Evaluating Collaborative Teaching under Taiwan Foreign English Teacher Program

1. Introduction

Taiwan’s ongoing commitment to English language education has produced one of Asia’s most extensive foreign teacher recruitment programs. The Taiwan Foreign English Teacher Program (TFETP), originally launched by the Ministry of Education in 2004 as the Foreign English Teacher (FET) program, was substantially expanded in 2021 in alignment with the Bilingual 2030 national policy—a government initiative seeking to establish a bilingual Chinese-English environment in education, government, and public life by the year 2030 (National Development Council, n.d.). By 2024, the program employed 1,096 foreign English teachers and teacher assistants across the country, representing a growth of over 500 percent since its inception (TFETP, n.d.-a).

The program represents a significant public investment. Foreign English teachers under TFETP receive competitive salary packages, housing allowances, flight reimbursements, and health insurance (TFETP, n.d.-b), making the program’s cost-effectiveness an important policy consideration. Central to the program’s design is a mandated collaborative teaching requirement: FETs are placed in elementary and junior high schools where they are required to co-teach with local English teachers for up to 20 periods per week (TFETP, n.d.-c). This collaborative teaching arrangement is conceived not merely as an administrative convenience but as the primary mechanism through which TFETP seeks to elevate students’ communicative English proficiency, enrich instructional practice, and support the professional development of local teachers.

Despite the program’s scale and the centrality of collaborative teaching to its stated goals, empirical research on how this model is enacted in everyday classroom practice remains limited. Most existing studies have focused on bilingual education recruitment, teacher identity, and program implementation rather than the lived dynamics of FET-LET collaboration (Lin & Wu, 2021; Wang & Lin, 2013; Yeh, 2024). This study responds to that gap by conducting a qualitative evaluation of the TFETP collaborative teaching model across five junior high schools in central Taiwan. The evaluation is guided by three research objectives: (1) to examine the patterns and influencing factors of the collaborative teaching model; (2) to identify the benefits and challenges to teachers’ professional development; and (3) to generate evidence-based recommendations.  

2. Literature Review

2.1. Defining Collaborative Teaching

Collaborative teaching refers broadly to a pedagogical arrangement in which two or more educators work together to deliver instruction to a diverse group of students in a shared educational space (Cook & Friend, 1995). Within this framework, Friend and Cook (2020) described co-teaching as a service delivery approach define collaborative teaching as an instructional arrangement in which specialized support is provided to students with disabilities or other special needs within the general education classroom, rather than through separate settings. While these definitions originated in special education contexts, their core elements—shared planning, joint instruction, and mutual professional responsibility—have been extended to bilingual and EFL settings, including programs like TFETP (Dove & Honigsfeld, 2018, as cited in Friend & Cook, 2020). 

Friend and Cook (2020) identity six principal co-teaching configurations: One Teaching, One Observing; Station Teaching; Parallel Teaching; Alternative Teaching; Teaming, and One Teaching, One Assisting. Each model distributes instructional roles and responsibilities differently, carrying distinct implications for teacher parity, planning requirements, and student grouping and instructional outcomes. The “One Teaching, One Assisting” model, wherein one teacher leads instruction, and the other provides in-class support, requires minimal joint planning and is commonly adopted when preparation time is scarce or when one teacher is less confident in delivering the subject content (Friend & Cook, 2020). More integrated models such as Teaming and Parallel Teaching require greater relational familiarity, shared planning, and negotiated role boundaries.

2.1. Taiwan-Specific Frameworks

Building on international co-teaching literature, Luo (2007, 2010, 2014) developed a contextually grounded model for collaborative EFL teaching between native English-speaking teachers (NESTs) and local teachers in Taiwan. Luo’s framework identifies four core components—lesson planning, collaborative teaching, monitoring, and reflective dialogue—and emphasizes eight relational and professional conditions for effective collaboration, which she terms R.E.F.L.E.C.T. Knowledge: respect, equality, flexibility, language sensitivity, empathy, collaborative culture, time allocation, and professional knowledge. Crucially, Luo (2014) described the framework as intentionally non-prescriptive, emphasizing that teachers should adapt it according to their professional contexts and needs.

