Research Design

This study employed a quasi-experimental design, selected due to the practical limitations of implementing randomized controlled trials in real-world school-based settings (Cook et al., 2002). The intervention was delivered within pre-existing classroom structures, where random assignment was not feasible due to administrative constraints and instructional scheduling (Mertens, 2019). Assignment occurred at the classroom level, with four intact sixth-grade classrooms assigned to conditions. All classrooms received their regularly scheduled physical education, with three classrooms additionally assigned to one of the tailored, multicomponent CSPAP intervention components (Nutrition, PADS, or FCE) and one classroom assigned to the control group. This approach aligned with the school’s instructional organization and allowed the intervention to be implemented without disrupting existing teaching schedules. However, the unit of analysis remained at the individual student level, with life satisfaction assessed for each participant.

Participants

Participants included 96 sixth-grade students from an elementary school located in Mokpo-si, South Korea. Four intact sixth-grade classrooms, each comprising 24 students, were assigned to study conditions based on pre-existing classroom structures. Three classrooms were assigned to the CSPAP intervention groups, while one classroom served as the control group. This assignment procedure was determined by administrative and scheduling constraints inherent to the school's organizational structure. Standard protocols for obtaining student assent and parental consent were rigorously followed. Institutional Review Board approval was obtained from Gwangju National University of Education, along with administrative approval from the participating school. The selection of the current study site is notable given the area's high concentration of socioeconomically underserved populations. According to the Regional Report, 2019, Mokpo-si accounted for the highest proportion of low-income and single-parent households among all administrative districts in Jeollanam-do. Of the 7,843 households in the region receiving support under the Single-Parent Family Support Act and the National Basic Livelihood Security Act, 24.3%, or 1,906 households, were located in Mokpo-si. The proportion of low-income and single-parent households in Mokpo-si (24.3%) exceeds that of other similarly populated cities, such as Yeosu-si (14.9%) and Suncheon-si (12.0%), suggesting that Mokpo-si is a particularly salient context for examining school-based interventions aimed at promoting students' mental health.

Multi-Component CSPAP Program Development

To develop a multi-component CSPAP program incorporating local context, this study employed a consensus-based approach that integrates diverse perspectives to enhance decision-making quality and provide nuanced interpretations of contextual needs (Michaelsen et al., 1989). A consensus-based approach that emphasizes collaborative interpretation through structured agreement was selected to capture the context-dependent needs of the student population (Miller et al., 1987). Specifically, this method can be utilized in the development of tailored PA interventions for participants and incorporates their feedback to ensure contextual relevance (Devereux et al., 2024; Johannes et al., 2024; Whooten et al., 2020).

In the first phase, to operationalize the consensusbased approach, a dedicated coding team was established. The coding team consisted of four individuals: one educational researcher and three inservice teachers familiar with the local context and student population. Data from a preliminary survey, administered to 96 students, were primarily used for the consensus coding process. This survey was informed by unstructured student interviews and basic student profile data, and it included comprehensive questions on PA participation, nutritional habits, and FCE. Key findings revealed that 91.7% of students reported experiencing hunger during class, 14.6% regularly participated in PADS, and 29.2% engaged in PA-related conversations with their family at home (see Table 1).

In the second phase, each team member independently reviewed the preliminary survey responses. During this phase, coders generated preliminary thematic categories and accompanying analytic memos based on their interpretations of student needs, shaped by both professional judgment and contextual experience. This step allowed for diverse interpretations of student needs to emerge, forming the foundation for later synthesis. This was followed by a series of three structured consensus-building sessions, during which the team reconciled divergent codes and negotiated the boundaries and meanings of emerging subthemes. To ensure analytic consistency, subthemes were retained only when 80% agreement among coders was reached, in line with established qualitative research standards (Miles & Huberman, 1994). Throughout these sessions, emphasis was placed on identifying overlapping codes, resolving semantic discrepancies, and addressing any ambiguities in interpretation. Raw data excerpts were frequently revisited to ensure that the codes remained closely aligned with student perspectives and the intent embedded in the original responses. This iterative process enabled the team to refine a shared understanding of the data while maintaining transparency in how interpretations evolved.

In the third phase, the team arrived at a shared consensus and interpretive structure, grounded in both individual coder perspectives and contextual knowledge. The collaborative nature of the process enhanced the analytic rigor of the study, contributing to the internal consistency of the coding structure and minimizing individual bias. Upon completion of the consensusbuilding phase, the final coding framework was synthesized into three key themes: Nutrition, PADS, and FCE. Notably, while Nutrition is not part of the original five-component CSPAP framework, it emerged as a critical component through the consensus process, reflecting priorities for comprehensive health promotion. This empirically-derived expansion aligns with the whole-of-school approach, which recognizes nutrition and PA as interconnected determinants of child health (World Health Organization [WHO], 2021). A summary of the consensus coding process used to incorporate local context into intervention design is provided in Table 2.

