Skip to content
Teaching and Education Research

An international forum for rigorous research on teaching, learning and teacher education

Research articleOpen access

Surface enactment of competency-based curriculum reform: Evidence from geography classrooms in Vietnam

Show affiliations and correspondence
  1. aHanoi National University of Education, Faculty of Geography, 136 Xuan Thuy, Hanoi, Viet Nam; University of Bayreuth, Didactics of Geography, Universitätsstraße 30, Bayreuth, 95447, Germany

DOI: https://doi.org/10.1016/j.tate.2026.105833

Outline

Abstract

Despite the global adoption of competency-based curriculum (CBC), a persistent gap exists between policy intentions and classroom practice. Empirical studies on CBC enactment are scarce, especially in geography education and Asian contexts. This study examines CBC enactment in Vietnamese geography classrooms through video analysis of 24 secondary-level lessons, integrating three frameworks: Curriculum Genre Theory, LUKAS competence-oriented task framework, and foundational CBC principles. The findings reveal a recurring pattern of surface enactment across the lessons analyzed, in which visible reform features coexist with unchanged core pedagogical structures. Across all analyzed dimensions, examination-oriented constraints steer instructional practice toward knowledge reproduction, contrary to reform objectives. The concept of surface enactment provides a theoretical lens to understand partial implementation as a coherent teacher response to conflicting systemic demands. These results suggest that substantive transformation requires addressing assessment culture alongside instructional practices, with significant implications for CBC implementation in centralized, exam-driven educational systems worldwide.

1. Introduction

Competency-based curriculum (CBC), now a leading global reform paradigm (OECD, 2018), reorients education from transmitting subject content toward students' demonstrated ability to apply integrated knowledge, skills, and attitudes in meaningful contexts. Despite global adoption, a significant gap persists between curricula and classroom practice. Systematic reviews consistently demonstrate that teachers encounter difficulties in translating competency frameworks into instructional change (Evans et al., 2020; Wawire et al., 2026) indicating the presence of systemic rather than solely context-specific barriers to reform. Vietnam's 2018 General Education Curriculum (GEC 2018) constitutes one of the most ambitious competency-based reforms in Southeast Asia, shifting the national curriculum from content transmission to the development of students' qualities and competencies (Duong & Nguyen, 2023; MOET, 2018a).

This transformation occurs within a complex implementation environment. Teachers' responses are shaped by centralized governance, Confucian traditions emphasizing teacher authority, and structural constraints such as high-stakes exams and limited professional development support (Pham et al., 2023; Tran & Le, 2025). Recent research identifies a pattern of partial transformation, in which surface-level changes coexist with deeper pedagogical continuities, raising questions about whether the adoption of new techniques reflects genuine competency-based transformation or merely superficial compliance (Ho & Dimmock, 2023). Examining how teachers enact reform requirements in the classroom, amid these competing demands, is critical to understanding CBC implementation in other centralized, exam-oriented systems.

Geography education provides a valuable perspective on these dynamics. The discipline emphasizes spatial thinking, systems analysis, and real-world problem-solving, which align closely with CBC principles (International Geographical Union Commission on Geographical Education, 2016). However, geography's educational value lies in a distinctive mode of reasoning that reforms may compromise if they reduce the subject to generic transferable skills (Lambert, 2014; Maude, 2016). In Vietnam, recent studies indicate that geography teachers often hold ambiguous understandings of spatial thinking (Nguyen et al., 2024) and that curriculum materials offer limited opportunities for argumentation (Nguyen & Budke, 2022), highlighting challenges at both the pedagogical and curricular levels.

Recent research has examined geography teachers' perceptions and Open access funding enabled and organized by Projekt DEAL. identified distinct profiles of reform engagement (Dang, 2026); however, no study has conducted a detailed, framework-guided analysis of their moment-to-moment instructional practices. This methodological gap is important because research consistently reveals discrepancies between teacher self-reports and observed practices (Koziol & Burns, 1986; Wiggs et al., 2023). To address this need for systematic classroom-level research, this study asks: How are CBC principles enacted in geography classrooms through teachers' pedagogical practices in Vietnam? To answer this question, the research uses systematic video analysis of 24 geography lessons and integrates three theoretical frameworks: Curriculum Genre Theory to analyze lesson structure; the LUKAS competence-oriented task framework to examine the cognitive potential of learning tasks; and foundational CBC principles to assess pedagogical alignment.

The study provides the first detailed analysis of geography teachers' CBC enactment in Vietnam, contributing to an understanding of subject-specific pedagogy and curriculum reform implementation in centralized educational systems.

2. Literature review

2.1. Global patterns in CBC implementation: The persistent gap between

policy and practice Over the past two decades, the global expansion of CBC has demonstrated a persistent trend: classroom practices remain largely unchanged despite substantial policy investment and reform discourse. The concept of “travelling policies” (Tahirsylaj & Sundberg, 2026) highlights how CBC reforms circulate internationally yet undergo significant recontextualization across national borders. Anderson-Levitt and Gardinier (2021) contend that these reforms are characterized by “polysemy, hybridity, and silences,” as the notion of competence shifts during adoption and adaptation. The persistent gap between policy intentions and classroom realities has been examined through various theoretical frameworks.

Nudzor's (2013) “implementation paradox” describes how substantial investment in policy development often yields limited practical change, as implementers adapt policies to established beliefs and practical constraints. From an institutional perspective, Meyer and Rowan (1977) explain this gap through “decoupling”: organizations adopt formal structures to demonstrate compliance with institutional expectations while continuing established operational practices. Consequently, CBC reforms are neither fully adopted nor entirely dismissed; they are filtered through teachers' existing pedagogical frameworks, resulting in hybrid practices that may appear reformed on the surface while maintaining traditional operations at their core.

Teacher capacity is frequently identified as a key mechanism behind this gap, yet systematic reviews indicate that the issue cannot be reduced to a “training deficit” (Evans et al., 2020; Wawire et al., 2026). Policy implementation is a cognitive process: teachers interpret reform messages through existing knowledge structures, so new policies are often assimilated into established practices rather than prompting genuine change (Spillane et al., 2002). Further barriers include the lack of implementation models, reluctance to relinquish control over learning, and the absence of shared definitions of competence (Evans et al., 2020). Assessment systems constitute a significant structural factor perpetuating this gap.

Alderson and Wall (1993) conceptualize this as the backwash effect (also termed washback): examinations determine the priorities of teachers and students, often irrespective of curricular intentions. Au's (2007) metasynthesis demonstrates that high-stakes testing consistently narrows the curriculum, as teachers abandon assignments requiring critical thinking in favor of test-aligned strategies. This tension between competency-oriented curricula and recall-oriented assessment is particularly evident in Asian contexts: Cave (2024) documents persistent conflicts between competency objectives and subject-content requirements in Japan, and Deng et al. (2024) found that Chinese teachers who conceptually supported CBC principles continued to teach “knowledge” rather than “competencies” under examination pressure.

In Confucian heritage cultures across East and Southeast Asia, hierarchical teacher-student relationships and collectivist classroom norms can conflict with the student-centered, inquiry-driven pedagogies that CBC promotes (Phuong-Mai et al., 2005). The relationship is not deterministic, however: Nguyen et al. (2006) found that Vietnamese teachers actively mediate between cultural expectations and reform requirements, producing hybrid practices, so cultural factors operate as constraints that shape implementation rather than as insurmountable barriers. Reform structures themselves may compound these challenges: So and Kang (2014) observe that South Korea's increasingly prescriptive national curricula have paradoxically “deskilled” teachers, eroding the professional judgment that CBC instruction requires.

Although contextual differences exist, intended competencyoriented curricula consistently diverge from classroom enactment. These studies share a methodological limitation: most address the gap through perception surveys, policy analyses, or interviews, with few directly investigating classroom practices (Evans et al., 2020; O'Donnell, 2008), even though only direct observation of enacted curricula can reveal the reality of classroom learning (Remillard & Heck, 2014) and self-reports repeatedly diverge from observed practices (Koziol & Burns, 1986; Wiggs et al., 2023). Without such analysis, efforts to support enactment remain speculative rather than evidence-based.

2.2. Subject-specific dimensions: The case of geography education

A further aspect remains insufficiently examined: how CBC reforms manifest within specific disciplinary contexts. Although competence is now embedded across curricula, Tahirsylaj and Sundberg (2026) find that subjects remain schooling's main organizing category. This gap matters because competence develops through engagement with distinct knowledge domains, each with its own epistemological structure and modes of knowing (Muller, 2009; Young, 2008). Geography offers a particularly instructive case. As a discipline bridging earth and social sciences, it integrates horizontal (contextual, applied) and vertical (hierarchical, theoretical) knowledge structures (Krause et al., 2025), and its educational value lies in cultivating “powerful geographical knowledge,” disciplinary knowledge that enables systematic analysis of spatial relationships and human-environment interactions (Lambert, 2014; Maude, 2016).

