Learning has shifted from an individual and teacher-directed process to a more interactive, group-oriented approach [1]. Knowledge is therefore constructed collaboratively to enhance knowledge acquisition, clinical reasoning, and teamwork skills [2-4]. The theoretical roots can be traced back to Piagetian and Vygotskian perspectives, both emphasizing the role of social interactions in learning [5,6]. In the Piagetian theory, sociocognitive conflicts drive learning by exposing individuals to alternative problem-solving strategies [5,7-10]. These conflicts encourage cognitive restructuring, deeper understanding, and skill development [11,12]. The Vygotskian perspective highlights the so-called zone of proximal development, where learning occurs through guided social interaction and peer support until learners achieve independence [6,7,10,13].

Contemporary educational research further highlights the importance of leadership, learning culture, and socially supportive educational environments in shaping successful learning processes [14-18]. Within modern “leadership for learning” approaches, educators are increasingly understood not merely as instructors but as mediators who create psychologically safe, collaborative, and student-centered learning environments [14-17]. More recent perspectives, such as socially shared regulation of learning, conceptualize learning as a process in which learners manage motivation, cognition, and learning activities through interaction, feedback, and shared responsibility within groups [19-21]. Similarly, the communities of practice theory emphasizes that learning and professional development occur through active participation in socially and professionally meaningful communities [22,23].

These developments are also reflected in competency-based medical education frameworks such as entrustable professional activities, which underline the acquisition of professional responsibility, self-regulated learning, feedback, and collaborative practice within authentic clinical contexts [24]. Together, these modern perspectives increasingly conceptualize learning not only as knowledge acquisition but as a socially embedded, collaborative, and professionally oriented developmental process.

Collaborative learning is a widely studied but inconsistently defined concept, often used interchangeably with terms such as group work or tutorial learning [25]. It refers to situations where 2 or more individuals learn together by joint problem-solving and pursuing shared goals [26,27]. Two forms can be distinguished: initiative collaborative learning, which is proactive, and subordinate collaborative learning, which is reactive and occurs when students seek help after encountering difficulties [28]. Collaborative learning offers numerous advantages across multiple layers and levels [29]. Cognitively, it promotes argumentation, questioning, and explanation, deepening understanding for both the explainer and the recipient [30-32]. Culturally, it fosters a learning environment that values professional exchange [33]. It encourages the identification of knowledge gaps, explanation, questioning, and clarification, leading to deeper learning [34-37]. Professionally, it supports communication and teamwork skills essential in fields such as medicine, where complex knowledge must be conveyed clearly and collaboratively [38,39].

In addition, collaborative learning influences cognitive, emotional, and motivational regulation, as conceptualized in socially shared regulation of learning [19-21]. Learners engage in self-shared, coshared, and socially shared regulation to refine strategies and sustain engagement [19,21,40]. These shared responsibilities and mutual support create a Sense of Community (SoC), which plays a key role in sustaining motivation and fostering a positive learning environment [1,41].

Despite extensive research, there is no universally accepted definition of the term “Sense of Community” [42-45]. Some definitions emphasize belonging and emotional connection, while others focus on interdependence and shared identity [46,47]. The study by McMillan and Chavis [44] describes SoC as a feeling of belonging, mutual significance, and confidence that members’ needs will be met through group commitment [45]. Core elements include trust, responsibility, shared goals, and mutual interdependence [42,48-51]. The expression of SoC, however, varies by context, particularly in educational settings [48,52,53].

In classrooms, community takes a distinct form shaped by learning goals, structured interactions, and time-limited membership [54]. It consists of 2 main components: connectedness, reflecting belonging and trust, and learning, representing shared academic goals [42,55]. Learners who experience a high sense of classroom community are better able to coordinate working together to achieve the group’s collective goals [56]. A strong classroom community promotes engagement, collaboration, and resilience [57,58]. Weak community ties instead increase the risk of disengagement and dropout [59].

Belonging appears to be a fundamental psychological need that directly influences motivation, engagement, and persistence in educational contexts [60,61]. This perspective is further extended by the concept of communities of practice, which frames learning as participation in a shared professional culture and the gradual development of identity through social interaction [22,23]. Educators therefore play a crucial role in fostering connectedness and collective learning to enhance student satisfaction and success [62,63].

