What is good design research in G3 Design Studies K376? It is the disciplined work of finding out what people actually need, what stops them, and which design choices could be justified by evidence. A polished poster or elegant room does not solve a research problem by itself. In Singapore's 2027 SEC Design Studies syllabus, students must identify requirements from a design brief, study users and stakeholders, select design strategies and explain the connection between research and a developed concept.
This independent eduKateSG Ministry of Education V3.0 Reference 380 focuses on Unit One: Design Research and Design Strategies for 2027 G3 Design Studies K376. It explains human/desirability, technology/feasibility and business/viability factors, neutral user investigation, source limitations, practical criteria, SWOT and genuine journal evidence. The official Design Project is 30 hours and carries 60% of the SEC grade; its journal includes a 12-mark Research and Design Strategies criterion.
The brief, interviews, survey numbers and situations used here are invented educational examples, not an assigned project brief, official examination question or actual research results. K376 project work is accepted only when carried out by the candidate in class under direct teacher supervision. This guide teaches how to reason about research; it does not provide findings, interview records or journal pages for a candidate to submit as authentic work.
The official 2027 K376 boundaries
| Component | Official arrangement | Why it matters |
| Paper 1: Written Examination | 2.5 hours, 40% overall; seven Section A, seven Section B and two Section C questions, all compulsory | Knowledge must be applied to scenarios and interpreted rather than only memorised |
| Paper 2: Individual Design Project | 30 hours, ten supervised sessions of three hours, 60% overall | Candidates select one of three official briefs and create original work under supervision |
| Design Journal | 60 of 100 project marks | Research/Strategies 12; Ideas/Concepts 12; Evaluation/Refinement 8; Detailed Sketches 10; Principles 18 |
| Digital Presentation and Oral | 28 and 12 of 100 project marks respectively | Final communication and oral explanation must trace decisions back to the genuine research |
SEAB lists Design Studies K376 as an Applied Subject for approved schools only, and its 2027 G3 catalogue states that 2028 is the last examination year for Applied Subjects. Reference 380 is an eduKateSG article number, not a new K380 syllabus. The examination subject remains K376 (formerly NP05).
What research needs to accomplish before sketching
- Translate the client's broad intention into an actual user problem without prescribing a solution too soon.
- Identify users, clients and operators separately, including who may be overlooked by an easy convenience sample.
- Collect only authorised, relevant information and distinguish observed facts from inference.
- Map a user journey, environmental conditions and essential constraints.
- Turn meaningful findings into explicit design criteria and strategies that can shape distinct concepts.
- Keep a transparent evidence trail and acknowledge limitations instead of inventing supporting data.
Twenty-eight research skills, with diagnostic repairs
1. Translate a brief into a problem
A design brief may combine wishes, requirements and assumptions. Rewrite its central problem without jumping to a predetermined product, room or building. Identify what the intended audience should be able to do when the design succeeds.
What can go wrong: A request for better museum orientation becomes 'make a touchscreen' before anyone checks whether a touchscreen is needed. Before correcting a whole design, ask the pupil which assumption introduced the problem. An error in defining a user need requires new questioning or research; an error in visual layout is a different stage. Treating both as 'needs more creativity' loses the reason.
Practice repair: Develop a neutral problem statement, such as how new visitors can find an exhibition independently, and test potential outputs against that purpose. Then give the pupil a fresh, deliberately fictional brief in another context. Have them identify the same research issue and describe how a strategy would change. Independent transfer is stronger evidence of learning than copying a polished research table.
2. Separate users from commissioners
The person funding or requesting a design is not necessarily the person using it. Staff who maintain it may have further constraints. Research can record all these perspectives without treating one as representative of the others.
What can go wrong: A sponsor wants a giant logo while visitors need a legible timetable. Before correcting a whole design, ask the pupil which assumption introduced the problem. An error in defining a user need requires new questioning or research; an error in visual layout is a different stage. Treating both as 'needs more creativity' loses the reason.
Practice repair: Map user, client, operator and affected third parties, identifying what each needs and where interests conflict. Then give the pupil a fresh, deliberately fictional brief in another context. Have them identify the same research issue and describe how a strategy would change. Independent transfer is stronger evidence of learning than copying a polished research table.
3. Distinguish needs from preferred solutions
A request for 'more screens' is a solution idea. A need might instead be to access reliable information without waiting for assistance. A researcher should ask what problem motivated the request.
What can go wrong: A client asks for a digital catalogue and the designer assumes paper is inherently inferior. Before correcting a whole design, ask the pupil which assumption introduced the problem. An error in defining a user need requires new questioning or research; an error in visual layout is a different stage. Treating both as 'needs more creativity' loses the reason.
Practice repair: Record the underlying task and compare digital, physical and hybrid approaches before commitment. Then give the pupil a fresh, deliberately fictional brief in another context. Have them identify the same research issue and describe how a strategy would change. Independent transfer is stronger evidence of learning than copying a polished research table.
4. Use contextual observations honestly
Observation can reveal the setting in which a problem occurs, but it does not reveal every person's motivations. An observation note should separate visible actions, times and conditions from interpretations.
What can go wrong: A visitor walks past a sign, so a pupil writes that the visitor found the sign confusing. Before correcting a whole design, ask the pupil which assumption introduced the problem. An error in defining a user need requires new questioning or research; an error in visual layout is a different stage. Treating both as 'needs more creativity' loses the reason.
Practice repair: Record what was actually seen and treat confusion as a hypothesis needing further evidence. Then give the pupil a fresh, deliberately fictional brief in another context. Have them identify the same research issue and describe how a strategy would change. Independent transfer is stronger evidence of learning than copying a polished research table.
5. Ask neutral questions
Research questions should not tell interviewees which answer is desirable. Open questions can reveal what a person attempted, where an obstacle arose and what alternatives they tried.
