---
name: research-proposal-writing
description: >
  Builds an assessed undergraduate or honours research proposal: problem statement,
  aligned aim, objectives and research questions, a preliminary literature section
  that establishes a gap, method justified against the question, explicit scope and
  delimitation, ethics at proposal stage, a realistic timeline, and a contribution
  claim pitched at the degree level. Use for "help me write my research proposal",
  "what is the difference between an aim and an objective", "my objectives do not
  match my aim", "how do I justify my methodology", "how long should my proposal
  timeline be", "my proposal was rejected", "how do I write a problem statement".
category: 15 Academic University Research
ref: "15.04"
tier: 3
inherits: [K2, K3, K4, K5]
---

# Research Proposal Writing

## 1. One-line description

A method for writing a research proposal that is assessed on what proposals are actually assessed on, which is whether the project is coherent and can be done, not whether it promises an interesting result: a problem stated as a tension, an aim and objectives and questions that visibly add up, a method justified against the question, and a timeline that survives contact with an ethics committee.

## 2. What this skill is used for

**The research problem it solves.** Students write proposals as advertisements for a result they hope to get, and are then marked on something else entirely. Proposals are assessed on feasibility and internal coherence: whether the pieces fit, whether the method could answer the question, and whether the work can be completed. Five failures do most of the damage, and all five are visible to a marker in minutes. The problem statement is a topic with a sentence of context in front of it, containing no tension, so nothing in the proposal has anything to be about. The objectives do not add up to the aim, which is the single most reliably penalised structural error in proposals and the one students are least aware of making. The literature section demonstrates reading rather than establishing a gap, so the reader finishes it not knowing why the study is needed. The methodology describes a method instead of justifying it against the question. And the timeline is a row of equal boxes that ignores ethics approval, recruitment and the fact that writing takes longer than everything else. This skill supplies an alignment matrix, a justification structure, and a backward-planned timeline.

**Where it sits in the research lifecycle.** After the topic is committed and the preliminary searching is done, before fieldwork. In most programmes it is the first formally assessed piece of the project and it functions as a gate: an approved proposal is permission to proceed.

**Typical use cases.**
- Writing a first research proposal to a departmental template.
- Repairing a proposal returned with "objectives not aligned" or "method not justified".
- Turning a committed topic and a preliminary literature search into a proposal document.
- Working out what an aim is, as distinct from an objective and from a research question.
- Producing a defensible timeline that accounts for ethics approval.
- Stating a contribution at the right level for the degree without overclaiming.
- Preparing for a proposal defence or supervisory approval meeting.

**Who uses it.** Undergraduate and honours students writing their first proposal, usually with a departmental template and little explanation of what the sections are for; supervisors reviewing proposals at volume; programme staff diagnosing why a cohort's proposals fail in the same way. Assume the template is fixed and the student's problem is knowing what to put in it.

## 3. When to use it

- A research proposal is due and you have a topic but no document.
- Your proposal has been returned with comments about alignment, coherence, justification or feasibility.
- You cannot articulate the difference between your aim, your objectives and your research questions, or you suspect they are the same thing written three times.
- Your literature section reads as a summary of what you have read and you cannot see how to make it argue for the study.
- You are writing a methodology section and can describe the method but not defend the choice.
- Your timeline is a set of equal-length blocks and you have not costed ethics approval.
- You are asked what your study will contribute and are either stuck or about to overclaim.
- You are preparing to present or defend the proposal.

