Impact of Service-Learning Among Adolescents: A Systematic Review

Abstract

Service-learning has emerged as a promising pedagogical approach to address social challenges while fostering students’ academic, socioemotional, and civic development. Despite the growing body of literature, there is a lack of systematic reviews evaluating its effectiveness in secondary education beyond the United States. Therefore, this systematic review aimed to synthesize the available empirical research on service-learning interventions implemented with secondary education students (grades 7–10) outside United States. Following The Campbell Collaboration guidelines, a comprehensive search was conducted across Web of Science and ProQuest, supplemented by other resources (general web search, hand searches, ongoing research, open access and relevant institutions and networks). Inclusion criteria required quantitative empirical studies (randomized controlled trials, quasi-experimental designs, or pretest–posttest designs) published from 2008 to June 2025 that measured academic, cognitive, socio-emotional, civic, and community outcomes using quantitative procedures. After screening, 10 studies met the inclusion criteria. The findings indicate that service-learning programs can generate positive outcomes on academic performance and non-cognitive skills. Nevertheless, methodological weaknesses and heterogeneity across studies prevent drawing robust conclusions. The review highlights the need for more rigorous research to strengthen the evidence base for service-learning among adolescents in systems worldwide.

1. Introduction

Global transformations associated with economic growth, digitalization, and unsustainable resource use are reshaping societies and worsening social inequalities [1,2]. However, education plays a key role by providing spaces to envision shared futures and developing the skills and competences needed to respond to these challenges, which requires substantive changes in curricular content and pedagogical approaches [1,3]. In this scenario, service-learning emerges as a promising strategy that integrates community service with curricular objectives to develop academic, socioemotional, and civic skills while tackling social problems [4].

Service-learning is a teaching method that involves students in community engagement as part of their academic course work [4]. According to Bringle et al. (2017) [5], service-learning is defined as follows:

“a course or competency-based, credit-bearing educational experience in which students (a) participate in mutually identified service activities that benefit the community and (b) reflect on the service activity in a way that enables them to gain a deeper understanding of course content, a broader appreciation of the discipline, and an enhanced sense of personal values and civic responsibility (p. 10)”.

Service-learning has its roots in Kolb’s experiential learning theory, which emphasizes the importance of practical experience in the learning process. Kolb’s model identifies four roles within the learning cycle: (a) concrete experience (i.e., active participation in an activity or situation); (b) reflective observation (i.e., connecting prior experiences or understanding); (c) abstract conceptualization (i.e., integrating observations and perceptions into abstract concepts); (d) active experimentation (i.e., shaping skills and testing new ideas). Concrete experience and abstract conceptualization represent the modes of grasping experience, whereas reflective observation and active experimentation are the methods for transforming that experience [6,7].

Building on this theory, Furco (1996) conceptualized service-learning as a specific form of experiential education characterized by an intentional balance between service and learning, such that both the service providers and recipients benefit from the activity, mutually reinforcing each other [8]. Accordingly, service-learning is an experiential pedagogical approach designed to address community needs while fostering a deeper understanding and essential skills development among students [9]. Students engage in a concrete experience in real-world community contexts and systematically reflect on these experiences within service-learning interventions. Structured reflection enables students to construct abstract concepts that can be tested and refined in new service-learning experiences [10].

This methodology can transform the teaching-learning process by immersing students in authentic social challenges, thereby promoting the acquisition of values and the enhancement of critical thinking skills [11]. Beyond the transmission of disciplinary knowledge, service-learning supports the development of key competencies, such as analytical and synthetic reasoning, as well as effective teamwork [12]. Service-learning is characterized by a deliberate pedagogical intent that integrates curricular objectives with community engagement, ensuring a reciprocal benefit for both students and the community [8,13]. Consequently, service-learning is conceptually distinct from related experiential practices and other methodologies in its intentional balance between academic learning and community service, as illustrated in

Table 1. Theoretical Distinctions between Service-Learning and Alternative Methodologies.

Methodology	Primary Focus	Beneficiary	Definition
Service-learning	Balance between service provision and learning	Students and their community partners	Students apply academic content to address real community needs.
Volunteering	Service	Community	Altruistic activities performed to benefit others without a learning goal.
Community Service	Service	Community	Organized activities in which students contribute to addressing a community cause or need without intentional learning.
Internships	Learning	Students	Structured work experiences in which students acquire professional skills relevant to their future career.
Field Education	Learning	Students	Supervised experiences in which students conduct observations, interviews, or data collection in real-world settings.
Project-Based Learning	Learning	Students	Extended inquiry into a complex question or challenge in which students develop understanding and skills through project design and completion.
Action Research	Research	Community members and researchers	Collaborative process in which researchers and community members collaborate to identify problems, conduct inquiry, and test solutions for social change.
Community-Based Research	Research	Community members and researchers	Partnership-based research approach in which academic researchers and community members generate knowledge and develop policy recommendations to address community needs.
Social Entrepreneurship	Service and business	Community members	Social enterprises that combine economic viability with explicit social impact goals to address community challenges are created and operated.

.

The implementation of service-learning assumes particular significance within the context of secondary education, given the distinctive psychosocial characteristics that define this developmental stage. Adolescence constitutes a critical period characterized by identity formation, exploration of personal purpose, and the establishment of value systems that fundamentally shape students’ subsequent academic and professional trajectories [14]. Within this developmental framework, service-learning provides adolescents with distinctive pedagogical opportunities to examine their potential roles as agents of social transformation while simultaneously developing integrated academic and personal competencies through experiential learning methodologies [15].

Secondary education constitutes a pivotal period for cultivating social awareness and civic responsibility. Through service-learning programs, exposure to social issues and public institutions enhances students’ engagement and fosters a heightened awareness of political and social dynamics within their communities [16,17]. Such experiences facilitate the development of abstract reasoning and critical reflection, which are essential for analyzing complex social challenges and making informed decisions. Notably, the heightened neuroplasticity of the adolescent brain renders it particularly receptive to external influences, positioning service-learning as a distinctive educational intervention capable of inducing structural brain changes with enduring effects that may persist into adulthood [15,18,19].

