This non-randomized controlled trial employed a pre-test/post-test design. This study was conducted adhered by the Undergraduate Nursing Program Professional Accreditation Standards (33) and was reported following relevant elements of the Strengthening the Reporting of Observational studies in Epidemiology (STROBE) and Consolidated Standards of Reporting Trials (CONSORT) guidelines, adapted for quasi-experimental designs, to ensure transparency and methodological rigor (34–36). The Template for Intervention Description and Replication (TIDieR) checklist was utilized to detail the SCOPE program (e.g., session details, instructor qualifications) and control group activities (37, 38). The Guideline for Reporting Evidence-based practice Educational interventions and Teaching (GREET) checklist ensured comprehensive reporting of educational intervention components, including participant recruitment and outcome assessment procedures (39). Ethical approval was obtained from the university where the intervention was implemented (Approval No: 2023-0067). All participants provided written informed consent, and data were anonymized to protect confidentiality.

A consecutive sample of 221 third-year undergraduate nursing students enrolled in community nursing courses at a university in China was recruited between February and September 2023. Participants were divided into two groups based on their course enrolment: those taking Community Nursing A were assigned to the intervention group receiving the Structured Community-Based Older People Education Program (SCOPE), while those taking Community Nursing B formed the control group receiving the Standard Community-Based Older People Education Program. Eligibility screening was conducted according to predefined inclusion and exclusion criteria. Inclusion criteria were Enrolment in Semester 5–6 (year 3) of the Bachelor of Science in Nursing (BSN) program; completion of prerequisite courses: Fundamentals of Nursing I-II (Semesters 2–3); Health Assessment Practice (Semester 3); Medical/Surgical Nursing I (Semester 4); Active registration in Community Nursing A (intervention group) or B (control group); Provision of written informed consent. Exclusion criteria were: Documented mental health conditions or communication impairments affecting community practice; Academic suspension or withdrawal during the study period; and refusal to provide participation consent.

Data were collected using a self-administered questionnaire comprising four sections: demographic characteristics (14 items), the 22-item Chinese College Student’s Empathy Questionnaire (CCSEQ; cognitive, emotional, and behavioral empathy dimensions; Cronbach’s α = 0.881) (44), the 33-item Chinese Emotional Intelligence Scale (EIS; four dimensions including emotional perception and regulation; Cronbach’s α = 0.897) (45), and the 37-item Chinese Caring Ability Inventory (CAI; three dimensions: acknowledge, patience, and courage; Cronbach’s α = 0.815) (46). All scales utilized Likert-type responses (5- or 7-point) and were grounded in Watson’s Theory of Human Caring (47). A pilot study (n = 30) confirmed robust psychometric properties: test-retest reliability over 1 month yielded intraclass correlation coefficients (ICC) of 0.82 (CCSEQ), 0.85 (EIS), and 0.79 (CAI). Exploratory factor analysis (EFA) demonstrated strong construct validity, with Kaiser-Meyer-Olkin (KMO) measures ≥ 0.846 and significant Bartlett’s tests (p < 0.001) for all scales. Varimax-rotated factor loadings ranged from 0.385 to 0.829, explaining 67.8–72.4% of total variance across scales. Internal consistency for subscales exceeded 0.70, except for self-emotion regulation (α = 0.595). Normality assessments revealed skewed distributions, necessitating non-parametric analyses. Pre-post pilot comparisons using Wilcoxon tests indicated significant improvements in cognitive empathy (p < 0.05) and caring behavior dimensions (p < 0.05), validating the instruments’ sensitivity to detect intervention effects.

Data were analyzed using SPSS 27.0 (IBM Corp.) with a two-tailed significance threshold of α = 0.05. Continuous variables approximating normality (Shapiro-Wilk test, p > 0.05) were expressed as mean ± SD, while categorical variables were summarized as frequencies and percentages. Group comparisons utilized independent t-tests for normally distributed scale scores. Chi-square/Fisher’s exact tests for demographic comparisons (gender, single-child status, residence type, etc.). Longitudinal changes in empathy, emotional intelligence, and caring behavior across baseline (T0), post-intervention (T1), and follow-up (T2) were assessed via Generalized Estimating Equations (GEE). The model incorporated baseline scores as covariates to control for initial imbalances and included three predictors: Group (intervention/control), Time (T0/T1/T2), and Group*Time interaction. The Least Squares (LS) estimates reflect the GEE-derived adjusted group differences, grounded in the Generalized Least Squares (GLS) theory extended by Vanegas et al. (48). By modeling the mean-variance relationship of the response variable, these estimates effectively isolate true intervention effects from correlated data variability. A stratified gate-keeping approach was implemented. Interaction effects were prioritized, with nonsignificant terms removed to test main effects. Robust standard errors accounted for within-subject correlations in repeated measures. Identity link function and exchangeable correlation structure were specified, given approximately normal residuals. Effect sizes were reported with 95% confidence intervals. All analyses adhered to intention-to-treat principles, with missing data handled via multiple imputations (five iterations, predictive mean matching). Sensitivity analyses confirmed model stability across complete-case and per-protocol cohorts.