## 4. When NOT to use it

- **The topic is not yet committed.** A proposal cannot be written around a moving question, and rewriting it after the question changes costs more than settling the question first. Go to **15.01 Academic Research Topic Selection** and commit, then return.
- **No preliminary searching has been done.** The literature section establishes a gap, and a gap cannot be established from impression. Run at least a first pass of **15.02 Academic Literature Search Strategy** first. A proposal claiming a gap that a marker disproves with one search is worse than a proposal that claims less.
- **This is a masters or doctoral proposal.** Those are longer, require explicit theoretical framing and a defended position within a paradigm, carry a much higher originality standard, and are frequently examined by people other than the supervisor. The structure here is the same shape but the depth and the contribution claim are not. Track 15B covers masters-level proposal work.
- **This is a systematic review protocol.** A protocol is a different document with different obligations: pre-specified eligibility criteria, outcomes, search strategy, screening design and analysis plan, written to a reporting standard and often registered before work begins. Use **15.06 Systematic Literature Review**.
- **This is a research ethics application.** Related, and not the same. Ethics applications are assessed on risk, consent, data handling, participant protection and withdrawal procedures, by a committee that is not marking your writing. This skill covers ethics at proposal level, which means identifying the issues and the approval route. It does not draft the application.
- **This is a grant or funding application.** Those are judged on value for money, applicant track record, impact beyond the academy and fit to a funder's priorities, none of which appear in an assessed student proposal. The commercial and applied analogue is **01.02 Business Problem to Research Question** with **01.04 Research Method Selection**.
- **The proposal is a formality and the project is already designed and approved.** Where the document merely records decisions already made and signed off, writing it as an argument wastes time the project needs.
- **You are asking for the proposal to be written for you.** See §12.1. This skill builds the matrix, interrogates the alignment, tests the timeline and critiques the draft. The prose is the student's, and the proposal defence assumes they wrote it.

## 5. Required inputs

**Required.**
- **The committed research question**, with its population, phenomenon and context.
- **The departmental proposal template or required section list, and the word count.** Proposals are marked against a structure, and writing a well-argued document in the wrong shape loses marks that have nothing to do with the research. If no template exists, ask for one; if there genuinely is none, use the structure in §9 and say so.
- **Preliminary literature**, meaning sources actually retrieved and read, sufficient to state what is known and what is not. Not a full review.
- **The submission date and the project end date**, because the timeline is a required element and it is built backwards from the second of these.

**Optional, and what each one adds.**
- **The marking rubric for the proposal specifically.** Proposal rubrics weight feasibility and coherence heavily and result-interest not at all, and seeing that in writing changes how a student allocates their effort.
- **Examples of approved proposals from the same programme.** The most efficient calibration available for length, depth and expected specificity, and usually obtainable by asking.
- **The institution's ethics categories and turnaround times.** Converts the ethics section from a paragraph of good intentions into a dated route, which is what the feasibility judgement actually rests on.
- **The supervisor's stated concerns.** If a supervisor has already said the scope is too large or the method is wrong, the proposal must answer that, and answering it explicitly is worth more than a stronger proposal that ignores it.
- **Any access already secured**, in writing. A named organisation that has agreed in writing converts the single largest feasibility risk into a resolved one, and it belongs in the proposal.
- **Whether the proposal is defended orally.** Changes what the document must carry: a defended proposal can leave some detail to the discussion, an undefended one cannot.

## 6. Questions to ask before starting

1. **What structure and word count does the department require?** Determines the shape and the depth of every section. Default if unanswered: use the structure in §9, state that it is a general structure, and flag that the departmental template must be obtained before submission.
2. **What is the proposal marked on, and in what weightings?** Determines effort allocation, and usually reveals that feasibility and coherence outweigh everything else. Default: assume feasibility, alignment and method justification carry the most weight, and say so as an assumption.
3. **What is the ethics route for this project, and how long does it take?** Determines whether the timeline is honest. Default: assume any human-participant study needs formal approval and budget six to ten weeks, stated explicitly as an assumption to be replaced.
4. **Has data access been secured, and is there written evidence?** Determines whether the feasibility claim in the proposal is a fact or a hope. Default: assume not secured, and write the contingency into the proposal rather than omitting the risk.
5. **What degree level is this, and therefore what contribution claim is appropriate?** Determines the anticipated contribution section, which is where overclaiming does the most damage. Default: assume the contribution is demonstrating competent research on a defensible question, not advancing the field.
6. **Is the proposal defended orally or assessed on the document alone?** Determines how much reasoning must be explicit on the page. Default: assume document-only, which is the more demanding case.
7. **What is your institution's AI use policy for this assessment, and what must be declared?** Governs how this skill may be used and what must be disclosed. Default: assume assistance must be declared and no generated text may be submitted (§12.1).