When reforms emphasize generic transferable skills over disciplinary content, this character is jeopardized. He et al.'s (2022) research in China exemplifies the tension: despite national reforms emphasizing core geographical competencies, significant gaps persist between policy-prescribed frameworks and teachers' ability to translate them into robust disciplinary knowledge. Despite these challenges, empirical research on CBC implementation in geography remains scarce. Whereas science and mathematics education use validated observation protocols such as RTOP (Sawada et al., 2002) and CLASS (Pianta et al., 2008), geography has not developed equivalent subject-specific instruments. Rather than proposing one, this study adapts established general frameworks for analyzing competence-oriented teaching and task design (Sections 3.1–3.3); because these were not devised for geography, they capture the structural features of competence-oriented practice but not, in themselves, the disciplinary substance specific to geographical reasoning, a limitation I take up in the discussion.

2.3. The Vietnamese context: An understudied case of rapid CBC reform

Vietnamese schooling has long been organized around content transmission and high-stakes examination within a centralized system shaped by Confucian heritage (Phuong-Mai et al., 2005). These conditions function as the institutional setting that teachers actively mediate rather than as fixed cultural determinants (Nguyen et al., 2006). Against this backdrop, GEC 2018, initiated by Resolution 29 of the Communist Party Central Committee (CPV, 2013), reoriented the national curriculum toward developing students' qualities and competencies, establishing explicit outcome standards aligned with global CBC frameworks (Duong & Nguyen, 2023; MOET, 2018a).

Implementation began in 2020, allowing approximately five years for initial adaptation patterns to emerge; however, comprehensive evaluation mechanisms remain underdeveloped (Duong & Nguyen, 2023). Within this reform framework, geography is integrated with history at lower secondary level (Grades 6–9) and taught separately at upper secondary level (Grades 10–12). The GEC 2018 mandates geography instruction to develop five qualities, general competencies, and three subject-specific competencies: geographical scientific cognition, geographical inquiry, and practical application (MOET, 2018b). Recent research on GEC 2018 implementation reveals a consistent pattern of partial transformation. Duong et al.'s (2025) longitudinal study documented ambivalence in interpreting competencies, with teachers shifting toward process-oriented approaches while remaining constrained by examination pressures.

Ho and Dimmock (2023) characterize Vietnamese schools as operating within a “performativity culture” that prioritizes visible compliance over substantive change, and Tran and Le (2025) found that teachers perceived assessment demands as burdensome rather than empowering. In geography, Dang (2026) identified three teacher profiles (passively adaptive, transformative-adaptive, and pioneering), indicating that engagement with CBC reflects the interaction of professional roles, reform-oriented beliefs, school culture, and training quality. Collectively, these studies highlight a paradox: policy grants teachers autonomy, yet insufficient preparation and support hinder meaningful change (cf. Ho & Dimmock, 2023).

These studies rely primarily on perception data, self-reports, and document analysis. Several have incorporated classroom observation, but as a supplementary source used to triangulate interview findings (Duong et al., 2025; Ho & Dimmock, 2023) rather than as the object of systematic analysis. Dang's (2026) typology offers nuanced insight into teacher orientations yet explicitly calls for classroom observation to establish whether and how these perceptions translate into practice. No study has yet conducted a detailed, framework-guided examination of geography teachers' enacted practices in Vietnamese classrooms.

3. Theoretical background

Curriculum enactment involves lesson organization, task design, and alignment with policy intentions. Because no single framework examines all three, this study integrates three perspectives on different properties of the same practice: Curriculum Genre Theory for lessonlevel organization, the LUKAS framework for learning task design, and five CBC principles for alignment with competency-based policy intentions.

3.1. Curriculum Genre Theory: Analyzing lesson-level organization

Curriculum Genre Theory (CGT), situated within the Systemic Functional Linguistics tradition (Martin & Rose, 2008; Rose, 2014, 2018), provides tools for analyzing the structure of pedagogic discourse in classroom contexts. Rose (2014) conceptualizes curriculum as realized through predictable configurations of meaning-making practices, termed “curriculum genres.” A curriculum genre integrates pedagogic and curriculum registers, with pedagogic registers comprising three interrelated dimensions: pedagogic relations (interpersonal), pedagogic activities (field), and pedagogic modalities (mode) (Rose, 2018). Pedagogic relations address interactional dynamics between teachers and students. Rose (2014, 2018) identifies two primary exchange types: knowledge exchange, in which information is negotiated between primary and secondary knowers (K1/K2), and action exchange, in which tasks are assigned and performed by primary and secondary actors (A1/A2).

Classroom knowledge exchanges often follow the delayed primary knower pattern (dK1^K2^K1): teachers pose display questions, students respond, and teachers evaluate. This structure, widely recognized as the Initiation-Response-Evaluation/Feedback (IRE/IRF) cycle (Sinclair & Coulthard, 1975), positions teachers as authoritative validators of knowledge, since “the key to pedagogic practice is continuous evaluation” (Bernstein, 2000, p. 42). The analysis considers both exchange types to determine how interactional control is distributed: whether students act as agents with decision-making rights or as respondents and executors within teacher-regulated sequences. Pedagogic activities refer to the structural organization of lessons through learning cycles consisting of core Focus, Task, and Evaluate phases, with optional Prepare and Elaborate phases (Rose, 2014, 2018).

The Task phase, where learners construct knowledge, is central. This dimension supports assessment of whether observed lessons align with the inquiry-oriented, student-centered structures promoted in CBC policy or retain transmission-oriented sequences. Pedagogic modalities encompass the semiotic resources through which knowledge is communicated, drawn from three main sources: environment, speaking, and records such as written texts, images, and diagrams (Rose, 2018). Because geography integrates diverse modalities, including maps, photographs, satellite imagery, and field data, this dimension is particularly relevant to how geographical knowledge is constructed in classrooms (B´eneker et al., 2024).

Beyond the pedagogic register, the framework includes the curriculum register, which concerns the knowledge and values exchanged in pedagogic discourse (Rose, 2018) and forms the fourth dimension in Fig. 1. It connects to debates about powerful knowledge (Lambert, 2014; Maude, 2016; Young, 2008), access to which CBC reforms may narrow by privileging generic skills. Because the cognitive quality of engagement with this knowledge is captured more precisely by task analysis, I operationalize it through the LUKAS type-of-knowledge and cognitive-process features (Section 3.2) rather than coding it separately. CGT is valuable here because it reveals the interactional structures and forms of classroom control that shape students' opportunities for agency, central to the research question.

However, it does not address the cognitive demands of learning tasks or evaluate alignment with policy intentions; these limitations motivate the additional frameworks introduced below.

3.2. LUKAS framework: Analyzing learning task design

The LUKAS framework (Lucerne model for the development of competency-enhancing task sets; Luthiger et al., 2014, 2018) analyzes the cognitive potential within individual task designs. It integrates research on task quality from subject didactics (Bruder, 2010; Büchter & Leuders, 2005), general didactics (Bl¨omeke et al., 2006; Maier et al., 2013), and instructional quality research (Hattie, 2013; Helmke, 2009). The framework identifies four competence-relevant feature areas, and its applicability to geography has been demonstrated by Reinfried (2016). The framework comprises ten features across four areas: Authenticity (competence representation, real-world connection), Cognition (work on pre-concepts, type of knowledge, and cognitive process, the last based on Anderson and Krathwohl's (2001) revised taxonomy), Complexity (task structuring, representation form), and Differentiation (task openness, learning support, variety of learning paths).

Each feature is scored on an ordered scale from lowest to highest competence-orientation, with level labels in Table 2 and boundary definitions and worked examples in Supplementary Material S1. Two features carry particular weight in geography: cognitive process, ranging from reproduction through near and far transfer to creative transfer, and representation form, given the subject's integration of textual, cartographic, and statistical representations (cf. Reinfried, 2016). Such task quality models offer valuable criteria for analyzing task design, but were not developed to evaluate the competence-relevant potential of tasks within specific disciplinary and policy contexts; the LUKAS framework addresses this limitation by focusing explicitly on competence-oriented features.

Although LUKAS originates in a European constructivist tradition, its features describe structural properties of tasks (the cognitive process required, the type of knowledge addressed, the authenticity of context) rather than culturally specific pedagogies, which makes them portable across national settings while leaving disciplinary substance unregistered, a limitation developed in Section 6.1. This study extends the framework from prescriptive task design to empirical analysis.