What can go wrong: 'Don't you prefer this cleaner layout?' elicits agreement that may reflect leading wording. Before correcting a whole design, ask the pupil which assumption introduced the problem. An error in defining a user need requires new questioning or research; an error in visual layout is a different stage. Treating both as 'needs more creativity' loses the reason.
Practice repair: Ask what happened in the user's recent experience and follow with a neutral question about difficulties. Then give the pupil a fresh, deliberately fictional brief in another context. Have them identify the same research issue and describe how a strategy would change. Independent transfer is stronger evidence of learning than copying a polished research table.
6. Understand interview limits
A short conversation can generate useful hypotheses, but its results may be highly specific to those individuals and their circumstances.
What can go wrong: Three friends say they like blue, and a project claims most teenagers prefer blue. Before correcting a whole design, ask the pupil which assumption introduced the problem. An error in defining a user need requires new questioning or research; an error in visual layout is a different stage. Treating both as 'needs more creativity' loses the reason.
Practice repair: Report the sample precisely and avoid population-level percentages that were never measured. Then give the pupil a fresh, deliberately fictional brief in another context. Have them identify the same research issue and describe how a strategy would change. Independent transfer is stronger evidence of learning than copying a polished research table.
7. Record consent and privacy
Research involving people, especially schoolchildren, should follow school rules, appropriate consent and data minimisation. Avoid publishing personal experiences that identify someone unnecessarily.
What can go wrong: A journal includes a named classmate's sensitive disclosure unrelated to the design decision. Before correcting a whole design, ask the pupil which assumption introduced the problem. An error in defining a user need requires new questioning or research; an error in visual layout is a different stage. Treating both as 'needs more creativity' loses the reason.
Practice repair: Use authorised, minimal records and anonymised examples where appropriate; consult the teacher supervisor. Then give the pupil a fresh, deliberately fictional brief in another context. Have them identify the same research issue and describe how a strategy would change. Independent transfer is stronger evidence of learning than copying a polished research table.
8. Define observational units
A count of actions is different from a count of people. A visitor can ask for directions twice, so 20 questions do not necessarily represent 20 different visitors.
What can go wrong: A desk logged 35 questions and the designer reports 35 confused people. Before correcting a whole design, ask the pupil which assumption introduced the problem. An error in defining a user need requires new questioning or research; an error in visual layout is a different stage. Treating both as 'needs more creativity' loses the reason.
Practice repair: Name what was measured and keep the unit consistent in the analysis. Then give the pupil a fresh, deliberately fictional brief in another context. Have them identify the same research issue and describe how a strategy would change. Independent transfer is stronger evidence of learning than copying a polished research table.
9. Recognise convenience sampling
People near the researcher's own classroom may not represent the entire user community. Research design should consider which groups are missing.
What can go wrong: Only confident English-speaking students are consulted about a multilingual sign. Before correcting a whole design, ask the pupil which assumption introduced the problem. An error in defining a user need requires new questioning or research; an error in visual layout is a different stage. Treating both as 'needs more creativity' loses the reason.
Practice repair: State the sample's limits and, where authorised, include perspectives relevant to the actual brief. Then give the pupil a fresh, deliberately fictional brief in another context. Have them identify the same research issue and describe how a strategy would change. Independent transfer is stronger evidence of learning than copying a polished research table.
10. Distinguish qualitative and quantitative evidence
A quotation explains one person's experience; a count describes a defined set of observations. Neither automatically proves causation or universal preference.
What can go wrong: One vivid comment is converted into an exact percentage of visitors. Before correcting a whole design, ask the pupil which assumption introduced the problem. An error in defining a user need requires new questioning or research; an error in visual layout is a different stage. Treating both as 'needs more creativity' loses the reason.
Practice repair: Label descriptions and counts separately and avoid converting one type into the other. Then give the pupil a fresh, deliberately fictional brief in another context. Have them identify the same research issue and describe how a strategy would change. Independent transfer is stronger evidence of learning than copying a polished research table.
11. Map the user journey
Many difficulties occur between steps rather than within a single product. Note entry, first decision, action, feedback, recovery and exit as relevant to the project.
What can go wrong: A digital check-in screen looks excellent but users cannot find it from the building entrance. Before correcting a whole design, ask the pupil which assumption introduced the problem. An error in defining a user need requires new questioning or research; an error in visual layout is a different stage. Treating both as 'needs more creativity' loses the reason.
Practice repair: Inspect the entire route to the task, not merely the attractive final interface. Then give the pupil a fresh, deliberately fictional brief in another context. Have them identify the same research issue and describe how a strategy would change. Independent transfer is stronger evidence of learning than copying a polished research table.
12. Build a task inventory
A user may need to find information, compare options, request assistance or complete a transaction. Different tasks imply different requirements and design criteria.
What can go wrong: The design optimises the appearance of a menu while omitting prices and opening hours. Before correcting a whole design, ask the pupil which assumption introduced the problem. An error in defining a user need requires new questioning or research; an error in visual layout is a different stage. Treating both as 'needs more creativity' loses the reason.
Practice repair: List critical user actions and identify what information each needs. Then give the pupil a fresh, deliberately fictional brief in another context. Have them identify the same research issue and describe how a strategy would change. Independent transfer is stronger evidence of learning than copying a polished research table.
13. Identify environmental conditions
Lighting, viewing distance, weather, noise, privacy and physical space can change whether a design functions. Contextual research should be proportional to the brief.
What can go wrong: A wall sign is assessed only at full brightness on a laptop screen. Before correcting a whole design, ask the pupil which assumption introduced the problem. An error in defining a user need requires new questioning or research; an error in visual layout is a different stage. Treating both as 'needs more creativity' loses the reason.
Practice repair: Evaluate the intended viewing environment before claiming legibility. Then give the pupil a fresh, deliberately fictional brief in another context. Have them identify the same research issue and describe how a strategy would change. Independent transfer is stronger evidence of learning than copying a polished research table.