Subsequent research confirmed the relevance of this framework while also exposing persistent implementation challenges. Teachers praised the model’s structure but consistently cited time constraints as a barrier to full enactment (Luo, 2014). Wang (2022) similarly found that co-teachers in Taiwanese bilingual schools struggled to coordinate lesson planning across demanding schedules, with local English teachers reporting that the additional responsibilities of bilingual curriculum development and co-teaching significantly their overall workload.

2.3. Key Conditions for Successful Collaborative Teaching

Research identified several conditions as fundamental to effective collaborative teaching partnership. Structured co-planning time is widely recognized as essential (Friend & Cook, 2020; Honigsfeld & Dove, 2010; Luo, 2014), as it enables the negotiation of roles, alignment of pedagogical goals, and integration of instructional styles. Equally important are mutual respect and professional trust, particularly in intercultural settings where differences in educational background, communication norms, and classroom expectations may generate misunderstanding or hierarchy (Carless & Walker, 2006; Luo, 2007). Research has shown that ambiguous role distribution frequently leads to marginalization of one co-teacher, with local teachers often reduced to assistants or translators rather than equal instructional partners (Kim & Moodie, 2023), underscoring the importance of clear and equitable role negotiation. Therefore, ongoing communication and flexibility are critical conditions for co-teachers to adapt instructionally and sustain collaboration (Honigsfeld & Dove, 2019; Simmons, 2018).

2.4. Challenges in Cross-Cultural Teaching

Research consistently documents significant challenges in cross-cultural co-teaching arrangements. Role ambiguity and power imbalance are particularly prevalent, with local teachers frequently cast in administrative or supportive functions while foreign teachers assume instructional leadership (Kim & Moodie, 2023; Luo, 2013). In Taiwan, Luo (2013) found through classroom discourse analysis that the NEST consistently assumed the role of primary initiator in co-taught lessons, while local English teachers largely took on responding and supporting functions, with little substantive interaction between the two teachers observed during instructional activities. Wang (2022) further documented how administrative downsizing and the lack of dedicated support structures increased the workload of local English teachers, who were expected to simultaneously manage co-teaching, curriculum design, and foreign teacher administration, while schools frequently leveraged foreign teacher recruitment as a strategy to attract student enrollment rather than to advance pedagogical goals.

Despite this body of research, empirical studies specifically examining how collaborative teaching is enacted under the current TFETP framework—launched in 2021 with significantly expanded scope—remain scarce. This study addresses that gap by offering a grounded, context-specific analysis of collaborative teaching as experienced by TFETP participants in junior high school settings.

3. Method

3.1. Research Approach

This study adopts a qualitative evaluation research approach, which is particularly suited to understanding complex educational programs through the lived experiences and contextual narratives of participants (Patton, 2015). Rather than measuring program outcomes through numerical indicators, qualitative evaluation seeks to illuminate the processes, dynamics, and contextual realities that shape implementation. This orientation is well-suited to the study of TFETP collaborative teaching, where interpersonal relationship, institutional structures, and cultural negotiation significantly influence how the program unfolds in practice.

Within this primary qualitative evaluation framework, the study incorporates Krueger’s (2015) using stories in evaluation as a supplementary interpretive tool. Using stories in evaluation (Krueger, 2015) captures the emotional dimensions, values, and meaning-making processes embedded in participants’’ accounts—elements that systematic coding alone may not fully surface. As Krueger observed, stories gathered across multiple participants allow patterns to emerge that reveal how individuals evaluate and make sense of their shared experiences (Krueger, 2015). In this study, representative narrative episodes from interview data are used to illustrate and deepen the interpretation of identified patterns.

3.2. Research Sites and Participants

Data were collected at five junior high schools ---pseudonymized as School A through E---all located within a single city in central Taiwan. At the time of data collection, nine junior high schools in this city were participating in TFETP during the 2021-2025 period. All nine were invited to participate in the study; five agreed and constitute the sample for this research. Limiting the study to one city controlled for regional policy variation while still allowing comparison across diverse institutional contexts, including schools of different sizes and urban-rural profiles.