3.3. CBC principles: Assessing alignment with policy intentions

Neither CGT nor the LUKAS framework directly assesses whether observed practices align with competency-based intentions. Because CBC is defined variously across systems (Anderson-Levitt & Gardinier, 2021), the present study operationalizes it through five principles that recur across authoritative syntheses of the literature (Gervais, 2016; Le et al., 2014; Levine & Patrick, 2019; Sturgis & Casey, 2018; Wesselink et al., 2003) and are observable in classroom practice, using them as evaluative benchmarks. First, clear and measurable learning objectives require explicit communication of learning intentions, success criteria, and targeted competencies derived from analyses that integrate knowledge, skills, and attitudes within specific contexts, so that students understand what they are expected to learn and how progress will be assessed (Levine & Patrick, 2019; Spady, 1977; Sturgis et al., 2011; Wesselink et al., 2003).

Second, active student agency and ownership position students as knowledge constructors who collaborate, select learning strategies, and reflect, with teachers acting as facilitators rather than transmitters Fig. 1. Analytical framework. (Gervais, 2016; Johnstone & Soares, 2014; Sturgis & Casey, 2018). Third, student-centered instructional strategies address diverse preferences, paces, and pathways through approaches such as inquiry, collaborative discussion, and authentic problem-solving (DeLorenzo & Battino, 2010; Tahirsylaj & Sundberg, 2020). Fourth, competency-aligned and differentiated materials connect explicitly to targeted competencies, reflect students' prior knowledge and readiness, and support self-directed learning through structured scaffolding (DeLorenzo & Battino, 2010; Gervais, 2016).

Finally, authentic assessment engages students in performance-based tasks that demonstrate competency application rather than knowledge recall, linking classroom learning to real-world contexts (Albanese et al., 2010; Gervais, 2016). The three perspectives examine different properties of the same practice (the distribution of control, the cognitive demand of tasks, and alignment with policy intent); they can therefore return different verdicts on the same task and converge only at the level of the five findings (Fig. 1). This convergence is an empirical result rather than a design feature, as examined in the divergent-task-level cases in Section 5.3.

4. Methodology

A qualitative multiple-case study design was used to investigate geography teachers' classroom practices under CBC reform in Vietnam. This approach suits “how” questions about complex phenomena in realworld settings, and multiple cases enhance the transferability of findings (Yin, 2018). Teachers were recruited by convenience sampling with purposive attention to contextual diversity across urban and suburban Hanoi, and rural Ninh Binh; the distribution of lessons (16 urban, 2 suburban, 6 rural) reflects access rather than a stratified design. Each lesson was Table 1 Contextual characteristics of study sites. Urban (n = 6) Suburban (n = 1) Rural (n = 4) Schools Lower secondary 3 schools 1 school 1 school Upper secondary 2 schools - 3 schools Integrated (6–12) 1 school - - Observations Lessons observed 16 2 6 Typical class size 40–50 students (one exception: 25–30) 45–50 students 35–40 students (lower) 40–50 students (upper) Resources Display tech Dual blackboards, TV, projector Blackboard, TV/ projector Blackboard, TV/ projector Computing Computer Computer Computer Climate control Air conditioning Air conditioning None Student-owned devices Student smartphones with internet access - - Table 2 Distribution of 92 learning tasks across LUKAS framework features, by lesson phase.

Area Feature Orientation (n = 22) Acquisition (n = 44) Practice (n = 14) Application (n = 12) Total % Authenticity Competence representation Singular 20 34 11 8 73 79.3 Additive 2 10 3 4 19 20.7 Integrative 0 0 0 0 0 0.0 Real-world connection None 13 34 8 3 58 63.1 Constructed 6 10 6 5 27 29.3 Authentic 3 0 0 4 7 7.6 Real 0 0 0 0 0 0.0 Cognition Work on (pre-)concepts None 17 41 9 7 74 80.4 Implicit 4 3 5 5 17 18.5 Explicit 1 0 0 0 1 1.1 Reflective 0 0 0 0 0 0.0 Type of knowledge Facts 17 13 3 1 34 37.0 Procedural 0 3 4 0 7 7.6 Concepts 5 28 7 11 51 55.4 Metacognition 0 0 0 0 0 0.0 Cognitive process Reproduction 15 36 7 7 65 70.6 Near transfer 5 8 7 4 24 26.1 Far transfer 2 0 0 1 3 3.3 Creative transfer 0 0 0 0 0 0.0 Complexity Task structuring Highly structured 16 40 11 9 76 82.6 Partially structured 6 4 3 3 16 17.4 Unstructured 0 0 0 0 0 0.0 Representation form Singular 22 36 12 10 80 87.0 Integrating 0 6 2 2 10 10.9 Transforming 0 2 0 0 2 2.2 Differentiation Task openness Explained and closed 14 38 9 9 70 76.1 Free and closed 3 5 3 2 13 14.1 Explained and open 3 1 2 1 7 7.6 Free and open 2 0 0 0 2 2.2 Learning support None 17 35 8 10 70 76.1 Integrated 5 9 6 2 22 23.9 Feedback-oriented 0 0 0 0 0 0.0 Variety of learning paths None 20 41 12 9 82 89.1 Compensatory 2 3 2 3 10 10.9 Profile-building 0 0 0 0 0 0.0 Self-differentiating 0 0 0 0 0 0.0 Note.

Phases follow the Orientation–Acquisition–Practice–Application sequence used in Vietnamese geography lesson planning. Within each feature, levels are ordered from the lowest (top) to the highest (bottom) competence-orientation, following the LUKAS framework (Luthiger et al., 2014, 2018). N = 92. regularly scheduled and taught as usual practice, recorded with a single fixed camera while I observed as a nonparticipant. The final sample comprised 24 lessons taught by 18 geography teachers across 11 secondary schools in varied socioeconomic and geographical contexts (Table 1). The lessons, each lasting 45 min, were observed at lower secondary (n = 17) and upper secondary (n = 7) between December 2024 and February 2025.

Five teachers were observed in two or three lessons, and the remaining 13 in one. General teaching experience ranged from 2 to 16 years (median 12), concentrated in the 11to 15-year band; 15 teachers were women and 3 were men, all holding degrees in Geography Education. Because the GEC 2018 reached secondary geography through a phased rollout completed by the data-collection period, tenure under the new curriculum was uniformly short, making general experience the meaningful axis of variation. All had completed the mandatory MOET curriculum-transition training. To preserve anonymity, teacher characteristics are reported only in aggregate, not linked to specific lessons. Video-based observation was the primary data-collection method, following established guidelines for educational video research (Derry et al., 2010).

Each lesson was also documented through field notes emphasizing teaching and learning activities, classroom interactions, and instructional strategies. Lesson plans were collected from teachers, along with midor final-semester examination papers from the observed classes where available, yielding 29 papers (21 lower-secondary, 8 upper-secondary) with 486 items. Each item was coded according to the three LUKAS dimensions applicable to written assessment: cognitive process, type of knowledge, and real-world connection. Written informed consent was obtained from participating teachers and school administrators, and students gave written assent. All participants were informed that participation was voluntary and could be withdrawn at any time.

Consent was not sought from parents or guardians, as the study analyzed teaching practice rather than student performance and no individual student was assessed or reported on. All names were pseudonymized, no identifying information was reported, and recordings were held under restricted access and used only for analysis. All recordings were transcribed verbatim and annotated for nonverbal interaction, and the participating teachers reviewed their own transcripts for accuracy. Together with field notes, lesson plans, and examination items, the transcripts constituted the corpus for analysis. Qualitative content analysis (Kuckartz & R¨adiker, 2023) integrated deductive coding from the theoretical frameworks with inductive category development from the data, facilitated by MAXQDA.

The analysis comprised four stages: task identification, deductive coding, inductive coding, and interpretive synthesis (see Fig. 1). In the initial stage, transcripts and field notes were segmented into discrete learning tasks, the basic unit at which instruction is enacted (Doyle, 1983) and an established unit of analysis in geography education (Simon & Budke, 2020). Each task was bounded by the teacher's assignment of an activity and the subsequent evaluation and consolidation, the point at which the enacted task is defined (Doyle, 1983). Tasks typically corresponded to one lesson objective; where a teacher chained several display questions toward one consolidation, the chain was coded as a single task, since it yielded one evaluated product.

The demarcation rule, with worked examples, is set out in Supplementary Material S1. This segmentation identified 92 learning tasks across the 24 lessons. In the second stage, tasks were deductively coded using categories from CGT (Section 3.1) and the LUKAS framework (Section 3.2). Two further categories, learning objectives and assessment practices, were developed from the CBC principles literature to address aspects not covered by these frameworks. In the third stage, subcategories were identified inductively, capturing manifestations of teaching practice and interaction not anticipated by the deductive scheme. In the fourth stage, coded data were synthesized to determine how each of the five CBC principles was enacted in classroom practice.

