Editorial
The landscape of education and…
by Boon Lishi Lisa, Hoe Sion Hee, Shawn Neo Zhe Ming
Introduction
“What if the red ink scattered across a student’s science worksheet isn’t a bridge to learning, but a dead end?”
For years, traditional marking meant diagnosing past errors in isolation—telling students what went wrong without giving them the compass to navigate forward. In the Science Department at Naval Base Secondary School (NBSS), we realized that passive copying and repetitive error loops were holding our learners back. We needed to transform feedback from a terminal judgment into a formative dialogue that actively fuels scientific inquiry. Over a five-year journey, our department shifted from standard Assessment of Learning to empowering Assessment for Learning and Assessment as Learning. Today, by fusing grounded pedagogical frameworks with AI-enhanced feedback tools, we are exploring a new frontier: one where students take ownership of their growth, and teachers leverage technology to make learning visible, personal, and truly purposeful.
Welcome to our feedback horizon.
This article is structured as a layered journey — you are welcome to read it from start to finish, or to navigate directly to the sections most relevant to your context.
Section 1 sets the scene with the school and departmental backdrop that shaped our “why.”
Section 2 unpacks the pedagogical frameworks and how they evolved across five years of practice.
Section 3 traces the tools and data strategies we adopted at each phase.
Section 4 turns inward to explore how we grew as a professional learning community.
Section 5 offers honest reflections on our leadership journey.
If you are a classroom teacher curious about practical tools, you might begin with Sections 2 and 3. If you are a department head or school leader interested in building assessment culture, Sections 1, 4, and 5 may resonate most.
Wherever you choose to start, we hope each section offers something useful to your own feedback horizon.
Section 1 – Journey of Science feedback pedagogy
Naval Base Secondary School (NBSS) has been on a deliberate, school-wide journey to build an assessment culture that goes beyond marking and grades to genuinely improve how students learn. Anchored in a three-tiered framework, NBSS’s approach operates the tiers of Key Personnel as Assessment Leaders, Teachers as Assessment Innovators, and Teacher Leaders as Research Practitioners — all with student learning at the centre. This runs as a continuous cycle: identifying departmental needs and analysing current reality using artefacts, identifying assessment literacy gaps, and applying assessment knowledge to enact new practices that close those gaps.
Teachers gained increased clarity on AfL and applied the Feedback Cycle to improve Feed Up, with a sharper focus on what students do with the feedback to feed forward. This common assessment language — grounded in the Triangulated Model of AfL (Tan, 2025) and the Four Boxes of Feedback Literacy (Tan, 2022)— became the shared foundation from which departments developed contextualised practices for their subjects.
For the Science Department, this translated into a progression from establishing assessment baseline standards using the CKS approach in 2022, to strengthening those with Feedback Cover Sheets (Arts et al, 2021) in 2023, enhancing student receptivity to feedback in 2024, and incorporating AI into the feedback process by 2025. It is within this broader school context that the Science Department’s AfL journey took root.
The “Why”
Within the Science Department, a closer look at our own classroom practices revealed three critical gaps that made a fundamental shift in our assessment culture not just desirable, but necessary. We needed to move beyond Assessment of Learning (AoL)—treating exams as static, end-of-unit verdicts—and actively embed Assessment for Learning (AfL), where teachers use ongoing diagnostics to shape instruction. Crucially, this evolution culminates in Assessment as Learning (AaL), an approach that hands the reins to students, empowering them to self-diagnose gaps, reflect on their thinking, and actively drive their own scientific inquiry.
From “Task-level Feedback” to “Formative Growth” (The Shift to AfL)
Book and file checks revealed that traditional teacher marking was highly task-oriented and directly told the student what was needed to be done next. As seen in the marking artifact, heavy teacher red ink (e.g., “this is true but you’re not answering the question”, “shouldn’t be making such mistakes now!”) focuses entirely on diagnosing past errors rather than providing a forward-looking roadmap for improvement. Transitioning to AfL ensures that feedback actively closes the conceptual gap. Instead of pointing out terminal errors, feedback is structured around transparent success criteria and diagnostic questions that guide the student to construct the right scientific explanation themselves.