Participants were selected through purposeful sampling and included: four to six foreign English teachers (FETs); four to six local English teachers (LETs); and four to six school administrators, each with direct oversight responsibilities for the TFETP. Selection criteria required FETs and LETs to have at least one year of experience within the TFETP collaborative teaching model, and administrators to have direct involvement in the program’s coordination. To protect participant confidentiality, all interview data were attributed using a school-based identification system. Participants are identified by their professional role — foreign English teacher (FET), local English teacher (LET), or administrator — and their school affiliation, designated School A through School E (e.g., FET interview, School A; Administrator interview, School C).

3.3. Data Collection

Three complementary data collection methods were employed. Semi-structured interviews served as the primary method, combining predetermined guiding questions with the flexibility to explore emerging themes (Galletta, 2013). Interview protocols were tailored to each participant group, covering collaborative teaching experiences, role perceptions, challenges, professional development impacts, and recommendations for improvement. All interviews were recorded and transcribed with participant consent.

Classroom observations provided an empirical lens to validate and contextualize interview narratives. At least one observation per FET-LET teaching pair was conducted during collaborative teaching sessions, using a structured observation form designed around the study’s research questions and aligned with established dimensions of effective collaborative teaching (Friend & Cook, 2020; Honigsfeld & Dove, 2019). Observational notes documented instructional roles, communication patterns, classroom management approaches, and student engagement.

Document analysis examined institutional and instructional materials, including lesson plans, school timetables, TFETP policy guidelines, and administrative reports. This method provided contextual grounding and enabled cross-referencing of participant accounts with formal program expectations.

3.4. Data Analysis

Data analysis followed an iterative, inductive process in which interview transcripts, classroom observation records, and institutional documents were repeatedly read and compared across participants, schools, and data sources. Through continuous comparison, patterns and models were identified and refined, forming the thematic structure through which the findings are organized.

Drawing on using stories in evaluation (Krueger, 2015) as an interpretive technique, the analysis attended to the evaluative judgement, values, and meaning-making processes embedded in teachers’ narratives. This technique enabled the researcher to surface implicit dimensions of collaborative experience—such as professional identity, relational dynamics, and cultural negotiation—that extend beyond coding categories. institutional documents to enhance credibility, with representative stories selected to illustrate broader patterns across participants and school contexts.

5. Results

5.1. The Common Institutional Landscape: Flexible Learning Periods^[1]

Across all five participating schools, a consistent administrative strategy positioned collaborative teaching sessions within Flexible Learning Periods—a school-developed curricular space established under Taiwan’s 12-Year Basic Education Curriculum Guidelines that grants schools autonomy to design contextually responsive learning experiences. This arrangement was not incidental; it reflected a deliberate institutional logic designed to insulate collaborative teaching from the exam-driven demands of the core English curriculum.

The primary driver of this separation was the Comprehensive Assessment Program (CAP)^[2], a high-stakes national exam for ninth-grade students in which English is a key subject. School administrators consistently framed their decision to position FET-led sessions outside the core curriculum as a protective measure that preserved instructional stability while allowing communicative goals to be pursued. As one administrator explained, placing the foreign teachers’ activities within Flexible Learning Periods ensured they would not interfere with the systematic progress required by core English classes.

This structural placement simultaneously defined the FET’s instructional role. In all five schools, FETs were explicitly informed by administration that their focus should be on motivating students to engage with spoken English rather than on examination content. One FET recounted being told that their goal was to help students enjoy using English, with all test-related instruction handled separately by local English teachers. Administratively, this logic also served to maintain LETs’ cooperation with the program: by ensuring FET activities reinforced rather than competed with LET-led grammar instruction, schools positioned the collaborative model as complementary rather than threatening to local English teachers’ professional role.