The intervention (N = 96) and control (N = 94) groups demonstrated comparable baseline profiles (Table 1). Participants were predominantly female (67.9%), not the single child (59.5%), and born in the same Province (87.9%). Over 93% had parents with high school or lower education, 50.5% voluntarily chose nursing majors, and 64.2% reported prior community service experience. No significant between-group differences were observed in gender distribution, voluntary major selection, cohabitation with older/disabled individuals, or regional origin (all p ≥ 0.05), confirming baseline equivalence for outcome attribution.

The SCOPE program led to significant and sustained improvements in empathy, emotional intelligence, and caring behavior among nursing students compared to standard training. According to Table 2, the total empathy score (CCSEQ) in the intervention group increased markedly from T0 (81.25 ± 0.85) to T1 (86.92 ± 0.84), and further to T2 (94.90 ± 1.24), while the control group showed a decline at T1 (74.24 ± 1.06) followed by partial recovery at T2 (87.39 ± 1.48). Between-group comparisons indicated significantly higher empathy scores in the intervention group at both T1 and T2 (p < 0.001). All three empathy subdimensions demonstrated similar trends: cognitive empathy increased from 22.91 to 26.23, emotional empathy from 30.15 to 34.55, and behavioral empathy from 29.70 to 34.11 in the intervention group. These improvements were statistically significant within groups over time (p < 0.001), and between-group comparisons at T1 and T2 also reached significance in most dimensions.

Emotional intelligence, as shown in Table 3, also significantly improved in the intervention group. The total EIS score rose from 122.39 ± 1.10 at T0 to 142.63 ± 1.31 at T2, while the control group’s increase was smaller, from 122.80 ± 1.16 to 129.29 ± 1.62. Significant between-group differences were found at both T1 and T2 (p < 0.001). Among EIS subdimensions, the Emotional Perception score improved notably from 41.61 to 51.66, with the intervention group significantly outperforming controls at T2 (p < 0.001). Self-emotion Regulation increased from 30.13 to 33.96 (p < 0.001), and the Understanding Others’ Emotions dimension rose from 23.03 to 26.21, while the control group showed no significant change in this dimension (p = 0.128). Additionally, the Use of Emotion increased from 27.61 to 30.80, a statistically significant change compared with the control group (p < 0.001).

Similarly, the CAI total score (Table 4) in the intervention group increased from 177.72 ± 1.60 at baseline to 204.77 ± 2.12 at T2, compared to the control group, which rose more modestly from 173.79 ± 1.42 to 191.22 ± 2.45. Significant between-group differences emerged at T1 (p = 0.024) and were even more pronounced at T2 (p < 0.001). Within the caring behavior subdimensions, the Patience score showed the most substantial increase (from 48.20 to 67.66, p < 0.001), followed by Acknowledge (from 72.39 to 77.33, p < 0.001). While Courage also increased (from 57.14 to 59.78), its early changes were less significant, particularly at T1 (p = 0.030), but became more apparent by T2 (p < 0.001). These multidimensional improvements across empathy, emotional intelligence, and caring behaviors provide strong evidence of the intervention’s effectiveness in enhancing nursing students’ professional competencies.

To examine the specific effect of the SCOPE program over time, GEE analyses were conducted for the intervention group (Tables 5–7). In empathy (Table 5), significant increases were observed from T1 to T3 (β = 9.55, 95% CI: 7.59–11.51, p < 0.001), with an even larger jump from T1 to T3 in the total score (Group T3: β = 12.15, p < 0.001). Emotional and behavioral empathy dimensions followed a similar trend, with T3T1 changes exceeding 3.4 points in both dimensions (p < 0.001). Emotional intelligence (Table 6) also showed a strong and consistent increase over time. The EIS total score rose by 13.44 points from T0 to T2 (β = 13.44, 95% CI: 10.95–15.92), with a group-time interaction effect (Group*t2) reaching β = 20.24 (p < 0.001). Subdimensions such as self-emotion regulation and understanding others’ emotions also showed robust growth, with β coefficients of 3.83 and 3.18, respectively (p < 0.001). Similarly, for caring behavior (Table 7), the intervention group exhibited a marked increase in CAI total score from T0 to T2 (β = 22.29, p < 0.001), and group-by-time interaction at T2 yielded a large effect size (β = 27.05, 95% CI: 22.42–31.69, p < 0.001). Across dimensions (acknowledge, patience, courage), most showed statistically significant and sustained improvement, confirming the long-term efficacy of the program.

