Grant Writing and Award Nomination Tips for PhD Students

These notes come from applying to doctoral funding during my own program and from reading the guidelines closely enough to notice where they disagree with each other. Please take the parts that fit your funder and leave the rest.

Fit before writing

The first thing I look at is the awardee list. Most funders publish one, and it does two jobs. It tells you whether your project belongs in the pile, and it shows you what a funded version of a project like yours looks like, including how narrow the question is and how much a person promises to do in one year. The NAEd/Spencer fellows page names its current cohort with their projects, and Grand Challenges publishes an awardee brochure each cycle. If nothing in two or three years of awardees resembles your project, the fit is probably wrong.

Some funders write the filter down. Harry Frank Guggenheim funds research on physical violence, and its advice to applicants says that “as many as a third of the proposals we receive in each round of applications are eliminated from further consideration because they don’t meet this relevance criterion.” The same document draws the line: “We do not fund research on projects that are about metaphorical extensions of the concept of violence, such as ‘symbolic violence’ or ‘structural violence.'” Newcombe asks how central questions of religion, ethics, values, or morals are to the project, and a dissertation where they are incidental does not clear that bar.

When the page is genuinely ambiguous, I email the program officer.

Early drafts and second attempts

What works for me is writing a rough draft fast and spending the time saved on feedback. A draft that is done three weeks out can go to your chair, to a second committee member, and to someone outside your area, and each of them will catch something different. The readers furthest from your subfield are the ones who catch assumptions you take for granted. A polished draft finished the night before has only been read by you.

This is why I write the short pieces first. A one-page summary or a 750-word description is something a busy person will read on the day you send it, and their reaction tells you whether the project is legible before you have written the long version.

These competitions mostly repeat every year, and several open twice. Fahs-Beck takes applications on April 1 and November 1, SPSSI on May 15 and October 15, and the Melanie Foundation says outright that “candidates may reapply for additional years.” Applying to the same grant in consecutive years is normal, and the second attempt starts from a document that already exists. NIH is the exception, since it allows a single resubmission and no more.

Review criteria before instructions

Funders publish what they are scoring, and reading that first changes what you spend your time on.

NIJ is the clearest case, because it publishes weights.

• Statement of the Problem, 5 percent
• Project Design and Implementation, 40 percent
• Capabilities and Competencies, 15 percent
• Potential Impact, 25 percent
• Budget, 15 percent

The budget carries three times the weight of the problem statement.

I write for a reviewer outside social work. The panels at Russell Sage, Harry Frank Guggenheim, Horowitz, AERA, and NAEd/Spencer are cross-disciplinary by design, so a proposal built on terms only a social work audience shares is harder for a sociologist or an economist to score.

IRB timing

Several funders want IRB approval in hand at submission, and one of them will not accept the version most students have.

Fahs-Beck’s guidelines require “evidence that the dissertation proposal has received official final approval from the IRB of the sponsoring institution,” and then specify that “proposals with provisional approval will not be reviewed.” Provisional or conditional approval is the stage most proposals arrive at first, and final approval can be months behind it.

The same guidelines bar “expenses incurred prior to faculty approval of dissertation proposal,” so money already spent is not recoverable by winning a grant afterward. Fahs-Beck also wants the whole IRB-approved proposal and every instrument in the package, which takes an afternoon of printing and collating.

I work backwards from this. If a February deadline needs final IRB approval, your submission to the IRB belongs in the previous fall, which means your proposal defense comes before that.

Letter writers

Nearly every application here needs a letter from your dissertation chair, and several need more than one. NAEd/Spencer requires two letters, one of which must be your chair. Fahs-Beck wants a confidential letter from your primary faculty sponsor, on letterhead, “placed in a sealed envelope with the faculty member’s signature written over the seal,” plus two institutional letters. Three separate people have to act, and the package goes by mail.

Best practice is to ask once, for the whole year. Each letter-writer gets a single packet (aims, timeline, CV, and the list of deadlines) instead of five separate requests over five months, with a note about which dates are real.

NIH fellowships handle this differently. You need at least three reference letters and no more than five. Referees submit them through a separate eRA Commons page, so they never appear inside your own application, and your sponsor cannot be one of them. NIH’s reference letter page says an application with fewer “will not be reviewed,” so it is rejected before anyone reads it.