14. Use secondary sources responsibly
Design precedents, published standards and official information can inform decisions, but the source's scope and date must be checked. A reference image is not automatically proof of performance.
What can go wrong: An online mood board is treated as evidence that a particular space meets access standards. Before correcting a whole design, ask the pupil which assumption introduced the problem. An error in defining a user need requires new questioning or research; an error in visual layout is a different stage. Treating both as 'needs more creativity' loses the reason.
Practice repair: Record provenance and distinguish inspiration, verified information and personal interpretation. Then give the pupil a fresh, deliberately fictional brief in another context. Have them identify the same research issue and describe how a strategy would change. Independent transfer is stronger evidence of learning than copying a polished research table.
15. Compare existing solutions fairly
A competitor analysis should identify features serving a similar user task and limitations under the current brief. Popularity alone does not establish suitability.
What can go wrong: The best-known app is copied because it has many downloads. Before correcting a whole design, ask the pupil which assumption introduced the problem. An error in defining a user need requires new questioning or research; an error in visual layout is a different stage. Treating both as 'needs more creativity' loses the reason.
Practice repair: Compare task completion, information hierarchy, access needs and relevant constraints rather than brand prestige. Then give the pupil a fresh, deliberately fictional brief in another context. Have them identify the same research issue and describe how a strategy would change. Independent transfer is stronger evidence of learning than copying a polished research table.
16. Write requirements as outcomes
A criterion such as 'users can identify the correct room quickly' is more actionable than 'make it cool'. Some outcomes need testing or qualified evidence.
What can go wrong: The strategy says 'modern, fresh, colourful' but cannot guide a design decision. Before correcting a whole design, ask the pupil which assumption introduced the problem. An error in defining a user need requires new questioning or research; an error in visual layout is a different stage. Treating both as 'needs more creativity' loses the reason.
Practice repair: State what success would mean and which observation or user review could evaluate it. Then give the pupil a fresh, deliberately fictional brief in another context. Have them identify the same research issue and describe how a strategy would change. Independent transfer is stronger evidence of learning than copying a polished research table.
17. Separate mandatory and negotiable constraints
Safety, accessibility, privacy or a fixed site condition may be non-negotiable; colour and decorative style may offer more flexibility. Do not let a weighted average rescue a violation of a fundamental requirement.
What can go wrong: A concept wins on novelty despite lacking a required exit route. Before correcting a whole design, ask the pupil which assumption introduced the problem. An error in defining a user need requires new questioning or research; an error in visual layout is a different stage. Treating both as 'needs more creativity' loses the reason.
Practice repair: Screen essential constraints first, then compare acceptable options on preferences. Then give the pupil a fresh, deliberately fictional brief in another context. Have them identify the same research issue and describe how a strategy would change. Independent transfer is stronger evidence of learning than copying a polished research table.
18. Understand desirability
The human factor asks whether the proposed design fits the user's situation and experience. It should not be reduced to personal liking or appearance.
What can go wrong: A viewer says the poster looks attractive but cannot find the registration link. Before correcting a whole design, ask the pupil which assumption introduced the problem. An error in defining a user need requires new questioning or research; an error in visual layout is a different stage. Treating both as 'needs more creativity' loses the reason.
Practice repair: Investigate the specific user task and what makes it understandable or convenient. Then give the pupil a fresh, deliberately fictional brief in another context. Have them identify the same research issue and describe how a strategy would change. Independent transfer is stronger evidence of learning than copying a polished research table.
19. Understand feasibility
Technology and implementation factors ask whether a concept can be produced and used under its conditions. A student need not certify engineering but must name assumptions honestly.
What can go wrong: A design requires a device or mounting system the fictional venue does not have. Before correcting a whole design, ask the pupil which assumption introduced the problem. An error in defining a user need requires new questioning or research; an error in visual layout is a different stage. Treating both as 'needs more creativity' loses the reason.
Practice repair: Identify equipment, output format, physical constraints and verification needs before ranking the concept. Then give the pupil a fresh, deliberately fictional brief in another context. Have them identify the same research issue and describe how a strategy would change. Independent transfer is stronger evidence of learning than copying a polished research table.
20. Understand viability
Business factors concern delivery, operating resources, maintenance and value in context. A school design may still have practical operating constraints even when it is not sold commercially.
What can go wrong: A pop-up exhibit needs expensive daily reassembly that staff cannot sustain. Before correcting a whole design, ask the pupil which assumption introduced the problem. An error in defining a user need requires new questioning or research; an error in visual layout is a different stage. Treating both as 'needs more creativity' loses the reason.
Practice repair: Include realistic workload, budget category and lifecycle implications as appropriate. Then give the pupil a fresh, deliberately fictional brief in another context. Have them identify the same research issue and describe how a strategy would change. Independent transfer is stronger evidence of learning than copying a polished research table.
21. Transform findings into strategies
Research is useful only if it changes what the designer chooses to do. Connect each significant finding to a criterion or specific strategy.
What can go wrong: A journal lists ten survey comments and then presents an unrelated final logo. Before correcting a whole design, ask the pupil which assumption introduced the problem. An error in defining a user need requires new questioning or research; an error in visual layout is a different stage. Treating both as 'needs more creativity' loses the reason.
Practice repair: Write a traceable chain from evidence to requirement to intended design response. Then give the pupil a fresh, deliberately fictional brief in another context. Have them identify the same research issue and describe how a strategy would change. Independent transfer is stronger evidence of learning than copying a polished research table.
22. Prioritise without false precision
A weighted comparison can support decisions, but arbitrary scores are judgments. Explain where weights originate and whether another weighting would change the choice.
What can go wrong: A score of 97.3 is presented as experimentally proven superiority with no test. Before correcting a whole design, ask the pupil which assumption introduced the problem. An error in defining a user need requires new questioning or research; an error in visual layout is a different stage. Treating both as 'needs more creativity' loses the reason.