Numerous systematic reviews and meta-analyses have examined the literature on service-learning in secondary education. White (2001) [20] conducted the first review to include studies that specifically focused on this population. Drawing on studies published from 1983 to December 2000, this meta-analysis explored the association between middle and high school students’ service-learning interventions and outcomes such as academic achievement, self-concept, and social or personal development. The findings indicated that service-learning programs impacted students’ academic performance, self-concept, and social and personal growth. The effect sizes reported were moderate, underscoring the significant impact of service-learning across these domains. Furthermore, the review identified the following key characteristics of effective service-learning programs: (a) clear integration with the academic curriculum; (b) opportunities for structured reflection; (c) active student involvement in the planning and implementation of activities; (d) meaningful collaboration with the community.

Conversely, the meta-analysis conducted by Conway et al. (2009) [21] examined the available evidence regarding the impact of service-learning on academic, personal, social, and civic outcomes across primary, secondary, and higher education. The authors reported moderately positive effects on academic achievement, significant but small effects on personal outcomes, small-to-moderate effects on social outcomes, and small effects on civic outcomes. The analysis revealed that service-learning programs incorporating structured reflection activities produced greater changes among students in terms of moderators and factors contributing to program success. Positive effects were consistently observed across all educational levels, and no clear differences were observed based on the duration or number of service hours.

Celio et al. (2011) [22] conducted a systematic review to determine the benefits experienced by participants in service-learning programs, specifically evaluating variables that may moderate student outcomes at the elementary, secondary, or postsecondary levels. This meta-analysis included 62 studies with a total of 11,837 students and found that students who participated in service-learning programs demonstrated significant improvements in five areas: (a) self-attitudes (i.e., self-esteem, self-concept); (b) school and learning; (c) civic engagement; (d) social skills; (e) academic achievement. The average effect sizes ranged from 0.27 to 0.43, indicating small to moderate but consistent and significant impacts across all domains. Consistent with previous meta-analyses, the authors highlighted curriculum integration, active student participation, community collaboration, and structured reflection as key factors for achieving significant outcomes. Positive effects were observed at all educational levels (i.e., elementary, secondary, and postsecondary).

Most recently, the review by Filges et al. (2022) [23], which included studies published up to 2020, focused on the effects of service-learning on academic achievement, NEET status, personal and social skills, and risk behaviors among primary and secondary school students in the United States. The findings indicated that research on service-learning in these educational stages remains inconclusive, except for a modest improvement in mathematics. No statistically significant effects were observed for other indicators, such as academic achievement, personal/social skills, or risk behaviors.

A systematic review using evidence synthesis summarizes the best available research on specific questions [24]. Following the guidelines of The Campbell Collaboration, a high-quality systematic review should include the following [24]: (a) clear inclusion and exclusion criteria; (b) an explicit search strategy; (c) systematic coding and analysis of the included studies; (d) meta-analysis (when possible); (e) an integrative summary of the findings.

Despite the variety of reviews evaluating the impact of service-learning in secondary education, certain limitations remain. Filges et al. (2022) [23] is the only updated systematic review of the impact of service-learning in secondary education. However, this review was limited to research conducted in the United States, thereby restricting the generalizability of its findings to other contexts. The remaining literature reviews are outdated, with the last one conducted by Celio et al. in 2011 [22]. Therefore, a systematic review and meta-analysis that examines the impact of service-learning interventions in secondary education in other countries is needed to establish the evidence quality in this geographical area. However, a narrative synthesis was conducted instead of a meta-analysis due to the limited number of eligible studies and the substantial heterogeneity in designs and outcome measures.

To this end, the following research questions were posed:

(RQ 1) What are the most relevant characteristics of the studies (i.e., country, type of publication, sampling procedures and characteristics, type of research design, and duration) on service-learning interventions implemented in secondary education?

(RQ 2) What are the main characteristics of service-learning interventions implemented in secondary education (i.e., contents, procedures, practices, strategies, techniques, resources, activities, and services provided)?

(RQ 3) What are the outcomes and what instruments have been used to assess the impact of service-learning interventions implemented in secondary education?

(RQ 4) Which service-learning interventions have proven effective in improving secondary school students’ academic, learning, socio-emotional, civic, and community outcomes?

2. Method

2.1. Criteria for Considering Studies for This Review

First, a systematic review protocol was elaborated according to Campbell Systematic Reviews Policies and Guidelines [24]. The protocol is registered in Open Science Framework repository at the following link: https://osf.io/27ycz/overview?view_only=14ee7f81526e4eb38ba91da16a68fcb6 (accessed on 8 January 2026). This study relied on secondary data; therefore, ethical approval was not required.

Types of interventions. Service-learning interventions are practices, products, policies, programs, projects, experiences, actions, and activities that connect academic activities with community service. For an intervention to be considered service-learning, it must address real community needs, be implemented during the academic year, integrate service and learning, assign students an active role, include activities of reflection or critical analysis of both the curricular content and the services provided, ensure the participation of students, community partners, and members of the secondary school institution, and generate at least one of the study outcomes. Interventions will not be of restricted duration or setting.

Types of outcome measures. Service-learning interventions must assess at least one of the following outcomes: (a) academic (e.g., grade point average); (b) cognitive (e.g., problem-solving); (c) socioemotional (e.g., social awareness); (d) civic (e.g., civic engagement); (e) community (e.g., services provided). Outcomes must be measured using quantitative procedures (e.g., questionnaires).

Types of studies. This review will include studies that adopt a randomized controlled trial design, a cluster-randomized trial, a quasi-experimental design with a control group, or a pretest–posttest design. This study aims to obtain empirical evidence that allows objective and systematic comparison of results through numerical indicators. Although qualitative studies are highly valuable for understanding outcomes, they do not provide directly comparable data and may introduce methodological heterogeneity that complicates the interpretation of results. Studies must be published in English or Spanish, given the expertise of the evaluation team. Studies conducted in the United States and those published before 2008 will be excluded because previous meta-analyses on service-learning programs—conducted by Filges et al. (2022) and Celio et al. (2011) [22,23]—included programs implemented in secondary schools in the United States up to 2007. No restrictions based on publication status or type will be applied.

Types of participants. This systematic review will include studies that directly target students aged 12–16 years, as this age group corresponds to compulsory secondary education in Spain. However, studies that include participants slightly outside this range (e.g., up to 18 years) due to grade retention will also be considered eligible. Studies involving participants from educational institutions in the United States will be excluded.