Budget exclusions

Budgets do not travel either, because funders disagree with each other on the same line items. Russell Sage lets you pay yourself a stipend of up to $7,500 out of a $15,000 grant, and requires any research assistant you hire to be paid at “the minimum wage in the state where the fiscal agent for the grant is located or at least $15.00 per hour, whichever is higher.” Fahs-Beck forbids “salary, fringe benefits, tuition, or living expenses for the applicant” entirely.

Funder	Cap	Indirect costs	Will not pay for	Paid to
Russell Sage	$15,000	“No indirect costs are allowed”	Personal computers “not generally approved.” Student fees and tuition remission	The university
NIJ GRF	$180,000 over up to 3 years	Inside a capped $16,000 per year allowance, alongside tuition and fees	International travel. Applications primarily to buy equipment or supplies	The university
NAEd/Spencer	$30,000	“No funding for overhead or indirect costs”	Nothing specified; spending is at the fellow’s discretion, and unspent funds are returned	Fellow, or the institution by choice
Fahs-Beck	$5,000	Up to 10 percent, counted inside the $5,000	Applicant salary, tuition, living expenses, major travel, conference attendance, major equipment	The applicant organization
Horowitz	$10,000, paid $7,500 then $2,500 on completion	“If processed through an institution, a waiver for overhead is required”	Awards are made to individuals	The individual

I build the budget backwards from the exclusion list, finding what the funder will not cover and then designing the study costs around what is left. Transcription, participant incentives, and statistical consultation survive almost everywhere. Your own time, tuition, conference travel, and equipment are each disallowed by at least one funder here and allowed by at least one other.

Indirect costs usually come out of your cap instead of being added on top of it. Fahs-Beck allows 10 percent and asks the institution to waive it, but if charged it comes out of the same $5,000. NIJ folds institutional overhead into the same $16,000 a year that covers tuition and fees. It is worth asking your sponsored programs office early whether they will waive, because Horowitz requires a waiver in writing and some offices need weeks to produce one.

Payment and taxes

Whether the check goes to your university or to you changes what happens after you win. Money routed through the university arrives as a subaward with institutional rules and payroll attached. Money paid directly to you is fellowship income that nobody withholds tax on, which Harry Frank Guggenheim states plainly: a recipient “may be liable for income taxes on funds awarded,” and the foundation “does not provide advice on tax matters.”

If you are an international student, I would ask your university’s tax or international student office before accepting a direct payment. Treaty and withholding treatment differs from a university-administered subaward, and no funder on this page answers that question.

Nomination deadlines

Some awards cannot be applied for at all, and the deadline that governs your calendar belongs to somebody else.

Award	Who nominates	Cap	Timing
SSWR Doctoral Fellows	A current SSWR member. Self-nomination is permitted	One nominee per nominator	Nomination ID requested by August 4, package due August 25, on the 2026 calendar
Google PhD Fellowship	The university. “Direct applications from students are not accepted”	Four per university in the US and Canada, three in most other regions	Opened March 5, closed April 30 in 2026
P.E.O. Scholar Awards	A local P.E.O. chapter	Not stated	Chapters nominate August 20 to November 20, then the nominee has 45 days
APF COGDOP	The graduate department	One nominee at departments of 100 students or fewer, two at 101 to 200, three above 200	June
GADE student awards	Not stated on the award pages	Not stated	Leadership and Service opens March 15; Research and Teaching open September 15

The SSWR two-step is easy to miss. Requesting a Nomination ID is a separate action three weeks before the real deadline, so a nominator who hears about the award on August 20 is already locked out, which is why I would raise it with an advisor in June.

A per-university cap means an internal competition you have to enter. Google takes four nominees from a US university, so somebody on campus is choosing those four in the fall for an April deadline, and the first task is finding out who makes that choice. It is worth asking your graduate program office which internal competitions exist, and when, before you need one.

AERA Dissertation Grants are the exception here. They appear on most nomination round-ups and are a direct application, with an advisor letter and no institutional gatekeeping step.

Cycles move and internal deadlines are rarely posted anywhere, so a published list of dates only gets you part of the way. What has worked for me is a calendar of my own, with a row for each award and columns for the things that are not the funder’s deadline: who actually submits, what has to happen on campus first, and what has to be finished before the package can go out. I fill it in once a year, in June, at the same time I ask letter writers.

