Research on the Flipped Classroom + Learning Community Approach and Its Effectiveness Evaluation—Taking College German Teaching as a Case Study

Abstract

The evolution of theoretical research on flipped classrooms has ushered in the 2.0 era, in which the inaugural iteration, narrowly focused on process inversion, has proven inadequate in fostering enhanced interaction within flipped classrooms. This shortcoming stems from its disregard for nurturing an efficacious collaborative environment, which has emerged as a pivotal roadblock impeding the sustainable development of the flipped classroom concept. To confront this challenge, this paper advocates for the seamless integration of flipped classrooms with learning communities, underscoring the essence of knowledge construction theory and the imperative of creating a collaborative environment. Furthermore, it introduces the groundbreaking “Tai Chi Pillar” flipped classroom + learning community model and meticulously scrutinizes its effectiveness in the context of college German courses through a case study lens, examining both student learning performance and perception development. A comparative analysis with the standalone flipped classroom model reveals that the proposed model boosts academic performance for 72.6% of students and notably enhances perception development for 72% of learners, albeit with limited success among underperforming students. This underscores the model’s strengths while also highlighting the necessity for continued research and improvement. This study serves as a reference point for sustainable Flipped Classroom 2.0 research and points to future directions for investigation.

1. Introduction

Since 2012, the flipped classroom has become widely embraced. The investigation into flipped classrooms has evolved from theoretical explorations and model constructions to a stage characterized by abundant empirical research, encompassing various disciplines. In terms of research trends, mirroring Zanna and Fazio’s [1] classification of research problem progression, studies on flipped classrooms have progressively shifted from the first-generation to second- and even third-generation questions, transitioning from examining “whether flipped classrooms are effective” to inquiries such as “under what conditions are they more effective?” and “how can they be made even more effective?” Bergmann and Sams [2] emphasize that the flipped classroom is constantly progressing in practice, entering a second phase where the primary challenge lies in maximizing the effectiveness of face-to-face interactions with students. At its core, the flipped classroom paradigm involves shifting direct instruction from the shared learning space to individual learning spaces, thereby transforming the shared space into a dynamic, interactive learning environment. The progression of the flipped classroom concept has propelled academic research beyond superficial procedural inversions and format changes, ushering in a new era that prioritizes its essence: the development of a “learner-centered” framework for constructivist learning.

Post-2014, the flipped classroom research has diversified significantly, centering around the integration or reimagining of the flipped classroom model with diverse teaching methodologies. Notably, key terms like collaborative learning, active learning, problem-centered learning, and instructional strategies frequently recur, with collaborative learning standing out for its robust interconnectivity with other keywords, underscoring its pivotal significance in contemporary flipped classroom investigations [3,4]. A study by Cheng et al. [4] analyzing 100 highly cited articles from 2013 to 2018 on the WOS platform revealed that the flipped classroom model is predominantly applied in foreign language courses at universities. Notably, 47% of these studies arranged collaborative interactions as in-class activities, and an additional 19% employed group projects during class time, emphasizing the pivotal role of collaboration and interaction in flipped classroom research. Zuber [5] further argues that while the flipped classroom is characterized by the inversion of traditional teaching methods and the use of instructional videos, its effectiveness may stem not so much from these features but from the in-class activities themselves. This is echoed by Fung et al. [6], who, through a review of the literature on flipped classroom effectiveness, found that all studies reporting positive results included in-class activities such as peer interaction, collaboration, and discussions, suggesting that interactivity is a cornerstone of an effective flipped classroom, which offers learners more opportunities for discourse, cooperation, and engagement than traditional classroom settings (e.g., [3,7,8,9,10,11]). Abeysekera and Dawson [12] posit that the flipped classroom methodology essentially leverages technology to promote social and collaborative learning within physical classrooms. However, existing flipped classroom models often confine collaborative learning to in-class group discussions, and the suboptimal efficiency of these discussions has been identified as a key contributor to the ineffectiveness of in-class activities [3].

Collaborative learning endeavors to facilitate learners’ internalization and profound understanding of knowledge and skills through social negotiation, striving for cognitive synergy (1 + 1 + 1 > 3) and fostering skills such as problem-solving, teamwork, communication, critical thinking, creativity, and self-directed learning. This differs fundamentally from traditional “classroom learning groups”, which serve as instructional tools for teachers to enhance teaching efficiency, with fleeting interactions, loose structures, and teacher-centric dynamics that often reduce “discussions” to mere “exchanges”, which are inadequate for genuine learning. This is because exchanges often recycle the “known” rather than exploring the “unknown”, leading to inefficiency in such collaborative efforts. Amidst the digitization of education and the advent of Flipped Classroom 2.0 [13,14], educational environments are becoming more interactive, collaborative, and interconnected, with learning evolving into a dynamic process involving the establishment of social, knowledge, and neural networks. This shift underscores the transition from individual to distributed cognition and autonomous to collaborative learning, emphasizing the paramount importance of crafting effective collaborative learning environments for the sustained growth of the flipped classroom and its progression into the 2.0 era.

2. Literature Review

Amidst the proliferation of research in education focused on collaborative groups, studies have meticulously examined their foundational elements, the virtues of cooperative learning, and the intricate factors that shape its effectiveness (e.g., [15,16,17,18,19]). Notably, the notion of “learning communities”, grounded in socioconstructivist theory, has emerged as a prominent pedagogical metaphor, attracting widespread attention [20]. This paradigm underscores the collaborative construction of knowledge, where members contribute their distinct expertise towards shared objectives and visions, thereby nurturing individual learning. Members engage in a continuous cycle of learning from each other, attentive listening, viewpoint integration, and collaborative problem-solving, fostering a rich learning ecosystem. Learning communities are not fleeting endeavors or loosely knit assemblies but rather cohesive and perpetually evolving collectives [21]. Under this definition, a learning community, framed within the constructivist approach to learning, is a learning organization characterized by both collaboration and personalization. It plays a pivotal role in facilitating knowledge construction and social skill development, fostering both collective and individual intellectual expansion. Furthermore, it promotes a culture that encourages learning about learning, which is in harmony with the theoretical underpinnings and practical principles of the flipped classroom. Consequently, it provides a solid foundation and a model for how the flipped classroom can harness collaboration to achieve in-class knowledge internalization.