Fig. 1 Traditional Marking artifact
From “Passive Copying” to “Cognitive Interaction” (The Shift to AaL)
Teacher reflections highlighted that students were merely passively copying correct answers from the whiteboard during class review sessions, resulting in little to no genuine cognitive interaction with their mistakes. Since students did not have to process why their original answers failed, they lacked the metacognitive awareness required to internalize the corrections. Shifting to AaL positions the student as the critical connector between assessment and learning.
Breaking the Cycle of Recurrent Errors
The ultimate catalyst for this shift was the lack of opportunity to internalise the feedback in order to feed forward. Students who appeared to “fix” their corrections in class consistently failed to apply that feedback in subsequent assignments, repeating identical errors in both similar and new assessment contexts. Traditional summative grading (AoL) treats a correction as the end of a unit. Without an integrated AfL/AaL cycle, feedback does not result in cognitive changes, leaving students, especially those who are lower-readiness, trapped in repetitive error loops. Moving toward AfL and AaL ensures that feedback is treated as an ongoing loop. By utilizing data-informed interventions like differentiated worksheets and learning logs, the department ensures that students are explicitly trained to internalise feedback and able to feed forward effectively to new scientific problems.
The 5-Year Evolution
To resolve these traditional marking gaps, the NBSS Science Department shifted its practice over five years:
- Year 1 – 2022: Established the assessment baseline approach Content-Keywords-Support (CKS) strategy and modified AfL framework to change instructional strategies to incorporate baseline
approach in answering science questions and enforcing a feedback cycle.
- Year 2 – 2023: Launched a pedagogical pilot study with the National Institute of Education (NIE) on Feedback Partnership to study the effectiveness of the assessment baseline approach and trialled Feedback Cover Sheets (FBCS) to establish feedback dialogue with students and administered simplified Receptivity to Feedback (RIF) survey
- Year 3 – 2024: Enhanced the FBCS to increase students’ ownership of learning and administered surveys to study students’ RIF via FBCS
- Year 4 – 2025: Integrated Artificial Intelligence tools in the FBCS to provide immediate and personalised feedback
- Year 5 – 2026: Experiment with AI in Feedback Pedagogy to allow students to actively monitor their progress and become self-regulated learners
The five-year sequence was not accidental — it reflects a deliberate, inside-out logic of capacity building. The journey began with establishing a common language and baseline (2022), before layering on structures for dialogue and feedback ownership (2023–2024), and only then introducing technology as an accelerator (2025–2026). This sequencing was significant for two reasons.
First, it ensured that both teachers and students developed genuine assessment literacy before AI tools were introduced — meaning that when technology arrived, it amplified an already sound pedagogical foundation rather than substituting for one. Second, the progression mirrors the shift in agency: in the early years, the teacher held the primary role in diagnosing and directing; by the later years, students were increasingly positioned as active participants in their own feedback loops, capable of self-diagnosing gaps, initiating feedback dialogues, and feeding forward into new tasks.
Each phase built irreversibly on the last — the Feedback Cover Sheets could not have worked without the CKS baseline, and the AI-enhanced feedback loops could not have been meaningful without the years of feedback literacy that preceded them. In this sense, the sequence was not simply chronological but cumulative and intentional, designed to grow both teacher capacity and student self-regulation in tandem.
Section 2 – Targeted Feedback: Applying Pedagogical Concepts
Underpinning this five-year progression were two complementary pedagogical models that provided the department with a coherent theoretical spine — guiding how tasks were designed, how feedback was structured, and how student ownership of learning was progressively built.
The Triangulated Model of AfL (Tan, 2025) conceptualises assessment for learning as operating along two axes: task design as the horizontal axis, which situates learning across past and future desired outcomes, and standards as the vertical axis, which makes the learning gap visible. Together, these axes map the trajectory of a student’s growth, giving both teachers and students a shared reference point for where learning currently stands and where it needs to go.