## 7. Step-by-step methodology

**Step 1. Establish what this proposal is assessed on, and write it at the top of your working file.**
Proposals are not marked on the promise of an interesting result. They are marked on whether the project is coherent, whether the method could answer the question, whether the scope fits the words and the months, and whether the researcher understands what they are undertaking. This is the reframing the whole document depends on, because a student who believes they are selling a result writes persuasion where justification is required, and inflates the contribution, which is the fastest route to a low mark. Get the rubric and its weightings. Where the rubric is unavailable, assume feasibility, alignment and justification dominate and record the assumption. *Correct result: the assessment criteria written down, with weightings, before any drafting.*

**Step 2. Write the problem statement as a tension, not as a topic with context.**
A topic names a subject. A problem names something wrong, unresolved, inconsistent or unknown, and it is what gives the study a reason to exist. The reliable form has four moves: establish the territory in one or two sentences (what is broadly known and accepted); establish the specific situation (what is happening, or what the evidence shows); state the tension, which is the sentence containing "but", "yet", "however" or "despite"; and state the consequence of leaving it unresolved. The tension is the load-bearing sentence in the entire proposal, and everything downstream, including the gap, the questions and the contribution, is a consequence of it. Test it by deletion: remove the tension sentence, and if the rest of the paragraph still reads as complete, there was no tension in it and you have written a topic description. *Correct result: a problem statement of one to three paragraphs whose central sentence contains a genuine tension, and which a reader could restate in one line.*

**Step 3. Build the alignment matrix, and build it before writing prose.**
This is the step that prevents the most reliably penalised failure in student proposals. Lay out four columns and check them against each other in both directions.

| Aim (one) | Objectives (three to five) | Research question or questions | How each will be answered |
|---|---|---|---|

Rules. The **aim** is a single sentence saying what the study seeks to achieve, in the language of purpose. The **objectives** are the pieces of work that get you there, each starting with a verb naming something you will actually do (identify, measure, compare, analyse, evaluate), each achievable and checkable. The **research questions** are interrogative sentences whose answers constitute the achievement of the aim. Then run the two tests markers run. The **upward test**: do the objectives, taken together and with nothing else, achieve the aim? If an objective could be deleted and the aim still be reached, it is decoration. If all could be completed and the aim still not be reached, something is missing, and the missing piece is usually the analytical or comparative objective students omit because it is the hard one. The **downward test**: does every question map to at least one objective, and every objective to at least one question? Orphans on either side are the symptom of a proposal assembled section by section. Check one error specifically: objectives that are all descriptive under an aim that is analytical or evaluative. Description does not add up to evaluation, and good writing does not hide the gap. *Correct result: a completed matrix with no orphans, passing both tests, drafted before any section is written.*

**Step 4. Write the preliminary literature section to establish a gap, not to demonstrate reading.**
Its job is to bring the reader to the point where your research question is the obvious next thing to ask. That is a different document from a summary of what you have read, and the difference is structural rather than stylistic. Organise by theme or by position, never by source, and never chronologically unless the chronology is itself the argument. Each subsection should do one of four things: establish what is settled (briefly, since settled matter needs little space); identify where the literature disagrees; identify a limitation shared across studies; or identify a population, context or condition not yet examined. Then close with an explicit gap statement in one or two sentences, phrased against what you actually searched. The commonest failure is a section that summarises ten studies competently and then asserts a gap that follows from none of them. The second is the gap claimed too widely: "no research has examined this" is almost never true and is easily disproved, while "studies have concentrated on X, and the question has not been examined in Y" is defensible and sufficient. Where the honest position is that the literature is contested rather than absent, say so, because a contested question is a stronger basis for an undergraduate study than a supposed void. *Correct result: a thematically organised section ending in a specific, defensible gap statement that the preceding text has actually earned.*

**Step 5. Justify the method against the question, rather than describing it.**
A description says what you will do. A justification says why this and not the alternatives, and it is what the criterion actually asks for. Structure the justification in four moves. First, name what the question requires: whether it needs measurement of prevalence or difference, understanding of meaning or process, a causal claim, or documentation of what exists. Second, name the approach that follows from that requirement, stated as a consequence of the first move rather than as a preference. Third, name at least one serious alternative and say why it was rejected, on grounds of fit to the question, not on grounds of difficulty, since "I am not confident with statistics" is a real constraint but not a methodological justification. Fourth, work down through the decisions with a one-line reason each: population and sampling approach, instrument or protocol, collection procedure, and analysis method. Analysis is where student proposals most often stop early, and a proposal that does not say how the data will be analysed has not shown that the question can be answered. Justify the sample size in the terms the design uses, and where a target is aspirational say so and give the minimum you could work with. *Correct result: every methodological decision carries a one-line reason tied to the question, and at least one rejected alternative is named.*