Scholars have begun to explore the integration of flipped classrooms with collaborative learning. Wu et al. [22], for example, introduced the Community of Inquiry (CoI) model into flipped classrooms, fostering online learning communities and empirically demonstrating the model’s effectiveness in enhancing learners’ English speaking abilities compared to traditional settings. Munir et al. [23] conducted an empirical study on Engineering courses, revealing that combining flipped classrooms with collaborative learning led to improvements in student performance, problem-solving, data analysis, and communication skills. Li and Qu [24] designed flipped classrooms grounded in theories of autonomous and cooperative learning, while Li and Li [25] integrated collaborative learning throughout the learning journey, encouraging group members to share knowledge and experiences to facilitate collective learning. However, these studies lack clarity on how collaborative learning environments are structured and how learners collaborate, treating it as a generalized approach rather than delving into specifics. Overall, research on flipped classrooms from a collaborative learning perspective is still in its infancy [23].

The intersection of flipped classrooms and learning communities remains understudied, with the first investigations emerging in 2015 ([26,27]). During the practical implementation of flipped classrooms, researchers observed that students frequently pursued self-directed learning in isolation during pre-class activities, facing challenges without prompt resolution. This prompted the suggestion of leveraging online learning communities to foster a collaborative extracurricular learning environment. Subsequently, further studies (e.g., [22,28,29]) focused on online learning communities across diverse university disciplines, empirically demonstrating improved student participation, satisfaction, and learning outcomes in flipped classrooms augmented by such communities, as compared to traditional setups. In essence, research exploring the combined “flipped + learning community” approach is scarce, with a predominant emphasis on online learning communities as a remedy for collaborative gaps in flipped classrooms. However, these studies fail to fully integrate the two models, treating online communities more as instructional facilitators within flipped classrooms, neglecting their inherent features of learning communities and autonomy promotion.

This study aims to unravel the mechanisms for seamlessly integrating the flipped classroom and learning community models, with a focus on preventing the latter from degenerating into a mere superficial entity. It examines the unique attributes of a learning community within the flipped classroom context and investigates whether this combined approach yields superior effectiveness compared to the standalone flipped classroom. Through empirical validation in practice, the study endeavors to provide a comprehensive assessment of the educational efficacy of this innovative teaching model. Specifically, this study will delve into the following questions:

• How can a learning community be constructed within a flipped classroom?
• What is the impact of the flipped classroom combined with a learning community model on learners’ academic performance, compared to the flipped classroom model alone?
• How does the flipped classroom combined with a learning community model influence learners’ perceptions, compared to the flipped classroom model alone?

3. The Construction of Flipped Classroom + Learning Community Model

There is a unified understanding within academia concerning the “flipped” or “inverted” nature of the instructional sequence in the flipped classroom approach. In contrast to traditional classrooms, where knowledge transmission predominantly takes place during face-to-face sessions and knowledge internalization occurs thereafter, the flipped classroom reverses this paradigm, with knowledge transmission preceding class sessions and knowledge internalization being the cornerstone of in-class activities. This distinction forms the cornerstone of the flipped classroom’s divergence from traditional models. However, this portrayal is not without its shortcomings. The dichotomy between pre-class and in-class stages mirrors the disconnection between extracurricular and in-class activities inherent in traditional teaching frameworks, artificially dividing knowledge transmission and internalization across temporal and spatial boundaries. Does the pre-class stage inherently constrain the process of knowledge internalization? Can the employment of pre-class instructional videos alone suffice for effective knowledge transmission? Is the in-class stage devoid of any element of knowledge dissemination? Or is the flipped classroom merely a traditional teaching approach that leverages pre-class video previews as a superficial novelty? To resolve this issue, research on flipped classrooms has advanced into the 2.0 iteration [30]. The Flipped Classroom 2.0 framework promotes a dynamic learning culture that emphasizes enhanced engagement, generative thinking, collaborative efforts, and digital fluency. It underscores proactive metacognition, piquing students’ interest in course materials and pertinent issues while amplifying their involvement in collaborative problem-solving endeavors. This progression leads to students translating their insights into instructional materials. The act and progression of content creation not only facilitate knowledge internalization but also inspire novel insights through practical application and introspection. Moreover, this process encourages students to integrate their aspirations or insightful ideas into educational discussions with peers and instructors, gradually assuming distinct expert personas. This fosters ‘interactive positional cognition’, empowering students to perceive themselves as autonomous agents actively contributing to group learning, thereby transforming the teacher-centric ideological framework of courses into a teacher-guided learning community. In this community, students’ comprehension and generation of knowledge content are cultivated through problem-driven, collaborative inquiry and socialization.

The integration of the flipped classroom and the learning community finds its foundation in the constructivist view of learning (e.g., [21,31,32,33,34]), which emphasizes fostering genuine, constructive, and socially interactive learning contexts for learners. This shared theoretical cornerstone serves as a crucial prerequisite, enabling both pedagogical approaches to harmoniously blend and support learners’ knowledge development. Embracing the preceding arguments, the flipped classroom model, embedded within the ambit of learning communities, discards the historical focus on segregated in-class and extracurricular dimensions. It now revolves around the epicenter of knowledge construction, nurturing a learning ecosystem that is both personalized and socially integrated. The flipped classroom construct, alongside the complementary elements of the learning community paradigm, transform into facilitating resources and contextual underpinnings that bolster this cardinal purpose, as eloquently portrayed in Figure 1.