Lastly, based on the three key themes identified through consensus coding above, the research team employed a co-design approach to develop targeted intervention strategies. Co-design is a participatory method that incorporates stakeholder perspectives to ensure relevance, contextual alignment, and program sustainability (Iniesto et al., 2022). The team held two in-person meetings (90 minutes each) to interpret the coding results and establish the overall intervention framework. Additionally, two online consultations (40 minutes each) with the school’s nutrition and counseling teachers supported the refinement of specific lesson content and scheduling. This co-design process localized the intervention to align with the school’s contextual realities and student needs, thereby enhancing both relevance and feasibility (Jago et al., 2023).

Program Overview and Implementation

The intervention maintained a multi-component CSPAP approach by integrating physical education as the foundational component across all groups, supplemented by one context-specific tailored component per group based on the program development process. The four-week intervention preserved the integrity of the multi-component CSPAP by consistently delivering three standardized physical education classes per week across all tailored programs. Each intervention aligned with the Korean National Elementary Physical Education Curriculum, ensuring fidelity to national standards (Ministry of Education, 2015). Building upon this common curriculum, each group received one tailored component (Nutrition, PADS, or FCE), while the control group received no supplemental activities. This design enabled clear isolation of effects attributable to each tailored intervention, maintaining CSPAP’s multi-faceted nature. Implementation rigor and fidelity were verified through daily classroom visits and reviews of teacher logs by the primary researcher, focusing on adherence to session plans, dosage consistency, and participant engagement. Program assignments and characteristics are summarized in Table 3.

Physical Education

Delivered three times per week over four weeks (12 classes total), the physical education component was structured to develop students' life satisfaction. In Week 1, students participated in health-enhancing activities such as yoga, stretching, and low-intensity endurance training, coupled with post-activity reflection to support mood awareness and emotional regulation. Week 2 emphasized mastery-oriented tasks, where students set individualized fitness goals (e.g., time trials, jumping distance) and tracked progress to build perceived competence and self-efficacy. During Week 3, the focus shifted to social development through small-sided team games and cooperative challenges that promoted empathy, peer support, and a sense of belonging. In the final week, students explored emotional expression through creative movement and improvisational dance, reflecting on internal emotional states and sharing interpretations with peers to deepen interpersonal awareness. Through this shared physical education curriculum, students across different tailored CSPAP programs were supported within a consistent environment designed to promote life satisfaction and a multi-component CSPAP framework.

Nutrition Program

The primary objective of the Nutrition Program was to improve students’ irregular eating habits, alleviate hunger during class, and enhance their basic knowledge of nutrition. The intervention consisted of two core components: the provision of healthy snacks and a structured nutrition education curriculum. Healthy snacks such as bananas, yogurt, and whole grain cookies were provided in classrooms every morning (8:30–8:45 AM), five days a week, for four consecutive weeks (20 sessions total), with careful consideration of nutritional balance. Snack distribution followed a standardized protocol: classroom teachers distributed pre-portioned snacks while briefly explaining the nutritional benefits (e.g., “Bananas provide sufficient vitamin C and energy for your morning activities”). Students were encouraged to consume snacks mindfully and to reflect on how proper nutrition affected their energy levels and classroom engagement.

The nutrition education sessions were jointly developed and delivered in collaboration with a school nutritionist and conducted once a week (40 minutes per session) for a total of four sessions integrated into PE class time. Each session followed a consistent instructional structure: (1) 5-minute introduction using visual aids (e.g., food pyramid posters, nutrient flashcards), (2) 20-minute interactive lesson with guided discussions and question-answer activities, (3) 15-minute hands-on application activity (e.g., students categorized food cards, practiced reading mock nutrition labels, or role-played making healthy food choices at a cafeteria), with the pedagogical approach emphasizing student autonomy and practical application to daily life. Instructional content covered topics such as the functions of essential nutrients (Week 1), constructing a balanced diet (Week 2), reading food labels (Week 3), and practicing healthy dietary choices through role-playing activities (Week 4). Developed through co-design with a local nutrition teacher, the age-appropriate and interactive instructional materials were intentionally structured to foster students’ intrinsic motivation and self-efficacy in making healthy dietary decisions. Implementation of fidelity was monitored using class participation logs and weekly face-to-face meetings.