Figure 3 presents a consolidated visualization comparing the effectiveness of the SCOPE program against the Standard program across the total scores among three scales. This grouped analysis reveals distinct patterns of divergence between the two pedagogical approaches. The SCOPE program demonstrates consistently superior outcomes in all three total scores, indicating holistic enhancement in geriatric nursing competencies. The total CCSEQ score and EIS score both showed steep upward trajectories for the intervention group, while the control group exhibited either minimal improvement or fluctuations. Similarly, the CAI total score steadily increased for the intervention group and only modestly improved in the control group. These graphical trends visually support the statistical findings, highlighting both the immediate and sustained impact of the SCOPE intervention on nursing students’ empathy, emotional intelligence, and caring behaviors.

The differing outcomes between the SCOPE program and the Standard Community-based Education program underscored contrasts in pedagogical philosophy and curriculum design. The SCOPE program produced immediate and sustained improvements in empathy and caring behavior. However, the temporary gains in emotional intelligence (EI) raise questions about the long-term impact of competency-based training. In contrast, the control group showed modest, less durable improvements, revealing the limitations of skill-centric nursing education. Nonetheless, the control group’s exposure to clinical environments still contributed to incremental growth, suggesting that experiential learning, while limited, remains valuable. The positive impact of the SCOPE program on empathy, emotional intelligence, and caring behavior is consistent with the broader trend in nursing education that experiential learning is more effective than traditional didactic methods. Prior research has consistently advocated for the integration of community-based experiences into nursing curricula (49). This study provides powerful support for these calls, demonstrating that a structured program combining community engagement, reflection, and skills application leads to significant and sustained improvements. This reinforcement of existing knowledge strengthens the argument for policy changes and curriculum reforms that prioritize experiential learning over passive instruction (50). Furthermore, the finding that caring behavior showed the most substantial gains at the 3-month follow-up (T2) aligns with theories suggesting that professional attributes require extended reinforcement and practice for full integration into a nurse’s repertoire (51).

The SCOPE curriculum aligns with constructivist learning theory, where repeated, scenario-based simulations help students internalize caring behaviors (52). In contrast, the standard program reflects a behaviorist approach, emphasizing procedural mastery rather than psychosocial development. The drop in EI scores at T2 in the SCOPE group suggests that emotional regulation may require more sustained or repetitive exposure. According to Salavera et al. (53), EI exhibits trait like stability, and short-term interventions may be insufficient to shift deeply ingrained emotional traits, especially under post-graduation stress.

Critics argue that simulated settings create artificial separations between training and real world care (54, 55). Although simulations enhance clinical decision making and safety (42), they may oversimplify the emotional and cognitive complexities of geriatric care. Standardized patients cannot replicate cognitive decline or emotional resistance, which may potentially inflate empathy assessments in controlled settings. These concerns mirror critiques by Lin et al. (56) regarding the limitations of decontextualized skills training.

A significant contribution of this research is that it addresses a critical gap: the lack of empirical data on scalable, theoretically informed, and community-based interventions. While many studies on empathy exist, they often fail to provide a replicable model for comprehensive implementation (57). The SCOPE program, with its clear structure, objectives, and sustainable effects, offers a concrete solution. This moves the conversation beyond theoretical appeals for empathy development and provides evidence for a practical and policy related educational strategy.

The intervention group’s long-term empathy gains support the neuroplasticity hypothesis, suggesting that repeated interaction with vulnerable populations may strengthen affective pathways (58). In contrast, the control group’s minimal progress in empathy highlights ethical concerns about relying solely on incidental exposure in clinical rotations. Gholamzadeh et al. (52) found that empathy training incorporating reflection and role-playing is more effective than unguided clinical experiences. This supports critiques of the hidden curriculum in nursing, where institutional priorities may inadvertently discourage compassion (59).

The control group showed no significant improvement in the “courage” dimension (CAI scale) or in “understanding others’ emotions” (EIS), exposing a critical gap in traditional training. Standard programs often lack content addressing moral distress or ethical decision-making. In contrast, SCOPE’s scenario-based simulations presented students with dilemmas involving autonomy and family conflict, potentially enhancing their moral reasoning. However, this benefit could be short-lived without ongoing mentorship, risking the perception of courage as a checklist item instead of an internalized value.