The Tai Chi pillar model features the following characteristics:

• The Seamless Merger of Personalized and Collaborative Learning Environments. The model conceptualizes the autonomous learning milieu in flipped classrooms and the collaborative learning setting within learning communities as analogous to the complementary polarities in Tai Chi, where they maintain their autonomy while interweaving seamlessly. Embedded within the personalized learning milieu, there exists the potential for collaboration through discussions, negotiations, and reciprocal learning frameworks within communities, accelerating learners’ cognitive progression. After independently engaging with learning materials, such as watching micro-lectures or completing tasks, learners can immediately share their insights and receive feedback from their community, significantly enhancing knowledge retention. Additionally, these knowledge-focused discussions contribute to the vitality of learning communities by providing a continuous flow of topics, fostering camaraderie among members, and gradually shaping a strong sense of collective identity and cohesion. In the collaborative setting of a learning community, individual development is not only tolerated but actively promoted, where differences in interests and strengths are embraced as strengths. This heterogeneity among members fosters a unique dynamic where each member can observe and learn from others, leading to the expansion of individual knowledge and competencies. Notably, personalized learning environments within this community can manifest in overt or covert ways. In overt instances, such as collaborative task completion, experienced learners on a given topic may assume leadership roles, sharing their expertise and guiding the evolution of the community’s collective knowledge. Meanwhile, other members engage in personalized learning through targeted questions, consultations, and feedback sessions. Conversely, implicit personalized learning occurs through subtle mechanisms of observation and imitation. For example, a learner facing challenges in vocabulary acquisition might unconsciously improve by observing and adopting strategies from proficient peers. Similarly, learners deficient in social skills or time management can also benefit from this covert form of learning. In essence, while personalized and collaborative learning environments exhibit distinct characteristics, they are not mutually exclusive entities defined by discrete learning phases or physical spaces (e.g., in-class versus out-of-class settings).
• The pivotal goal revolves around the constructive evolution of both individual and collective knowledge frameworks. This teaching model meticulously blends personalized learning milieus with collaborative learning ecosystems within the learning community, geared towards nurturing and accelerating the construction of both individual learners’ knowledge and the collective knowledge of the group, akin to the converging gray area depicted in the Tai Chi diagram. It underscores that neither personalized nor collaborative learning environments can be neglected in achieving the goal of knowledge construction, as their synergy is vital. Moreover, the model incorporates the process of knowledge construction within the framework of Bloom’s Taxonomy of Cognitive Domains, illustrating that this process does not adhere to a rigid linear progression, nor do the cognitive levels in Bloom’s Taxonomy necessarily escalate sequentially. Instead, knowledge construction progresses in a spiral manner, with potential intersections or skipping across cognitive objectives. The confinement of knowledge construction to static learning environments is a misconception. While it is commonly assumed that memorization and comprehension primarily occur in individual settings, this notion is challenged by the fact that these processes can also be facilitated through collective endeavors within learning communities, such as shared recitation and mutual clarification. Furthermore, the belief that analysis, application, reflection, and innovation are inherently collaborative undertakings is also contested, as these cognitive objectives can equally be cultivated within personalized learning environments, demonstrating the versatility and interconnectedness of different learning modalities. It is essential to emphasize that individual and collective knowledge are not independent constructs or aspirations, but rather complementary aspects of the learning ecosystem. The collective knowledge within a learning community arises from the harmonious blend of individual efforts, as members collaborate, share insights, and engage in deep analysis to reach shared comprehension and develop a distinctive learning culture. This process is fundamentally driven by the social negotiation and construction of individual knowledge, where each individual’s cognitive landscape is enriched by the collective, and vice versa. The intertwined dynamic between individual and collective knowledge within a learning community fosters a continuous upward spiral, nurtured by both personalized and collaborative learning modalities. Consequently, the design of educational content, instructional methodologies, and assessment frameworks must encompass the dual imperatives of advancing both individual and collective knowledge development.
• The holistic learning environment is fortified by a diverse array of resources, seamlessly integrated and ubiquitous in their support of learning objectives. The proliferation of information-enabled educational technologies has significantly enriched the learning resource landscape. This encompasses a comprehensive spectrum, including flipped classroom resources, online resources, social resources, and the shared repertoire within a learning community, all of which are designed to cater to the varying demands of constructivist learning across multiple scenarios. Flipped and online resources range from curated micro-lectures and accompanying instructional materials to a panoply of digital assets, such as visuals, audios, videos, written content, websites, and innovative tools for interaction, documentation, and content creation (e.g., instant messaging platforms, massive open online courses, blogs, vlogs, and open-source platforms fostering user-generated knowledge). Social resources are multifaceted, encompassing teachers, renowned experts and scholars in specific domains, and individuals who reside within learners’ immediate learning ecosystems, offering invaluable assistance. The learning community’s shared skill repository represents a rich tapestry woven from the diverse talents and contributions of its members, intertwined with the collective intelligence embodied in their collaborative endeavors. This includes tangible outputs like project accomplishments and discussion archives, as well as intangible elements like the community’s discourse norms, learning ethos, and sense of belonging. This collective knowledge, which is both a product and a catalyst of community development, stands as a vital resource underpinning its continued growth. Additionally, there are intersections between these resource categories, with flipped resources, for instance, often incorporating online resources like MOOC videos, articles, and various digital platforms and tools. The endorsement from experts in certain disciplines, as part of social resources, often manifests as learners’ ability to procure pertinent resources via online platforms. Consequently, the categorization of resource types outlined in this roadmap aims to form an exhaustive list, embracing all potential avenues that can bolster learners’ knowledge construction endeavors. This underscores the model’s inclusivity and adaptability towards various resource types. Notably, these resources are not confined to a singular learning phase or environment, transcending the boundaries of both in-class and out-of-class contexts. Instead, they are seamlessly integrated across both personalized and collaborative learning ecologies, aligned with the model’s core mission to support constructive learning that transcends the traditional classroom paradigm.
• Repositioning the Teacher’s Role. This model eschews rigid definitions of teacher and student roles, mirroring the spontaneous nature of teaching and learning in a constructivist environment where the where, how, and by whom these activities occur are not predetermined. The flipped classroom model reverses the traditional script, pushing knowledge transmission and explanation beyond the confines of the classroom, encouraging teachers to descend from their authoritative perches and morph into instructional designers and supporters, intimately involved in identifying and supporting struggling students. In a learning community, membership and roles evolve organically in response to the learning themes, fostering an environment of equity and mutual respect. Core roles are not externally imposed but emerge naturally within the community’s fabric; for instance, when a learning theme aligns with a member’s domain expertise, they naturally ascend to the role of an expert, while others, through observation and participation, gradually attain core status. This cycle repeats as the community tackles new learning themes. Within the flipped classroom + learning community approach, teachers’ roles are not monolithic, but rather fluid, shifting within each community context. The teaching–learning dynamic extends beyond the classroom walls and lecture-style interactions; learners can accumulate knowledge through observing their peers within the community or engaging directly in practical experiences. In this framework, which aligns with contemporary educational structures, teachers retain a pivotal organizing role, particularly at the onset, by providing robust infrastructure for personalized and collaborative learning environments. This encompasses resource curation, community facilitation, guidance, and assistance with overcoming practical obstacles. Additionally, teachers serve as invaluable social resources and domain experts, offering learners accessible knowledge support that is constructive rather than prescriptive. They can also engage as active members within the learning community, forming the cornerstone of a collaborative teacher–student community. Although this study primarily explores the construction of student peer learning communities, the model’s applicability extends to fostering teacher–student learning communities as well.
• Positioning in-class and out-of-class activities as contextual backdrops. The flipped classroom paradigm harmoniously fuses traditional and digital learning modalities, enhancing educational outcomes. Nevertheless, the foundational model’s clear delineation of these phases, while emphasizing the novelty of the “flipped” approach, fails to capture the intricate, spiral dynamics of learning construction and cognitive elaboration central to constructivist perspectives. This segmentation may inadvertently foster disjointed learning experiences due to insufficiently nurtured learning environments and inadequate resource provisioning. The inflexible dichotomy between extracurricular and in-class learning, as embodied in the traditional flipped classroom model, poses a significant challenge for courses already burdened with inadequate class hours. By allocating extracurricular time primarily for self-study through video lectures, aiming at foundational knowledge acquisition, and reserving in-class sessions for collaborative endeavors that should ideally encompass application, analysis, reflection, and innovation, students may find their extracurricular tasks insufficiently enriching, fostering a limited understanding. Moreover, the condensed in-class period may prove inadequate for the thorough exploration of higher-order cognitive objectives, while the ad hoc collaborative environment may not adequately support effective collaboration. This formalistic adherence to the flipped classroom framework, neglecting the need for adaptability, has emerged as a critical obstacle hindering its further advancement. With the advancement of educational theories and practices, the cornerstone of the constructivist-oriented teaching model shifts from the mere dissemination of knowledge to the cultivation of an environment that nurtures learners’ knowledge construction. Learning is a perpetual, boundless endeavor that transcends temporal and spatial constraints, mirroring the essence of constructivist learning and the learning culture embraced in the digital age. Hence, the conventional dichotomy between in-class and extracurricular segments in the flipped classroom paradigm fades into the background, transforming into a contextual framework. In-class interactions embody the synchronous engagement of teachers and learners, empowering teachers, as experts, to efficiently observe and guide learning communities and individuals during direct encounters. Additionally, the mutual observation among learning communities sparks introspection and deliberation within each collective, emphasizing that in-class and extracurricular experiences, akin to learning resources, intertwine and permeate the entire learning ecology.