**Step 6. State scope and delimitation explicitly, and treat it as strength.**
Delimitations are the boundaries you chose: population, setting, period, constructs included and excluded, and what the study deliberately does not address. Limitations are the constraints you did not choose. Students conflate them, and they omit delimitations because stating a boundary feels like admitting inadequacy. It is the reverse: an explicit scope statement demonstrates control of the project and pre-empts the marker's objection, whereas an unstated boundary reads as an oversight. Write scope as a short list of inclusions and exclusions, each with a one-line reason. Include the deliberate non-questions: the adjacent thing a reader will wonder why you are not studying, named and excluded on purpose. *Correct result: a scope statement a reader can use to predict what your findings will and will not cover.*

**Step 7. Address ethics at proposal level, which means the route and the issues, not the application.**
Identify the approval route and its category, the participants and any vulnerability, the consent process, how withdrawal will work, how data will be stored and for how long, how anonymity or confidentiality will be maintained and where it cannot be, and any risk of distress and the response to it. Two points do more work than the rest. Name the approval date you are working to, because the ethics timeline drives the whole schedule. And state honestly where anonymity is limited: a study of one department in one institution cannot promise participants that colleagues will not recognise them, and saying so is a mark of competence. Even where a project uses only secondary or published data, say so and state why approval is not required, rather than omitting the section. *Correct result: an ethics section naming the route, the date, the specific issues for this study, and the mitigation for each.* `RESEARCHER SIGN-OFF REQUIRED` where participants may be vulnerable or the topic is sensitive (K5 §2.4).

**Step 8. Build the timeline backwards from the deadline, with stages and slack.**
Forward-planned timelines are optimistic because they start where enthusiasm is. Start from the submission date and work back: submission, final proofread and formatting, revision after supervisor feedback, complete draft, analysis, data preparation or transcription, data collection, recruitment, ethics approval including possible amendments, instrument development and piloting, and the literature review, which runs alongside. Each stage gets a start and an end date, not a duration. Then apply three corrections. Writing takes longer than students expect and should not absorb overruns; give it a third of the remaining time. Ethics approval is a dependency, not a task, and nothing involving participants begins until it clears. And build in at least two weeks of visible slack, because a plan with no slack fails on its first delay. If the backward plan runs off the front of the calendar, the scope is wrong, and it is far cheaper to discover that now than in month four. *Correct result: a dated stage plan that ends at the deadline, contains visible slack, and shows ethics as a dependency.*

**Step 9. State the anticipated contribution at the level of the degree.**
Undergraduate and honours contribution is not doctoral contribution, and pitching it there is a distinct and common failure: a proposal claiming to fill a significant gap in the field invites a marker to ask whether it does, and it will not. Legitimate claims at this level include evidence about a population or setting not previously examined, a test of whether an established finding holds in a different context, a documented account of something not yet described, or a synthesis useful to a specific practice or policy audience. State the contribution in two or three sentences, in the conditional, and say who would find it useful. Then state what the study will not establish, which costs three lines and inoculates the proposal against the objection that it overclaims. *Correct result: a modest, specific, conditional contribution claim, with an explicit statement of what is out of reach.*

**Step 10. Run the coherence audit as an examiner would, then say the proposal will change.**
Read the finished document once, quickly, checking only the golden thread: does the problem produce the gap, does the gap produce the questions, do the questions produce the method, does the method produce data that would answer the questions, and would answering them achieve the aim? Break the thread anywhere and the proposal fails on coherence regardless of the quality of any individual section. Then check three specific joins that break most often: a question that the proposed data cannot answer; an objective with no corresponding method; and a contribution claim that does not follow from the questions. Finally, state in the document, in one sentence, that the design may be refined as the literature review develops and access is confirmed. This is not hedging. Proposals are expected to change, supervisors know it, and a proposal written as though it were an unbreakable contract produces students who conceal necessary changes rather than reporting them. *Correct result: an unbroken thread from problem to contribution, and an explicit acknowledgement that detail will change.*

