The Scholar's Vault · DOCX skeleton
The page that decides it, paragraph by paragraph.
Specific Aims page template: word budgets, reader model, and the three NIH review factors
On most project grants of ordinary size, the panellist assigned to your application reads this page first, forms a position, and reads the rest looking for reasons to keep it. The other twelve pages argue. This one decides.
No ghostwriting, ever · Human-led, PhD-signed
The page is the content. The download is the working copy — a DOCX with the five slots as styled headings, each carrying its own word budget, and the factor-mapping table on page two. One email covers all six Vault resources.
What this is, and what it is not
What this skeleton does.
It gives you five paragraph slots, a word budget for each, the job each one has to do, and what the reader is deciding while they read it.
It does not give you sentences to fill in. A Specific Aims page assembled from somebody else’s phrasing reads exactly like what it is, and panellists read forty of these a cycle.
The failure it is built against is not weak science. It is a strong project whose page never states, in one sentence, what will be true at the end that is not true now.
The reader you are writing for
Assume three things about the person holding your page, because all three are usually true. They are reading it at the end of a working day, seventh in a stack. They are not in your subfield, and they will present your application to people who have not read it. Their first pass takes about 90 seconds.
That is not cynicism about peer review. It is the operating condition, and under that condition the page is an engineering problem rather than a writing problem.
The geometry
Five slots, and what each one costs you.
One page. No more, and a page that stops three lines short reads as confident rather than thin.
| Slot | Job | Words |
|---|---|---|
| P1 | The opening | 90–130 |
| P2 | The gap and the need | 80–120 |
| P3 | The pivot: goal, objective, hypothesis, rationale | 100–140 |
| Aims | Two or three aims, parallel in shape | 170–240 |
| P4 | The close: what changes | 70–100 |
| Total | 510–650 |
The ceiling is a page, not a word count. At the smallest formatting a typical funder allows (11pt, half-inch margins), one page holds roughly 650 words with headings and white space, and less once the aims are bolded and spaced. Write to the budgets, then set the page up in the real template and cut until it fits with room to breathe. A page that runs to a second line on a second page is read as an author who cannot prioritise.
Formatting that a tired reader needs: the aims in bold, one line of white space between slots, no paragraph over five lines on screen. Nothing decorative, because decoration on this page reads as space you did not have anything to put in.
The five slots
The skeleton, in full.
Each one with the job it has to do, what the panellist is deciding while they read it, the shape, and the ways it goes wrong.
P1 · The opening
- What it has to do
- Establish the problem in terms the panellist already accepts, and end on the single sentence that states what is not known.
- What the reader is deciding
- Whether this person is inside the field or outside it. The tell is not the science; it is whether the framing matches the one they use.
Shape
[SENTENCE 1: the problem, in its consequence rather than its mechanism — who is affected, at what scale, with the figure and the citation.][SENTENCES 2–3: what is established. The two or three findings the field agrees on, cited, stated as settled.][SENTENCE 4: what is not established. One sentence. This is the hinge of the whole page.]
Constraints
- Open on the consequence, not on the history of the field. A panellist who reaches line four still in 1987 has learned that you cannot prioritise.
- Cite the settled findings, and cite work the panel would recognise. This is the paragraph where an assigned reviewer checks whether you know who is in the room.
- The last sentence must be a statement of absence, not a statement of intent. “The mechanism by which [X] produces [Y] is unknown” works. “We therefore aim to investigate” does not, because it moves to your plan before the gap has landed.
Fatal versions: the textbook opening · four citations to your own lab and none to anyone else’s · a first sentence containing no number
P2 · The gap and the need
- What it has to do
- Convert the unknown from P1 into a cost, and explain why the gap has survived.
- What the reader is deciding
- Whether the gap is real or manufactured. Every application claims a gap; a panellist is looking for the reason nobody has closed this one.
Shape
[SENTENCE 1: the critical need — what cannot be done, decided, treated or built while this is unknown.][SENTENCE 2: why the gap persists. A methodological limit, a missing dataset, a measurement problem, a scale nobody has had.][SENTENCE 3: what has changed to make it addressable now — the new method, cohort, instrument, model or collaboration.]
Constraints
- Sentence 2 is the one applicants skip, and skipping it is why a page reads as ambitious rather than credible. A gap with no explanation for its persistence looks either trivial or intractable, and both readings cost you.
- Sentence 3 is where your preliminary data, your access, or your team earns its place on this page. One clause, not a paragraph.