Complementing this is the Four Boxes of Assessment Feedback Literacy (Tan, 2022) which shifts attention from feedback as a teacher act to feedback as a student responsibility. The four boxes ask: what teachers know about assessment (Box 1), what teachers do in terms of feedback actions (Box 2), what students do with the feedback they receive (Box 3), and what students ultimately get in terms of outcomes from acting on that feedback (Box 4). Crucially, the model redirects professional focus toward Boxes 3 and 4 — recognising that feedback only has value when students actively engage with and act upon it.
With these two models as the conceptual anchor, the department began translating theory into classroom practice.
Year 1 – 2022
To apply the triangulated model of AfL in our school context, we modified the AfL framework to change instructional strategies into three stages (pre-assessment stage, assessment stage, post assessment stage) to incorporate our assessment baseline approach (CKS) in answering science questions and enforcing a feedback cycle (Fig. 2).
Fig. 2 Modified AfL framework
Across the three instructional stages, teacher actions focus on transitioning from diagnostics to targeted task design, and finally to iterative feedback loops. Student actions transit from foundational application to independent self-assessment, and finally to collaborative peer evaluation and feeding forward. Hence developing assessment literacy in both teachers and students (Fig. 3).
Fig. 3 below, shows Lower Secondary Students were tasked to draw a cell based on a C-E-R graphic organizer. They used this to do self-assessment followed by peer assessment.
Year 2 & 3 – 2023 & 2024
To build department capacity, our instructional stages are aligned directly with the Four Boxes of Feedback Literacy. Following positive adoption of the baseline approach across the department, we institutionalized it within Feedback Cover Sheets (FBCS). As shown in Fig. 4, this strategy shifts ownership to students by requiring them to self-diagnose learning gaps, actively initiate feedback dialogues with teachers, and respond accordingly to teachers’ feedback. An artifact of the FBCS is shown in Fig. 5.
Fig. 4 Integration of the Four Boxes of Feedback Literacy within the AfL framework, featuring the embedded feedback request process.
Fig. 5 An artifact showing a chemistry student able to self-diagnose learning gap, initiate feedback dialogue with teacher and respond to the FB given by teacher.
Year 4 & 5 – 2025 & 2026
To provide immediate, personalized feedback, we integrated Artificial Intelligence tools within the FBCS process, leveraging three years of accumulated departmental assessment literacy as the foundation for designing tech-enabled lessons driven by AI-enhanced feedback pedagogy. Fig. 6 illustrates the use of AI-enhanced feedback pedagogy in designing a lesson and Fig. 7 shows the artifact of such a lesson.
Fig. 6 Use of AI-enhanced feedback pedagogy in designing a lesson.
Fig. 7 An artifact showing immediate and personalised feedback given by AI tool (SAFA), students diagnose their learning gaps, request for FB, and differentiated activities designed to close their learning gaps in chemistry
The progression across these three phases can be summarised in the table below, which captures how the role of the teacher, the student, and the tools evolved in tandem over five years.
What the table makes visible is a clear and deliberate progression across three phases. In Year 1, the priority was establishing a shared foundation — a common language and baseline approach that all teachers and students could work from. By Years 2 and 3, the focus shifted toward making feedback a two-way dialogue, with students taking on greater responsibility for identifying their own gaps and actively engaging with teacher feedback through the FBCS. The introduction of the RIF survey also marked a growing commitment to evidence-based refinement, allowing the department to track not just academic outcomes but students’ dispositions toward feedback itself. By Years 4 and 5, the department was ready to harness AI as a pedagogical tool — not as a shortcut, but as a means of providing the kind of immediate, personalised feedback that would have been impossible to scale manually. Across all three phases, the consistent thread is a gradual but intentional transfer of agency: from teacher to student, and from static feedback to a dynamic, self-regulated learning loop.
Section 3 – Choosing the Right Tools for Data-Informed Action (evidence-based approach)
Over the past four years, the Science Department has progressively transformed its pedagogical and feedback ecosystem. By giving teachers autonomy to explore diverse platforms—such as SALis, SAFA, AFA, Data Assistant, Mizou, and Sidekick—we transitioned from static, paper-based tracking to an agile, AI-enhanced feedback loop.