The flipped classroom + learning community model showcases an integration of personalized and collaborative learning environments, where individual and collective knowledge flourish in tandem within the learning community. It adeptly weaves together diverse learning resources and seamlessly integrates the in-class and out-of-class dimensions of the traditional educational system, fostering a multifaceted and inclusive learning landscape. This model embodies the cyclical and upward trajectory of knowledge construction, aligning perfectly with the constructivist vision of lifelong learning. Thus, it possesses remarkable adaptability and scalability, transcending the boundaries of a single course or learning phase. The conclusion of a course does not signal the end of learning; rather, it serves as a stepping stone for the model’s continuous ascension, with the ongoing process of knowledge construction persisting indefinitely.

4. The Effectiveness of Flipped Classroom + Learning Community Model

To validate the effectiveness of the teaching model, this study embedded it in a German language course at a university in southwestern China. Utilizing a quasi-experimental design featuring a single-group time-series methodology, we conducted pre- and post-tests and gathered data on the same set of learners, who were first exposed to the traditional flipped classroom model and then to the model of a flipped classroom combined with a learning community. This enabled a comparative analysis of the learners’ academic performances and perceptions across the two instructional approaches.

4.1. Participants

The entirety of the teaching practices and data collection endeavors for this study were grounded in the Second Foreign Language (German) curriculum offered at a university situated in southwestern China. This curriculum primarily targets third-year English majors who opt for German as their second foreign language, yet it also extends its reach to students from diverse disciplines as an interdisciplinary elective. Annually, a single class cohort participates in this program, which spans three academic semesters, each featuring 18 weeks of instruction. Within each week, four hours of face-to-face instructional time are allocated.

The study’s data were amassed from the initial two semesters of the second foreign language course, specifically tailored for the 2019 intake, during the autumn of 2021 and spring of 2022. The inaugural semester embraced a solitary flipped classroom approach, transitioning to a hybrid model incorporating both flipped learning and a learning community in the subsequent semester. Focusing on third-year students, this cohort exhibits autonomy in learning and a solid foundation in information technology, and all the students possess a background in foreign language acquisition. Encompassing 26 English majors, the experimental group’s members briefly collaborated solely through diverse group activities during the first semester, fostering a sense of class camaraderie. Commencing their German studies from scratch, the gender distribution was relatively equitable, with an 11:17 male-to-female ratio.

4.2. Instruments and Procedure

The data were gathered over two consecutive semesters from September 2021 to June 2022, each lasting 18 weeks. Initially, only two students had flipped classroom experience. Thus, the first semester initiated with a training program introducing the concept, format, and requirements. Utilizing an e-learning platform, the instructor curated a course and uploaded all flipped resources, allowing students to self-study videos and quizzes out of class. In-class sessions emphasized interactive activities to deepen understanding and facilitate application. Collaboration occurred through voluntary and randomized groupings, with diverse tasks recorded on the e-learning platform. Five interim assessments and a final exam evaluated grammar, vocabulary, sentence structures, and texts. The final grade comprised 70% regular assessments (26 exercises and 5 tests) and 30% final exam scores. Post-semester, 25 out of 26 students completed a questionnaire.

The second semester replicated the platform, with 21 exercises and 5 interim assessments contributing 70% of the grade total. Beforehand, students underwent learning community training, fostering understanding of its principles and differentiating it from other collaborations. Following training, students voluntarily formed communities of 3–4, which the instructor fine-tuned for heterogeneity. Nine communities emerged, eight with 3 members and 1 with 4 (due to cross-major enrollments, totaling 28, but analysis focused on 26 consistent participants). Communities devised visions, rules, and portfolios, engaging in weekly discussions. Post-semester, a 100% response rate was achieved for the questionnaire.