A consequence of this positioning was the emergence of a “one-to-many” staffing structure. Because Flexible Learning Periods are distributed across class timetables, administrators needed to allocate the FET across multiple time slots to fulfill the program’s mandated 20-period weekly teaching load. The result was that each FET rotated across numerous classes and collaborate with different LETs responsible for each class, rather than being paired with a single co-teacher. One FET at School A described this reality vividly: in Grade 7 alone, the FET was required to work with every local English teacher in the school, with no fixed partner (FET interview, School A).

5.2. The Localized “One Teaching, One Assisting” Model

Despite the theoretical availability of six co-teaching configurations described in the literature, classroom observations and interview data revealed that all five schools systematically enacted a localized version of the “One Teaching, One Assisting” model. In all observed sessions, the FET assumed primary responsibility for instructional delivery—leading communicative activities, modeling pronunciation and vocabulary, and managing pedagogical flow—while the LET performed a set of functional support roles.

Critically, this uniformity persisted within the Flexible Learning Periods, which had been specifically designed to be free from examination constraints. Classroom observation at School E documented lessons centered on communicative themes—such as occupational vocabulary and career discussions—that bore no relationship to examination content (Classroom observation, School E). This empirical finding challenges a simple exam-pressure explanation for model uniformity and points to deeper structural and psychological factors.

Three primary mechanisms were identified as driving the persistence of this model. The first is the structural consequence of the one-to-many pairing arrangement. More integrated co-teaching models, such as Station Teaching, Parallel Teaching, and Teaming, require not only joint planning time but a degree of relational familiarity—shared knowledge of students’ proficiency profiles, established communication routines, and a mutually negotiated pedagogical language—that can only develop through sustained, repeated collaboration.

When an FET-LET pair co-teaches the same class only once per week and the FET is simultaneously managing relationships with up to twenty different local teachers, the interpersonal and professional foundation required for more integrated models cannot realistically develop within a single academic year.

The second mechanism is the role identity that LETs internalized through administrative briefings and program design. Across all schools, LETs consistently described their functions in terms of linguistic support, classroom management, and technical facilitation---not co-instructional leadership. This self-positioning reflected a professional identity that conflicted with the role demands of alternative co-teaching configurations. As one LET at School E acknowledged, they did not feel it appropriate to interrupt the FET’s delivery, positioning themselves as a non-interruptive auxiliary figure. This role boundary, once established, effectively foreclosed the cognitive and organizational prerequisites for models requiring independent instructional authority from both partners (LET interview, School E).

The third mechanism was a stated pedagogical concern—that students’ limited English proficiency made independent engagement with FET-led instruction without LET linguistic mediation untenable. However, this rationale did not always align with observational evidence. At School E, notes recorded that approximately 80 percent of students were able to follow FET instructions without individual LET assistance during one speaking activity, and that no Chinese-language mediation was required for two consecutive demonstration rounds (Classroom observation, School E). This disjuncture suggests that the “students won’t understand” rationale may function more as much as a professional justification for low-burden role arrangement than as an accurate diagnostic of student need.

5.3. The Local English Teacher’s Functional Roles

Within the localized One Teaching, One Assisting model, LETs enacted three functional roles: linguistic translation, classroom management, and technical facilitation.

Classroom management was the most consistently observed LET function. LETs drew upon their established relationships with students as their English teachers to maintain order and attention during FET-led sessions and activities. Because the FET frequently lacked knowledge of individual students’ behavioral patterns and could not communicate with students in Chinese, the LET served as the primary disciplinary anchor. An administrator at School B acknowledged that early program implementations had encountered serious classroom management breakdowns when LETs failed to actively engage in managing the learning environment, including situations where students refused to participate or delivered inappropriate language to FETs without mediation or correction (Administrative interview, School B).

Linguistic translation formed the second core LET function. When students encountered difficulty understanding FET instructions—especially those involving unfamiliar vocabulary or longer utterances—LETs provided Chinese clarifications to enhance comprehension. Similarly, when students responded in Chinese due to language barriers, LETs translated student contributions for the FET. In one observed lesson, an LET immediately interpreted a student’s Chinese reference to a cultural landmark so that the FET could meaningfully engage with the student’s contribution (Classroom observation, School A). This real-time linguistic bridging was frequently reactive—triggered by student hesitation, blank expressions, or raised hands—rather than pre-planned.