Practice repair: Disclose that any scoring is illustrative unless an authentic evaluation has been conducted. Then give the pupil a fresh, deliberately fictional brief in another context. Have them identify the same research issue and describe how a strategy would change. Independent transfer is stronger evidence of learning than copying a polished research table.
23. Use SWOT for meaningful distinctions
Strengths and weaknesses relate to internal concept features; opportunities and threats involve external context. Each entry should affect a design decision.
What can go wrong: The matrix contains 'good, bad, nice, not nice' without connection to criteria. Before correcting a whole design, ask the pupil which assumption introduced the problem. An error in defining a user need requires new questioning or research; an error in visual layout is a different stage. Treating both as 'needs more creativity' loses the reason.
Practice repair: Describe an observed feature, its consequence and the refinement it suggests. Then give the pupil a fresh, deliberately fictional brief in another context. Have them identify the same research issue and describe how a strategy would change. Independent transfer is stronger evidence of learning than copying a polished research table.
24. Recognise causation limits
A before–after comparison may contain several changes. Improved outcomes do not prove the designer's favourite change was the cause.
What can go wrong: New labels and new staff instruction are added simultaneously; the improved wayfinding count is attributed to labels alone. Before correcting a whole design, ask the pupil which assumption introduced the problem. An error in defining a user need requires new questioning or research; an error in visual layout is a different stage. Treating both as 'needs more creativity' loses the reason.
Practice repair: Report what changed and identify alternative explanations rather than inventing experimental certainty. Then give the pupil a fresh, deliberately fictional brief in another context. Have them identify the same research issue and describe how a strategy would change. Independent transfer is stronger evidence of learning than copying a polished research table.
25. Plan fair evaluation questions
A research method should ask whether the solution serves the specified task in plausible conditions. Avoid questions designed to flatter a concept.
What can go wrong: Participants are asked whether the new sign is 'much better' rather than whether they can find a destination. Before correcting a whole design, ask the pupil which assumption introduced the problem. An error in defining a user need requires new questioning or research; an error in visual layout is a different stage. Treating both as 'needs more creativity' loses the reason.
Practice repair: Ask a neutral task question and record the outcome appropriately. Then give the pupil a fresh, deliberately fictional brief in another context. Have them identify the same research issue and describe how a strategy would change. Independent transfer is stronger evidence of learning than copying a polished research table.
26. Document what remains uncertain
A good strategy may acknowledge missing data, low sample coverage, untested materials or assumptions requiring specialist verification.
What can go wrong: A concept is declared accessible to all from a single sketch. Before correcting a whole design, ask the pupil which assumption introduced the problem. An error in defining a user need requires new questioning or research; an error in visual layout is a different stage. Treating both as 'needs more creativity' loses the reason.
Practice repair: State a clear limitation and the appropriate next form of investigation. Then give the pupil a fresh, deliberately fictional brief in another context. Have them identify the same research issue and describe how a strategy would change. Independent transfer is stronger evidence of learning than copying a polished research table.
27. Keep the journal authentic
K376 Paper 2 only accepts project work conducted under direct teacher supervision. Public examples cannot replace the candidate's own research record.
What can go wrong: An externally written interview transcript is inserted as if the pupil conducted it. Before correcting a whole design, ask the pupil which assumption introduced the problem. An error in defining a user need requires new questioning or research; an error in visual layout is a different stage. Treating both as 'needs more creativity' loses the reason.
Practice repair: Use teacher-approved research and preserve the provenance of actual notes and changes. Then give the pupil a fresh, deliberately fictional brief in another context. Have them identify the same research issue and describe how a strategy would change. Independent transfer is stronger evidence of learning than copying a polished research table.
28. Prepare for scenario questions
Paper 1 Sections B and C reward application and synthesis in unfamiliar circumstances. Students need to choose relevant research rather than repeat generic definitions.
What can go wrong: An exam scenario about a crowded exhibition is answered only with a definition of empathy. Before correcting a whole design, ask the pupil which assumption introduced the problem. An error in defining a user need requires new questioning or research; an error in visual layout is a different stage. Treating both as 'needs more creativity' loses the reason.
Practice repair: Name the observed problem, its user consequence and a testable design response. Then give the pupil a fresh, deliberately fictional brief in another context. Have them identify the same research issue and describe how a strategy would change. Independent transfer is stronger evidence of learning than copying a polished research table.
Original worked research exercise: three ways a library sign can fail
Imagine a fictional school library whose staff repeatedly direct visitors to the return desk. The head librarian asks for 'a brighter and more modern sign'. It is tempting to treat colour as the problem immediately. A research-led student instead asks where visitors enter, what they are looking for, what sign they first encounter, whether the sign is obscured, and how staff currently respond. These questions create a clearer account of the task before committing to any particular visual solution.
| Fictional observation | What it actually establishes | What it cannot establish |
| Six direction requests in one afternoon | Six requests were recorded during a stated period | That six different people were lost |
| Two visitors walked past the existing sign | Those visitors passed a location without stopping | That both were unable to read the sign |
| One respondent called the layout confusing | One person described their own experience | That every visitor shares the same difficulty |
| A revised layout was preferred in a sketch discussion | A subjective preference was expressed | That the revision improves navigation in the real space |
One research strategy could be to test whether first-time visitors can locate the return desk using the information presented at the entrance, with appropriate consent and supervision. Another could be to compare visibility at different viewing positions. The specific plan must be approved by the teacher for any actual candidate work. The fictional data above are for conceptual critique; they cannot be copied into a live project as if measured.