2.2. Search Methods for the Identification of Studies

Following The Campbell Collaboration guidelines on study search procedures [25] and based on the review conducted during the exploratory search, a comprehensive and diverse search strategy was used to identify relevant studies across a range of electronic databases. This search was conducted in June 2025. Therefore, the period covered in this study extends from 2008, the end-date of the studies reviewed by previous meta-analyses on service-learning programs, onwards until 2025.

The following electronic platforms and key databases were searched (primary search): (a) Web of Science: Science Citation Index Expanded, Social Science Citation Index, Current Contents Connect, Korean Journal Database, Preprint Citation Index, and SciELO Citation Index; (b) Proquest: APA PsicoArticles, APA PsicoExtra, APA PsicoInfo, ProQuest Central Premium (ERIC, ProQuest One Education, ProQuest One Psychology, ProQuest One Social Sciences, ProQuest One Social Sustainability), ProQuest Dissertations & Theses Global, ProQuest One Academic (Academic Complete, and ProQuest Central: Australia & New Zealand Database, Continental Europe Database, East & South Asia Database, East Europe, Central Europe Database, Education Database, Environmental Science Database, India Database, Latin America & Iberia Database, Middle East & Africa Database, Psychology Collection, Research Library, Social Science Database, Sociology Database, Turkey Database, UK & Ireland Database), ProQuest Dissertations & Theses Global, ProQuest One Education, Psychology Collection, and Social Science Premium Collection (Education Collection, International Bibliography of the Social Sciences, Social Science Database, Sociology Collection).

Likewise, resources selected for complementary search were the following: (a) a general web search (Google Scholar); (b) hand searches: reference lists of included studies were also searched; (c) ongoing research: Community Research and Development Information Service; (d) open access (gray literature): Directory of Open Access Repositories, and National Technical Information Service; (e) relevant institutions and networks: Center for Information and Research on Civic Learning & Engagement, Centro Latinoamericano de Aprendizaje y Servicio Solidario (Latin American Center for Solidarity Service-Learning), Centro Promotor de Aprendizaje y Servicio Solidario en Cataluña (Promotion Center for Service-Learning in Catalonia), Centro Promotor del Aprendizaje y Servicio Solidario en Euskadi (Promotion Center for Service-Learning in Euskadi), International Association for Research on Service-learning and Community Engagement, and National Youth Leadership Council; (f) manual search in additional key journals: Collaborations, A Journal of Community-Based Research and Practice, Engaging Pedagogies in Catholic Higher Education, Engaged Scholar Journal: Community-Engaged Research, Teaching, and Learning, International Journal of Research on Service-Learning and Community Engagement, and International Journal for Service Learning in Engineering.

The search terms were selected using the Education Resources Information Center Thesaurus. The following terms were searched in the title, abstract, and keywords (indexing) search fields:

Types of interventions: (“service-learning” OR “e-service-learning” OR “service-learning” OR “online service-learning” OR “community learning” OR “experiential learning” OR “experiential education” OR “practical learning” OR “competency-based learning” OR “competency-based teaching”).

Types of outcomes: (a) academic and learning: (academic OR learn* OR “academic performance” OR “grade point average” OR GPA OR achievement OR attendance OR dropout OR retention OR repetition OR graduation OR “problem solv*” OR “critical think*” OR “moral reason*” OR “self-regulated learn*” OR “sense of belong*” OR attitude* OR motivation); (b) socio-emotional: (socio-emotion* OR social OR emotion* OR value* OR abilit* OR behavior* OR self-awareness OR self-perception OR self-concept OR self-esteem OR self-efficacy OR “growth mindset” OR “social awareness” OR empath* OR self-control OR resilience OR will OR discipline OR perseverance OR “organizational strateg*” OR “planning strateg*” OR “relationship skill*” OR “active listen*” OR “seeking help” OR “offering help” OR “conflict resolution” OR “resisting negative social pressure” OR “working in groups” OR teamwork OR “responsible decision-making” OR well-being OR “drug use” OR “alcohol use” OR “sexual risk taking” OR “problem gambling”); (c) civic: (civic* OR politic*); (d) career and employment: (career OR employ*); (e) community: (community OR servic* OR infrastructur* OR “human capital” OR engagement OR satisfaction).

Types of settings: (“secondary education” OR “secondary school*” OR “high school*” OR “K-12” OR “middle school”).

Types of studies: (experiment* OR randomi* OR quasi-experiment* OR effect* OR efficacy OR evaluat* OR impact* OR assess* OR compar* OR control OR “regression discontinuity” OR “trial*” OR intervention* OR practice* OR product* OR polic* OR program* OR project* OR experience* OR action* OR activit*).

The search strategy was modified according to the specifications of each platform, database, and website. For example, the terms and strings used for the Proquest search were the following: TITLE,ABSTRACT,IF (“service-learning” OR “e-service-learning” OR “service-learning” OR “online service-learning” OR “community learning” OR “experiential learning” OR “experiential education” OR “practical learning” OR “competency-based learning” OR “competency-based teaching”) AND TITLE,ABSTRACT,IF (academic OR learn* OR “academic performance” OR “grade point average” OR GPA OR achievement OR attendance OR dropout OR retention OR repetition OR graduation OR “problem solv*” OR “critical think*” OR “moral reason*” OR “self-regulated learn*” OR “sense of belong*” OR attitude* OR motivation OR socio-emotion* OR social OR emotion* OR value* OR abilit* OR behavior* OR self-awareness OR self-perception OR self-concept OR self-esteem OR self-efficacy OR “growth mindset” OR “social awareness” OR empath* OR self-control OR resilience OR will OR discipline OR perseverance OR “organizational strateg*” OR “planning strateg*” OR “relationship skill*” OR “active listen*” OR “seeking help” OR “offering help” OR “conflict resolution” OR “resisting negative social pressure” OR “working in groups” OR teamwork OR “responsible decision-making” OR well-being OR “drug use” OR “alcohol use” OR “sexual risk taking” OR “problem gambling” OR civic* OR politic* OR career OR employ* OR community OR servic* OR infrastructur* OR “human capital” OR engagement OR satisfaction) AND TITLE,ABSTRACT,IF (“secondary education” OR “secondary school*” OR “high school*” OR “K-12” OR “Middle School”) AND TITLE,ABSTRACT,IF (experiment* OR randomi* OR quasi-experiment* OR effect* OR efficacy OR evaluat* OR impact* OR assess* OR compar* OR control OR “regression discontinuity” OR “trial*” OR intervention* OR practice* OR product* OR polic* OR program* OR project* OR experience* OR action* OR activit*).