Grade levels were analyzed within a single frame because the GEC 2018 sets one continuous competency standard across Grades 6–12, and the research question concerns enactment of that standard rather than developmental appropriateness as such. On that standard, reproductiononly tasks are not the developmental norm for any secondary grade, so the combined frame does not convert age-appropriate practice into apparent reform failure. The analysis coded the geography components throughout, including at lower secondary, where geography is taught within an integrated history-and-geography subject (MOET, 2018c). Fig. 1 presents the analytical framework, integrating theoretical perspectives and analytical stages.

I conducted all coding to maintain consistency. To enhance credibility (Lincoln & Guba, 1985), and following procedures for solo-researcher contexts (Kawamoto et al., 2023), two experienced geography education specialists reviewed the coding scheme, category definitions, and preliminary coded data across multiple rounds, with the scheme refined iteratively until agreement was reached on definitions and coding consistency. Because a single coder applied the scheme, no intercoder reliability statistic is reported. In its place, the coding manual (Supplementary Material S1) constitutes the audit trail, providing operational definitions for every category, level-boundary decision rules, and worked examples, including a case in which the three frameworks diverge.

As a geography educator within the Vietnamese system, I was aware that insider familiarity could lead me to read visible reform features as substantive change; I kept reflexive memos to record and check such judgments and grounded each decision in the manual's explicit definitions and boundary rules. I assumed sole responsibility for the final analytical interpretations.

5. Findings

Analysis of the 24 geography lessons identified a gap between CBC principles and enacted practice that recurred across five dimensions. The subsequent sections detail findings related to learning objectives, pedagogical interaction, strategies and learning tasks, learning materials, and assessment.

5.1. Objectives as display rather than dialogue

In all observed lessons, learning objectives were present but seldom functioned as genuine orienting tools. Objectives were briefly introduced at the start of the lesson and referenced only occasionally, with limited integration into pedagogical interactions. Two distinct patterns emerged.

5.1.1. Pattern 1: Objectives as content inventory

In the majority of lessons, objectives were presented during the opening phase as content announcements aligned with the textbook structure. Excerpt 1: Grade 10, February 6, 2025, Topic: World population distribution and urbanization. Teacher: In this lesson, we will have two main contents. First, we will learn about population distribution, including its current global distribution and the factors that influence it. And second, we will learn about the urbanization process, including its concept, influencing factors, and impacts on socio-economic development. Now let's move to the first content. (Students opened textbooks and notebooks and took notes attentively as the teacher wrote on the board.

After nearly one minute, the class moved on to a task in which students read the textbook and answered the teacher's questions.) Three aspects of this episode illustrate the gap between CBC principles and enacted practice. The teacher's statement functions as a textbook content inventory rather than a competency expectation, lacking action verbs and measurable criteria. No pedagogical action follows: the teacher moves directly to the content, so the objective serves as a procedural formality rather than a foundation for dialogue. The students' behaviors, opening textbooks and taking notes, reflect transmissionbased instruction. The objective announcement thus operates as an institutional symbol rather than a pedagogical tool.

5.1.2. Pattern 2: Objectives displayed but not operationalized

In a limited number of lessons, teachers articulated objectives using measurable action verbs, thereby formally aligning with CBC requirements. However, a consistent gap emerged between the stated objectives and their implementation. At a basic level, teachers employed verbs such as “describe” and “present” without specifying quality criteria, while classroom tasks required only recognition and recall. In contrast, some teachers explicitly clarified assessment criteria: Excerpt 2: Grade 10, January 10, 2025, Topic: World population distribution and urbanization. Teacher: Let's all observe the lesson objectives together. (Projected slides) Spend 1 min reading the lesson objectives. (After students read for 1 min).

Teacher: Notice the objectives I have bolded. First is to analyze the impact of natural and socio-economic factors on population distribution. Second is to present the concept and analyze factors affecting urbanization … With this lesson's objectives, please note the terms you confused in the midterm exam, which are “analyze and present.” When a task requires analysis, we not only describe but also need to provide reasons along with specific evidence to prove it (for example, data, charts, maps, or real-life examples). While at the presentation level, we can describe and interpret without necessarily needing specific evidence. (The lesson continued. Students were not asked to restate objectives or criteria in their own words.

During task completion, many students still did not provide specific evidence when performing "analysis" tasks.) This example demonstrates clear progress: the teacher allocates time for reading objectives, differentiates cognitive levels, and provides quality criteria. The process nonetheless remains monologic: students have no opportunity to discuss objectives, rephrase criteria in their own words, or self-assess, and their failure to provide evidence when “analyzing” indicates that criteria were transmitted but not internalized. Across both patterns, CBC objectives were implemented symbolically, possessing form but lacking substantive impact: they appeared in lesson plans and opening procedures but were absent from later instructional stages, not integrated into assessment, not used for student selfassessment, and not revisited at lesson close.

This reflects a formalistic understanding of objectives as elements to be written rather than resources to be used.

5.2. Pedagogical interaction: participation without agency

Students' roles were examined through two interaction types: knowledge exchange (negotiation of learning content) and action exchange (assignment and execution of tasks). In both, power and decision-making remained concentrated with teachers, while students functioned primarily as respondents and executors.

5.2.1. Address forms and interactional control

Classroom talk consistently opened with the address form “Con th ưa cˆo/th ầy" (literally, child respectfully addresses teacher), a routine feature recurring across the observed lessons. This convention is not in itself evidence of constrained agency; students' limited participation is better located in the interactional structure it accompanied, namely the teacher's control over who spoke, what counted as an acceptable contribution, and how exchanges were closed.

5.2.2. Knowledge exchange: Students as respondents

This concentration of control manifested in knowledge exchange through a dominant IRE/IRF pattern. Excerpt 3: Grade 9, February 7, 2025, Topic: Population of North central region. Teacher: Read section III in the textbook page 155, find information about population, population density, population distribution characteristics and ethnic composition of the North central region. (Students read for 1 min). Teacher: So, what is the population of this region? Student 1: Respectfully teacher, about 11.2 million people. Teacher: Thank you, sit down. Next, what is the population density? Student 2: Child respectfully reports to teacher, the population density is 218 people/km2. (The teacher continued calling on other students in turn to complete all content, confirmed answers, and wrote knowledge on the board; students took notes.) The teacher retains authority to initiate discourse, select respondents, and change topics, while “find” denotes “locate” rather than “discover,” marking a closed task with predetermined answers.

Questioning segments knowledge into discrete facts, each student supplying one item without integrative analysis; no “why” questions, connections between ideas, or counter-arguments arise. The classroom thus appears active while engagement remains performative, and the closing “Thank you, sit down” and writing on the board confirm that only teachersanctioned knowledge gains official status.

5.2.3. Action exchange: Students as executors

The same concentration of control extended to the regulation of actions. The central issue was not whether students participated but the extent of decision-making rights afforded to them. Excerpt 4: Grade 8, February 7, 2025, Topic: Vietnam's soil. Teacher: Now you will work in groups, the whole class is divided into 6 groups, every three tables form 1 group. Your tasks are as follows: read section 1 in the textbook pages 126-128, observe the map Figure 9.1 page 127, complete the tasks below in 5 min. Groups 1, 2: learn about characteristics of Feralit soil; Groups 3, 4: learn about distribution characteristics; Groups 5, 6: learn about use value. Have you grasped your tasks? (Silence. Teacher asked groups to begin.) (After 5 minutes, Group 2 presented.

Teacher asked for comments.) (Silence; no hands raised. Teacher invited Group 1, which agreed with Group 2.) Teacher: Everyone agrees? Nothing different? Then I'll comment, I agree with the characteristics, but I'll supplement about the explanation of this soil's color … (Teacher provided information, projected slides. Students took notes.) Three aspects marked this action exchange: task assignment was highly prescriptive, with group structure, content, sources, and timing predetermined; silence was the default response to invitations for opinion, converting voluntary engagement into obligation; and the teacher supplemented and finalized knowledge, so student contributions remained incomplete until validated.

This assignment-execution model dominated the observed lessons: students acted as “executors” rather than “designers” of learning. Across both exchange types, agency was enacted only at the surface. Participation increased without a transfer of power, as students could not initiate topics, co-design tasks, or set assessment criteria, so that altering interactional forms without addressing the underlying power structure left the traditional model intact.

5.3. Strategies and learning tasks: diverse in form, uniform in nature

Analysis of tasks across observed lessons indicates that teachers employed a range of strategies: structured questioning, open-ended questioning, group collaboration, quiz games, direct instruction, and skill practice. Although this variety suggests dynamic classrooms, cognitive-level analysis reveals a contrasting pattern: 65 of 92 tasks (70.6%) required reproduction, 24 (26.1%) involved near transfer, and only 3 (3.3%) reached far transfer; no task demanded creative transfer, a paradox of diverse strategies yet uniform cognitive demand. Table 2 reports the distribution of the 92 tasks across the ten LUKAS features. In eight features, no task reached the highest competence-oriented level; in two further features, the highest level appeared only as isolated cases (n ≤ 2).