4.3. Data Analysis

To address the second research question posed in this study, the researcher conducted a comparative analysis of learners’ academic performance, examining the impact of two instructional models: the traditional flipped classroom (FC) and the flipped classroom augmented with a learning community (FC + LC), leveraging descriptive statistics and the paired t-test. This comprehensive evaluation encompassed a weighted assessment comprising 40% regular assignments, 30% unit tests, and 30% final examinations, all tailored to reinforce German language proficiency across domains such as pronunciation, vocabulary, sentence patterns, grammar, translation, reading, and writing. These evaluations were consistently administered and recorded throughout the semester on a dedicated learning platform, ultimately yielding a comprehensive performance summary. This assessment strategy, as a process-driven evaluation focused on language development, provided insights into learners’ progressive mastery and application of linguistic knowledge.

For the third posted question about the development of learner perception, the researcher drew on Barkhuizen’s [35] definition and research methodology of learner perception. This study devised questionnaires from three dimensions: feelings (e.g., I prefer the combination of flipped classroom and learning community over the flipped classroom alone from the previous semester), evaluations (e.g., the flipped classroom model, coupled with a learning community, has been of immense assistance in strengthening my ability to use German language effectively), and predictions (e.g., I believe that the combination of flipping classroom and learning community will be very beneficial to my future studies in other disciplines). One survey was administered after the FC model, and another following the FC + LC model. Responses were recorded on a Likert scale ranging from “1 = Strongly Disagree” to “5 = Strongly Agree”, enabling quantifiable analysis based on the scores obtained.

Based on the classification criteria of the Likert 5-point scale by Oxford and Burry-Stock [36], a mean score of 3.5 or above indicates a high level, scores between 2.5 and 3.5 suggest a medium level, and scores below 2.5 represent a low level. Both surveys targeted all 26 participants in the course, yielding 25 valid responses (96.15% response rate) for the first survey and 26 responses (100% response rate) for the second. However, to ensure a comparison between the two modes, questionnaires from individuals who did not participate in the first survey were excluded from the second survey’s data analysis, resulting in a final count of 25 surveys. Reliability and validity analyses using SPSS revealed Cronbach’s α coefficients of 0.802 and 0.841, respectively, with validity confirmed by Bartlett’s test of sphericity (p = 0.01), indicating significant correlation and high validity. The survey results were scored according to the Likert scale, assigning 1 point to option E (“extremely disagree/dissatisfied/ineffective or counterproductive”) and incrementally adding 1 point for each preceding option up to 5 points for option A (“strongly agree/satisfied/very helpful”). Descriptive statistics and a paired t-test were conducted on all respondents’ scores.

To gain a comprehensive insight into the progression and determinants of learners’ cognitive development, the researcher expanded the research design to include interviews with 26 learners. The interview protocol, anchored in the questionnaire topics, comprised 10 semi-structured questions, designed to elicit detailed accounts of learners’ cognitive journeys with the teaching model. Following transcription and systematic coding, the researcher conducted a qualitative analysis, presenting the findings in accordance with the tripartite framework proposed by Barkhuizen [35] for learners’ perceptions, namely, the expression of sentiments, the formation of opinions, and the making of anticipations.

4.4. The Results

4.4.1. Academic Performance

The descriptive statistical analysis (

Table 1. The descriptive statistics of learners’ academic performance in FC and FC + LC.

VN.	N.	Max.	Min.	Mean	SD	Med.	Var.	P25	P75
FC	26	88.92	55.71	75.91	8.95	77.76	80.18	68.97	81.58
FC + LC	26	95.50	54.60	81.09	11.30	82.75	127.57	74.93	90.50

) highlighted a significant improvement in the average performance score upon the introduction of the FC + LC model, with an increase from 75.91 to 81.0. The implementation of a learning community model within the flipped classroom framework yielded a general elevation in the students’ academic achievements. Notably, the highest score achieved under the FC + LC surpassed that of the FC by 6.58 points, whereas the lowest score declined by 1.11 points, hinting at a potentially more pronounced enhancement for high achievers and a possible negative influence on lower achievers. To corroborate this inference, as presented in

Table 2. The percentile statistics of learners’ learning performance in FC and FC + LC.

VN.	P2.5	P5	P10	P25	P27	P33	P50	P67	P73	P75	P90	P95	P97.5
FC	55.71	57.23	61.50	68.97	69.43	74.30	77.76	80.23	81.07	81.58	87.97	88.89	88.92
FC + LC	54.60	56.32	60.41	74.93	76.11	78.42	82.75	88.44	90.23	90.50	92.76	95.08	95.50

, a percentile-based statistical analysis was conducted, revealing that the FC + LG scores surpassed the FC scores above the 25th percentile, but fell short below the 10th percentile, signifying a decrement in academic performance among the bottom 10% of the students upon the integration of the learning community.

The aforementioned descriptive statistics merely reflect data variations, and their validity necessitates a paired t-test, as presented in

Table 3. Paired t-test statistics for learners’ academic performance in FC and FC + LC.

PV	Mean ± SD			t	df	p	Cohen’s d
	PG1:FC	PG2:FC + LC	DBPV
FC/FC + LG	75.91 ± 8.95	81.09 ± 11.30	−5.18 ± −2.34	−3.69	25	0.001 ***	0.72

*** represents significance level of 1%.

. In comparing the academic performances of learners under two modes, the FC + LC model (M = 81.09, SD = 11.30) yielded higher scores than the FC model (M = 75.91, SD = 8.95), with a significant p-value of 0.001 ***, rejecting the null hypothesis and confirming a statistically significant difference. The magnitude of this difference, as measured by Cohen’s d, was moderate, at 0.72.

A further analysis, categorizing the learners by their respective learning communities (

Table 4. Statistical table of academic performance of nine community members.