## 8. Analytical framework

**The golden thread**, which is what coherence means in an assessed proposal:

    Problem (a tension) → Gap (what is not known) → Aim (what you will achieve) →
    Objectives (the work that achieves it) → Questions (what you will answer) →
    Method (how you will answer them) → Data (what you will have) →
    Contribution (what it adds, at this level)

Applying it: the thread is tested at the joins, not in the boxes. A proposal can have a good problem statement and a good methodology section and still fail, because the method answers a question the problem did not raise. Read the document backwards as well as forwards: from the contribution, ask what would have to be found to justify it; from the data, ask which question it answers. Backwards reading finds breaks that forwards reading conceals, because forwards reading is generous to the writer's intentions.

**The alignment matrix**, which is the operational form of the thread's middle:

    Aim ← (do these add up?) ← Objectives ↔ Questions → (can these be answered?) → Method

Applying it: run the upward test on the objectives and the downward test on the questions, as in Step 3. The single most common structural defect in student proposals is a set of descriptive objectives under an analytical aim, and it is invisible in prose and obvious in the matrix. That is why the matrix is built before the prose and not extracted from it afterwards.

## 9. Output format

Where a departmental template exists it governs. In its absence, or as a working structure beneath it:

**1. Title.** A working title naming the phenomenon, the population and the context.

**2. Problem statement.** Territory, situation, tension, consequence.

**3. Preliminary literature and the gap.** Thematically organised, ending in a specific gap statement phrased against what was searched.

**4. Aim, objectives and research questions.** Presented in a way that makes the alignment visible, with the matrix included where the template allows.

**5. Methodology and its justification.** Approach and why, alternatives rejected and why, then population and sampling, instrument, procedure, and analysis, each with a reason.

**6. Scope and delimitation.** Inclusions and exclusions with reasons, and the deliberate non-questions.

**7. Ethics.** Route, category, approval date being worked to, specific issues, and mitigations. Or an explicit statement of why approval is not required.

**8. Timeline.** Dated stages, backwards-planned, with ethics shown as a dependency and slack visible.

**9. Anticipated contribution.** Conditional, level-appropriate, with the audience named and the limits stated.

**10. Preliminary references.** Every entry retrieved and opened, per **15.03**.

**When the evidence or the access is thin, the format must not force fabrication (K4 §1).** A gap statement is written against the searches actually run, never as a general claim that nothing exists. Where access is not yet secured, the proposal says so and states the contingency, rather than presenting a hoped-for arrangement as an existing one, which is a misrepresentation of feasibility and the thing a proposal is most centrally assessed on. Where a section of the template does not apply, say why in one line rather than filling it.

## 10. Quality checks

Run before submission. These sit on top of K4 §8.

1. Does the problem statement contain a sentence with a genuine tension, and does the paragraph collapse if that sentence is deleted?
2. Do the objectives, taken together, achieve the aim, with nothing decorative and nothing missing?
3. Does every research question map to an objective, and every objective to a question?
4. Are the objectives descriptive while the aim is analytical, and if so has the analytical objective been added?
5. Does the gap statement follow from the literature actually presented, and is it narrow enough to be defensible?
6. Is the gap phrased against the searches performed rather than as a claim about all literature?
7. Does every methodological decision carry a reason tied to the question, including the analysis method?
8. Is at least one alternative approach named and rejected on grounds of fit rather than difficulty?
9. Could the proposed data actually answer each research question?
10. Is scope stated as deliberate delimitation, separate from limitations?
11. Does the ethics section name a route, a date and the specific issues for this study?
12. Is the timeline dated by stage, planned backwards, with ethics as a dependency and visible slack?
13. Is the contribution claim conditional and pitched at the level of the degree?
14. Has every preliminary reference been retrieved and opened?
15. Read backwards from contribution to problem, does the thread hold?