2.3. Data Collection and Analysis

The Zotero software (7.0.11, Corporation for Digital Scholarship, Vienna, VA, USA) was used to manage and document the search process. Bibliographic information on the studies was imported into Zotero, and the search records were retained.

The selection process included the following stages: (a) the first stage involved identifying and excluding duplicates and applying date and language filters; (b) the second stage involved removing studies that were clearly ineligible for inclusion according to the title and abstract; (c) the third stage consisted of reading each full-text document to determine whether it met all the inclusion criteria. The authors independently conducted the selection process. Any disagreements were resolved by consensus.

Data extraction was conducted independently by two reviewers using a standardized coding sheet in Excel. Discrepancies in coding (e.g., study design, outcome classification, or effect direction) were discussed and resolved by consensus. Each study was coded according to the following descriptors [26]: (a) contextual characteristics (i.e., reference, country, year of publication, and type of publication); (b) type of research design; (c) sampling procedure and characteristics (i.e., sample size, mean age and age range, and grade level); (d) name and description of the service-learning program implemented (i.e., setting, program provider, duration, intervention content, procedures, practices, strategies, techniques, resources, and activities); (e) outcome measures (i.e., outcome variable and measurement instrument); (f) results and conclusions. Effect sizes were extracted as reported in the original studies when available, without additional calculation or transformation into a common metric. The effect size metrics included standardized mean differences (Cohen’s d, Hedges’ g), eta-squared (η²), odds ratios, and Pearson correlation coefficients (r). This heterogeneity in effect size metrics further supported the decision to rely on a narrative synthesis rather than a meta-analysis.

Finally, a narrative content analysis approach was adopted for the data analysis [27,28], given that the number of studies included in the review was small and that the studies were too heterogeneous in terms of research designs and the set of dependent variables or outcomes analyzed [29,30]. A structured narrative synthesis was undertaken. The studies were grouped according to their primary outcome domain (academic, cognitive, socioemotional, civic, and community outcomes). Within each domain, specific outcomes, measurement instruments, and the direction and magnitude of effects were organized.

All materials used in this review (including template data collection forms, data extracted from the included studies, or any other relevant materials) are available from the authors upon request.

3. Results

Figure 1 presents a flow diagram visually summarizing the search and screening processes. After applying the search filters, the two main search strategies yielded 4342 results. At the first screening level, 1550 studies were excluded because they were duplicates or published before 2008, leaving 2792 studies. At the second screening level, after further examination of the title and abstract, 1955 studies were removed, resulting in 837 studies. Finally, 827 studies were excluded after reading the full text versions for having irrelevant outcomes, research design, or participant samples at the third screening level. As a result, 10 studies fully met the inclusion criteria.

Table 2. Characteristics of the Included Studies.

Study	Country	Sampling	Sample Size	Age	Design	Duration
Cedena-de-Lucas et al. (2024) [31]
	Spain	Convenience sampling. Classes were randomly assigned to experimental and control groups	60 EG = 31 CG = 29	Mean age of 14.9 years (SD = 0.52)	Cluster-randomized controlled trial	3 months
Fernández-Bustos et al. (2024) [11]
	Spain	Conveniencia. Classes were randomly assigned to experimental and control groups	202 EG = 135 CG = 67	Mean age of 15	Quasi-experimental with control group, pretest, posttest 1 (phase 1), and posttest 2 (phase 2)	Two semesters
Brozmanová-Gregorová et al. (2024) [32]
	Spain, Austria, Ireland, Slovakia, and Croatia	Convenience sampling	232 EG = 138 CG = 94	Mean age of 13.7 years (SD = 0.30)	Quasi-experimental design with a non-equivalent control group	An academic year, 8–9 months
Cabello-Sanz & Muñoz-Parreño (2023) [33]
	Spain	Convenience sampling	30 EG = 15 CG = 15	Mean age of 13.74 years (SD = 0.54; range = 13–16)	Quasi-experimental design (pre–post with control group)	10 weeks (one 50 min session per week)
Gutiérrez et al. (2019) [34]
	Spain	Convenience sampling	28 EG = 28 CG = NA	Range = 15–16	Pretest–posttest	2 months (15 ringo sessions + final 4 h interschool event)
Rauschert & Byram (2017) [35]
	Germany and India	Convenience sampling	59 EG = 32 CG = 27	Range = 15–16	Quasi-experimental design (pre–post with control group)	One semester (20 lessons)
Ochoa-Cervantes & Pérez-Galván (2019) [36]
	Mexico	Convenience sampling	117 EG = 117 CG = NA	Range = 13–16	Pretest–posttest	6 months
Ng et al. (2020) [37]
	Singapore	Convenience sampling	359 EG = 359 CG = NA	Mean age of 15 years (range = 13–17)	Pretest–posttest	6 months (12 visits per team)
Tan et al. (2008) [38]
	Singapore	Convenience sampling	279 EG = 279 CG = NA	Mean age of 15.64 years (SD = 0.84; range = 14–18)	Pretest–posttest	2 months
Wong et al. (2012) [39]
	China	Convenience sampling. Classes were randomly assigned to experimental and control groups	180 EG = 50 CG = 130	Mean age of 15.18 years (SD = 0.62)	Randomized Controlled Trial	6 months (4 h per month)

Note. EG = experimental group; CG = control group; NA = not applicable.

summarizes the contextual and methodological characteristics of the ten reviewed studies. All included studies were published as articles in indexed journals. The interventions were conducted in Europe (Spain, Germany, Austria, Ireland, Slovakia, and Croatia), Latin America (Mexico), and Asia (Singapore and China) with sample sizes ranging from 28 to 359 students. The majority employed pretest–posttest designs, although some implemented quasi-experimental designs and randomized controlled trials. Studies with the highest methodological quality were found in Spain, where 3 of the 5 included studies employed quasi-experimental designs and 1 study used a randomized controlled trial design. The duration of interventions varied widely, from short-term programs lasting a few weeks to academic year implementation.

Table 3. Characteristics of the Intervention.