From research to strategies: the decision chain
A research note becomes useful when it alters the design criteria. If the established problem is that important information is hidden by a queue, a plausible strategy is to improve information placement or change the flow. If visitors struggle with unclear category names, the strategy may involve labels and hierarchy rather than a new screen. In each case, evidence informs a requirement, which then informs a design response. The chain should be visible in the learner's explanation.
| Identified issue (fictional) | Research-derived requirement | Possible strategy to explore | Evaluation question |
| Visitors cannot see the timetable behind a queue | Essential times visible at first decision | Separate queue and information area | Can a new visitor identify the current time and location? |
| A public message uses internal abbreviations | Language understandable to first-time readers | Plain-language labels and consistent terms | Can an unfamiliar reader explain the intended action? |
| A proposed screen requires unreliable connectivity | Information accessible under site conditions | Physical alternative or hybrid display | Can a visitor complete the task without a connected phone? |
| Several user groups have different priorities | Avoid privileging one group by default | Compare needs and maintain flexible alternatives | Which essential tasks remain unsupported? |
This is not a ready-made assessed research plan. It illustrates the kind of reasoning the 12-mark research and strategies criterion is designed to recognise: information that is relevant to the design brief and genuinely used. A large collection of images may look diligent while doing little to establish a strategy. By contrast, a smaller set of well-sourced and carefully interpreted findings can directly change the design.
Eighteen fictional research mistakes and how to repair them
Research case 1: The missing directions
A fictional school museum receives complaints that visitors cannot find the newest gallery. Two parents and one pupil describe different confusing junctions.
Better approach: Map specific decision points and investigate whether signs, routes or entrance information are responsible.
Evidence limit: A three-person convenience sample cannot establish a universal failure rate. A candidate should describe what their actual observation or source supports, identify the unanswered question and name an appropriate next investigation. An imagined participant quotation is not acceptable assessed evidence.
Transfer question: change the intended users or environmental constraints and repeat the analysis. Ask what the revised research should investigate before proposing a new visual or spatial feature. A design criterion should follow the reason for the decision, not an aesthetic preference disguised as fact.
Research case 2: The owner wants an app
A shop owner asks for an app because other shops have apps, but customers mostly need clear opening times.
Better approach: Separate the proposed technology from the underlying information need before comparing options.
Evidence limit: Market popularity is not proof the solution is desirable. A candidate should describe what their actual observation or source supports, identify the unanswered question and name an appropriate next investigation. An imagined participant quotation is not acceptable assessed evidence.
Transfer question: change the intended users or environmental constraints and repeat the analysis. Ask what the revised research should investigate before proposing a new visual or spatial feature. A design criterion should follow the reason for the decision, not an aesthetic preference disguised as fact.
Research case 3: Miscounted visitors
Reception records 40 direction requests, some made repeatedly by the same people.
Better approach: Report 40 requests rather than 40 unique confused visitors.
Evidence limit: The unit of measurement determines the claim's scope. A candidate should describe what their actual observation or source supports, identify the unanswered question and name an appropriate next investigation. An imagined participant quotation is not acceptable assessed evidence.
Transfer question: change the intended users or environmental constraints and repeat the analysis. Ask what the revised research should investigate before proposing a new visual or spatial feature. A design criterion should follow the reason for the decision, not an aesthetic preference disguised as fact.
Research case 4: The biased survey
A pupil asks, 'Isn't our colourful poster much better?' and gets positive responses.
Better approach: Use neutral questions and relevant tasks instead of encouraging agreement.
Evidence limit: Leading questions can change responses without changing user experience. A candidate should describe what their actual observation or source supports, identify the unanswered question and name an appropriate next investigation. An imagined participant quotation is not acceptable assessed evidence.
Transfer question: change the intended users or environmental constraints and repeat the analysis. Ask what the revised research should investigate before proposing a new visual or spatial feature. A design criterion should follow the reason for the decision, not an aesthetic preference disguised as fact.
Research case 5: An unreadable QR code
A large QR code is offered as the only route to information in a place with unreliable connectivity.
Better approach: Research device and connectivity assumptions and consider appropriate alternatives.
Evidence limit: A digital feature cannot substitute for the information it is supposed to provide. A candidate should describe what their actual observation or source supports, identify the unanswered question and name an appropriate next investigation. An imagined participant quotation is not acceptable assessed evidence.
Transfer question: change the intended users or environmental constraints and repeat the analysis. Ask what the revised research should investigate before proposing a new visual or spatial feature. A design criterion should follow the reason for the decision, not an aesthetic preference disguised as fact.
Research case 6: The glare problem
A sign is tested in a shaded classroom, then intended for a sunlit walkway.
Better approach: Identify the environmental difference before claiming comparable legibility.
Evidence limit: A successful indoor demonstration is not a test under all viewing conditions. A candidate should describe what their actual observation or source supports, identify the unanswered question and name an appropriate next investigation. An imagined participant quotation is not acceptable assessed evidence.
Transfer question: change the intended users or environmental constraints and repeat the analysis. Ask what the revised research should investigate before proposing a new visual or spatial feature. A design criterion should follow the reason for the decision, not an aesthetic preference disguised as fact.
Research case 7: One user's accessibility need
A researcher records a visitor's difficulty with small text and assumes every user has the same visual profile.
Better approach: Use the specific report to identify a possible requirement, not an exhaustive representation of people.
Evidence limit: An individual's perspective is important but cannot be generalised automatically. A candidate should describe what their actual observation or source supports, identify the unanswered question and name an appropriate next investigation. An imagined participant quotation is not acceptable assessed evidence.
Transfer question: change the intended users or environmental constraints and repeat the analysis. Ask what the revised research should investigate before proposing a new visual or spatial feature. A design criterion should follow the reason for the decision, not an aesthetic preference disguised as fact.
Research case 8: Wrong audience
An event flyer contains inside-school acronyms that the pupil designers understand but outside visitors do not.
Better approach: Rewrite the brief around first-time readers and unfamiliar terminology.