## 11. Common failure modes

| Failure | How to recognise it | How to prevent it |
|---|---|---|
| **Topic instead of problem** | The problem statement has no "but" or "yet" in it | Write the tension sentence and test it by deletion (Step 2) |
| **Objectives that do not add up** | All objectives could be met and the aim still not achieved | Run the upward test on the matrix (Step 3) |
| **Descriptive objectives, analytical aim** | Every objective starts with identify, describe or explore | Add the analytical or comparative objective (Step 3) |
| **Aim, objectives and questions identical** | The three sections say the same thing in three grammars | Purpose, work and interrogative are different registers (Step 3) |
| **Literature as demonstration of reading** | Organised by source, one paragraph per study | Organise by theme or position, ending in a gap (Step 4) |
| **Unearned gap** | Ten studies summarised, then a gap that follows from none | Each subsection identifies disagreement, limitation or absence |
| **Gap claimed too widely** | "No research has examined this" | Narrow to population, context or condition, phrased against the search |
| **Method described, not justified** | The section reads like a textbook entry on the method | Name the requirement, the consequence, the rejected alternative (Step 5) |
| **No analysis plan** | Collection is specified, analysis is a sentence | The analysis method needs its own justification |
| **Missing scope statement** | The reader cannot tell what the study excludes | State delimitations as choices with reasons (Step 6) |
| **Ethics as good intentions** | "Participants will be treated ethically" and no route or date | Name route, category, date, issues, mitigations (Step 7) |
| **Equal-box timeline** | Every stage the same length, no ethics dependency | Plan backwards with dated stages and slack (Step 8) |
| **Overclaimed contribution** | The proposal promises to fill a gap in the field | Pitch to the degree level and state the limits (Step 9) |
| **Result-selling** | The proposal argues that the findings will be important | Proposals are assessed on feasibility and coherence (Step 1) |
| **Invented literature or access** | Sources or agreements asserted but not held | Nothing enters unverified; see §12.2 and 15.03 §12 |

## 12. AI guardrails

Skill-specific. The universal prohibitions in K4 apply in full and are not repeated.

1. **Academic integrity.** These skills assist a researcher's thinking, structure and rigour. They do not produce work to be submitted as the student's own unaided output. The user must comply with their institution's AI use policy and its declaration requirements, which vary by institution and by assessment. Where an institution prohibits AI assistance for a task, this skill must not be used for it. The skill never writes a passage for submission as though the student wrote it; it interrogates, structures, critiques and teaches. Operationally in this skill: building the alignment matrix, testing objectives against the aim, stress-testing a timeline and critiquing a draft against the golden thread are legitimate. Drafting the problem statement, the justification or the gap statement as submittable prose is not, and a proposal is frequently defended orally, where authorship becomes immediately apparent.

2. **Never assert a gap in the literature.** A gap is a claim about what has been published and it requires a search to support it. Where no search has been run, the honest statement is what remains unknown to this student, not what remains unknown to the field. Phrase every gap against the searches actually performed, and name them (K4 §2.4, and **15.02** §12).

3. **Never supply references, prior studies or "existing work in this area" from model knowledge.** The preliminary reference list is built only from documents retrieved and opened. Fabricated references are frequently first caught at proposal stage, where a supervisor reads the list closely for the first time. Full rule in **15.03** §12.

4. **Never estimate an ethics turnaround, a departmental word count, a template structure or a marking weighting.** These vary by institution and by year, and a plausible wrong figure produces a timeline that is confidently infeasible. Mark them unknown and say where to find the real values.

5. **Never write a timeline that assumes approval, access or recruitment will succeed.** A proposal timeline is a feasibility claim. Where access is not secured, the proposal states that and gives the contingency. Presenting a hoped-for arrangement as an existing one misrepresents the single thing the proposal is most centrally assessed on.

6. **Never inflate the contribution claim to make the proposal sound stronger.** Overclaiming is penalised, and it is the most common route by which a competent project receives a mediocre proposal mark. Where a student wants a bigger claim than the design supports, say what the design can actually establish and offer the honest version (K4 §9, K3 §4.2).

7. **Never assume the discipline's methodological conventions.** What counts as an adequate justification, an acceptable sample size, or a legitimate design differs substantially between fields and paradigms. Ask; where you cannot ask, handle the common cases and state which one you are assuming.

8. **Where a proposal rests on an unresolved feasibility risk, flag it rather than smoothing it** (K5 §2.7). Access not secured, ethics category unknown, a method the student has never used: each is a decision for the student and supervisor, and each changes the proposal if answered differently.