Both SoC and collaborative learning contribute to learning and may lead to academic success. These concepts share commonalities, may be linked, or build a communal foundation for social learning. Till date, there is no research connecting these 2 concepts or investigating both constructs in the same study population. This study aims to unveil the interrelations between SoC and collaborative learning based on 2 validated questionnaires [64-66]. The research questions (RQs) can be formulated as follows:

First-semester medical students from the winter semester 2021-2022 and summer semester 2022 cohorts were invited to participate in a paper-based survey exploring their learning strategies and perceived SoC. By this time, participants had completed all courses for the first semester of their medical studies. The initial invitation to participate was extended personally by the first author. Students received a €5 (€1=US $1.15 as of June 8, 2026) expense allowance for their participation.

Prior to the study, written informed consent was obtained from all participants, who were thoroughly informed about the study’s opportunities, risks, rights, obligations, and voluntary nature. Data collection was conducted in a pseudonymized format, with all participants consenting to anonymized data publication. Students retained the right to withdraw consent at any time without facing any disadvantages. Ethics approval for the study was granted by the ethics committee at Eberhard Karls University of Tuebingen (reference number 086/2022BO2). The study was carried out in accordance with the Declaration of Helsinki.

Following a descriptive analysis, Pearson correlation coefficients were calculated, and linear regression analyses were conducted. Subsequently, a principal component analysis (PCA) with varimax rotation and Kaiser normalization was conducted to verify the underlying structure and confirm the division into subordinate collaborative learning and initiative collaborative learning of the learning with fellow students scale. The PCA used listwise deletion, and items with factor loadings ≤ |0.30| on one or both dimensions were retained [68,69]. Prior to the PCA, the Kaiser-Meyer-Olkin measure of sampling adequacy and Bartlett test of sphericity were assessed. The PCA results served as the foundation for subsequent stepwise regression analyses. The significance level for evaluating results was set at .05.

SPSS Statistics (version 29; IBM Corp) was used for collecting questionnaire responses, data extraction, and statistical analysis.

A total of 344 first-semester medical students were surveyed. With a total number of 210 students per semester, a maximum of 420 participants could have been included, resulting in a response rate of 81.9% (344/420). Participants with missing data were excluded, resulting in a final sample of 331 respondents. The study sample reflected the gender distribution of the cohort, with 68.9% (n=228) of participants being female, and the ages of the participants ranged from 18 to 35 years, with a mean age of 20.72 (SD 2.72) years. The majority of participants were medical students (320/331, 96.7%), followed by dentistry students (10/331, 3%) and students from molecular medicine (1/331, 0.3%).

Descriptive statistics for the individual items, as well as the total CCS-D score and its subscale scores, are presented in Table S1 in Multimedia Appendix 1. The total CCS-D demonstrated good internal consistency (Cronbach α=0.87) as a measure for reliability. The connectedness subscale (Cronbach α=0.84) also showed good internal consistency, while the learning subscale (Cronbach α=0.77) showed acceptable internal consistency. Similarly, Table S2 in Multimedia Appendix 1) presents descriptive statistics of the LIST questionnaire along with the total score on the learning with fellow students scale and its subscales. The total learning with fellow students scale showed good internal consistency (Cronbach α=0.86). The subscales for initiative collaborative learning (Cronbach α=0.82) and subordinate collaborative learning (Cronbach α=0.84) demonstrated high reliability.

To measure the relationship between SoC and collaborative learning, Pearson correlation was conducted. The correlation coefficient for the relationship between the SoC and the learning with fellow students scale was found to be r=0.466; P<.001. Higher levels of SoC were associated with greater use of collaborative learning strategies (Figure 1).

To further examine this relationship, the next step was to examine the association between SoC and collaborative learning through linear regression analysis. The regression model explained 21.7% of the variance in the use of collaborative learning strategies (R²=0.217, adjusted R²=0.214; F_1,329=91.020; P<.001). The unstandardized coefficient was B=0.038 and the standardized coefficient was ß=0.466, both statistically significant at P<.001. SoC was significantly associated with the use of collaborative learning strategies.

To determine whether connectedness or learning is more strongly associated with collaborative learning, a stepwise regression analysis was conducted using CCS-D subscales connectedness and learning as predictors, with the learning with fellow students scale as the dependent variable (Table S3 in Multimedia Appendix 1).

The model including only connectedness yielded significant results: R²=0.257, adjusted R²=0.255; F_1,329=113.990; P<.001. The unstandardized coefficient B=0.070 and the standardized coefficient ß=0.507 were both significant at P<.001. The learning subscale was excluded from the model (β=0.023; 2-tailed t₃₂₉=0.382; r=0.021, T=0.651; P=.70).