Evidence limit: A designer's familiarity is not the audience's familiarity. A candidate should describe what their actual observation or source supports, identify the unanswered question and name an appropriate next investigation. An imagined participant quotation is not acceptable assessed evidence.
Transfer question: change the intended users or environmental constraints and repeat the analysis. Ask what the revised research should investigate before proposing a new visual or spatial feature. A design criterion should follow the reason for the decision, not an aesthetic preference disguised as fact.
Research case 9: Too many colour preferences
A concept survey asks which of six colours is favourite, while the brief requires users to find the correct venue.
Better approach: Replace preference-only research with a meaningful information task.
Evidence limit: Attractiveness and practical navigation are not identical outcomes. A candidate should describe what their actual observation or source supports, identify the unanswered question and name an appropriate next investigation. An imagined participant quotation is not acceptable assessed evidence.
Transfer question: change the intended users or environmental constraints and repeat the analysis. Ask what the revised research should investigate before proposing a new visual or spatial feature. A design criterion should follow the reason for the decision, not an aesthetic preference disguised as fact.
Research case 10: The fake percentage
Two respondents prefer Concept A and one prefers Concept B; the report says 67% of Singaporeans prefer A.
Better approach: State the observed two-of-three split and severe sampling limitations.
Evidence limit: A percentage does not make a tiny nonrepresentative sample authoritative. A candidate should describe what their actual observation or source supports, identify the unanswered question and name an appropriate next investigation. An imagined participant quotation is not acceptable assessed evidence.
Transfer question: change the intended users or environmental constraints and repeat the analysis. Ask what the revised research should investigate before proposing a new visual or spatial feature. A design criterion should follow the reason for the decision, not an aesthetic preference disguised as fact.
Research case 11: Source without date
An online infographic makes a claim about public signage but has no stated source, method or date.
Better approach: Treat it as a lead to investigate, not established evidence.
Evidence limit: An elegant chart is not independently verified data. A candidate should describe what their actual observation or source supports, identify the unanswered question and name an appropriate next investigation. An imagined participant quotation is not acceptable assessed evidence.
Transfer question: change the intended users or environmental constraints and repeat the analysis. Ask what the revised research should investigate before proposing a new visual or spatial feature. A design criterion should follow the reason for the decision, not an aesthetic preference disguised as fact.
Research case 12: Photography provenance
A pupil inserts images from a social platform into a research board without identifying where they came from.
Better approach: Record source and lawful use, separating inspiration from original candidate work.
Evidence limit: Source acknowledgement does not automatically grant all usage rights. A candidate should describe what their actual observation or source supports, identify the unanswered question and name an appropriate next investigation. An imagined participant quotation is not acceptable assessed evidence.
Transfer question: change the intended users or environmental constraints and repeat the analysis. Ask what the revised research should investigate before proposing a new visual or spatial feature. A design criterion should follow the reason for the decision, not an aesthetic preference disguised as fact.
Research case 13: Competing staff needs
A ticket kiosk reduces queue time for visitors but creates complicated support requests for staff.
Better approach: Evaluate visitor desirability and operational viability together.
Evidence limit: Improving one metric does not establish whole-system success. A candidate should describe what their actual observation or source supports, identify the unanswered question and name an appropriate next investigation. An imagined participant quotation is not acceptable assessed evidence.
Transfer question: change the intended users or environmental constraints and repeat the analysis. Ask what the revised research should investigate before proposing a new visual or spatial feature. A design criterion should follow the reason for the decision, not an aesthetic preference disguised as fact.
Research case 14: Unrealistic selection matrix
A concept is rated 10/10 on novelty and 2/10 on essential accessibility, then wins through weighted averaging.
Better approach: Treat essential requirements as constraints before comparing acceptable concepts.
Evidence limit: Aesthetic preference cannot erase a fundamental usability condition. A candidate should describe what their actual observation or source supports, identify the unanswered question and name an appropriate next investigation. An imagined participant quotation is not acceptable assessed evidence.
Transfer question: change the intended users or environmental constraints and repeat the analysis. Ask what the revised research should investigate before proposing a new visual or spatial feature. A design criterion should follow the reason for the decision, not an aesthetic preference disguised as fact.
Research case 15: An unsupported conclusion
Visitors complete a wayfinding task faster after lighting and sign placement both change.
Better approach: Describe the improvement but do not attribute it solely to the new sign.
Evidence limit: More than one intervention prevents an isolated causal inference. A candidate should describe what their actual observation or source supports, identify the unanswered question and name an appropriate next investigation. An imagined participant quotation is not acceptable assessed evidence.
Transfer question: change the intended users or environmental constraints and repeat the analysis. Ask what the revised research should investigate before proposing a new visual or spatial feature. A design criterion should follow the reason for the decision, not an aesthetic preference disguised as fact.
Research case 16: Anonymised learning note
A student wants to use a classmate's personal difficulty verbatim and by name in the journal.
Better approach: Follow school consent and minimisation rules; record only what is appropriate and authorised.
Evidence limit: Educational value does not eliminate privacy obligations. A candidate should describe what their actual observation or source supports, identify the unanswered question and name an appropriate next investigation. An imagined participant quotation is not acceptable assessed evidence.
Transfer question: change the intended users or environmental constraints and repeat the analysis. Ask what the revised research should investigate before proposing a new visual or spatial feature. A design criterion should follow the reason for the decision, not an aesthetic preference disguised as fact.
Research case 17: Conflicting design evidence
A visitor likes the concept but struggles to find the call-to-action during a task.
Better approach: Record both the subjective preference and the observed task difficulty.
Evidence limit: Two different evidence types need not agree. A candidate should describe what their actual observation or source supports, identify the unanswered question and name an appropriate next investigation. An imagined participant quotation is not acceptable assessed evidence.
Transfer question: change the intended users or environmental constraints and repeat the analysis. Ask what the revised research should investigate before proposing a new visual or spatial feature. A design criterion should follow the reason for the decision, not an aesthetic preference disguised as fact.