This evolution spans four distinct phases of tool adoption and data usage:
Phase 1 (2023): Foundation of Student Artefacts & Feedback Readiness
Before any new intervention could be meaningfully introduced, the departmentfirst needed to understand where students were starting from. Using student artefacts as the primary source of evidence alongside a simplified Receptivity to Feedback (RIF) survey, this phase focused on establishing a reliable baseline of student mindsets, dispositions, and overall openness to feedback. The deliberate choice to begin with artefacts rather than tools was significant — it grounded the department’s understanding of student readiness in actual classroom evidence rather than assumption. This baseline proved essential in shaping the targeted feedback strategies that would follow in subsequent phases.
Phase 2 (2024): Interactive Feedback Coversheets (FBCS)
Building on the baseline established in Phase 1, the department transitioned to Interactive Feedback Cover Sheets applied directly onto physical student work. The FBCS shifted the focus toward documenting specific learning difficulties on actual assignments, enabling teachers to check for student understanding and prompt students to summarise key insights from the feedback they received. Integrating the RIF survey with Weighted Assessment test scores also allowed the department to begin triangulating affective and academic data, providing a richer picture of student progress. While this phase successfully established initial dialogic feedback habits between teachers and students, real-time tracking remained limited by the manual nature of the process — a constraint that would be addressed in the phases that followed.
Phase 3 (2025): Expansion of AI Platforms & Structured Learning Logs
With dialogic feedback habits now established, the department was ready to scale its impact through technology. Structured Student Learning Logs (Fig. 8) were introduced alongside a suite of early AI tools — Student Learning Space (SLS) Short Answer Feedback Assistant (SAFA), SLS Data Assistant, Student AI Learning Space (SALis), and Mizou chatbot — each serving a distinct pedagogical purpose. Learning Logs helped students consolidate their mistakes and the AI’s suggested solutions, reducing the recurrence of errors while giving teachers visibility into student progress in real time. SAFA and SALis provided immediate, rubric-aligned feedback on short-answer questions, prompting students to log errors and transfer key concepts directly into their learning logs (Fig. 10a). The SLS Data Assistant automatically clustered misconceptions, freeing teachers from routine error-marking and enabling more targeted, differentiated intervention (Fig. 7). AI chatbots such as Sidekick and Mizou guided dialogic problem-solving without prematurely giving away answers, keeping students cognitively engaged (Fig. 9a). Data triangulation was also significantly expanded in this phase, drawing on the full RIF survey, Weighted Assessment scores, and Focus Group Discussions to cross-check and validate findings — allowing the department to make better-informed decisions about teaching and learning.
Fig. 8 Use of learning logs in LSS and Chemistry lesson
Phase 4 (2026): Advanced Multi-Tool Stacking & Prompt Engineering
In the most advanced phase of the department’s evolution, the focus shifted from using AI tools individually to stacking them strategically within a coherent feedback ecosystem. Sidekick chatbots, the SLS Annotated Feedback Assistant, and structured prompting frameworks such as COSTAR and Bloom’s Taxonomy were deployed in combination, with students trained to construct targeted prompts using cognitive keywords and to evaluate AI responses critically within their learning logs. The SLS Annotated Feedback Assistant enabled precise error annotation for chemical formulas and structural logic — a level of specificity that earlier tools could not provide. Comprehensive data triangulation across RIF surveys, Focus Group Discussions, and Weighted Assessment performance was maintained throughout. The cumulative impact of this phase was the establishment of a truly self-regulated feedback cycle, where AI serves as the immediate first layer of scaffolding while teachers leverage real-time data to provide deeper conceptual guidance and the relational support that technology alone cannot offer.
The progressive tool adoption and data-informed practices described in Section 3 did not emerge in isolation — they were the product of a sustained and deliberate investment in teacher professional development. Understanding how that culture of learning was built and sustained is key to appreciating why the Science Department’s practices were able to evolve so coherently over five years.
Section 4 – Growing Together: Cultivating Internal Professional Development
The Learning Culture
At NBSS, professional development is not treated as a series of isolated workshops but as an ongoing, structured commitment to building a school-wide culture of learning. School leaders play a critical role in this by intentionally resourcing and designing the conditions for teacher growth — bringing in external expertise such as Prof. Kelvin Tan to provide relevant assessment training, structuring macro-level approaches to school-wide PD, and crucially, protecting dedicated time and space for departments to do the deep work of translating frameworks into practice. Rather than managing through top-down compliance, school leaders built trust by empowering IP Heads with meaningful leadership responsibilities and demonstrating direct ownership by actively participating in PD sessions themselves.
