementation uses evidence as a continuing learning process.

Conclusion

Social and emotional learning in schools is a structured field of developmental and educational intervention concerned with the acquisition, exercise, and institutional support of emotional and interpersonal capacities. Its scientific object includes the learner's competence, the interactions through which skills are practiced, and the school environments shaping their expression. Historically, the field integrates emotional development, prevention research, classroom instruction, and whole-school approaches. Contemporary frameworks such as CASEL (n.d.) provide a useful classification of intended competencies, but empirical interpretation depends on precise outcomes and valid measures. The constructs of emotional knowledge, self-management, relationship behavior, school climate, academic learning, and mental health must remain analytically distinguishable.

Meta-analyses and program trials establish a substantial favorable average evidence base for several short-term social-emotional and educational outcomes, while documenting variation across interventions, contexts, populations, and timescales. Durlak et al. (2011), Cipriano et al. (2023), and Taylor et al. (2017) represent major stages in the cumulative literature. Outcome-specific reviews such as Ha et al. (2026), Hung et al. (2026), and Cai et al. (2026) extend that picture while refining the questions being answered. The negative or limited results of particular randomized evaluations, including Humphrey et al. (2018) and Hennessey & Humphrey (2020), demonstrate why causal effects must be attributed to identified interventions and settings rather than to a universal label. Claims about sustained prevention, exact active components, and transfer across different social environments require stronger and more explicit evidence.

The scientific priority is to identify mechanisms, improve measurement, extend independent replication, and test which educational conditions allow skills to generalize while protecting inclusion and student agency. Artificial instructional systems introduce additional methods of rehearsal and evaluation whose actual effects require controlled study and sound data governance. The history and current evidence of SEL support its serious consideration as an educational domain, accompanied by differentiated assessment, transparent reporting, and recognition of institutional as well as individual responsibility. In the Artificial Era, the central research commitment remains the same: identify what learners can understand and do, how educational relations shape those capacities, and which consequences can be demonstrated through reproducible inquiry.