Research case 18: The refined research question
A client says the display must feel premium, while users need to compare ticket tiers quickly.
Better approach: Define an information hierarchy criterion and treat premium appearance as a secondary design quality.
Evidence limit: Requirements can coexist but must be explicitly prioritised. A candidate should describe what their actual observation or source supports, identify the unanswered question and name an appropriate next investigation. An imagined participant quotation is not acceptable assessed evidence.
Transfer question: change the intended users or environmental constraints and repeat the analysis. Ask what the revised research should investigate before proposing a new visual or spatial feature. A design criterion should follow the reason for the decision, not an aesthetic preference disguised as fact.
How to distinguish strong research from excessive paperwork
The official journal rubric does not award marks for the sheer number of screenshots, interview forms or printed pages. Its higher levels place emphasis on meaningful research information and its use in developing design strategies. Students must therefore know why they collected something. A photograph of a reference poster might show a typographic treatment; without context it does not tell us that visitors understood the message. An interview quotation can reveal a difficulty; it cannot be extended automatically to an entire community.
In a fictional practice critique, a tutor can provide four research notes: one relevant visitor comment, one outdated source with no provenance, one observational count with a clear unit and one unrelated mood-board image. The learner should rank them by relevance, reliability and the questions they can actually answer. A useful design journal would not hide the limitations of each source. Any genuine candidate journal must be compiled only from the student's authorised supervised work.
A three-pupil research tutorial
Three fictional pupils receive the same non-assessed brief. Pupil A immediately proposes an app and needs help separating a solution from the problem. Pupil B interviews friends but generalises their preferences to all users. Pupil C gathers a large folder of precedents but cannot explain how any finding changes the concept. Their errors are not interchangeable. Each needs a different repair: reframing the task, sampling cautiously or extracting a design strategy from evidence.
The tutor rotates roles as user advocate, evidence examiner and feasibility critic. Each pupil must name a finding, its boundary and a possible design implication, then attempt another unfamiliar context independently. The tutorial should not produce fake research material for the real Paper 2 journal. Parents can ask which decision improved and whether the learner was able to apply it later, rather than expecting research marks from a polished template.
Twelve-stage research-to-strategy protocol
Stage 1: Read the real brief
Identify required output, people, context and non-negotiable conditions.
At each stage, the learner should be able to explain one error the action prevents, one piece of evidence needed and one conclusion that would exceed the available information. This makes the process usable in a new problem rather than a memorised list of headings.
Stage 2: State the problem without a solution
Describe the user difficulty before proposing an app, poster or building.
At each stage, the learner should be able to explain one error the action prevents, one piece of evidence needed and one conclusion that would exceed the available information. This makes the process usable in a new problem rather than a memorised list of headings.
Stage 3: Map the stakeholders
Separate user, client, operator and potentially overlooked groups.
At each stage, the learner should be able to explain one error the action prevents, one piece of evidence needed and one conclusion that would exceed the available information. This makes the process usable in a new problem rather than a memorised list of headings.
Stage 4: Choose ethical methods
Obtain the necessary supervision, minimise personal information and protect privacy.
At each stage, the learner should be able to explain one error the action prevents, one piece of evidence needed and one conclusion that would exceed the available information. This makes the process usable in a new problem rather than a memorised list of headings.
Stage 5: Define research questions
Ask what information would distinguish plausible explanations.
At each stage, the learner should be able to explain one error the action prevents, one piece of evidence needed and one conclusion that would exceed the available information. This makes the process usable in a new problem rather than a memorised list of headings.
Stage 6: Collect accurately
Record observations, participants and units without inventing values.
At each stage, the learner should be able to explain one error the action prevents, one piece of evidence needed and one conclusion that would exceed the available information. This makes the process usable in a new problem rather than a memorised list of headings.
Stage 7: Evaluate sources
Check date, provenance, relevance and what each observation can support.
At each stage, the learner should be able to explain one error the action prevents, one piece of evidence needed and one conclusion that would exceed the available information. This makes the process usable in a new problem rather than a memorised list of headings.
Stage 8: Identify patterns and limits
Describe recurring issues cautiously and avoid unsupported universal claims.
At each stage, the learner should be able to explain one error the action prevents, one piece of evidence needed and one conclusion that would exceed the available information. This makes the process usable in a new problem rather than a memorised list of headings.
Stage 9: Translate findings to criteria
Express user outcomes and essential constraints in actionable terms.
At each stage, the learner should be able to explain one error the action prevents, one piece of evidence needed and one conclusion that would exceed the available information. This makes the process usable in a new problem rather than a memorised list of headings.
Stage 10: Build design strategies
Choose promising directions linked to criteria, not only visual preferences.
At each stage, the learner should be able to explain one error the action prevents, one piece of evidence needed and one conclusion that would exceed the available information. This makes the process usable in a new problem rather than a memorised list of headings.
Stage 11: Compare trade-offs
Assess human, technology and business factors, disclosing assumptions.
At each stage, the learner should be able to explain one error the action prevents, one piece of evidence needed and one conclusion that would exceed the available information. This makes the process usable in a new problem rather than a memorised list of headings.
Stage 12: Test independent transfer
Apply the reasoning to a new fictional brief without reproducing a project journal.
At each stage, the learner should be able to explain one error the action prevents, one piece of evidence needed and one conclusion that would exceed the available information. This makes the process usable in a new problem rather than a memorised list of headings.
Four-week study sequence for Paper 1 and conceptual learning
- Week 1 — separate needs, wishes and solutions in several unfamiliar briefs; identify users and stakeholders.
- Week 2 — compare qualitative comments, observational counts and secondary research; state privacy and sampling limits.
- Week 3 — transform fictional findings into testable criteria and distinct design strategies across visual, interior and architectural examples.