It is within this enabling environment that the strategic integration of school-wide PD platforms and dedicated department white space became the foundation for the Science Department’s collaborative learning culture. This protected time and space empowers teachers to do more than attend sessions and receive information — it creates the conditions for genuine professional inquiry. Teachers are able to deeply internalise frameworks such as the modified Triangulated Model of AfL and the Four Boxes of Feedback Literacy, collaborate as a PLT unit to design lessons that integrate the department’s baseline approach with AI-enhanced pedagogy, routinely analyse student artefacts and performance data to evaluate the efficacy of their approaches, and engage in peer-led critique alongside structured self-reflection. All classroom experimentation is documented on the department’s Science PD Site (Fig. 13), turning professional development into actionable department plans that accumulate and build on one another over time.
Fig. 13 A snapshot of the synopsis of a Physics lesson in Science PD Site and peer reflection on the various PLT projects across the different science disciplines
Central to this culture of documentation and shared learning is the department’s Science PD Site, which functions as a living repository of the department’s collective professional knowledge. Rather than allowing insights from individual classroom experiments to remain siloed, the PD Site ensures that every lesson design, reflection, and iteration is captured and made accessible to all members of the department. Teachers document their lesson synopses, share annotated student artefacts, and record peer reflections on PLT projects across the different science disciplines — creating a growing archive that colleagues can draw on when designing their own lessons or embarking on new action research cycles. This is particularly valuable for teachers who are newer to a particular approach, as they can trace the reasoning and refinements behind a practice rather than starting from scratch. Beyond its archival function, the PD Site also serves as a platform for professional dialogue — colleagues can review one another’s documented experiments, build on what has worked, and collectively identify areas for further refinement. In this way, the PD Site transforms individual classroom experimentation into a shared departmental asset, ensuring that professional growth is cumulative, transparent, and owned by the whole team rather than residing in any single teacher.
The HOD & ST Partnership
Head of Departments (HODs) and Lead/Senior Teachers (L/STs) occupy distinct but complementary roles within the professional learning ecosystem. KPs such as Heads of Department are primarily responsible for setting strategic direction, building department culture, and aligning instructional practices with broader school goals. They operate at the level of systems and structures — designing the conditions under which good teaching can flourish. Lead and Senior Teachers, on the other hand, are specialists in pedagogy and practice. Their strength lies in their deep classroom expertise and their ability to translate theoretical frameworks into concrete, subject-specific routines that teachers can immediately apply. Where HODs set the compass, L/STs help colleagues navigate the terrain.
In the Science Department at NBSS, these two roles are deliberately leveraged in tandem to create a learning environment that is both strategically coherent and practically grounded. The HOD anchors the department’s assessment leadership by setting clear directions, aligning PLT units with a shared departmental purpose, and coaching team leaders toward greater autonomy. The ST complements this by serving as the bridge between theory and classroom reality — working alongside teachers in the white space to translate frameworks such as the Triangulated Model of AfL and the Four Boxes of Feedback Literacy into manageable, subject-specific routines. While the HOD normalises experimentation by openly trialling new assessment methods and sharing vulnerabilities from her own classroom, the ST drives the action research cycle — collecting perception and performance data, evaluating the efficacy of new practices objectively, and mentoring experienced teachers to step up and lead their own PLT units. Together, like the complementary strands of a DNA molecule, both roles co-exist and work in tandem to support professional learning and build a shared vocabulary around student self-regulation.
Multi-Platform Teacher Empowerment
As teachers gain confidence through this safe learning culture, they are empowered to scale their impact and share their classroom insights across multiple professional learning platforms. This outward sharing is not merely a performative exercise — it represents a significant milestone in a teacher’s professional journey. For many teachers, the prospect of presenting their classroom practice to peers, let alone to external audiences, can feel deeply exposing. It requires not only confidence in one’s practice but the ability to articulate the reasoning behind pedagogical decisions, defend findings with evidence, and engage critically with questions from fellow professionals. The fact that NBSS teachers are able to do this is itself a testament to the depth of assessment literacy and research rigour that has been cultivated within the department over five years.