Three dimensions, examined below, characterize this uniformity.

5.3.1. Dimension 1: Cognitive uniformity—Knowledge reproduction

Structured questioning constituted the largest proportion of tasks, yet all focused on knowledge reproduction rather than construction. As in the knowledge exchange pattern analyzed in Section 5.2.2, tasks were puzzles with predetermined answers: teachers posed display questions and students located and reported textbook information. No question required explanation (“Why is the density 218 people/km2?”), comparison and evaluation (“Is this density high or low compared to other regions?”), or application (“How does this density affect economic development?”). When combined with direct instruction and quiz games, which were likewise reproduction-oriented, the dominant strategies uniformly elicited recall.

Although concept-level knowledge predominated (51 tasks, 55.4%), these concepts were engaged reproductively, and no task reached the metacognitive level.

5.3.2. Dimension 2: Power uniformity—No choice

While structured questioning is characteristic of traditional teaching, group collaboration is widely regarded as learner-centered. Yet even within group work, students were not afforded choice, a fundamental component of ownership: topics were assigned by lesson content rather than student ability or interest, products were predetermined with no choice of expression, and students could not select their own research topics. In a Grade 7 lesson on African environments, students designated as “environmental experts” merely read textbooks and converted the information into mind maps; the appealing label did not change the task's reproductive nature. When combined with individual tasks assigned uniformly to the entire class, 70 of 92 tasks (76.1%) were “explained and closed,” prescribing both procedure and answer, and 76 (82.6%) were highly structured, affording students no decision-making authority over content, methods, or outcomes.

Across the three frameworks, the task diverges sharply. Against the CBC principle of student-centered strategy, the group format and the expert role register as reform features, yet the CGT analysis shows that control of the topic, product, and criteria remained with the teacher. A holistic observation would record only the visible feature, whereas the multi-framework reading shows it decoupled from cognition and from the distribution of control. The authentic scenarios in Section 5.3.3 show the same divergence on the meaning dimension.

5.3.3. Dimension 3: Meaning uniformity—Pseudo-authenticity

Tasks were equally constrained in meaning: even when teachers fostered authenticity through real-world situations, institutional structures often undermined these efforts. Excerpt 5: Grade 6, January 23, 2025, Topic: Earth's atmosphere, Air masses, Atmospheric pressure and winds. Teacher: Now I have two situations. Situation 1: Play the role of scientists, explain why airplanes usually fly at altitudes above 10 thousand meters? Find 2 reasons related to atmospheric characteristics. Situation 2: Play the role of environmental expert, explain how burning honeycomb coal on Hanoi streets is harmful? (Students answered situation 1: "crash into mountains," "crash into high-rise buildings," "avoid birds flying into engines," "avoid probe missiles." Teacher reminded: "Does anyone find reasons related to today's lesson?" One student mentioned thin air; no one answered the follow-up; teacher explained everything.) Teacher: Next, question 2, if we later have families, will we use the method of burning honeycomb coal?

Student: No, because honeycomb coal is extremely toxic, has smoke and dust, has carcinogens. Teacher: So in honeycomb coal there is carbon dioxide, released into the environment makes the atmosphere how? Whole class: Polluted. Teacher: Hanoi, the capital of peace, but we see Hanoi is one of the very polluted places … This task appears authentic, incorporating role-playing, real-world scenarios, and a local connection. However, two limitations diminished its authenticity. First, the atmospheric question presented a genuine geographical opportunity, since explaining why aircraft operate at altitude requires reasoning about the physical properties of the atmosphere. The task offered no scaffold through which students could undertake this reasoning, so they responded with speculation (“crash into mountains,” “avoid probe missiles”), and the teacher's instruction to “find reasons related to today's lesson” closed the exchange around compliance rather than inquiry.

Second, the honeycomb coal scenario raised a real human-environment relation, the effect of domestic fuel use on urban air quality, yet it was enacted as confirmation of a conclusion the teacher then supplied rather than as an analysis the students carried out. Of the 27 tasks (29.3%) using constructed scenarios, most operated at a hypothetical level without connection to students' lived experience; only 7 tasks (7.6%) reached the authentic level, and no task addressed a real problem. This produced meaning uniformity: despite teachers' attempts to foster authenticity, tasks failed to establish personal significance. Across the three dimensions, diverse strategies masked uniform practice: tasks remained closed in cognition (reproduction, not construction), in power (execution, not co-design), and in meaning (hypothetical situations lacking real-world significance).

When tasks adhere to the traditional model in which teachers control, students execute, and knowledge is reproduced, new strategies function as the surface enactment of student-centered pedagogy rather than its realization. This configuration was not confined to any single instructional moment. Even in the Application phase, where the curriculum framework anticipates competency demonstration, 7 of 12 tasks (58.3%) required only reproduction and only 4 (33.3%) carried an authentic real-world connection. Nor was it confined by level: reproduction was comparable at lower and upper secondary (69.8% and 72.4% of tasks).

5.4. Learning materials: diverse in type, monotonous in function

The observed learning materials were classified into four categories: text-based (textbooks, worksheets, tables), visual (maps, images, charts, mind maps), audiovisual (video clips, audio), and interactive tools (online resources, quiz applications). This diversity suggests wellresourced classrooms, but analysis of functional use reveals a contrasting pattern: most materials served illustration, fewer facilitated inquiry, and none supported self-directed learning.

5.4.1. Illustration instead of inquiry

Materials predominantly illustrated predetermined conclusions rather than supporting analysis or interpretation: teachers typically projected images, charts, or maps after delivering explanations, as confirmation rather than as evidence for reasoning. In a Grade 8 lesson on Vietnam's soil, the teacher projected a soil profile image and immediately explained it (“We can observe … this soil type has a red-yellow color … the reason is due to the accumulation of iron and aluminum oxides”), without inviting questions such as “What do you see?” or “What can you infer from this color?” Students remained passive, copying information when prompted; the image served as a decorative graphic rather than a cognitive tool.

In disciplinary terms, the soil profile offered a characteristic geographical opportunity, namely inferring an environmental formation process from observable visual evidence, but this opportunity was closed once the teacher supplied the explanation in advance. A contrasting, infrequent pattern showed materials used for inquiry rather than illustration: Excerpt 6: Grade 10, January 10, 2025, Topic: World population distribution and urbanization. Teacher: Listen to the following two pieces of music, and imagine the living scene occurring in these two pieces. Then share your imagination. (Teacher played audio 1: melodious, slow music with bamboo flute. Then audio 2: vibrant, fast music with car horns.

Replayed both.) Student 1: The first piece, I imagine a high mountain scene, not too many people. The second will be a neighborhood at night, sparkling with lights and bustling. Student 2: I imagine the first scene quiet and peaceful. Music piece 2 is bustling and noisy, vibrant but also makes people more tired. Student 3: I imagine people living in mountains will want to come to the city because it's too boring, and people in the city want to go to the mountains because it's too cramped. Teacher: You have shared very authentically about the difference between the two scenes … Those symbols help us relate to population distribution. This example demonstrates materials used for inquiry rather than illustration: the teacher played the audio without prior explanation, the open-ended prompt allowed diverse interpretations, and three students reasoned from auditory evidence through description, interpretation, and analysis, each acknowledged without being labeled correct or incorrect.

More broadly, where materials entered extended question sequences involving maps, videos, and data tables, more tasks reached near and far transfer. What made this episode geographically substantive, rather than only pedagogically open, was that students reasoned about the contrast between rural and urban settlement and about why people move between them, the kind of regional comparison and human-environment reasoning that the wider corpus rarely required. However, such instances were insufficient to change the overall trend.

5.4.2. Lack of differentiation and self-direction

The materials also lacked two essential CBC functions. None supported self-directed learning: there were no self-assessment checklists, rubrics for product quality, or step-by-step instructions for independent work. Nor were materials differentiated by student ability or learning style; all students engaged with identical resources in the same way. Material type thus varied while function remained constant, reflecting an underlying belief that knowledge requires transmission and illustration rather than discovery and construction, and leaving students reliant on teachers to determine “what I need to do” and “have I done well.”

5.5. Assessment: confirmation rather than learning support

Analysis of formative activities and summative tests reveals a parallel paradox: assessment occurred frequently but primarily confirmed learning rather than supported it. Formative assessment relied on oral questioning with praise or bonus points, rarely on product-based tasks. Summative assessment consisted exclusively of written tests, mainly multiple-choice questions. Judged against authentic assessment indicators (real-world context, performance tasks, transparent criteria, student selfand peer assessment, and descriptive feedback), most assessments lacked these characteristics.