PADS Program

The goal of the PADS Program was to expand opportunities for in-class PA during the school day. This intervention included two primary components. First, students engaged in a structured PA session lasting five minutes before the start of each morning class, five days a week for four weeks (20 sessions total). These sessions were based on a nationally developed routine, the New Millennium Health Exercise program, which was co-developed in 1999 by the Korea Sports Promotion Foundation and the Korean Broadcasting System. Second, PA equipment was provided and utilized to encourage more movement. Equipment such as two stability balls, two yoga mats, and five dumbbells was placed in classrooms for use between academic tasks, while additional items, such as two badminton sets, three soccer balls, and five baseball gloves, were made available during recess to promote autonomous participation in PA. Participation rates in the exercise routine and equipment usage were monitored through weekly meetings with teachers.

FCE Program

The FCE Program aimed to strengthen students' psychological and emotional support systems by enhancing the involvement of families and local communities in health-promoting activities, grounded in the premise that increased family engagement and social connectedness would enhance perceived social support and life satisfaction. Key components included home-school communication, family-inclusive physical education sessions, and community sports event participation.

Weekly health information materials were distributed for four weeks via printed newsletters and digital messages (KakaoTalk, 91% delivery confirmation), covering balanced eating habits, hydration, outdoor safety, and family workouts. Each newsletter included practical tips and discussion prompts to facilitate parent-child health conversations. Two family participation PE sessions (Weeks 2 and 4, Saturday mornings, 2 hours each) achieved participation rates of 53% (Session 1) and 45% (Session 2). Each session was structured as: (1) 10-minute warm-up emphasizing family PA benefits; (2) 50-minute collaborative planning workshop where families designed their own sports activities; (3) 60-minute cooperative games requiring parent-child teamwork (e.g., relay races, obstacle courses, cooperative ball games); and (4) 10-minute reflection circle for sharing experiences. The instructional approach emphasized enjoyment, cooperation over competition, and autonomy-supportive coaching to strengthen family bonds and create positive shared experiences hypothesized to enhance life satisfaction. One community sports event visit (Week 3) to a municipal youth festival achieved 57% family participation (n = 14). Students received a 20-minute orientation on event logistics and safety, with teachers facilitating social interactions during the event to foster community belonging and expanded social networks. Fidelity was ensured by collecting parent acknowledgment forms (100% return rate), family participation worksheets, and participant rosters. The intervention overview is summarized in Table 4.

Measures

Students' life satisfaction was measured using the Korean Version of the Satisfaction With Life Scale (K-SWLS; Cho & Cha, 1998), adapted from Diener et al.'s (1985) original scale. The K-SWLS has demonstrated strong psychometric properties in Korean adolescent populations, with Cronbach's α = .86 in a validation study of 437 adolescents aged 15–18 (Lim, 2012).

Given that the current sample comprised upper elementary students, younger than the typical validation population, two educational experts (i.e., one Ph.D. and one M.Ed.) independently reviewed the items for developmental appropriateness. Both experts confirmed that the original Korean wording was suitable for this age group. However, to reduce cognitive burden and facilitate more accurate responding, the response format was modified from a seven-point to a five-point Likert scale. This adaptation is consistent with best practices in survey research with children and early adolescents, as fewer response options enhance clarity and reliability without altering the underlying construct (Coombes et al., 2021; Mellor & Moore, 2014).

The K-SWLS consists of five items assessing global cognitive judgments of life satisfaction (e.g., “My life is close to my ideal,” “The conditions of my life are excellent,” “I am satisfied with my life”). Items were rated on a five-point scale (1 = strongly disagree to 5 = strongly agree), with higher scores indicating greater satisfaction. Internal consistency was strong at both time points (pretest α = .83; posttest α = .85). The scale was administered twice following a pretest-posttest design. At pretest, surveys were distributed during the week immediately preceding the intervention to establish baseline life satisfaction levels. At posttest, the same instrument was administered during the final week of the intervention period to assess changes in life satisfaction. Both administrations employed a standardized whole-class, paper-and-pencil format to ensure consistency across participants. During each session, the classroom teacher and primary researcher were both present to supervise administration, provide standardized instructions, and address any procedural questions. Each administration required approximately 10 minutes.

Data Analysis

To assess the effects of the CSPAP intervention on student-reported life satisfaction, K-SWLS data were securely managed and screened for accuracy, missing values, and outliers. Descriptive statistics summarized participants' baseline characteristics. Means and standard deviations for K-SWLS scores postintervention are calculated. A t-test and one-way analysis of variance (ANOVA) were conducted to assess differences in life satisfaction among groups, followed by Scheffé post hoc analyses to identify specific inter-group differences. Statistical analyses were performed using SPSS version 27.0, with significance set at p < .05.

Differences in Program Participation

An independent samples t-test was conducted to examine whether participation in a tailored, multicomponent CSPAP integrating context significantly improved students’ life satisfaction. Results indicated that students in the CSPAP intervention group (M = 3.58, SD = 0.83) reported significantly higher life satisfaction than those in the control group (M = 2.69, SD = 0.71), t(94) = 4.71, p < .001 (See Table 6).