Community nr.	Member 1			Member 2			Member 3			Average Scores in FC	Average Scores in FC
	FC	FC + LC	Difference (FC + LC/FC)	FC	FC + LC	Difference (FC + LC/FC)	FC	FC + LC	Difference (FC + LC/FC)
G1	82.78	89.8	7.02	69.01	73.2	4.19	82.71	88	5.29	78.17	83.67
G2	66.39	75.5	9.11	78.75	91	12.25	78.06	72.7	−5.36	74.4	79.73
G3	88.92	94.3	5.38	75.91	81.7	5.79	62.12	60.8	−1.32	75.65	78.93
G4	80.76	90.4	9.64	68.84	85.7	16.86	64.67	78.5	13.83	71.42	84.87
G5	87.6	91.6	4	76.87	90.8	13.93	88.84	95.5	6.66	84.44	92.63
G6	75.43	92.1	16.67	77.46	79.9	2.44	55.71	63.5	7.79	69.53	78.5
G7	80.05	88.3	8.25	74.68	82.4	7.72	81.2	82.6	1.40	78.64	84.43
G8	80.18	82.9	2.72	70.48	54.6	−15.88	60.06	59.85	−0.56	70.24	65.67
G9	87.39	85.4	−1.99	/	/	/	78.7	77.6	−1.10	83.05	81.50

), revealed that among the nine community samples, only two had negative average performance differences, indicating a decline in scores upon joining the community. Conversely, the remaining seven communities showed overall performance improvements. Scrutinizing the performance evolution among the constituents of individual learning communities unveils a remarkable consistency, wherein the vast majority, or indeed all, of the members exhibited congruent shifts in their performance profiles. Notably, six communities manifested a unified pattern of either rising or declining performance among their members, whereas in three communities, a single member’s performance trajectory diverged from the collective pattern exhibited by the other two. This phenomenon underscores the profound influence that learning communities wield over the direction and magnitude of learners’ performance changes, demonstrating the direct and consequential impact they have on fostering or altering the performance outcomes of their respective members.

Collectively, the integration of the flipped classroom and learning community paradigm had a moderately positive effect on learners’ overall language learning achievement, enhancing scores within the upper 75% percentile but paradoxically diminishing those in the lowest 10%. This signifies that while the model fosters improvement among mid-to-high performers, it may inadvertently hinder those in the lower performance bracket, potentially exacerbating their scores. Within the experimental context, 76.9% of the learners witnessed a performance boost upon engaging with the learning community, whereas 23.1% experienced a decline. Assessing the degree of improvement, 57.69% of the learners achieved substantial gains, with the respective improvements surpassing the overall mean performance difference (5.18) achieved by the cohort. Conversely, 15.4% displayed negligible changes (performance difference between −2 and 0), and a mere 7.7% encountered moderate-to-significant declines, attesting to the model’s prevalent effectiveness in enhancing learner performance. From a community-specific perspective, Group G4 manifested the greatest efficacy in performance elevation, Group G9 demonstrated marginal effectiveness, and Group G8 exhibited a counterproductive influence.

4.4.2. Learner Perception Development

Based on the results presented in

Table 5. Descriptive statistics of the scores from the questionnaire on learners’ perception development.

VN.	N.	Max.	Min.	Mean	SD	Med.	CV	t-Test	p	Cohen’s d
FC	25	4.36	2.91	3.69	0.41	3.91	0.11	−1.33	0.195	0.27
FC + LG	25	4.89	2.83	3.86	0.53	3.83	0.14

, notably, both surveys exhibited low standard deviations and coefficients of variation below 0.15, indicating minimal data dispersion and suitability for mean comparison. Both surveys yielded average scores above 3.5, signifying a high level of learner perception development in both modes. Nevertheless, the FC + LC model achieved a slightly higher mean score than the flipped classroom alone, albeit with a marginal increase in effect size (d). Notably, the flipped classroom + learning community model also exhibited a wider range of extreme values, with a higher maximum and a lower minimum score, suggesting greater variability in learner perception development. This underscores the potential for greater enhancement among students who adapt to or prefer this mode, as well as the potential inhibition of those with lower acceptance levels.

To further investigate the association between learners’ perception development and disparate community efficiencies, the researchers aggregated the scores by learning community, examining disparities in effects from this vantage point. As depicted in

Table 6. Statistical table of questionnaire results for learners’ perceptions under FC and FC + LC.

Community nr.	Member 1			Member 2			Member 3			Average Scores in FC	Average Scores in FC
	FC	FC + LC	Difference (FC + LC/FC)	FC	FC + LC	Difference (FC + LC/FC)	FC	FC + LC	Difference (FC + LC/FC)
G1	4.09	4.89	0.8	3	4.06	1.06	4	4.56	0.56	3.7	4.5
G2	3.45	3.5	0.05	3.91	4.06	0.15	3.36	3.72	0.36	3.58	3.76
G3	3.91	4	0.09	3.27	3.5	0.23	3.91	3.78	−0.13	3.7	3.76
G4	3.91	4.67	0.76	/	/	/	4.09	4.17	0.08	4	4.42
G5	3.91	3.28	−0.63	4.36	4.67	0.3	3.36	3.83	0.47	3.88	3.93
G6	3.91	4	0.09	4.09	3.78	−0.31	3	4.06	1.06	3.67	3.94
G7	2.91	3.83	0.92	3.73	3.94	0.22	3.18	4.28	1.1	3.27	4.02
G8	4	2.83	−1.17	3.27	3.33	0.06	3.82	3.56	−0.26	3.7	3.24
G9	3.91	3.22	−0.69	/	/	/	3.91	2.89	−1.02	3.91	3.06

, 72% of the learners exhibited positive progressions, and a substantial proportion of 46.2% of the learners attained a superior level of development, achieving or surpassing the average difference (0.17) in their progression. Remarkably, the average scores diverged significantly across the nine learning communities, while the changes within the individual members of these communities exhibited a fundamental uniformity, suggesting a disparity in the effectiveness of different learning communities.

In summary, both the FC and the FC + LC models garnered substantial recognition from the learners. Importantly, the latter model achieved a higher overall satisfaction average, indicating its superior capacity to enhance learners’ cognition over the FC model. However, this also led to more apparent cognitive disparities among the members, with dissatisfied respondents registering lower satisfaction scores than the least satisfied in the FC model, underscoring the potential for more intense negative impacts. These disparities are directly correlated with the variations in the effectiveness of the learning communities. Of the nine surveyed communities, seven reported increased overall satisfaction after adopting the FC + LC model, and two experienced declines. This highlights the pivotal role of learning community efficacy in this model, as a highly effective community can significantly enrich members’ understanding and motivation towards the flipped classroom, while an ineffective one can significantly hinder their learning experience and perception development.