Study	Intervention (Content, Procedures, Practices, Strategies, Techniques, Resources, and Activities)	Service Component
Cedena-de-Lucas et al. (2024) [31]
	The intervention combined training, action, and evaluation. First, the partner organizations and teachers were trained, the contents were adapted, and the workshops were prepared. Afterwards, the students in the intervention group participated in three interactive workshops, which addressed general issues related to online hate speech. Subsequently, the students designed a service activity within the Spanish Language subject to apply what they had learned and then carried out its implementation in the school community.	The students engaged in workshops and awareness-raising activities in which they reflected on the impact of hate speech and created educational campaigns targeting their school and community environments. Their contribution to the community consisted of promoting a more respectful and safer digital culture, helping to reduce discriminatory and violent attitudes on social media.
Fernández-Bustos et al. (2024) [11]
	The intervention consisted of the implementation of a hybrid program combining service-learning and the Sport Education Model, developed in two phases. In Phase 1, the game of ringo was taught through instructional units in Primary Education (14–18 sessions) and Secondary Education (18–21 sessions), incorporating SEM practices (responsibility roles, affiliation, record keeping, celebration, and final event) along with SL activities (Secondary students visiting Primary schools, joint training, and a culminating event). In Phase 2, in the following academic year, the intervention was consolidated with a 10-session program, preceded by 6 preparatory sessions, using modified net and wall games. Activities combined technical and tactical instruction, structured competition, and collective reflection to promote sports, social, and coexistence learning.	High school students acted as sports instructors, teaching the game of ringo to younger peers and organizing competitions. They contributed to the community by promoting values such as cooperation, responsibility, and coexistence through sport, providing a positive educational and recreational experience for both the recipients and the providers of the service.
Brozmanová-Gregorová et al. (2024) [32]
	The intervention was based on the implementation of the PLACE service-learning model within the Erasmus+ SLUSIK project in secondary schools in Europe. The process followed five stages: preparation and design of the experience by teachers and university students who participated as role models; engagement with community partners and definition of roles; action, meaning the implementation of the project addressing local needs; celebration through a public presentation of the results to families, the community, and policymakers; and effect, which included final reflection on learning outcomes and contributions to the community. The strategies involved collaborative work, experiential learning, real-world problem solving, and critical reflection	The students developed service projects tailored to local needs such as environmental care, social inclusion, and the promotion of democratic values. They planned, implemented, and publicly presented their initiatives together with community partners, which allowed communities to receive direct support while young people acquired key social and civic competencies to prevent exclusion and early school leaving.
Cabello-Sanz & Muñoz-Parreño (2023) [33]
	The intervention consisted of the implementation of the Universo Emocionante program, with weekly sessions based on the pentagonal model of emotional competencies. The practices and strategies included participatory dynamics, emotional expression, group work, decision-making, and community service. Techniques such as experience sharing, role-playing, critical reflection, and relaxation exercises were used, supported by educational resources including digital materials, activity guides, and final products. These products were designed to raise awareness and contribute to the educational community by highlighting the importance of emotional education.	The students developed workshops and emotional education activities for their peers. These activities were carried out in the classroom through group dynamics and practical sessions, allowing the learning to extend to the whole class group and improving school coexistence. In this way, they provided a service to their educational community by reducing conflict and enhancing collective coexistence and well-being.
Gutiérrez et al. (2019) [34]
	The intervention integrated service-learning into the Sport Education Model. A ringo season of 15 sessions plus a culminating event was designed, structured in phases: organization, preseason (i.e., learning the sport and responsibility roles), regular season (i.e., intragroup and interschool competition), and final stage (i.e., formal competition and celebration). Strategies included affiliation, roles, micro-teaching, and seasonal progression, with techniques such as friendly matches, training sessions, socialization activities, and fair play dynamics. Finally, the culminating event at the school integrated welcoming activities, training sessions, and formal competition.	Secondary school students organized school sports leagues for primary education pupils. In doing so, they acted as coaches and role leaders and organized the final sports event. They contributed to the educational community by transmitting knowledge and fostering the skills of primary students. In addition, social ties were strengthened, which facilitated the transition from primary to secondary education.
Rauschert & Byram (2017) [35]
	The intervention was carried out within English as a Foreign Language class, integrating linguistic, intercultural, and citizenship content. German students collaborated in producing a bilingual magazine together with young people from India: they conducted local interviews and analyzed video clips from their Indian peers, wrote background articles to contextualize cultural responses, and engaged in guided reflection sessions. Cooperative learning strategies, interview techniques, and critical analysis were employed, along with audiovisual and written resources. The central activity was the production of the magazine, the proceeds of which were allocated to the construction of a school in India.	The students worked with local and international communities to address social needs such as migrant integration, social justice, and environmental sustainability. Their service consisted of collaborating with diverse groups using language as a tool for communication, which contributed to fostering intercultural understanding and strengthening global citizenship while simultaneously responding to real community issues.
Ochoa-Cervantes & Pérez-Galván (2019) [36]
	The intervention was implemented in several phases: motivation, diagnosis, planning, implementation, and evaluation. The contents were linked to subjects such as Science, Spanish, Technology, Mathematics, and Civic and Ethical Education, integrating academic knowledge with real-world issues. Participatory strategies were employed, including social mapping, brainstorming, debates, project design, and teamwork, along with resources such as murals, brochures, collection campaigns, school gardens, and reforestation projects. The activities were designed to address community and school needs through inquiry-based techniques and group reflection.	The students carried out a participatory diagnosis to identify the needs of their school and community environments, and based on this, they designed and implemented concrete actions such as waste separation campaigns, reforestation of green areas, support for children with kidney failure in a hospital, and animal protection projects.
Ng et al. (2020) [37]
	The Tri-Generational HomeCare program was implemented through an interprofessional and intergenerational service-learning model. First, secondary school students and university students from health-related disciplines received preparatory training sessions (i.e., workshops, role-play, and practical activities). Afterwards, mixed teams (2–3 university students and 3–5 secondary students) carried out biweekly home visits for six months to older patients with frequent hospitalizations. The university students guided the subsequent reflections and facilitated multidisciplinary meetings with health professionals to discuss cases and receive feedback.	Teams of secondary school and health sciences university students conducted biweekly home visits to older adults with frequent hospitalizations. They provided companionship, assessed their needs, and presented cases to healthcare professionals to improve patient care. Their contribution to the community involved reducing loneliness among older adults, strengthening intergenerational networks, and lowering the number of hospitalizations.
Tan et al. (2008) [38]
	The intervention program consisted of a short-term international service-learning experience (3 to 5 days), preceded by two months of preparation in Singapore, during which students received training in relevant skills and processes. During the stay, they worked in host communities following an inquiry-based approach: they selected a local public issue, investigated it, analyzed the situation, proposed creative solutions, and carried out civic actions. The experience integrated community service, academic learning, collaborative work, and student leadership, supported by daily reflection activities and facilitated feedback, in which participants recorded emotions and learning outcomes.	The students collaborated in foreign communities by engaging in social and educational activities (e.g., supporting local schools, cultural activities, and community service). They contributed to the community by providing support in contexts of need, while at the same time strengthening their creativity, self-efficacy, and personal well-being, creating a rich cultural exchange.
Wong et al. (2012) [39]
	The intervention was a leadership program with a service-learning approach called “Hand in Hand Serves the Community.” It included training in leadership and teamwork, with activities designed to develop creativity, social communication, resilience, and self-esteem. Students took part in monthly training meetings, a leadership workshop, and community service activities. In addition, reflection and final evaluation were promoted through a closing ceremony, where students shared their experiences and consolidated their learning.	The students supported children with special needs and older adults in vulnerable situations. They organized visits, workshops, and festive events, providing companionship and emotional support to these groups. The community received valuable assistance and care, while the young participants strengthened their self-esteem, self-efficacy, and social commitment.