Fatal versions: “little is known about” as the whole justification · a need stated in field terms rather than in consequence terms · a gap that exists because nobody thought it was worth closing
P3 · The pivot
- What it has to do
- State, in four labelled moves, what you will do and on what basis.
- What the reader is deciding
- Whether there is a hypothesis or only a plan of activity. This is the paragraph that separates a project from a programme of work.
Shape
Long-term goal. [One clause. The direction of the research programme, beyond this application.]Objective of this application. [One sentence. What this specific work delivers.]Central hypothesis. [One sentence, falsifiable, naming the direction of the expected effect or relationship.]Rationale. [One sentence: what becomes possible once the hypothesis is tested, either way.]
Constraints
- Use the labels. Panellists scan for them, and the two seconds it saves are two seconds spent on your science instead.
- The hypothesis has to be capable of being wrong. “We hypothesise that [X] plays a role in [Y]” cannot be wrong and therefore cannot be tested, and an experienced reviewer will say so in one line.
- The rationale must hold if the hypothesis fails. If a negative result makes the project worthless, the design carries a risk you have not disclosed, and disclosing it here with a mitigation is stronger than having it found.
Fatal versions: an objective that restates the long-term goal at greater length · a hypothesis that is a topic · a rationale that is a benefit to the field rather than a consequence of the test
Aims · Two or three, parallel
- What it has to do
- Show that the objective decomposes into independent, finishable pieces of work.
- What the reader is deciding
- Whether this can be completed with this budget and this team, and whether one failure takes the rest down.
Shape
Aim [N]. [Verb phrase, 8–14 words, starting with a strong verb: Determine · Quantify · Test · Identify · Map · Compare.]Working hypothesis: [One sentence, specific to this aim.]Approach: [One or two sentences: design, sample or model, key measure, comparison.]Expected outcome: [One sentence: what you will know, stated as a result rather than as an activity.]
The four constraints on the aim block
- Aims must not be sequential. If Aim 2 cannot start until Aim 1 succeeds, you have one aim with three stages and a single point of failure. Reviewers read dependency as risk, and they are right to.
- Aims must be parallel in shape. Same length, same four-part structure, same verb strength. A short third aim reads as an afterthought, whatever it contains.
- Two strong aims beat three, one of which is padding. The third aim should broaden the contribution, not fill the page.
- Never open an aim with “Investigate” or “Explore”. Both describe an activity with no endpoint. A panellist cannot picture the result, so they cannot picture the paper.
On the aim that carries the risk. If one aim is genuinely more exposed than the others, name the alternative in the approach line, in one clause. Disclosed risk with a stated fallback is a strength on this page; discovered risk is a score.
P4 · The close
- What it has to do
- State what is true at the end that is not true now, and why that matters beyond your own project.
- What the reader is deciding
- What sentence they will say out loud when they present your application to the panel. Write that sentence for them.
Shape
[SENTENCE 1: expected outcomes — the concrete products: the parameter, the model, the dataset, the validated measure, the mechanism established or excluded.][SENTENCE 2: what is new here, in one clause, and why that newness changes what the work is worth rather than only how it is done.][SENTENCE 3: impact — what the field, the practice or the next study can do that it cannot do today.]
Constraints
- The outcome sentence has to survive the question “and then what”. A dataset that nobody uses is not an outcome; a dataset that resolves a named disagreement is.
- Do not write sentence 2 as a separate case for novelty. Under the NIH framework in force since January 2025, innovation is no longer scored on its own; it is folded into the first factor, Importance of the Research, alongside significance. Novelty that does not change the importance of the answer now argues for nothing, so write the newness as a reason the result matters rather than as a claim about technique.
- Do not restate the aims. A close that summarises the page tells the reader the page was too complicated to hold, which is a judgement on the page.
- Never end on gratitude, on funding language, or on a promise about future applications.
Fatal versions: impact stated as “will advance our understanding” · novelty that describes a technique which is standard in a neighbouring field · a close longer than P1
The 90-second test
Four answers, from a non-specialist.
Give the page to somebody outside your subfield. Take it back after 90 seconds and ask four questions.
| Question | If they cannot answer |
|---|---|
| What is not known? | P1’s last sentence is buried or missing |
| Why has nobody solved it? | P2 sentence 2 is missing. This is the most common gap |
| What are the two or three things this project will do? | The aims are not parallel, or the verbs are weak |
| What will be true afterward? | P4 restates instead of concluding |
Four answers, from a non-specialist, in 90 seconds. That is the specification the page is built to.