Internally, teachers lead peer learning sessions, update the department’s Science PD Site, and cross-pollinate best practices across different PLT units during the annual World Café. This internal sharing is significant because it builds a culture where professional knowledge is treated as a collective resource rather than individual expertise — normalising the habit of learning from one another and ensuring that good practices spread organically across the department rather than remaining confined to individual classrooms.
Externally, empowered by their evidence-based action research data, teachers are supported to present their findings at LEP, Cluster workshops, AST, and the Singapore International Science Teachers Conference. This external reach matters for two reasons. First, it establishes NBSS teachers as credible and confident advocates of assessment leadership within the broader education fraternity — contributing to a field of practice that extends well beyond their own school. Second, and perhaps more importantly, it creates a virtuous cycle within the department itself: when teachers see their work valued and recognised beyond the school gates, it deepens their sense of professional ownership and reinforces the belief that what they are doing in their classrooms genuinely matters.
Section 5 – Conclusion & Reflection
Reflection on the Leadership Journey
Yet behind every structural innovation and pedagogical milestone lies a more personal story — one of leadership lessons learned, assumptions challenged, and approaches refined through honest reflection. It is to this dimension of the journey that we now turn.
2022 & 2023: Nurturing the Core (Growing the Learning Community):
The early years of the department’s assessment journey were characterised by energy and ambition, but also by the inevitable growing pains of building something new. The HOD’s primary focus was on standardising good assessment practices, cultivating a collaborative culture, and building a strong “can-do” spirit across the team. In hindsight, however, this phase also revealed an important leadership blind spot. The HOD’s heavy top-down direction of every PD session and follow-up action, while well-intentioned, created an unintended dependency — teachers grew accustomed to receiving ideas and direction from the HOD rather than developing the habit of thinking critically and creatively for themselves. Equally, the attempt to push all staff toward identical deliverable outcomes failed to account for the reality that teachers were at very different points in their professional readiness. These were not failures, but necessary lessons — ones that would fundamentally reshape the HOD’s leadership approach in the years that followed.
2024 & 2025: Shifting to Autonomy (Building Capacity):
Recognising the limitations of the previous approach, the HOD made a deliberate and significant shift in leadership style — moving from controlling outcomes to building long-term capacity. Rather than directing every step, the HOD focused on setting clear directions and aligning the various PLT units with a larger department purpose, while sharing structural responsibilities and heavily coaching team leaders to autonomously guide their respective units. This decentralised model proved to be far more sustainable than its predecessor. Teachers who had previously relied on the HOD for ideas began to develop their own professional voice, designing and leading action research projects with growing confidence. PLT leaders took ownership of their units’ directions, and the department’s collective capacity grew in ways that would not have been possible under a more centralised approach. The shift was not without its challenges — letting go of control requires trust, and building that trust takes time — but the results affirmed that sustainable department growth is ultimately built on empowered people, not directed compliance.
Support from School Leaders (SLs)
None of the Science Department’s achievements over these five years would have been possible without the intentional, multi-layered support of School Leadership. School leaders played a far more active role than simply approving plans and allocating resources — they were genuine partners in the department’s professional learning journey. By identifying and bringing in external experts such as Prof. Kelvin Tan, they ensured that the department’s PD was grounded in rigorous, relevant knowledge rather than generic training. By empowering IP Heads with meaningful leadership responsibilities rather than managing through compliance checks, they built the foundation of trust that made genuine experimentation possible. Crucially, school leaders demonstrated direct ownership by sitting in on PD sessions and engaging in active professional discourse — signalling to the entire department that assessment leadership was a school-wide priority, not just a departmental one. The SSD and ACs played an equally vital role as the nexus between the department, school leaders, and external consultants — driving sustainable ground implementation, proposing concrete next steps, and ensuring that the momentum of each PD milestone was carried forward rather than lost between sessions. Together, this layered support created the conditions in which the Science Department’s assessment culture could not only take root, but genuinely thrive.