Small awards

A $1,000 grant covers participant incentives for a qualitative study and a $3,000 award covers transcription. The application is a few pages, the applicant pool is limited to people in your field, and the CV entry counts as external funding at any size.

I think the order matters. Winning a small award early gives you a funded-project line, a real budget you have administered, and a reference who has seen you deliver, which are the three things a larger application asks you to demonstrate.

The NIH fellowship

If your work is health-related and you are a citizen or permanent resident, the F31 is the largest thing on this list. One thing to check before reading any guide to it: NIH rewrote its fellowship application and review process for due dates on or after January 25, 2025. Five scored criteria became three, two required attachments disappeared, and grades are no longer allowed in the biosketch. Guides written before that date describe a review that no longer happens.

The reviewer guidance names the three: Candidate’s Preparedness and Potential, Research Training Plan, and Commitment to Candidate. Your sponsor writes the third one, producing five statements inside a six-page limit. So the most useful thing you can do for your own score is make that section easy to write, early. I would ask two months out, hand over your aims page and training timeline and the list of five statements they owe, and draft the factual parts for them: how many students they have mentored, what those students went on to do, what the department actually provides.

The Specific Aims page is one page and is not a scored criterion on its own. It is still the first thing every reviewer reads, and it sets how they read the six-page Research Strategy behind it. I write it first and circulate it alone. A single page is short enough that busy people will actually read it and tell you the project is unclear, and month one is the right time to hear that.

NIH also accepts one resubmission, the A1, and no more. Your score appears in eRA Commons within days of the review meeting, and the summary statement with the actual critiques follows about a month later, so the score arrives weeks before the critiques that explain it. I would wait for the statement before deciding what to do about it.

Advisors will tell you the F31 requires candidacy and a defended proposal. The parent announcement does not say that; its eligibility text asks only that the candidate “be currently enrolled in a PhD or equivalent research degree program.” The bar is real anyway, set by sponsors who will not sign a thin application and by study sections that expect a defended project. The requirement is informal, so the people to convince are your sponsor and the study section.

AI disclosure

This is recent enough that most advice predates it. NOT-OD-25-132, effective for applications submitted on or after September 25, 2025, says NIH “will not consider applications that are either substantially developed by AI, or contain sections substantially developed by AI, to be original ideas of applicants,” with referral to the HHS Office of Research Integrity possible. The same notice caps a principal investigator at six applications per calendar year.

Foundations are asking for disclosure instead. Newcombe’s phase one already includes an AI-use disclosure among its required components, so it is worth deciding your own answer before you are drafting under deadline.

Funded examples

Reading a funded application taught me more than any guide. NIAID posts full funded applications with their summary statements, including three F31s, each with the reviewers’ own comments attached.

All of them are biomedical. There is no publicly posted funded application from a social work doctoral student. The structure still transfers (how an aims page is built, how a sponsor statement reads, what reviewers said about something that got funded) and the science does not. If your program keeps successful applications from previous students, that file is worth more than anything public, so it is worth asking for it.

The rest of the free reading: NIH’s general grant writing tips, its advice on application sections, NIMH’s grant writing page, the All About Grants podcast, SSWR’s doctoral student writing page, and Purdue OWL on grant writing. Success rates by mechanism and year are in the NIH Data Book, and open opportunities across every federal agency are searchable at grants.gov.

Reference tables

NIH fellowship page limits, from the official table for the current FORMS-I package.

Section	Limit
Project Summary/Abstract	30 lines
Project Narrative	3 sentences
Research Training Project Specific Aims	1 page
Research Training Project Strategy	6 pages
Candidate’s Goals, Preparedness, and Potential	3 pages
Training Activities and Timelines	3 pages
Training in the Responsible Conduct of Research	1 page
Sponsor(s) Commitment	6 pages
Introduction to Resubmission	1 page

Your own three pages hold four personal statements: professional and fellowship goals, fellowship qualifications, a self-assessment, and a statement of scientific perspective. Training in the Responsible Conduct of Research is required and, per the reviewer guidance, explicitly not considered in the overall impact score.

FY2026 amounts, from NOT-OD-26-044, for awards made on or after October 1, 2025.

Item	Amount
Predoctoral stipend	$29,364 per year, or $2,447 per month
Tuition and fees	60 percent of actual cost, up to $16,000 per year
Institutional allowance	$4,750 at non-federal public, private, and nonprofit institutions