Which paragraph feeds which review factor
The mapping almost nobody has rewritten for.
For applications to due dates from 25 January 2025, NIH scores most research project grants against three factors rather than five.
Significance and Innovation are merged into Factor 1; Approach becomes Factor 2; Investigator and Environment become Factor 3, which is judged sufficient or not sufficient rather than scored. This changes what each of your paragraphs is doing.
| Factor | What the reviewer is answering | The slots that feed it | What that means for your page |
|---|---|---|---|
| 1 · Importance of the ResearchSignificance and Innovation, scored 1–9 | Should it be done? | P1, P2, and sentences 2 and 3 of P4 | Your gap statement and your close carry the whole of Factor 1. Novelty is no longer a separate case to be made; it counts only where it changes how much the answer matters. |
| 2 · Rigor and FeasibilityApproach, scored 1–9 | Can it be done well? | P3 and the aims block | The falsifiable hypothesis, the independence of the aims, and the expected-outcome line in each aim are Factor 2 on one page. This is where a page with three sequential aims loses. |
| 3 · Expertise and ResourcesInvestigator and Environment, judged sufficient or not | Are the expertise and resources in place? | Sentence 3 of P2 | The “what has changed to make this addressable now” clause is the only place the Aims page touches Factor 3. Name the method, cohort, instrument or collaboration that makes you the person who can do this. |
The practical consequence. Sentence 3 of P2 was always the sentence applicants skipped. It is now the sentence that opens your only Factor 3 argument on the page that gets read first, and Factor 3 is judged as a threshold rather than scored — so a reviewer who cannot answer it has to justify a finding of insufficiency in writing. Give them the sentence.
Check the current notice of funding opportunity for your mechanism before you write. Fellowship and career-development awards keep their own criteria, where the aims are read alongside a training plan and a career goal, so the aims block there carries a second job this skeleton does not cover.
Scheme mapping
Where else this page is the same page.
The structure holds where the label changes. Check the call text for the word limit and the required headings, and keep the five slots underneath whatever the scheme calls them.
| Scheme convention | The page it maps to |
|---|---|
| Specific Aims (NIH-style) | This skeleton, one page, with the factor mapping above |
| Case for support summary, or summary of proposed research | P1–P4 compressed, aims become objectives |
| Excellence section opener (European schemes) | P1, P2 and P3 expand into Section 1.1 (Objectives & Ambition) and 1.2 (Methodology) under the Horizon Europe V5.1 template. Aims map to scientific objectives, which then translate to Section 3.1’s independent work packages where the dependency rule is critical. |
| Lay or non-technical summary | P1 sentence 1, P2 sentence 1, P4 sentence 3, rewritten without terms of art |
The dependency rule is the one that transfers everywhere. Any scheme that funds work packages is looking for work packages that fail independently.
The thirteen-point check
Run these on the finished page.
- P1 ends on a statement of what is not known.
- P2 says why the gap has survived.
- The four labels appear in P3, in bold or italics.
- The hypothesis can be wrong and still state a direction.
- No aim depends on another aim succeeding.
- Every aim opens with a verb from the strong list.
- Every aim has an expected outcome stated as a result.
- The aims are within 15 words of each other in length.
- P4 names a concrete product and a consequence.
- The page has no sentence that could appear on any other application in the call.
- A non-specialist answers the four questions in 90 seconds.
- Sentence 3 of P2 names what makes this addressable now, and by you.
- The page fits on one page in the real template, with white space left.
Where this stops, and what we do
Where the page usually fails.
The most common structural failure is a page that is entirely credible and never says what would change.
It passes every readability check, describes competent work, and gives the panellist nothing to advocate with. That is not a writing problem, and rewriting will not fix it. It means the objective in P3 has not been decided yet, and the page is honestly reporting that.
Grant Surge — the application is read against the criteria the panel will score it on, with the objections ranked by what moves a decision.
- Seed€649· 5 business days
- Major€999· 7 business days
Carries the Fit Guarantee: before work starts, a full refund or a credited rescope, your choice; after delivery, a free revision, then half the fee back if the work still misses the brief we agreed.
Or take the free route first. A written strategy brief: tell us where you are stuck, and within 3 working days we send a three-step plan, including which service fits, or that none of ours does. No follow-up sequence.
We do not guarantee funding, and we have no relationship with any panel or funder. What we do is tell you which of your aims a reviewer will attack first, and we say when the honest answer is that the page is not ready.