Technical facilitation constituted the third functional role. Because FETs rotated across multiple classrooms and often lacked access to school-specific digital infrastructure—including Wi-Fi credentials, platform login details, and device obtained and charging logistics—LETs were responsible for pre-arranging the technological resources and on-site support required for FET-led activities. One LET in School A described managing software updates and group assignments on the school’s digital platform before each lesson, noting that the failure to complete this preparatory work would disrupt instructional flow and student engagement (LET interview, School A). Far from being incidental, this pre-lesson coordination represented a form of invisible labor that made communicative activities in FET-led teaching sessions operationally feasible.

Collectively, these three functional roles reveal that the LET’s contribution, though framed as “assistance,” constitutes a professional scaffolding function without which the collaborative teaching sessions would struggle to operate effectively. LETs served not as subordinate aides but as institutional bridges—sustaining instructional coherence across the linguistic, cultural, and logistical boundaries that the TFETP’s cross-cultural FET-LET pairing creates.

5.4. Structural Constraints on Model Diversification

The findings illuminate several interlocking structural factors that constrained model diversification beyond the One Teaching, One Assisting configuration.

Limited co-planning time was identified as the most pervasive structural barrier. Across all five schools, dedicated, scheduled co-planning sessions between FETs and LETs were absent. Coordination was reduced to informal hallway conversations, brief exchanges before class, or digital messaging—interactions insufficient to support the joint curriculum design, role negotiation, and shared pedagogical visioning that more integrated models require. One LET at School B described typically seeing the FET’s lesson materials—slides and activity sheets—during the short break immediately preceding the class, precluding any meaningful co-construction of instructional content (LET interview, School B).

Administrative workload compounded this constraint. The majority of LETs held concurrent homeroom responsibilities, managing student daily affairs, administrative paperwork, and examination logistics alongside their teaching duties. One administrator at School B noted that a formal co-planning initiative had once been implemented at the school but eventually discontinued because LETs’ competing responsibilities left them unable to use the allocated time productively (Administrative interview, School B). The three-monthly examination cycle in each semester further intensified this pressure in recurring waves, consuming the time and energy that might otherwise be devoted to collaborative planning and professional reflection.

A power asymmetry rooted in employment conditions added a further constraint. Because FET’s contract renewals were significantly influenced by school administration decisions, some FETs were observed to accommodate existing arrangements rather than advocate for alternative configurations. One tended to “go along with things” because the school held substantial influence over their continued employment. This institutional vulnerability created an incentive for FETs to conform to the LET’s and school’s preferred arrangement, reinforcing the persistence of the minimal-coordination One Teaching, One Assisting model.

5.5. Benefits and Challenges to Teacher’s Professional Development

Alongside these structural constraints, the study also reveals that collaborative teaching generates meaningful, if largely incidental, benefits for teachers' professional development—particularly in oral proficiency, pedagogical repertoire, and intercultural awareness.

5.5.1. Benefits

Teachers across school sites reported meaningful professional benefits arising from the collaborative arrangement, though these were often incidental rather than structurally designed.

For LETs, the most consistently noted was improved oral English proficiency. One LET from School E observed that his spoken English improved markedly after a year of daily interaction with the FET, noting that prior to this collaboration, his last sustained contact with a foreign speaker had been during university (LET interview, School E). The LET also described adapting his instructional approach after observing the FET’s use of body language, classroom commands, and oral assessment—techniques he subsequently incorporated into his own lessons.

For FETs, the collaborative context offered exposure to Taiwanese pedagogical norms and institutional culture that would have been inaccessible without co-teacher mediation. The FET from School C described how a senior LET had informally mentored during the first year—demonstrating classroom techniques within shared lessons, providing feedback, and offering memorable guidance and encouragement (FET interview, School C). The FET credited these interactions with reshaping his curriculum design from a theme-based approach toward one more systematically linked to the content his students were studying in their regular English classes and more relatable to everyday life.