- Week 4 — answer original scenario questions, explaining what is known, what remains uncertain and how research changes a concept.
Frequently asked questions
1. Is Reference 380 a new examination subject?
No. It is an article number; Design Studies remains G3 K376.
2. What part of the syllabus does it cover?
Unit One's Design Research and Design Strategies, with related design-thinking factors.
3. How much is the research rubric worth?
Twelve of the Design Journal's sixty project marks.
4. Can a student collect data from classmates without permission?
Any real research involving people must follow the school's supervision, consent and privacy rules.
5. Does a user interview prove the whole population prefers a concept?
No. Sample limitations and the nature of the evidence matter.
6. Should the design be chosen before research?
The brief and early ideas can guide investigation, but research should be able to challenge the first preferred solution.
7. Does the project need quantitative research?
Use methods appropriate to the approved brief; the syllabus calls for relevant research and its use, not arbitrary survey quotas.
8. What does human-centred design mean here?
Prioritising genuine user tasks and needs while checking technical feasibility and business viability.
9. Is SWOT an automatic scoring formula?
No. It helps analyse concept strengths, weaknesses and external conditions; criteria and judgment remain necessary.
10. May an AI generate authentic project interviews?
Invented interviews must not be presented as real candidate research. Assessed journal evidence must be authentic and supervised.
11. What does Paper 1 test?
Basic knowledge, application to scenarios, and interpretation and synthesis, across its three compulsory sections.
12. Is K376 available at all schools?
No. It is approved-school only, and SEAB states 2028 is the final Applied Subjects examination year.
Independent research exercises
Practice 1: The missing directions
Fictional challenge: A fictional school museum receives complaints that visitors cannot find the newest gallery. Two parents and one pupil describe different confusing junctions.
Answer-check route: Map specific decision points and investigate whether signs, routes or entrance information are responsible. Explain the difference between observed evidence and inference. Then identify one question the designer still needs to investigate before committing to a concept.
Practice 2: The owner wants an app
Fictional challenge: A shop owner asks for an app because other shops have apps, but customers mostly need clear opening times.
Answer-check route: Separate the proposed technology from the underlying information need before comparing options. Explain the difference between observed evidence and inference. Then identify one question the designer still needs to investigate before committing to a concept.
Practice 3: Miscounted visitors
Fictional challenge: Reception records 40 direction requests, some made repeatedly by the same people.
Answer-check route: Report 40 requests rather than 40 unique confused visitors. Explain the difference between observed evidence and inference. Then identify one question the designer still needs to investigate before committing to a concept.
Practice 4: The biased survey
Fictional challenge: A pupil asks, 'Isn't our colourful poster much better?' and gets positive responses.
Answer-check route: Use neutral questions and relevant tasks instead of encouraging agreement. Explain the difference between observed evidence and inference. Then identify one question the designer still needs to investigate before committing to a concept.
Practice 5: An unreadable QR code
Fictional challenge: A large QR code is offered as the only route to information in a place with unreliable connectivity.
Answer-check route: Research device and connectivity assumptions and consider appropriate alternatives. Explain the difference between observed evidence and inference. Then identify one question the designer still needs to investigate before committing to a concept.
Practice 6: The glare problem
Fictional challenge: A sign is tested in a shaded classroom, then intended for a sunlit walkway.
Answer-check route: Identify the environmental difference before claiming comparable legibility. Explain the difference between observed evidence and inference. Then identify one question the designer still needs to investigate before committing to a concept.
Practice 7: One user's accessibility need
Fictional challenge: A researcher records a visitor's difficulty with small text and assumes every user has the same visual profile.
Answer-check route: Use the specific report to identify a possible requirement, not an exhaustive representation of people. Explain the difference between observed evidence and inference. Then identify one question the designer still needs to investigate before committing to a concept.
Practice 8: Wrong audience
Fictional challenge: An event flyer contains inside-school acronyms that the pupil designers understand but outside visitors do not.
Answer-check route: Rewrite the brief around first-time readers and unfamiliar terminology. Explain the difference between observed evidence and inference. Then identify one question the designer still needs to investigate before committing to a concept.
Practice 9: Too many colour preferences
Fictional challenge: A concept survey asks which of six colours is favourite, while the brief requires users to find the correct venue.
Answer-check route: Replace preference-only research with a meaningful information task. Explain the difference between observed evidence and inference. Then identify one question the designer still needs to investigate before committing to a concept.
Practice 10: The fake percentage
Fictional challenge: Two respondents prefer Concept A and one prefers Concept B; the report says 67% of Singaporeans prefer A.
Answer-check route: State the observed two-of-three split and severe sampling limitations. Explain the difference between observed evidence and inference. Then identify one question the designer still needs to investigate before committing to a concept.
Practice 11: Source without date
Fictional challenge: An online infographic makes a claim about public signage but has no stated source, method or date.
Answer-check route: Treat it as a lead to investigate, not established evidence. Explain the difference between observed evidence and inference. Then identify one question the designer still needs to investigate before committing to a concept.
Practice 12: Photography provenance
Fictional challenge: A pupil inserts images from a social platform into a research board without identifying where they came from.
Answer-check route: Record source and lawful use, separating inspiration from original candidate work. Explain the difference between observed evidence and inference. Then identify one question the designer still needs to investigate before committing to a concept.
Official and connected reading
SEAB 2027 Design Studies K376 syllabus — research, strategies and rubric
SEAB G3 school-candidate list and Applied Subject entry conditions
Reference 376 — complete K376 Design Studies guide
Reference 377 — visual communication
Reference 379 — site analysis and architectural design
Immutable eduKateSG tutorial benchmark
Design research becomes meaningful when a learner can show why a particular user need matters, which evidence supports it and which design strategy follows. K376's research criterion is therefore not a request for attractive paperwork. It is a demand for accountable thinking that can guide authentic concepts, responsible decisions and clear explanations.