presents the main characteristics of the service-learning interventions. The programs were implemented across diverse academic areas. The service components ranged from social awareness campaigns and intergenerational care to sports training, environmental projects, and intercultural collaborations, demonstrating the versatility of service-learning in addressing both educational and community goals.

Table 4. Summary of Results and Conclusions.

Study	Academic	Cognitive	Socioemotional	Civic	Community
Cedena-de-Lucas et al. (2024) [31]
	NA	NA	Outcomes: Perpetration of online hate speech. Instruments: Justification of Cyberbullying Scale [40]. Risk Perception Scale [41]. Spanish version of the Coping with Cyberhate Questionnaire [42]. Motivations for Hate Speech Perpetration Scale [43]. Effect Size: d = 0.42 Conclusions: The study did not report statistically significant improvements.	NA	NA
Fernández-Bustos et al. (2024) [11]
	Outcomes: learning in physical education (i.e., technical performance in the game of ringo, tactical decision-making in the game, theoretical knowledge). Instruments: Game Performance Assessment Instrument adapted to the game of ringo, rubric format (execution of reception and decision-making in throwing). Ad hoc questionnaire on theoretical knowledge of ringo (15 questions divided into: technique, tactics, rules, fair play, and health). Effect size: NA Conclusions: The study reported statistically significant improvements in all outcomes.	NA	Outcomes: Social competence (i.e., prosocial behavior, cognitive reappraisal, social efficacy, social adjustment, normative adjustment). Instruments: Adolescent Multidimensional Social Competence Questionnaire [44]. Effect size: Providers: d = 0.37–0.66 Conclusions: The study reported statistically significant improvements in all outcomes.	NA	Outcomes: Social competence (i.e., prosocial behavior, cognitive reappraisal, social efficacy, social adjustment, normative adjustment). Instruments: Adolescent Multidimensional Social Competence Questionnaire [44]. Effect size: Recipients: d = 0.24–0.25 Conclusions: The study did not report statistically significant improvements, except for slight improvements in prosocial behavior and normative adjustment.
Brozmanová-Gregorová et al. (2024) [32]
	Outcomes: School engagement and motivation, intention to drop out of school. Instruments: Student Engagement, Motivation and Beliefs Survey [45]. Single item on intention to drop out of school. Effect size: Sense of belonging: CLES = 0.40 Intention to drop out of school: CLES = 0.43 Conclusions: The program generated significant improvements in all outcomes.	NA	Outcomes: Social competencies (i.e., emotional intelligence, empathy, regulation, social skills). Instruments: Trait Emotional Intelligence Questionnaire [46]. Effect size: Well-being: CLES = 0.37 Overall emotional intelligence: CLES = 0.34 Conclusions: The program generated significant improvements in all outcomes.	Outcomes: Civic competencies (i.e., responsibility, values, civic efficacy, participation). Instruments: Civic Responsibility Survey [47]. Civic Values and Beliefs Scale [48]. Effect size: Civic responsibility: CLES = 0.22 Civic efficacy: CLES = 0.23 Social responsibility: CLES = 0.29 Conclusions: The program generated significant improvements in all outcomes.	NA
Cabello-Sanz & Muñoz-Parreño, (2023) [33]
	NA	NA	Outcomes: Emotional awareness, emotional regulation, emotional autonomy, social competence, and competencies for life and well-being. Instruments: Emotional Development Questionnaire for Adolescents [49]. Effect size: Emotional awareness: ŋ2 = 0.39 Emotional competence (total) ŋ2 = 0.34 Emotional regulation: ŋ2 = 0.22 Social competence: ŋ2 = 0.05 Conclusions: The study reported statistically significant improvements in emotional awareness, emotional regulation, social competence, and overall emotional competence, although it did not achieve significant changes in emotional autonomy, competences for life and well-being.	NA	NA
Gutiérrez et al. (2019) [34]
	Outcomes: Responsibility, autonomy, and fair play. Instruments: Physical Education Season Survey [50]. Interviews. Effect size: NA Conclusions: The study reported statistically significant improvements in all outcomes.	NA	Outcomes: Social competence, teamwork Instruments: Physical Education Season Survey [50]. Self-reports. Interviews. Effect size: NA Conclusions: The study reported statistically significant improvements in all outcomes.	Outcomes: Civic competence. Instruments: Self-reports. Interviews. Effect size: NA Conclusions: The study reported statistically significant improvements in all outcomes.	NA
Rauschert & Byram (2017) [35]
	NA	NA	NA	Outcomes: Intercultural competence, and civic responsibility. Instruments: Ad hoc questionnaires on intercultural competence. Effect size: NA Conclusions: The study reported statistically significant improvements in all outcomes.	NA
Ochoa-Cervantes & Pérez-Galván, (2019) [36]
	NA	NA	Outcomes: Socio-emotional competencies. Instruments: Self-Assessment Rubric for Service-Learning Experiences. Effect size: NA Conclusions: The study reported statistically significant improvements in all outcomes.	Outcomes: Participation and group cohesion, school coexistence, civic competences. Instruments: Child Participation and School Coexistence Questionnaire. Organized Language Comprehension Logbook. Self-Assessment Rubric for Service-Learning Experiences. Effect size: NA Conclusions: The study reported statistically significant improvements in all outcomes.	NA
Ng et al. (2020) [37]
	Outcomes: Knowledge acquisition, and knowledge application. Instruments: Fund for the Improvement of Postsecondary Education Survey [51]. Effect size: NA Conclusions: The study reported statistically significant improvements in all outcomes.	Outcomes: Critical thinking. Instruments: Fund for the Improvement of Postsecondary Education Survey [51]. Effect size: NA Conclusions: The study reported statistically significant improvements in all outcomes.	Outcomes: Attitudes toward older adults, knowledge about aging, leadership, communication, teamwork. Instruments: Kogan’s Attitudes towards Old People Scale [52]. Palmore Facts on Aging Quiz [53]. Fund for the Improvement of Postsecondary Education Survey [51]. Effect size: NA Conclusions: The study reported statistically significant improvements in all outcomes.	Outcomes: Identification of social problems, responsibility. Instruments: Fund for the Improvement of Postsecondary Education Survey [51]. Effect size: NA Conclusions: The study reported statistically significant improvements in all outcomes.	NA
Tan et al. (2008) [38]
	NA	Outcomes: Perceived creative self-efficacy. Instruments: Creativity-relevant efficacy scales [54,55]. Effect size: Creative self-efficacy—cognitive style: Cluster 1: d = 0.02 Cluster 2: d = −0.02 Work style: Cluster 1: d = 0.11 Cluster 2: d = 0.02 Conclusions: The study reported statistically significant improvements in all outcomes.	Outcomes: Positive and negative affect, life satisfaction, and subjective happiness. Instruments: Positive and Negative Affect Schedule [56]. Satisfaction With Life Scale [57]. Subjective Happiness Scale [58]. Effect size: Positive affect: Cluster 1: d = 0.16 Cluster 2: d = 0.18 Life satisfaction: Cluster 1: d = 0.28 Cluster 2: d = 0.58 Subjective happiness: Cluster 1: d = 0.20 Cluster 2: d = −0.00 Conclusions: The study reported statistically significant improvements in all outcomes.	NA	NA
Wong et al. (2012) [39]
	NA	NA	Outcomes: Self-esteem and perceived self-efficacy. Instruments: Rosenberg Self-Esteem Scale [59]. General Self-Efficacy Scale [60]. Effect size: NA Conclusions: The program did not generate significant improvements in self-esteem or self-efficacy when considering the entire sample, although it did produce statistically significant effects among girls.	NA	NA