With the qualitative interviews, under the rubric of “Expressing Feelings”, a meticulous analysis revealed 224 reference codes, which were subsequently categorized into three refined sub-categories: opinions exclusively on the flipped classroom approach (42), sentiments towards the FC + LC model (149), and attitudes and personal responses to collaborative learning dynamics (33). The learners responded favorably to the FC model, citing its empowering effect on individual learning, which offers unprecedented autonomy in managing study time and progress. Additionally, the departure from traditional teaching modalities elicited a refreshing sense of novelty among the learners. For example, G7S3 stated the following:

During the first semester, I found the flipped classroom quite novel because I could arrange my own time. Also, since it’s a brand new language, being able to listen and watch repeatedly is of great help to my learning. I wouldn’t worry about not understanding in class and how to make up for it after class. If I didn’t understand something, I could always go back and review it. I felt this was extremely beneficial for laying a solid foundation. In general, I have a relatively high acceptance of the flipped classroom model.

It is noteworthy that numerous participants expressed discomfort initially, which can be distilled into four main factors: the significant demands placed on individuals’ autonomous learning abilities, insufficient personal discipline, the lack of instantaneous feedback, and the inefficiency of collaborative efforts undertaken during classroom sessions. The following examples illustrate this point:

G4S2: The main difficulty I faced during the first semester was with the classroom activities. Since everyone was still unfamiliar with each other, but we were expected to interact frequently, it felt quite awkward.

G3S4: Although this method can enhance learning, I often find it challenging to fully prepare for class because I rely heavily on self-study. Unfortunately, I’m not particularly disciplined, so there are times when I don’t do as much advance preparation as I should. Consequently, during class, I sometimes struggle to fully understand the material.

In comparison, the interviewees voiced overwhelmingly positive perceptions of the flipped classroom merged with the learning community approach, emphasizing the crucial roles it played in fostering mutual oversight, accelerating the pace of autonomous learning, and ensuring the instantaneous feedback. For example, G5S2 noted the following:

With the encouragement and oversight from the learning community in the second semester, I was able to better adapt to the flipped classroom model, and both my study time and efficiency have improved. I think the flipped classroom is working quite well now.

Classified under the primary code, “Formulating Judgments”, there are three subsidiary classifications: appraisals of individual learning communities (56), the repercussions of the flipped classroom + learning community framework for German language proficiency (73), and its implications for miscellaneous skills (32), amounting to 161 coded instances.

In assessing their own learning communities, a unanimous affirmation emerged, with all the respondents acknowledging them as equitable and mutually supportive environments, even those who expressed reservations over other aspects. Regarding the effect on German acquisition, two respondents cited a minimal influence, whereas the vast majority (24) documented positive outcomes, primarily stemming from the following: (1) the community’s pivotal role in overseeing vocabulary development; (2) the enhancement of grammatical comprehension through peer-to-peer explanations, facilitating the identification and rectification of knowledge gaps; (3) the facility for instant discussions post-video viewing, enabling prompt feedback and optimizing learning efficacy; (4) the ability to address unnoticed learning deficiencies through the perspectives of peers and the enriching exchange of study methodologies. In terms of the impact on other capabilities, the majority of the respondents mentioned advancements in self-learning abilities (encompassing self-discipline, the cultivation of study habits, the enhancement of learning capacity, reflective thinking, and critical thinking abilities), as well as advancements in team collaboration skills, interpersonal communication skills, and, to a certain extent, innovative capabilities. An example was provided by G2S1, as follows:

Especially when it comes to grammatical points, it’s incredibly helpful. It also significantly aids in vocabulary acquisition and enhances my listening, speaking, reading, and writing skills. Initially, when I started memorizing German vocabulary, I had a hard time remembering the gender of words—masculine, feminine, and neuter. I would often mix them up. However, I mentioned this issue in our learning community, and Zhang (G2S2) suggested associating and grouping the words together. I adopted her method, and it really helped me retain the information more firmly. Regarding listening, speaking, reading, and writing, we regularly practice these skills after each unit. For instance, we’ll dictate or write down the words after completing a unit. Additionally, during our discussions, we often read out loud from the texts, and my peers kindly correct my pronunciation whenever I mispronounce something. I find these practices incredibly useful.

The predictive evaluation yielded 48 open-coded responses, delving into the learners’ perspectives on the future of the flipped classroom + learning community model and strategies for its improvement. Notably, a lone voice emerged, expressing dissatisfaction with the combined model but admiration for the flipped classroom concept, preferring solo learning endeavors. In stark contrast, 25 interviewees conveyed their willingness to continue embracing the model, even those who disclosed negative sentiments and assessments, suggesting they would proceed with adjustments to team configurations. The remaining 23 participants enthusiastically endorsed the model’s continuation, expressing a desire to preserve their collaborative dynamics with current team members. For example, G4S1 stated the following:

I believe that language courses are generally quite suitable for FC + LC model. Since we’re learning a foreign language, and normally in school, we only speak Chinese, there’s a lack of language environment, especially when it comes to learning German. Therefore, I think the learning community model is quite fitting for language courses. I hope to continue using the flipped classroom combined with the learning community model in the next semester, and I’m willing to continue forming a community with the other two community members.

5. Discussions

The flipped classroom is a teaching model developed based on constructivism, within which Vygotsky’s sociocultural theory underscores the significance of a collaborative learning environment in the flipped classroom [37,38]. Having evolved beyond its initial iteration, which merely inverted the conventional teaching sequence, the flipped classroom paradigm has shifted towards prioritizing knowledge construction, subsequently necessitating a socially collaborative framework that nurtures this process (e.g., [14,39]). This study proposes a “Tai Chi Pillar” teaching model combining the flipped classroom and the learning community, in which personalized learning environments and community-based collaborative learning environments are seamlessly integrated, forming the core components of the model. This teaching model organically integrates the concepts of the flipped classroom and the learning community, embodying a constructivist approach to learning, breaking down barriers between in-class and out-of-class phases, and exhibiting sustainability and scalability.