Further analysis of the control group confirms these shortcomings. The lack of significant gains between T1 and T2 in the courage dimension of caring behavior (in CAI) (p = 0.128; p = 0.086; p = 0.750) aligns with prior findings by Gholamzadeh et al. (52). Traditional training failed to induce meaningful emotional or behavioral change, although it may still help develop technical skills, particularly in low-resource settings lacking simulation infrastructure. On the other hand, Bauchat et al. (54) argue that caring behavior relies heavily on practical communication skills and that simulation training provides the opportunity to improve those skills. Bearman et al. (55) confirm that students make fewer mistakes and improve safety in actual patient care by identifying and correcting errors in a simulation environment. It was also confirmed that simulation training allows students to receive immediate feedback and self-assessment after performing caring behavior, which is critical for continuous improvement.

SCOPE participants showed initial improvements in EI, but these gains regressed at follow-up. This aligns with Salavera et al. (53), who argued that emotional intelligence, while somewhat malleable, requires ongoing reinforcement. Zeidner et al. (60) further contend that EI development stagnates without structural support. In hierarchical hospital settings, expressing emotional vulnerability may be discouraged, reducing the practical value of EI training. This underscores the disconnect between individual level interventions and institutional norms.

The control group’s minimal improvement in EI challenges the belief that clinical exposure naturally builds emotional competence. Lönn et al. (61) warns that unsupervised patient interaction may encourage emotional detachment rather than empathy, particularly in high-stress fields like geriatrics. SCOPE’s guided debriefings may buffer students against this emotional numbing by creating space for reflection.

Caring behavior was the most distinct domain between the two programs. The SCOPE group benefited from structured simulations that encouraged practice and feedback, helping students internalize caring acts. This aligns with Bauchat et al. (54) and Bearman et al. (55), who found that simulation fosters clinical empathy and reduces errors. In contrast, the standard program assumed caring would emerge organically through practical notion that lacks empirical support. Cho and Kim (42) emphasize that caring must be developed through deliberate, repeated exposure.

The elderly people’s feelings of stiff joints, blurred vision, hearing loss and back pain were given to the nursing students to experience the physiological decline of the older people, which had already been tested in previous studies, that help with their cognitive development (62, 63). The nursing students reflected that they felt helpless and hurt physically when they got a stiff back and how hard it was to climb stairs when they stepped into the older people’s shoes, the same feeling described by Weekes and Phillips (64). Nevertheless, simulations have limitations. They may fail to capture the emotional unpredictability of real-life care settings (65). Over-reliance on simulations could lead to unrealistic expectations of clinical interactions, where outcomes are not always predictable or structured.

The findings suggest that an integrated curriculum could maximize the strengths of both approaches. SCOPE’s structured empathy and caring exercises could be embedded within community rotations to blend simulation with real-life complexity. However, this integration requires trained faculty and resources, which may not be feasible in all settings. Nursing students watch the caring behavior of the teacher in action and perform simulated experiments. As the study revealed, the emotions evoked by personal care activities could significantly improve their attitude toward older people and trigger their empathetic emotions (66). At the same time, their confidence was also supported when they focused on the patients’ needs (67). Referring to the scenarios based on the older peoples’ life in community, these critical life situations impact one’s emotional control and can trigger stress (68). Nevertheless, this is one of the styles of the cognitive transactional model of stress-appraisal-coping, which has been indicated in nursing students’ clinical placement process to make sense of life (69). It has been proved that the role of simulation in nursing education is well-established for facilitating skill acquisition, critical thinking, and confidence in a controlled environment, which is also aligns with the research of Alanazi et al. (70).

Importantly, neither this study nor that of Gholamzadeh et al. (52) included long-term follow-up beyond T2. Without this, it remains unclear whether brief interventions can foster enduring professional identity or only temporary behavioral change. As nursing education shifts toward competency-based models, future research must explore how to sustain affective learning over time.

This study has several strengths. The SCOPE program adopted a well-rounded approach that combined classroom learning, simulation, and hands-on community experience, offering nursing students both theoretical knowledge and practical skills. It focused on developing empathy, emotional intelligence, and caring behavior-key components of high-quality geriatric care. Data was collected at multiple points, helping to observe changes over time. Involving both campus and community settings also increased the practical relevance of the findings.

However, there are some limitations. The 4-week intervention period was relatively short, which may not have been enough to create lasting changes. Limited opportunities for real-life application in community settings may have reduced the depth of learning. Since the study was conducted within a specific cultural and educational setting, the findings may not be easily applied elsewhere.

Future research should explore ways to integrate community-based programs like SCOPE more deeply into nursing curricula, either as dedicated modules or through existing courses, while addressing barriers such as limited faculty resources and time constraints. Tailoring educational strategies based on students’ backgrounds and incorporating tools like reflective practices or digital feedback platforms could further enhance empathy and emotional intelligence development. Methodologically, future studies would benefit from randomisation and qualitative approaches to provide more robust and well-rounded insights.