5.5.1. Formative assessment: Confirmation rather than support

In most lessons a consistent pattern emerged: a student provided an answer, the teacher confirmed its correctness and awarded points, and the lesson proceeded without elaboration. In a Grade 6 lesson, after a student explained why the hottest air temperature occurs at 1 p.m. rather than noon, the teacher responded, “Very accurate. You absolutely deserve a score of 10,” then advanced to the next topic. Such feedback was evaluative rather than descriptive, leaving other students uncertain about the criteria for a strong response; no explicit scoring criteria were stated; and there was no opportunity for uptake through rephrasing, follow-up questions, or peer perspectives. A more advanced pattern appeared in a few lessons: Excerpt 7: Grade 8, January 23, 2025, Topic: Red River Delta and Mekong Delta Civilizations (After two groups presented on water regimes, teacher asked for peer comments) Student: Group 2's presentation was coherent, however, I think the group should add a summary section … Also, the images weren't very sharp.

Teacher: Basically, both groups performed quite well in content, complete, prepared on time. However, in structure, Group 2's was more explicit … Group 1 wrote quite generally. In format, Group 1 presented more attractive slides … regarding visuals, both groups had maps but only stopped at illustration without analysis. Group 1 had a chart about the water regime but didn't analyze it—if you had, the river water regime would have been extremely clear. (After feedback, teacher moved to next section. Groups had no opportunity to revise.) This example moves toward authentic assessment, integrating a product, peer assessment, and detailed teacher feedback, yet three limitations reduce its authenticity.

Criteria are referenced but not operationalized into rubrics: the teacher mentions content, structure, format, and visuals without level descriptors, leaving terms such as “coherent,” “attractive,” and “explicit” subjective. Feedback emphasizes form over academic content, focusing on whether slides are “attractive” or images “sharp” rather than on depth of understanding. And without a revision cycle, assessment serves as a final judgment instead of a tool for ongoing learning.

5.5.2. Summative assessment: Standardized but lacking authenticity

No alternative summative formats such as projects, portfolios, or presentations were used to capture the full range of CBC competencies. Across the 29 papers, the 486 items were largely selected-response (386 items, 79.4%, predominantly multiple-choice), with constructedresponse items accounting for the remaining 100 (20.6%) and rarely extending beyond short-answer formats. At the item level, the examination corpus closely paralleled the cognitive and authenticity profile of the enacted tasks (Table 3). Reproduction accounted for 392 of 486 items (80.7%), no item reached far or creative transfer, and 416 items (85.6%) carried no real-world connection. Reproduction was more pronounced at lower secondary (90.3% of items) than at upper secondary (68.7%).

Where items did establish a real-world connection, the form was authentic rather than constructed: the data typically concerned a named place and called for a calculation or an observation. One Grade 11 item, for example, supplied mean monthly temperatures for Hanoi and Yangon and asked students to find the difference in annual temperature range between the two cities. The contrast was recognizably geographic, yet the question required arithmetic subtraction rather than any account of why the two regimes differ; coded against the framework, it registered an authentic connection while remaining at near transfer and procedural knowledge. Constructed, hypothetical framings were almost entirely absent (2 of 486 items).

6. Discussion

6.1. Surface enactment as a system-wide configuration

The five dimensions reported above do not constitute five separate gaps between policy and practice. Instead, they combine into a single, multidimensional configuration in which the visible features of the reform coexist with the pedagogical routines that preceded it: objectives are stated but not operationalized; participation is active but constrained; strategies appear varied but call for reproductive work; materials illustrate rather than prompt inquiry; and assessment confirms knowledge rather than supports learning as it develops. Across the 92 tasks analyzed in 24 lessons spanning diverse institutional settings, this pattern recurred consistently enough to be read as system-wide rather than as variation across individual teacher repertoires.

I term this configuration surface enactment: the selective adoption of a reform's visible forms together with the retention of the cognitive and assessment routines that predate it, so that the appearance of change is separated from its pedagogical substance. Surface enactment shares conceptual ground with an established family of accounts of the policy-practice gap, but cannot be reduced to any one. Meyer and Rowan's (1977) decoupling describes organizations adopting formal structures to signal conformity while keeping everyday routines separate from them. Nudzor's (2013) implementation paradox names the disconnect between policy commitment and practical change. Tahirsylaj and Sundberg (2026) frame CBC as travelling policies that circulate across systems, and Anderson-Levitt and Gardinier (2021) describe the “polysemy, hybridity, and silences” such reforms acquire.

Ho and Dimmock (2023) place Vietnamese reform in a performativity culture that rewards visible compliance over genuine change. Bernstein's (2000) recontextualization explains how knowledge is selectively transformed as it moves from production into pedagogic reproduction. These accounts operate at the organizational or discursive level and leave unspecified where, in the act of teaching, the gap between form and substance is produced. Surface enactment adds precision at the level of practice: it identifies where in the instructional sequence reform forms and reproductive routines converge, and shows their coexistence to be not accidental but a consistent pattern across the pedagogical chain, from objectives to assessment.

It therefore does not compete with these accounts but operationalizes them for the classroom, making visible the task-level regularity they leave unspecified. Three interpretations of this pattern merit consideration: deliberate resistance to reform, a coherent adaptive response to conflicting systemic demands, or a transitional stage toward fuller implementation. The evidence aligns most closely with the second. The teachers I observed did not reject CBC features. They displayed objectives, used group work and varied materials, and conducted formative assessment regularly. But they paired these visible elements with the cognitive and assessment routines the examination system still rewards.

This pattern is better understood as professionals managing contradictory accountability signals (cf. Spillane et al., 2002) than as a refusal to change. The transitional-stage interpretation cannot be ruled out, given the cross-sectional design, and only longitudinal data could settle the question. Even so, the structural constraints identified here, namely assessment-driven incentives, institutional rewards for visible compliance, and limited subject-specific support, suggest the configuration is unlikely to shift on its own. Surface enactment is therefore better described as a stable equilibrium than as a preparatory stage. The geographical setting gives this configuration a distinctive character.

The analysis above specifies what, here, falls on the substance side of the gap. The operations distinguishing geographical thinking were available but seldom required. Only 2 of 92 tasks required moving across cartographic, textual, and statistical representations, an integration central to geographical reasoning (Reinfried, 2016). The deeper reasoning that regional comparison and human-environment analysis demand, registering as far transfer, was reached by only 3 tasks (3.3%) (Table 2). The data show not an absence of geographical content but the narrowing of geographical operations to reproducing given knowledge. This is why generic competency frameworks, including the one adapted here, can leave the subject dimension unexamined.

They record whether a task integrates representations or reaches transfer, not whether the concept is spatial or the reasoning geographical. That content-agnosticism allows disciplinary substance, rather than reform form, to be hollowed out while the visible markers of competence-oriented teaching remain. Surface enactment is therefore not only a matter of where the gap lies but of what it consists in, disciplinarily. The difficulty was not that the content was irrelevant but that the reasoning through which it acquires meaning was seldom enacted. Yet powerful geographical knowledge is valued partly for its reach beyond students' immediate experience (Lambert, 2014; Maude, 2016). That these operations were reached at all, and were curriculum-expected, indicates that their scarcity reflects the incentive structure rather than the limits of teacher capability.

This logic extends beyond geography to any subject whose distinctive affordances meet assessment backwash, examined in the next section.

6.2. Mechanisms driving surface enactment

Three interconnected mechanisms explain the emergence and persistence of surface enactment, operating across the structural, cultural-institutional, and disciplinary dimensions identified in this study. Assessment systems represent the most influential structural mechanism. Across the examination corpus, the cognitive and real-worldconnection profile of summative assessment closely paralleled that of the enacted classroom tasks and, on reproduction and decontextualization, exceeded it (Table 3). This convergence aligns with a backwash effect, whereby high-stakes examinations shape classroom practice (Alderson & Wall, 1993; Au, 2007): the assessed curriculum defines the real task (Doyle, 1983) and sets a reproductive floor that enacted practice only marginally exceeds.

Comparable dynamics are reported in Japan and China (Cave, 2024; Deng et al., 2024). The cross-sectional design does not permit a stronger causal claim, but the pattern is consistent with surface enactment operating as a system-level response to misaligned incentive structures rather than a teacher-level phenomenon. Cultural and institutional dynamics represent a second mechanism. Table 3 Cognitive, knowledge, and real-world-connection profiles of enacted learning tasks (N = 92) and summative examination items (N = 486). Feature Level Learning tasks Examination items N % N % Type of knowledge Facts 34 37.0 209 43.0 Procedural 7 7.6 68 14.0 Concepts 51 55.4 209 43.0 Metacognition 0 0.0 0 0.0 Cognitive process Reproduction 65 70.6 392 80.7 Near transfer 24 26.1 94 19.3 Far transfer 3 3.3 0 0.0 Creative transfer 0 0.0 0 0.0 Real-world connection None 58 63.1 416 85.6 Constructed 27 29.3 2 0.4 Authentic 7 7.6 68 14.0 Real 0 0.0 0 0.0 Note.