5.5.2. Challenge

Despite these benefits, both FETs and LETs identified constraints that limited deeper professional growth.

The FET at School A expressed concern that two years of simplifying his language for students had weakened his English vocabulary (FET interview, School A). The FET noted that the communicative demands of his role required him to consistently reduce lexical complexity. The one exception was an IGP (gifted) class, with whom he had experimented with idiom instruction—an experience he described as personally motivating and consistent with her professional aspirations.

The LET at School A identified repetition as a structural limitation. She observed that the FET’s tendency to review prior content at the start of each session, while pedagogically intentional, could feel redundant for higher-ability students who were also encountering the same material in their regular English classes (LET interview, School A). This overlap between the FET’s review-oriented approach and the LET’s own instruction sometimes created a sense of low challenge for stronger learners.

For the LET at School B, the primary professional challenge was structural uncertainty at the outset. Having joined the school in August as a new teacher, she had no advance knowledge of the FET's curriculum sequence or instructional style. She identified this as the most significant thing she would change in retrospect—seeking earlier information from the academic director about the FET's teaching rhythm and expectations so that she could have been better prepared before the first class (FET interview, School B).

Beyond individual readiness, power dynamics also shaped FET professional agency in ways that constrained development. At School A, the administrator acknowledged that the FET tended to defer to existing institutional arrangements rather than advocate for preferred configurations—such as grouping students for differentiated activities—partly because contract renewal decisions rested with school administration. This institutional vulnerability discouraged professional risk-taking (Administrative interview, School A).

6. Discussion

The findings of this study illuminate a significant gap between the collaborative teaching aspiration embedded in TFETP policy and the instructional realities that emerge when those aspirations encounter the structural constraints of Taiwan’s junior high school context. This gap has both theoretical and practical implications.

Theoretically, the findings affirm and extend the existing literature on challenges in cross-cultural collaborative teaching. The role ambiguity, power imbalance, and limited planning time documented in earlier studies (Kim & Moddie, 2023; Luo 2013; Wang, 2022) are not only replicated in TFETP’s expanded iteration but are shown to be actively produced through the school-level arrangements that emerge in response to the program’s implementation—including one-to-many teacher pairing, Flexible Learning Periods curriculum scheduling, and unequal division of instructional responsibility. The one-too-many staffing arrangement, while administratively efficient, systematically prevents the relational continuity that meaningful collaboration requires. This finding responds to Luo’s (2014) call for inquiries into the development and implementation of viable co-teaching models in context, and to her recommendation that school-level administrative arrangements be taken into account in such development by demonstrating that the “context” in the TFETP setting operates not merely at the level of individual teacher preferences or cultural differences, but at the level of the institutional structures that determine what forms of collaboration are possible in the first place.

The persistence of the One Teaching, One Assisting model across all five sites—within examination-free Flexible Learning Periods—reveals that exam pressure alone is an insufficient explanation. The deeper driver is the intersection of role identity, relational infrequency, and institutional power asymmetry. This finding advances the field by demonstrating that model adoption in cross-cultural co-teaching is determined not primarily by pedagogical deliberation but by structural conditions that constrain the range of viable professional choices available to teachers.

From a language acquisition perspective, the pervasive reliance on Chinese mediation in the observed sessions also raises questions about pedagogical alignment with TFETP’s communicative goals. Drawing on principles of comprehensible input and communicative language teaching, optimal language acquisition environments are characterized by exposure to input that is slightly beyond learners’ current competence, supported by scaffolding strategies that preserve the target language as the primary medium of communication (Krashen, 1982). When systematic translation becomes the default response to student comprehension difficulty, the productive struggle through which language acquisition develops may be reduced. The study’s observations at School E—where FET-led communicative activities succeeded without Chinese mediation—suggested that alternative scaffolding approaches are pedagogically viable and could better serve TFETP’s communicative objectives (Classroom observation, School E).