## 13. Best-practice principles

- **Proposals are assessed on coherence and feasibility, not on the promise of a result.** Say this to a student first, because it redirects effort from persuasion to alignment, and alignment is where the marks are.
- **The tension sentence is the whole document in miniature.** If you cannot write it, you do not yet have a project, and no amount of literature will supply one.
- **Build the matrix before the prose.** Misalignment written into fluent paragraphs is expensive to find and painful to fix; misalignment in a four-column table is obvious in a minute.
- **The literature section argues for the study.** Every paragraph should be answerable to the question "why does this make my study necessary?", and a paragraph that cannot answer it belongs in the review, not the proposal.
- **Justify by rejection.** Naming the alternative you did not choose, and why, demonstrates methodological understanding more efficiently than any amount of description of the method you did choose.
- **The analysis plan is part of the method.** A proposal that collects data without saying how it will be analysed has not demonstrated that the question can be answered, and this is the most commonly thin section in student proposals.
- **State your boundaries on purpose.** An explicit delimitation reads as control. The same boundary left unstated reads as an oversight, and markers cannot tell the difference from the document.
- **Plan the timeline backwards, and protect the writing.** Writing is the stage that always overruns and the stage students always compress. Give it a third of what remains and put the slack before it, not after.
- **Ethics approval is a dependency, not a task.** Everything involving participants waits on it, and a timeline that treats it as a parallel activity is not a timeline.
- **Claim small and claim precisely.** A modest contribution stated exactly is stronger than an ambitious one stated vaguely, and it is the version a marker can agree with.
- **Say the proposal will change.** It will. Writing it as a contract encourages students to hide necessary changes, which is worse for the project than the changes.

## 14. Worked example

Generic fictional scenario, academic, environmental science.

**INPUT**

An honours student in environmental science must submit a 3,000 word proposal in four weeks, for a project due nine months later. Their committed question concerns household water use behaviour during municipal restrictions in a mid-sized urban area. Their first draft has a two-page background section, an aim, five objectives, three research questions, a paragraph naming a survey as the method, and a timeline of six equal monthly blocks.

**PROCESS**

*Step 1.* The rubric is obtained. Feasibility and coherence carry 40% between them, method justification 25%, literature and gap 20%, and presentation 15%. Nothing rewards the interest of the anticipated result. This changes the student's allocation immediately, since most of their effort had gone into the background section.

*Step 2.* The problem statement is rewritten. The draft version described water scarcity for two paragraphs and then announced the study. The tension is located and stated: restriction policies assume that households reduce consumption in response to information and price signals, yet metered data during recent restriction periods shows highly uneven reduction across similar household types, and the sources of that variation are not established. Deleting that sentence collapses the paragraph, which is the test passing.

*Step 3.* The matrix exposes the main defect. The aim is to explain variation in household response to restrictions. The five objectives are: identify restriction measures in force; describe household water use practices; identify sources of information households use; describe attitudes to restrictions; and summarise findings. Every one is descriptive. Run upward, all five could be completed and the aim, which is explanatory, would still not be achieved. Two changes follow: the fifth objective, which is a task rather than an objective, is removed, and a comparative objective is added, to compare reported practices between households with high and low reported reduction. The aim is also softened from "explain" to "examine and account for", because an explanatory aim would require a design this project cannot support. Three questions are then mapped to four objectives with no orphans.

*Step 4.* The literature section is reorganised. The draft summarised nine studies in nine paragraphs. It is restructured into three themes: what is established about price and information effects, where studies disagree about the role of household composition, and the near-absence of work in mid-sized municipalities under short-duration restrictions, which is where the study sits. The gap statement is narrowed from "little research exists on household water behaviour" (untrue and easily disproved) to a claim about the specific setting and restriction duration, phrased against the databases searched.

*The judgement call.* The student wants to include a metered consumption dataset held by the municipality alongside the survey, which would be a materially stronger design because it would let reported behaviour be compared against measured behaviour. They have had an encouraging conversation with a municipal officer and nothing in writing. The temptation is to write the proposal around the stronger design. The resolution is to write the survey as the design the proposal commits to, since the proposal is a feasibility claim and an unsecured dataset is not a feasibility claim; and to include the metered data as a stated enhancement, conditional on a written data-sharing agreement by a named date, with the analysis it would enable described in two sentences. This is honest, it protects the project if access fails, and it demonstrates exactly the awareness of feasibility risk that the rubric rewards.