Note. NA = not applicable.

presents the included studies’ outcomes, instruments, and main conclusions. Given the diversity of outcomes, results are organized according to five domains: (a) academic: Four studies assessed academic related outcomes (e.g., physical education performance, knowledge acquisition, and knowledge application), and all four reported statistically significant improvements associated with service-learning; (b) cognitive: Two studies explicitly examined cognitive outcomes, particularly critical thinking and creative self-efficacy, and both studies reported statistically significant improvements in these cognitive variables following participation in service-learning programs; (c) socioemotional: Nine studies evaluated socioemotional outcomes such as emotional awareness, emotional regulation, social competence, self-esteem, self-efficacy, empathy, resilience, and subjective well-being, and eight of these studies found statistically significant improvements in at least one socioemotional indicator, whereas two studies reported non-significant changes for some variables; (d) civic: Five studies focused on civic outcomes, including civic responsibility, democratic values, participation, and intercultural civic competence, and all five reported statistically significant positive changes in at least some civic indicators; (e) community: One study reported community-level outcomes, such as social competence in service recipients. Overall, the interventions showed positive effects on socioemotional competencies (e.g., emotional awareness, social skills, empathy), civic competence, intercultural competence, sense of belonging, intention to drop out from school, and well-being, with some studies reporting improvements in academic performance and physical education learning. However, the magnitude of the effects varied considerably, with some programs showing small or gender-differentiated effects. However, only 4 of the 10 studies reported effect sizes.

4. Discussion

This systematic review aimed to identify, describe, and synthesize the characteristics and effectiveness of service-learning interventions on the academic, cognitive, socioemotional, civic, and community outcomes of secondary education students. Regarding RQ1, the studies included in this systematic review show considerable geographical diversity, with most interventions conducted in Europe, followed by Asia and Latin America. Unlike the review published by Celio et al. (2011) [22], which identified unpublished works that met the criteria, all the studies in this case were published in scientific journals. Most of the studies involved small samples, and six included a control group. Only three of the ten selected articles had a minimum of 30 participants per group, as recommended in the specialized literature [61]. In addition, four studies relied on pretest–posttest designs, four implemented quasi-experimental designs (pre–post with control group), and two were randomized controlled trials. These methodological weaknesses highlight the limited amount of available evidence. The lack of randomized trials, the predominance of quasi-experimental designs, and the small sample sizes make it difficult to obtain consistent and generalizable results, contributing to the variability and fragility of the conclusions drawn from the reviewed studies.

The duration of the interventions varied notably in the reviewed studies, ranging from very short programs of just 2 months to experiences lasting an entire academic year. Unlike the findings of Celio et al. (2011) and Filges et al. (2022) [22,23], who pointed out that many studies did not systematically report the duration of interventions, all studies in this review reported the duration of interventions (e.g., duration of the service provided). However, in many cases, the information provided was limited, making it difficult to determine the intervention intensity, which restricts the possibility of drawing solid conclusions.

In relation to RQ2, few studies systematically reported on curricular integration, including both curricular and service objectives, as also noted by Celio et al. (2011) [22]. This represents a major challenge in service-learning research, given the importance of linking curricular content with real-world issues. Not only is such integration associated with better outcomes, but the lack of systematic reporting also limits the replicability of studies [23]. Concerning reflection activities, an essential component of service-learning to enhance its effectiveness, only a few studies described how these activities were carried out, and they were not always implemented consistently, as similarly highlighted by Filges et al. (2022) [23]. The community services implemented in the included studies covered a wide range of activities, including anti-hate campaigns, support for older adults, emotional workshops, reforestation projects, and sports leagues. Following Filges et al. (2022) [23], many studies tended to focus on needs within the context of their own school. However, service-learning interventions are often designed to provide services to vulnerable populations and communities. Therefore, the diversity of interventions illustrates that service-learning interventions can be adapted to different educational contexts and community needs, contributing to the variability of outcomes.