To determine which items are most strongly associated with collaborative learning, a stepwise regression analysis was conducted using the CCS-D items as independent variables, with the scale learning with fellow students as the dependent variable.

The final model, which included item 15, item 3, item 14i, item 9i, and item 19, yielded significant results: R²=0.338, adjusted R²=0.328; F_5,325=33.145; P<.001. The unstandardized coefficients were B=0.211 (P<.001) for item 15, B=0.130 (P=.007) for item 3, B=0.131 (P=.002) for item 14i, B=0.128 (P=.009) for item 9i, and B=0.138 (P=.01) for item 19. The standardized coefficients ß ranged from ß=0.139 to ß=0.244, indicating that item 15 (“I feel that members of this course depend on me”) showed the strongest association with collaborative learning (ß=0.244), followed by item 14i (ß=0.156) and item 19 (ß=0.139; Table S6 in Multimedia Appendix 1).

This study aimed to explore the relationship between SoC and collaborative learning among first-semester medical students. The results of this study show that perceived SoC is significantly associated with the use of collaborative learning strategies. The underlying structure of subordinate collaborative learning and initiative collaborative learning was confirmed. Connectedness showed a stronger association with collaborative learning than the learning dimension. Connectedness was more strongly associated with subordinate collaborative learning than with initiative collaborative learning. Item 15 of the CCS-D (“I feel that the members of this course depend on me”) showed the strongest association with collaborative learning.

The findings resonate with theoretical perspectives that link social and emotional bonds to cooperative learning [5,6]. A stronger SoC is associated with a greater engagement in cooperative learning, supporting prior research highlighting the importance of social belonging in promoting student engagement [42,45].

This is particularly relevant for first-semester medical students who must navigate both academic and social transitions [72]. Up to this point of their studies, many students have little experience with self-organized group learning and must first familiarize themselves with new learning demands and environments [73]. School-based learning strategies must be adapted to the new university requirements at the beginning of studies [74]. In this transitional phase, social connections can help by reducing uncertainties and lowering the threshold for interaction [75]. Numerous studies show that both learning environment and interaction with fellow students play a central role in the selection and adaptation of learning strategies [76-79]. Collaborative learning helps reduce exam anxiety and related stress, which are known to hinder academic success [80,81]. Students who feel well-prepared are more likely to remain motivated and persist through academic challenges [82,83].

In this study, only connectedness was significantly associated with collaborative learning, while learning was not retained from the analytical models. This suggests that interpersonal connectedness is more strongly associated with collaborative learning behavior than sharing common academic goals. One possible explanation for this lies in the specific structure of medical studies, which is more characterized by external learning objectives (eg, exams and clinical competencies). These standardized and homogeneous goals may be less relevant for collaborative behavior. Studies further show that SoC serves as a mediator between motivation and student contribution during collaborative learning, with the mediation effect being stronger for intrinsic motivation than for extrinsic motivation [84]. Therefore, interpersonal connectedness not only appears to facilitate social integration but also strengthens intrinsic motivation, which in turn significantly contributes to active participation in group learning [84]. The first term of medical school is a phase in which students need to orient themselves both academically and socially as written above. In this orientation phase, feelings of belonging and trust are often more relevant than clearly defined common learning goals. It is therefore plausible that interactions in early study phases primarily serve social exchange and network building, while learning goals gain secondary importance.

Another interpretation is that social distancing during the COVID-19 pandemic may have increased the importance of interpersonal connection over content motivation in developing collaborative learning. During the pandemic, opportunities to build personal relationships and engage in face-to-face interactions were severely restricted, as universities and schools shifted to digital learning environments [85-89]. The participants in this study had to complete high school during that time under remote learning conditions [90]. Many students reported missing in-person interaction and expressed a need to meet and learn together to stay motivated [91,92]. These prior experiences may have intensified the value of social connectedness as a foundational support in their early university learning. In this context, feeling connected to peers may serve not only as a motivator for collaborative learning but also as a compensatory mechanism for previously unmet social and educational needs.

Interestingly, connectedness is more strongly associated with subordinate collaborative learning than initiative collaborative learning. This difference can be understood by considering their distinct natures. Initiative collaborative learning is proactive when students actively engage with peers to work through tasks, discuss lecture content, and question one another to strengthen understanding [28]. In contrast, subordinate collaborative learning is reactive, arising when students encounter challenges such as unclear concepts or incomplete notes and turn to peers for support to fill those gaps and improve comprehension [28]. This differentiation and our findings suggest that social factors are particularly relevant when students rely on support, whereas initiative is more likely driven by individual traits such as motivation or self-efficacy [93-95].