To ascertain its effectiveness, this model was implemented in real-world teaching settings, with learners’ academic performance and perceptions analyzed statistically. In terms of the academic performance, the outcomes reveal that 76.9% of the learners experienced academic gains under the FC + LC model, with 57.69% achieving substantial improvements, signifying a robust effect. This conclusion reinforces the notion that emphasizing collaborative environments in Flipped Classroom 2.0 intensifies the model’s effectiveness in promoting learning outcomes [39,40,41], emphasizing the intertwining relationship between flipped classrooms’ sustainability and deeply embedded social construction. Nevertheless, our research outcomes indicate that while the FC + LC approach generally improves personal learning performance among most learners, it paradoxically hinders the progress of the bottom 10% performers. Additionally, the functionality of individual learning communities directly correlates with their members’ achievements, implying that any impediments to their formation or development can significantly detract from their members’ success. This finding underscores the complexity in cultivating collaborative environments, as social construction moves beyond gathering and talking; the establishment and growth of learning communities are challenged by a myriad of influencing variables [42].

In terms of the aspect of the development of learners’ perceptions, both instructional paradigms elicited substantial satisfaction, with the FC + LC configuration yielding notably higher satisfaction ratings than the FC model. This underscores the FC’s known superiority over traditional methodologies in promoting learners’ perceptions [43,44] and simultaneously substantiates the efficacy of the FC + LC model explored in this study. Notably, 72% of the learners attested to improvements in their German language proficiency, adaptability to the flipped classroom environment, and personal development subsequent to incorporating the learning community approach. Their perception development and satisfaction levels escalated to a greater extent than under the FC. The learners eloquently illuminated how the learning community reinforced and complemented the FC, enhancing learning efficacy, productivity, and not merely advancing linguistic competencies but also cultivating self-directed learning, social interaction, teamwork, and innovation skills. According to the interview findings, despite the FC model’s merits, challenges such as cultivating learner autonomy, the scarcity of immediate feedback during self-study, inefficient collaborative activities within class, and social awkwardness during interactions, as highlighted in prior research (e.g., [3,45,46]), remained prevalent. However, the integration of the LC transformed the learning landscape by concurrently ensuring personal learning spaces and fostering a collective collaboration sphere that surpassed the physical classroom. This collaboration transcended traditional barriers, necessitating both individual problem-solving and joint project work, both in and out of class. The FC + LC models harmoniously supported and complemented each other, maximizing learning efficacy. Notably, the LC’s provision of instant feedback to learners mitigated a common grievance under the FC model, the lack of timely feedback, thereby enhancing the overall learning experience. Unlike other collaborative frameworks, the learning community promotes enduring, consistent, and regular self-driven collaboration, where members engage in equitable dialogue to drive mutual progress and shape their individual roles and identities. This process gradually cultivates a strong bond of mutual trust, reliance, and a profound sense of belonging, ensuring that collaboration is substantive and that it continuously evolves. This, in essence, lays a solid foundation for the model’s continued advancement. Importantly, similar to the fluctuations in learning performance, learners’ perception development is intimately tied to the growth and development of their learning community. When the community’s development stagnates, this leads to a significant drop in the satisfaction levels of its members. Although the study recorded notable shifts in learners’ perception development, these changes were relatively modest when juxtaposed against the substantial gains in learning performance. Interestingly, a minority of the learners reported marginally lower satisfaction with the FC + LC model compared to the FC model, yet their academic achievements soared. This underscores that learner satisfaction and learning outcomes are not necessarily congruent, as learners’ perceptions of a pedagogical model are multifaceted and influenced by diverse factors [47,48].

6. Limitations and Future Research

This research endeavors to present a flipped classroom teaching model, framed within the constructivist paradigm and metaphorically conceptualized as a “Tai Chi pillar”, which harmoniously integrates the pivotal elements of both the flipped classroom and learning community. This synthesis imparts a universal character to the model. However, to ascertain its validity and foster continuous enhancement, the model necessitates iterative experimentation within a panoply of disciplinary landscapes and educational milieus. By positioning its educational design and experimentation within the context of college German courses, this study provides a tangible illustration of the model’s application in foreign language education. Nevertheless, the exploration of its applicability and effectiveness across diverse academic domains remains an open frontier for future research endeavors.

Additionally, in the context of empirical research evaluating the model’s validity, the inherent constraints imposed by external factors resulted in a comparatively small sample size. Despite employing a rigorous quasi-experimental methodology and taking meticulous steps to neutralize the confounding effects of individual-specific variables during the data analysis, the potential interference from uncontrollable variables inherent in a limited sample pool remained.

Furthermore, the study’s exclusive reliance on the researcher as the sole implementer precludes a nuanced examination of how different teachers’ prior experiences with flipped classrooms and their unique teaching styles might contribute to or detract from the model’s efficacy. Additionally, the restriction of the study’s focus to English majors within a comprehensive university oversimplifies the complexities that might arise from the model’s application across broader educational strata and disciplines.

To build upon the current research limitations, future research avenues could encompass:

• Experimentation and iterative refinement of the FC + LC teaching model across a spectrum of disciplines, academic levels, majors, and cultural milieus, fostering a more nuanced comprehension of its practical implications.
• Strengthening the empirical foundation by augmenting sample sizes and introducing control groups, thereby offering a more compelling argument for the superiority of the FC + LC model over the FC model.
• Delving deeper into the mechanisms that govern learning community effectiveness in flipped classrooms, recognizing its direct bearing on learners’ academic achievement and perception development, with a view to mitigating the potential for the Matthew effect by elucidating the influencing factors and dynamic evolution of such communities, ultimately fostering sustainable learning outcomes within the FC + LC model.

7. Conclusions

This research presents an innovative flipped classroom model, intertwined with a learning community framework, specifically tailored for Tai Chi pillar instruction. It advances beyond the flipped classroom model by intensifying the cultivation of a collaborative learning environment and shifting the model’s core from a mere procedural reversal to the fostering of knowledge construction. By achieving a seamless fusion of the various components within the flipped classroom, it paves the way for its sustainable progression. An empirical assessment of the model’s efficacy in a college German course reveals marked improvements in both learning performance and student perceptions for a substantial proportion of learners, underscoring the augmented benefits of integrating a learning community into the flipped classroom paradigm. Nevertheless, the study also underscores that disparities in effectiveness among learning communities can exert a direct influence on their members’ academic achievements and the evolution of their learning perceptions. Consequently, if the construction and development of learning communities are unsuccessful, they can significantly impede these positive outcomes. Thus, this study recommends expanding the application of this teaching model across various educational contexts for further experimentation and refinement, ultimately enhancing its effectiveness.