Life Satisfaction Differences Across Tailored CSPAP Programs

A one-way ANOVA was conducted to examine whether students' life satisfaction differed by participation in the CSPAP programs. Results indicated a statistically significant difference among the four groups, F(3, 92) = 15.91, p < .001 (see Table 7). Post hoc comparisons using the Scheffé test revealed that students in the PADS program (M = 4.13, SD = 0.55) reported significantly higher life satisfaction scores than those in the Nutrition program (M = 3.22, SD = 0.71), the FCE program (M = 3.39, SD = 0.90), and control group (M = 2.69, SD = 0.14). Additionally, students in the FCE program had significantly higher life satisfaction than those in the control group (see Table 8). These findings indicate that participation in different tailored CSPAP programs had varying effects on students' life satisfaction, with the PADS reporting the highest level of life satisfaction among all groups.

Strengths and Limitations

Although there are numerous examples of school-based PA intervention approaches, many studies continue to rely on highly standardized intervention protocols without actively promoting local adaptation of intervention components (Krishnaswami et al., 2012; Langford et al., 2015). In contrast, this study contributes to the CSPAP literature by examining mental health outcomes, particularly life satisfaction, within underserved contexts. To our knowledge, this is the first tailored, multi-component CSPAP intervention to examine mental health outcomes in an underserved student population. A key strength of this study was the deliberate alignment of intervention components with identified local context, achieved by actively involving teachers in consensus coding and co-design processes. This collaborative approach provides evidence for the feasibility and potential efficacy of tailored CSPAP initiatives, even within resourcelimited settings. Furthermore, the significant improvement observed in life satisfaction highlights CSPAP’s potential as a practical tool for enhancing students’ mental health and broader well-being.

First, classroom-level assignment created substantial clustering effects that were unaddressed in the analysis. Students were nested within classrooms taught by single teachers, yet analysis treated observations as independent, violating t-test and ANOVA assumptions and potentially biasing standard errors. Teacher-specific factors such as instructional style and rapport remain confounded with intervention effects. Additionally, control group dissatisfaction may have suppressed their life satisfaction scores. Future research must employ cluster-randomized designs or multilevel modeling to account for nested data structures.

Second, the four-week intervention duration severely limits conclusions about sustained effects. This brief timeframe raises questions about whether improvements represent meaningful change or transient responses. Without follow-up assessments, durability remains unknown. The compressed timeline may also have favored components with immediate impact over those requiring cumulative learning. Extended interventions with multiple follow-up points are needed to establish sustainability.

Third, the posttest-only design and unmeasured contextual factors limited analytical rigor and generalizability. Analyzing only post-intervention scores without baseline covariates reduced statistical power, particularly given the small sample (n = 96) and potential regression to the mean. Moreover, critical individual-level moderators, including SES indicators, family structure, parental education, and access to recreational facilities, were not collected, preventing examination of for whom the intervention was most effective. The absence of parent and community stakeholder engagement further limited ecological integration. Future studies should employ ANCOVA or mixed-effects models with baseline controls, systematically measure contextual covariates to identify key moderators, and adopt participatory approaches involving families and local resources to enhance sustainability in underserved settings.

Finally, while we treated life satisfaction as a unidimensional construct based on previous validation research with adolescents (Lim, 2012), some studies have found that factor structure may vary across populations. Future research could examine whether elementary school students' life satisfaction comprises distinct subdimensions and whether CSPAP components differentially affect these potential facets.

Implications and Future Directions

The present study demonstrated the value of consensus-based, co-design approaches that actively engage local educators to ensure interventions reflect contextual realities. Rather than applying a standardized protocol, this collaborative framework enabled tailored CSPAP implementation aligned with school-specific needs, resources, and cultural considerations. A key methodological contribution lies in systematically operationalizing context through replicable procedures, moving beyond abstract conceptualization to concrete measurement. The findings confirm that collaboratively developed interventions can effectively promote life satisfaction among underserved youth, offering a promising avenue for addressing concurrent challenges of physical inactivity and mental health in schools.

Future research should continue exploring tailored approaches through portfolio-based frameworks that emphasize flexibility and local adaptation while maintaining fidelity to core intervention principles (Porter et al., 2024). Such methodologies will refine strategies for integrating context-specific considerations into CSPAP and foster more inclusive PA programming. Given the positive associations with life satisfaction of PADS observed in this study, schools may prioritize brief, peer-led activity breaks that integrate seamlessly into existing routines. For example, 5-minute movement sessions led by trained student leaders at the start of morning classes can provide consistent PA opportunities while promoting social interaction, requiring minimal teacher burden and no additional resources. Fidelity monitoring through simple weekly logs or brief check-ins during regular meetings ensures sustainable implementation without increasing staff workload, making this approach particularly viable for resource-constrained settings.