In highly selective programs such as medicine in Germany, where all students have been admitted based on outstanding academic achievement, asking for help is not only a question of competence but also social safety [96]. When all peers may be considered intellectually capable, students may hesitate to seek help from someone if they fear being judged or feeling help seeking as evidence of incompetence [96-99]. In this context, interpersonal trust becomes a crucial factor for starting collaborative behaviors [100]. Particularly in subordinate collaborative learning, students are maybe more likely to turn to peers who create a nonthreatening environment and signal mutual respect. This supports the finding that social connectedness, more than academic motivations and shared learning goals, is associated with collaborative learning strategies. Trust creates a psychologically safe space in which students can take academic risks without fear of judgment [101]. Feeling embedded in a supportive peer community may lower the psychological threshold for admitting uncertainty and promotes help seeking without fear of negative evaluation [100,101].

Particularly noteworthy is the role of item 15 of the CCS-D (“I feel that the members of this course depend on me”), which showed the strongest association with collaborative learning. This highlights the importance of perceived responsibility within a learning community and aligns with the concept of a “Sense of Community Responsibility” in collaborative environments [50,51].

Accordingly, psychological experiences of responsibility and commitment among members of a community are significantly linked to their attitudes and behaviors within collaborative processes [102]. It has also been demonstrated that collaborative learning fosters a sense of accountability, as students feel responsible not only for their own performance but also for that of their peers [103]. Research has shown that individuals tend to emphasize group cohesion when the value of the group is threatened, with such reactions being most pronounced by those members who already feel a strong sense of commitment toward other group members [102,104]. Feeling that others depend on one’s contribution may strengthen identification with the group, which in turn increases the motivation to participate actively and contribute meaningfully. This aligns with social identity theory, which posits that individuals derive part of their self-concept from group membership [105]. Within learning contexts, a sense of connectedness and perceived responsibility toward peers can foster intrinsic motivation and engagement, thereby promoting collaborative behaviors and support [106].

It is also worth discussing whether the present finding may, in part, reflect a form of “helper syndrome,” in which individuals derive a sense of self-worth from being needed by others and perceive their peers as dependent on their contributions [107]. In the context of health care professionals and medical education, where altruism, care, and accountability are core professional values, this perceived responsibility may be reinforced by internalized norms of helping and supporting others. The statement “they depend on me” likely captures a positive perception of having a meaningful role within the group, which can be intrinsically motivating for collaborative learning as individuals experience social worth when they feel needed [108,109]. While such attitudes can strengthen engagement and cooperative behaviors, they may also be associated with a tendency to prioritize others’ needs over one’s own well-being, or to underestimate personal limits, which are characteristic aspects of helper syndrome [110].

This study has several strengths. Data were collected across multiple cohorts, providing a broader and more generalizable view of first-semester medical students. The high response rate reduces selection bias, and the use of validated instruments strengthens construct validity for both SoC and collaborative learning measures. Robust statistical analyses, including principal component and regression analyses, further enhance methodological rigor.

However, some limitations should be noted. The LIST questionnaire may not fully capture the range of collaborative learning strategies actually used, and self-reported data are susceptible to recall and response bias. Furthermore, all students participated in fixed-group laboratory courses and mentoring programs that fostered collaboration and community, which may have elevated CCS-D scores. Still, as all participants experienced the same conditions, a consistent baseline across the sample can be assumed. A moderate proportion of variance in collaborative learning behavior was explained by the models, while a substantial proportion remained unexplained. This may be attributed to the use of stepwise regression analyses, which are prone to chance findings, thereby potentially compromising the stability and generalizability of the identified predictors. Similarly, the cross-sectional design does not allow for causal inferences, meaning that the observed relationships can only be interpreted as associations. The results should therefore be interpreted with appropriate caution.

A strong SoC appears to be closely associated with collaborative learning in medical education. Social connectedness showed a stronger association with collaborative learning than shared academic goals. Both initiative collaborative learning and subordinate collaborative learning benefited from stronger community ties, with the effect being most pronounced for reactive peer support. The feeling that course members depend on one’s own contributions showed the strongest association for collaborative learning.

These findings suggest that deliberately cultivating SoC through structured group work, peer networks, and mentoring may support not only learning outcomes but also teamwork skills and collaboration. Beyond the academic context, fostering a strong sense of belonging and mutual support may better prepare future health care professionals for the collaborative demands of clinical practice, where trust, communication, and cooperation are essential.