At the policy level, the findings suggest that TFETP’s collaborative teaching mandate, while sound in principle, requires more deliberate structural support to function as intended. Mandating collaboration without providing the institutional conditions—dedicated planning time, manageable caseloads, clear role framework, and targeted professional development—risks producing compliance-driven arrangements that resemble collaboration superficially without achieving its professional or pedagogical benefits.

7. Conclusion

The study offers the first empirically grounded, qualitative examination of collaborative teaching under the expanded TFETP framework across multiple junior high school sites. The findings reveal that despite significant government investment and explicit collaborative teaching mandates, TFETP’s implementation is systematically shaped by institutional structures that constrain pedagogical integration. The localized One Teaching, One Assisting model that prevails across all five schools is not a failure of individual teachers but a predictable outcome of structural conditions: one-to-many pairing arrangements, absent co-planning infrastructure, compounded administrative workloads, and power asymmetries that discourage professional risk-taking—all of which collectively limit teachers’ capacity to experiment with more integrative collaborative teaching models that could better serve students’ communicative and linguistic learning needs. 

Based on these findings, this study offers five recommendations for strengthening collaborative teaching under TFETP. First, schools should establish structured, formally timetabled co-planning sessions for each FET-LET pair—time that is protected from competing administrative demands and embedded within the school calendar as a professional responsibility rather than a voluntary add-on. Second, administrative briefings should consciously reframe the LET’s role from linguistic support provider to co-instructional partner, providing language and conceptual tools for LETs to understand and perform more active and equitable roles in collaborative sessions. Third, the one-to-many staffing model should be reconsidered in favor of a smaller, stable pairing arrangement that allow relational continuity and gradual development of shared pedagogical practice. Fourth, targeted professional development programs should equip both FETs and LETs with the practical skills needed to implement alternative co-teaching configurations---not as prescriptions but as a professional repertoire from which teachers can select according to instructional contexts and student needs.

This study is subject to several limitations. The research is confined to junior high schools in a single city in central Taiwan; findings may differ in elementary school contexts, in other regions, or in private bilingual schools operating under different institutional arrangements. The data were collected in schools that participated voluntarily, potentially over-representing sites with relatively stable program implementation. Future research should extend the inquiry to elementary school settings, examine the perspective of students as program participants, and investigate whether professional development interventions targeting role recalibration and co-planning skills produce measurable shifts in collaborative teaching patterns over time.

The TFETP represents a significant and continuing commitment to bilingual English education in Taiwan. Realizing its collaborative teaching potential requires not only policy mandates but the structural conditions and professional scaffolding that allow teachers to move beyond functional compliance toward genuine pedagogical partnership.

Author Contributions: Conceptualization, H.H.W.; Methodology, H.H.W.; Investigation, H.H.W.; Formal Analysis, H.H.W.; Data Curation, H.H.W.; Writing – Original Draft Preparation, H.H.W.; Writing – Review & Editing, H.H.W. All authors have read and agreed to the published version of the manuscript.

Funding: This research received no external funding.

Conflicts of Interest: The author declare no conflict of interest.

Informed Consent Statement: All participants provided informed consent prior to their involvement in the study. Data collection procedures, including interviews and classroom observations, were conducted in accordance with the ethical guidelines of National Taiwan Normal University. Participant confidentiality was maintained throughout the research process through the use of anonymized identifiers.

Declaration of Generative AI and AI-assisted Technologies: This study has not used any generative AI tools or technologies in the preparation of this manuscript.

^[1] The Flexible Learning Periods are part of the school-developed curriculum under Taiwan’s 12-Year Basic Education Curriculum Guidelines. It grants schools autonomy to design flexible and context-responsive learning experiences based on students’ needs, school characteristics, and local contexts.

^[2] The Comprehensive Assessment Program (CAP) is a nationwide exam for ninth-grade students in Taiwan, typically held in May. English is one of the core subjects assessed. The results significantly influence students’ admission to senior high schools, making it an important academic milestone.