Task figures are drawn from Table 2; examination figures are based on item-level coding of 29 papers collected from the observed classes. Within each feature, levels are ordered from lowest (top) to highest (bottom) competenceorientation, following the LUKAS framework (Luthiger et al., 2014, 2018). The examination corpus comprised 21 lower-secondary (Grades 6–9) and 8 upper-secondary (Grades 10–12) papers and is weighted toward lower secondary, consistent with the lesson sample. Deferential address is a routine feature of the observed interaction, not in itself evidence of constrained agency. It limited the power redistribution CBC requires only in combination with the teacher-controlled questioning and examination pressure discussed above, which together kept exchanges under teacher direction.

This aligns with research on tensions between student-centered pedagogies and hierarchical teacher-student relationships (Phuong-Mai et al., 2005), and with Ho and Dimmock's (2023) account of a performativity culture. Yet cultural explanations alone are insufficient: the observed teachers attempted to incorporate student-centered features, and prior research documents genuine engagement with reform alongside persistent uncertainty (Dang, 2026; Duong et al., 2025). In Dang's (2026) typology, even “transformative-adaptive” teachers remained constrained by examination pressures and institutional expectations, indicating that willingness alone is inadequate without systemic support. Cultural factors thus act as constraints shaping implementation rather than insurmountable barriers.

Institutional environments that restrict pedagogical flexibility, what So and Kang (2014) call “deskilling” through prescriptive curricula, exacerbate them. The third mechanism is disciplinary. As argued in Section 6.1, the substance that surface enactment hollows out is subject-specific, yet teachers lacked established models of competency-based geography teaching. Roberts' (2017, 2023) conceptualization of geographical inquiry as the discipline's signature pedagogy offers a pathway, but it remained largely unrealized in the observed lessons, constrained by the same assessment pressures. This disciplinary support gap, distinct from the assessment and cultural-institutional mechanisms, indicates that CBC reform in geography depends on subject-specific pedagogical models that generic frameworks do not supply.

6.3. Implications, limitations, and future research

The findings present three implications. The first concerns assessment. In a centralized system, the central examination is the principal lever. The item-level analysis (Section 5.5.2) shows it is dominated by selected-response recall, with no item reaching far or creative transfer, setting a reproductive floor that classroom practice only marginally exceeds. A feasible reform within this governance structure would recompose the centrally set instruments, capping recall items and requiring a minimum share of items that demand transfer and real-world application, so that the central signal that teachers respond to rewards higher-order performance. Consistent with the hedged backwash argument above, this is a necessary condition for pedagogical change, not a guarantee.

The second concerns professional development. The most feasible approach is to repurpose structures already present in Vietnamese schools, namely school-based subject groups and lesson study, rather than import individualized models of reflective practice. These forums are currently dominated by evaluative, hierarchical observation that sorts rather than supports teachers (Tran et al., 2023). The implication is therefore not to build new infrastructure but to reorient existing forums toward shared, formative inquiry into competence-oriented task design. The third concerns subject-specific support. Because generic frameworks cannot supply disciplinary substance (Section 6.1), geography teachers require models grounded in the discipline's signature pedagogy.

Geographical inquiry (Roberts, 2017, 2023) offers one, equipping teachers as curriculum makers (Lambert & Biddulph, 2015) who build students' structured reasoning through maps, data, and fieldwork rather than applying generic techniques. Several limitations must be acknowledged. Convenience sampling, despite purposive efforts toward contextual diversity, restricts generalizability; the sample is weighted toward lower secondary (17 versus 7 lessons), and the focus on geography limits transferability. Most teachers were observed once, so for them, the data offer a snapshot that cannot distinguish a characteristic pattern from a single performance. The findings are therefore framed at the level of the tasks and lessons analyzed: their recurrence across 92 tasks in 24 lessons establishes breadth, while stability within a teacher's practice would require sustained observation.

For the five teachers observed more than once, whose lessons showed the same configuration, this offers limited support for within-teacher stability. Observation captures enacted practice, not the pre-lesson reasoning behind it; that reasoning is examined separately (Dang, 2026). These limitations point to several research directions. Longitudinal studies tracking teachers over time could determine whether surface enactment remains stable or evolves with sustained support, and intervention studies targeting pedagogical beliefs and assessment alignment could test whether such efforts move teachers beyond surface enactment. Comparative studies across subjects could clarify whether the pattern is specific to geography or reflects broader systemic tendencies, and studies pairing observation with pre-lesson reasoning within a single teacher sample could examine whether teachers' reasoning translates into their enacted practice.

Declaration of generative AI use

manuscript preparation process During the preparation of this work, the author used Claude AI to improve vocabulary and grammar. After using this tool, the author reviewed and edited the content as needed and takes full responsibility for the content of the published article. Funding sources This work was supported by the Open Access Publishing Fund of the University of Bayreuth.

Funding

Universität Bayreuth

Declaration of competing interest

The author reports that there is no conflict of interest.

Acknowledgements

I am deeply grateful to Gabriele Schrüfer (University of Bayreuth) for her scholarly guidance and critical feedback, which shaped the analytical framework and interpretation of findings in this study. I also thank Kati Barthmann (University of Bayreuth) for her methodological expertise during the coding and analysis process. Special thanks are extended to the 18 Vietnamese geography teachers who generously opened their classrooms across the 24 lessons observed for this research. I appreciate the anonymous reviewers for their constructive suggestions that improved this manuscript.

Data availability

Data will be made available on request.