To address RQ3, the interventions showed positive effects across a broad variety of outcomes, with particularly consistent improvements in socioemotional (e.g., emotional awareness, regulation, empathy, and resilience) and civic competences (e.g., responsibility, participation, democratic values, and sense of belonging). Some programs also reported improvements in academic achievement, school coexistence, intercultural competence, and physical education learning outcomes, highlighting the diversity of benefits observed. Most studies employed validated and reliable self-report measures, aligning with the pattern reported by Celio et al. (2011) [22]. In addition, some studies complemented quantitative scales with qualitative data (e.g., interviews and open-answered questions), allowing for a richer understanding of the processes and perceived impact. Unlike Filges et al. (2022) [23], who mainly focused on academic outcomes, the present evidence suggests that service-learning interventions in secondary education tend to prioritize the development of socioemotional and civic competencies.

The effect sizes of RQ4 varied from small to moderate, with stronger impacts typically observed in the socioemotional and civic competences. Particularly noteworthy are the studies by Cabello-Sanz and Muñoz-Parreño (2023), Fernández-Bustos et al. (2024), Brozmanová-Gregorová et al. (2024), Cedena-de-Lucas et al. (2024), and Tan et al. (2008) [11,31,32,33,38], while the remaining studies did not provide effect size estimates but reported significant results.

When study results are analyzed by outcome type (i.e., academic, cognitive, socioemotional, civic, and community), distinct patterns emerge. The 4 studies measuring academic outcomes consistently reported significant improvements across knowledge acquisition in diverse disciplines, motivation, school dropout rates, and academic performance, with notable effect sizes. The 2 studies evaluating cognitive outcomes demonstrated significant gains in critical thinking and creativity, although the effects were small. The most extensively studied domain was socioemotional outcomes, with seven studies reporting significant improvements in socioemotional competencies and well-being. However, the effect sizes ranged widely from small to large. Notably, 2 studies yielded non-significant socioemotional outcomes (i.e., self-esteem, self-efficacy, emotional autonomy, and competences for life and well-being). All six studies assessing civic outcomes reported significant improvements in responsibility, civic participation, and intercultural competence, with effect sizes ranging from small to moderate. Conversely, community outcomes were evaluated in only 1 study, demonstrating non-significant results.

Nevertheless, the lack of studies reporting effect sizes in their findings creates difficulties in comparing the actual magnitude of the benefits and limits the possibility of conducting quantitative synthesis. Therefore, the lack of systematic reporting and the inconsistency of effects weaken the conclusions about the impact of service-learning in secondary education.

5. Limitations

The findings of this systematic review should be interpreted with caution because of several limitations. First, the analysis was conducted narratively rather than through a meta-analysis because of the small number of eligible studies and the lack of consistent reporting across many of them. Second, the literature search was completed in June 2025, indicating that subsequent studies were not considered. Third, the possibility of publication bias cannot be ruled out because the inclusion of gray literature was limited, even though multiple well-established sources were consulted. Fourth, a risk of bias assessment was not conducted, which limits the ability to evaluate the methodological quality of the included studies and may affect the reliability of the synthesized findings (e.g., most interventions relied on non-randomized designs and convenience samples, which increases the likelihood of selection bias and limits the generalizability of the findings; outcome measures were predominantly based on self-report questionnaires administered to students, making the results vulnerable to social desirability and common-method variance). Although two reviewers independently conducted the selection process to reduce potential bias, the influence of reviewer subjectivity on the final sample cannot be excluded.

Several limitations related to the characteristics of the included studies should also be acknowledged. Most adopted pretest–posttest designs without control groups or quasi-experimental designs with small samples (i.e., fewer than 30 participants per group), which undermines internal validity and reduces the statistical power of the findings. In addition, many studies relied on measurement instruments with limited or insufficient evidence of validity and reliability. Another limitation is the scarcity of studies reporting effect sizes, which restricts meaningful cross-study comparison and prevents an accurate estimation of the true magnitude of intervention effects. Taken together, these methodological limitations constrain the strength of the conclusions that can be drawn about the impact of service-learning interventions in secondary education outside the United States.

6. Conclusions

This study synthesizes the most recent evidence on service-learning interventions conducted with secondary education adolescents outside the United States, a field that has been insufficiently examined in previous reviews [22,23]. This systematic review suggests that service-learning interventions in secondary education can generate positive effects, particularly in socioemotional and civic competences, fostering empathy, emotional regulation, responsibility, and a sense of school belonging. Although the academic effects appear more inconsistent, some programs report improvements in specific areas, such as physical education, indicating the impact of service-learning on different outcomes. However, cognitive and community outcomes were assessed less frequently across the studies, indicating that the impact of service-learning on these domains remains unclear. Therefore, more interventions should systematically evaluate these outcomes in order to draw stronger conclusions about their effects.

Several of the interventions included in this systematic review explicitly targeted students in vulnerable contexts, such as those at risk of school dropout, living in disadvantaged communities, or experiencing social exclusion. Service-learning was associated with improvements in school engagement, sense of belonging, social competence, and civic responsibility, which are key protective factors against educational exclusion, across these programs. These findings suggest that well-designed service-learning initiatives may contribute to educational equity by strengthening students’ connections to school and community, particularly in settings characterized by socioeconomic disadvantage or a high prevalence of early school leaving.

Nevertheless, these findings should be interpreted with caution because many of the reviewed interventions present methodological limitations, including small sample sizes, frequent absence of effect size reporting, and predominance of pretest–posttest designs without control groups. As noted in previous reviews, this situation reveals that the available evidence base remains weak and heterogeneous, making it difficult to draw firm and generalizable conclusions.

Future research on service-learning in secondary education should address several methodological priorities. First, more rigorous designs, preferably randomized or well-matched quasi-experimental designs with adequate sample sizes, should be employed to reduce selection bias and strengthen causal inference. Second, to enhance comparability and enable future meta-analyses, researchers should use standardized, psychometrically robust instruments across studies and systematically report effect sizes, confidence intervals, and attrition data. Third, to mitigate common method bias, multi-informant and multi-method assessment strategies (e.g., combining student self-reports with teacher ratings, objective academic records, and community partner feedback) are needed. Fourth, longitudinal designs and follow-up assessments would allow examination of the sustainability of effects to be examined over time. Finally, future studies should intentionally target and report on high-vulnerability and high-diversity contexts, explicitly analyzing how service-learning may contribute to educational equity and the reduction in school dropout.