References

  • Cai, L., Wu, D., Fan, H., & Guo, S. (2026). Promoting social and emotional learning in K-12 students through digital-based interventions: A systematic review and meta-analysis. BMC Psychology, 14, Article 744. https://doi.org/10.1186/s40359-026-04434-4
  • Cipriano, C., Ha, C., Wood, M., Sehgal, K., Ahmad, E., & McCarthy, M. F. (2024). A systematic review and meta-analysis of the effects of universal school-based SEL programs in the United States: Considerations for marginalized students. Social and Emotional Learning: Research, Practice, and Policy, 3, Article 100029. https://doi.org/10.1016/j.sel.2024.100029
  • Cipriano, C., Strambler, M. J., Naples, L. H., Ha, C., Kirk, M., Wood, M., Sehgal, K., Zieher, A. K., Eveleigh, A., McCarthy, M., Funaro, M., Ponnock, A., Chow, J. C., & Durlak, J. A. (2023). The state of evidence for social and emotional learning: A contemporary meta-analysis of universal school-based SEL interventions. Child Development, 94(5), 1181–1204. https://doi.org/10.1111/cdev.13968
  • Collaborative for Academic, Social, and Emotional Learning. (n.d.). What is the CASEL framework? CASEL. https://casel.org/fundamentals-of-sel/what-is-the-casel-framework/
  • Conduct Problems Prevention Research Group. (1999). Initial impact of the Fast Track prevention trial for conduct problems: II. Classroom effects. Journal of Consulting and Clinical Psychology, 67(5), 648–657. https://pubmed.ncbi.nlm.nih.gov/10535231/
  • Denham, S. A., Blair, K. A., DeMulder, E., Levitas, J., Sawyer, K., Auerbach-Major, S., & Queenan, P. (2003). Preschool emotional competence: Pathway to social competence? Child Development, 74(1), 238–256. https://doi.org/10.1111/1467-8624.00533
  • Domitrovich, C. E., Cortes, R. C., & Greenberg, M. T. (2007). Improving young children's social and emotional competence: A randomized trial of the preschool “PATHS” curriculum. The Journal of Primary Prevention, 28(2), 67–91. https://doi.org/10.1007/s10935-007-0081-0
  • Durlak, J. A., Weissberg, R. P., Dymnicki, A. B., Taylor, R. D., & Schellinger, K. B. (2011). The impact of enhancing students' social and emotional learning: A meta-analysis of school-based universal interventions. Child Development, 82(1), 405–432. https://doi.org/10.1111/j.1467-8624.2010.01564.x
  • Goodman, R. (1997). The Strengths and Difficulties Questionnaire: A research note. Journal of Child Psychology and Psychiatry, 38(5), 581–586. https://doi.org/10.1111/j.1469-7610.1997.tb01545.x
  • Greenberg, M. T., Weissberg, R. P., O’Brien, M. U., Zins, J. E., Fredericks, L., Resnik, H., & Elias, M. J. (2003). Enhancing school-based prevention and youth development through coordinated social, emotional, and academic learning. American Psychologist, 58(6–7), 466–474. https://doi.org/10.1037/0003-066X.58.6-7.466
  • Ha, C., McCarthy, M. F., Strambler, M. J., & Cipriano, C. (2026). Disentangling the effects of social and emotional learning programs on student academic achievement across grades 1–12: A systematic review and meta-analysis. Review of Educational Research, 96(5), 1839–1880. https://doi.org/10.3102/00346543251367769
  • Hagelskamp, C., Brackett, M. A., Rivers, S. E., & Salovey, P. (2013). Improving classroom quality with the RULER approach to social and emotional learning: Proximal and distal outcomes. American Journal of Community Psychology, 51(3–4), 530–543. https://doi.org/10.1007/s10464-013-9570-x
  • Hennessey, A., & Humphrey, N. (2020). Can social and emotional learning improve children's academic progress? Findings from a randomised controlled trial of the Promoting Alternative Thinking Strategies (PATHS) curriculum. European Journal of Psychology of Education, 35(4), 751–774. https://doi.org/10.1007/s10212-019-00452-6
  • Hua, Y., Godor, B. P., Nederhand, M., Smeets, G., & Van der Hallen, R. (2026). Programs and practices influencing socio-emotional learning in primary schools: A meta-analysis. Journal of Applied School Psychology, 42(4), 346–369. https://doi.org/10.1080/15377903.2026.2701647
  • Humphrey, N., Hennessey, A., Lendrum, A., Wigelsworth, M., Turner, A., Panayiotou, M., Joyce, C., Pert, K., Stephens, E., Wo, L., Squires, G., Woods, K., Harrison, M., & Calam, R. (2018). The PATHS curriculum for promoting social and emotional well-being among children aged 7–9 years: A cluster RCT. Public Health Research, 6(10). https://doi.org/10.3310/phr06100
  • Hung, C., Brass, N. R., Brockmeier, L., Bergin, C., Imler, M., & Luper, S. (2026). Social and emotional learning programs and students’ prosocial behavior: A meta-analysis. Review of Educational Research. https://doi.org/10.3102/00346543261438462
  • Jennings, P. A., & Greenberg, M. T. (2009). The prosocial classroom: Teacher social and emotional competence in relation to student and classroom outcomes. Review of Educational Research, 79(1), 491–525. https://doi.org/10.3102/0034654308325693
  • Jones, S. M., & Bouffard, S. M. (2012). Social and emotional learning in schools: From programs to strategies and commentaries. Social Policy Report, 26(4), 1–33. https://doi.org/10.1002/j.2379-3988.2012.tb00073.x
  • Li, A., Miller, F. G., & Williams, S. C. (2025). Cultural adaptations to social-emotional learning programs: A systematic review. School Psychology, 40(2), 108–120. https://doi.org/10.1037/spq0000649
  • Rivers, S. E., Brackett, M. A., Reyes, M. R., Elbertson, N. A., & Salovey, P. (2013). Improving the social and emotional climate of classrooms: A clustered randomized controlled trial testing the RULER approach. Prevention Science, 14(1), 77–87. https://doi.org/10.1007/s11121-012-0305-2
  • Shi, J., & Cheung, A. C. K. (2024). Effective components of social emotional learning programs: A meta-analysis. Journal of Youth and Adolescence, 53(4), 755–771. https://doi.org/10.1007/s10964-024-01942-7
  • Sklad, M., Diekstra, R., De Ritter, M., Ben, J., & Gravesteijn, C. (2012). Effectiveness of school-based universal social, emotional, and behavioral programs: Do they enhance students’ development in the area of skill, behavior, and adjustment? Psychology in the Schools, 49(9), 892–909. https://doi.org/10.1002/pits.21641
  • Taylor, R. D., Oberle, E., Durlak, J. A., & Weissberg, R. P. (2017). Promoting positive youth development through school-based social and emotional learning interventions: A meta-analysis of follow-up effects. Child Development, 88(4), 1156–1171. https://doi.org/10.1111/cdev.12864
  • Zhao, Y., & Sang, B. (2025). The effect of social-emotional learning programs on elementary and middle school students’ academic achievement: A meta-analytic review. Behavioral Sciences, 15(11), Article 1527. https://doi.org/10.3390/bs15111527