References

  1. 1Albanese (2010). Building a competency-based curriculum: The agony and the ecstasy. Advances in Health Sciences Education, 15, 439. https://doi.org/10.1007/s10459-008-9118-2
  2. 2Alderson (1993). Does washback exist?. Applied Linguistics, 14, 115. https://doi.org/10.1093/applin/14.2.115
  3. 3(2001).
  4. 4Anderson-Levitt (2021). Introduction contextualising global flows of competency-based education: Polysemy, hybridity and silences. Comparative Education, 57, 1. https://doi.org/10.1080/03050068.2020.1852719
  5. 5Au (2007). High-stakes testing and curricular control: A qualitative metasynthesis. Educational Researcher, 36, 258. https://doi.org/10.3102/0013189X07306523
  6. 6Béneker (2024). Exploring 'Future three' curriculum scenarios in practice: Learning from the GeoCapabilities project. Curriculum Journal, 35, 396. https://doi.org/10.1002/curj.240
  7. 7Bernstein (2000).
  8. 8Blömeke (2006). Analyse der Qualität von Aufgaben aus didaktischer und fachlicher Sicht: Ein allgemeines Modell und seine exemplarische Umsetzung im Unterrichtsfach Mathematik. Unterrichtswissenschaft, 34, 330.
  9. 9Bruder (2010). Lernaufgaben im Mathematikunterricht.
  10. 10Büchter (2005).
  11. 11Cave (2024). School curriculum reform in contemporary Japan: Competencies, subjects, and the ambiguities of PISA. Comparative Education, 60, 278. https://doi.org/10.1080/03050068.2023.2208455
  12. 12Dang (2026). Understanding geography teachers' varied responses to competency-based curriculum reform in Vietnam: A qualitative typology. Teaching and Teacher Education, 172. https://doi.org/10.1016/j.tate.2025.105365
  13. 13DeLorenzo (2010).
  14. 14Deng (2024). ‘Pursuing competencies’ or ‘pursuing scores’? High school teachers' perceptions and practices of competency-based education reform in China. Teaching and Teacher Education, 141. https://doi.org/10.1016/j.tate.2024.104510
  15. 15Derry (2010). Conducting video research in the learning sciences: Guidance on selection, analysis, technology, and ethics. The Journal of the Learning Sciences, 19, 3. https://doi.org/10.1080/10508400903452884
  16. 16Doyle (1983). Academic work. Review of Educational Research, 53, 159. https://doi.org/10.3102/00346543053002159
  17. 17Duong (2023). Vietnam's basic education quality in the wake of standardization policy and national curriculum reform. Journal of Southeast Asian Economies, 39, S6. https://doi.org/10.1355/ae39-Sb
  18. 18Duong (2025). Complexities in teaching competencies: A longitudinal analysis of Vietnamese teachers' sensemaking and practices. Pedagogy, Culture & Society, 33, 349. https://doi.org/10.1080/14681366.2023.2262473
  19. 19Evans (2020). Making sense of K–12 competency‐based education: A systematic literature review of implementation and outcomes research from 2000 to 2019. The Journal of Competency-Based Education, 5. https://doi.org/10.1002/cbe2.1228
  20. 20Gervais (2016). The operational definition of competency‐based education. The Journal of Competency‐Based Education, 1, 98. https://doi.org/10.1002/cbe2.1011
  21. 21Hattie (2013).
  22. 22He (2022). Exploring the gap between academic geography and school geography: Knowledge transformation of the competencies-based curriculum making in China. Geography, 107, 128. https://doi.org/10.1080/00167487.2022.2114162
  23. 23Helmke (2009).
  24. 24Ho (2023). Changing teachers' beliefs and practices towards learner-centred education: Experiences and lessons from Vietnam's education system reforms. Practice, 5, 200. https://doi.org/10.1080/25783858.2023.2177191
  25. 25Johnstone (2014). Principles for developing competency-based education programs. Change: The Magazine of Higher Learning, 46, 12. https://doi.org/10.1080/00091383.2014.896705
  26. 26Kawamoto (2023). Proposal of a qualitative content analysis process for a solo researcher. Libri - International Journal of Libraries and Information Services, 73, 139.
  27. 27Koziol (1986). Teachers' accuracy in self-reporting about instructional practices using a focused self-report inventory. The Journal of Educational Research, 79, 205. https://doi.org/10.1080/00220671.1986.10885678
  28. 28Krause (2025). The barriers and enablers of curriculum thinking and teacher agency in geography education: A multinational study. International Research in Geographical & Environmental Education, 34, 220. https://doi.org/10.1080/10382046.2025.2466160
  29. 29Kuckartz (2023).
  30. 30Lambert (2014). Subject teachers in knowledge-led schools.
  31. 31Lambert (2015). The dialogic space offered by curriculum-making in the process of learning to teach, and the creation of a progressive knowledge-led curriculum. Asia-Pacific Journal of Teacher Education, 43, 210. https://doi.org/10.1080/1359866X.2014.934197
  32. 32Le (2014).
  33. 33Levine (2019).
  34. 34Lincoln (1985).
  35. 35Luthiger (2014). Entwicklung von kompetenzorientierten Aufgabensets: Prozessmodell und Kategoriensystem. Journal für LehrerInnenbildung, 14, 56.
  36. 36Luthiger, H., Wilhelm, M., Wespi, C., & Wildhirt, S. (Eds.). (2018). Kompetenzforderung mit Aufgabensets: Theorie-Konzept-Praxis. hep, der Bildungsverlag.
  37. 37Maier (2013). Allgemeindidaktische Kriterien für die Analyse von Aufgaben.
  38. 38Martin (2008).
  39. 39Maude (2016). What might powerful geographical knowledge look like?. Geography, 101, 70. https://doi.org/10.1080/00167487.2016.12093987
  40. 40Meyer (1977). Institutionalized organizations: Formal structure as myth and ceremony. American Journal of Sociology, 83, 340. https://doi.org/10.1086/226550
  41. 41(2018).
  42. 42Muller (2009). Forms of knowledge and curriculum coherence. Journal of Education and Work, 22, 205. https://doi.org/10.1080/13639080902957905
  43. 43Nguyen (2024). Exploring geography teachers' perceptions towards spatial thinking in Vietnam. International Research in Geographical & Environmental Education, 33, 210. https://doi.org/10.1080/10382046.2023.2298071
  44. 44Nguyen (2006). Culturally appropriate pedagogy: The case of group learning in a Confucian heritage culture context. Intercultural Education, 17, 1. https://doi.org/10.1080/14675980500502172
  45. 45Nguyen (2022). Significance of argumentation tasks in Vietnamese geography textbooks following the competency-based curriculum reform. Journal of Curriculum and Teaching, 11, 28. https://doi.org/10.5430/jct.v11n5p28
  46. 46Nudzor (2013). Exploring the policy implementation paradox: Using the free compulsory universal basic education (fCUBE) policy in Ghana as an exemplar. International Journal of Qualitative Studies in Education, 26, 933. https://doi.org/10.1080/09518398.2012.705043
  47. 47O'Donnell (2008). Defining, conceptualizing, and measuring fidelity of implementation and its relationship to outcomes in K–12 curriculum intervention research. Review of Educational Research, 78, 33. https://doi.org/10.3102/0034654307313793
  48. 48Pham (2023). Curriculum reform in Vietnam: Primary teachers' views, experiences, and challenges. Education, 51, 440.
  49. 49Phuong-Mai (2005). Cooperative learning vs Confucian heritage culture's collectivism: Confrontation to reveal some cultural conflicts and mismatch. Asia Europe Journal, 3, 403. https://doi.org/10.1007/s10308-005-0008-4
  50. 50Pianta (2008). Classroom effects on children's achievement trajectories in elementary school. American Educational Research Journal, 45, 365. https://doi.org/10.3102/0002831207308230
  51. 51Reinfried (2016). Kompetenzorientierte Lernaufgaben–mehr als alter Wein in neuen Schläuchen. Geographie aktuell & Schule, 38, 4.
  52. 52Remillard (2014). Conceptualizing the curriculum enactment process in mathematics education. ZDM Mathematics Education, 46, 705. https://doi.org/10.1007/s11858-014-0600-4
  53. 53Roberts (2017). Geographical education is powerful if. Teaching Geography, 42, 6.
  54. 54Roberts (2023). Geography through enquiry: Approaches to teaching and learning in the secondary school (2nd ed.). Geographical Association.
  55. 55Rose (2014). Analysing pedagogic discourse: An approach from genre and register. Functional Linguistics, 1. https://doi.org/10.1186/s40554-014-0011-4
  56. 56Rose (2018). Languages of schooling: Embedding literacy learning with genre-based pedagogy. European Journal of Applied Linguistics, 6, 59. https://doi.org/10.1515/eujal-2017-0008
  57. 57Sawada (2002). Measuring reform practices in science and mathematics classrooms: The reformed teaching observation protocol. School Science and Mathematics, 102, 245. https://doi.org/10.1111/j.1949-8594.2002.tb17883.x
  58. 58Simon (2020). How geography textbook tasks promote comparison competency—An international analysis. Sustainability, 12. https://doi.org/10.3390/su12208344
  59. 59Sinclair (1975).
  60. 60So (2014). Curriculum reform in Korea: Issues and challenges for twenty-first century learning. The Asia-Pacific Education Researcher, 23, 795. https://doi.org/10.1007/s40299-013-0161-2
  61. 61Spady (1977). Competency-based education: A bandwagon in search of a definition. Educational Researcher, 6, 9. https://doi.org/10.3102/0013189X006001009
  62. 62Spillane (2002). Policy implementation and cognition: Reframing and refocusing implementation research. Review of Educational Research, 72, 387. https://doi.org/10.3102/00346543072003387
  63. 63Sturgis (2018). Designing for equity: Leveraging competency-based education to ensure all students succeed. iNACOL.
  64. 64Sturgis (2011). It's not a matter of time: Highlights from the 2011 competency-based learning summit. iNACOL.
  65. 65Tahirsylaj (2020). The unfinished business of defining competences for 21st century curricula—A systematic research review. Curriculum Perspectives, 40, 131. https://doi.org/10.1007/s41297-020-00112-6
  66. 66Tahirsylaj (2026). Five visions of competence-based education and curricula as travelling policies: A systematic research review 1997–2022. Journal of Curriculum Studies, 58, 1. https://doi.org/10.1080/00220272.2025.2492605
  67. 67Tran (2023). Sorting or supporting teachers? An exploration of the imbalanced role of classroom observation in the development and assessment of Vietnamese secondary schoolteachers. Practice, 5, 220. https://doi.org/10.1080/25783858.2023.2177187
  68. 68Tran (2025). Curriculum reform in Vietnam: Teacher autonomy and the reality of its implementation in classrooms. International Journal of Learning, Teaching and Educational Research, 24, 1. https://doi.org/10.26803/ijlter.24.1.1
  69. 69Wawire (2026). Competence-based curriculum implementation in Africa: A scoping review of pedagogical and assessment practices. Journal of Curriculum Studies, 58, 556. https://doi.org/10.1080/00220272.2025.2523429
  70. 70Wesselink (2003). Competence-based education: An example from vocational practice [Paper presentation].
  71. 71Wiggs (2023). Convergence between teacher self-report and school administrator observation ratings using the classroom strategies assessment system. Assessment for Effective Intervention, 48, 113. https://doi.org/10.1177/15345084221112858
  72. 72Yin (2018).
  73. 73Young (2008). From constructivism to realism in the sociology of the curriculum. Review of Research in Education, 32, 1. https://doi.org/10.3102/0091732X07308969