*Step 5.* The method section is rewritten as justification. What the question requires: variation across households, which needs comparison across a reasonable number of cases, which points to a survey rather than interviews. The rejected alternative, a small interview study, is named, with the reason that it would produce rich accounts but could not establish variation across household types, which is what the question asks. Sampling, instrument, procedure and analysis each get a reason, with the analysis specified as group comparison on defined household categories rather than left implicit.

*Steps 6 to 8.* Scope is stated: single municipality, one restriction period, household-level only, explicitly excluding agricultural and commercial use, and explicitly not evaluating policy effectiveness, which is the adjacent question a reader will wonder about. Ethics is low-risk but requires approval, and the route and target approval date are named, with the note that the survey collects no identifying data. The timeline is rebuilt backwards from the deadline: writing is given ten weeks, ethics is shown as a dependency with a six to eight week window and an amendment allowance, and two weeks of slack sit before analysis. The backward plan shows the survey must be in the field by a specific week, which becomes the project's controlling date.

*Step 9.* The contribution is stated conditionally: evidence on within-setting variation in household response for a municipality size and restriction duration where little is documented, useful to municipal water demand teams. Followed by what it will not establish, namely causal effects of restriction measures and generalisability beyond comparable municipalities.

**OUTPUT**

A proposal whose four columns visibly align, whose gap is narrow and defensible, whose method is justified by rejection, whose scope is stated as choice, whose timeline is dated backwards with ethics as a dependency and two weeks of visible slack, and whose contribution claim a marker can agree with. The conditional enhancement is written in a form that strengthens rather than weakens the feasibility case.

`RESEARCHER DECISION REQUIRED` on whether to pursue the municipal dataset, which depends on institutional relationships and the supervisor's judgement about the officer's authority to agree (K5 §2.7). `RESEARCHER SIGN-OFF REQUIRED` on the ethics section before submission (K5 §2.4).

## 15. Advanced usage

**Repairing a rejected proposal.** Rejections cluster on three causes: misalignment, an unearned gap, and an infeasible timeline. Diagnose before rewriting, because each has a different and localised fix. Misalignment is fixed in the matrix and then propagated to the prose. An unearned gap is fixed by more searching, not by more writing. An infeasible timeline is fixed by reducing scope, which means changing an objective, which means returning to the matrix. Rewriting the whole document without diagnosing which of the three applies is the usual student response and it usually reproduces the fault.

**Preparing for a defence.** Where the proposal is defended, prepare three answers in advance, because they are asked almost every time: why this method and not the obvious alternative, what happens if recruitment or access fails, and what you will do if the results are null or the opposite of what you expect. A student who has an answer to the third question has understood that the proposal is not a promise of a result.

**Where the standard approach does not fit.** In practice-based and design disciplines, the method section documents a process rather than a protocol, and the justification shifts to why this process suits the enquiry. In secondary-data and computational projects, the ethics section shrinks and the data provenance, licensing and documentation sections expand, and feasibility turns on data availability and skill rather than on recruitment. In both cases the golden thread test is unchanged.

## 16. Skill chain

**Recommended previous skills:**
- **15.01 Academic Research Topic Selection.** Hands over the committed question, the problem's tension, the contribution form, and a feasibility table that supplies most of the scope and timeline reasoning.
- **15.02 Academic Literature Search Strategy.** Hands over the retrieved sources and the search log from which a defensible gap statement can be written.

**Recommended next skills:**
- **15.05 Academic Writing Structure and Argumentation.** Turns the proposal's sections into paragraphs that argue, and carries the same discipline into the dissertation itself.
- Fieldwork and analysis follow the proposal's approved design.

**Runs well alongside:**
- **15.03 Referencing and Citation Management**, since the preliminary reference list is where fabricated references are most often first caught.
- **01.02 Business Problem to Research Question** and **01.04 Research Method Selection**, where the project is set by an external organisation and the question arrives as a business need, or where the method choice needs a fuller design comparison than a student proposal requires.

---
A Yazi Supplied Skill and resource.