How to present a research proposal: a 12-slide outline, timing, defense questions and a worked example

Disclosure: NOXI publishes this guide and is one of the tools it mentions; the method works in any presentation software.

This guide covers the three versions people search for: the thesis or dissertation proposal defense (also called the prospectus defense) in a master's or PhD programme, the research proposal presented for a module grade in an MSc, MPH or MBA, and the short internal grant pitch. It gives what three published programme handbooks actually require, a minute-by-minute timing table for 10, 15 and 20-minute slots, a 12-slide outline, the 12 questions committees ask mapped to the slide that answers each, a sample-size slide with the finite population correction worked through, a feasibility slide and one fictional MPH proposal built end to end. If you are defending a finished thesis instead, our thesis defense presentation guide is the better fit.

What a research proposal presentation has to prove

A research proposal presentation has to prove two things to a committee: that the research question is worth answering, and that you can answer it with the design, data, time and money available to you. Everything on the slides should serve one of those two proofs, and most of the time should go to the second one.

The published handbooks describe the same job in their own words. The MIT–WHOI Joint Program's marine geology and geophysics handbook lists the issues a proposal discussion should cover: the scientific merit, originality and viability of the research, whether the student's background is adequate, and whether the necessary financial, logistical and intellectual resources will be available (MIT–WHOI Joint Program, Thesis Proposal Defense/Presentation). Three of those five issues are about feasibility, not about the topic. The University of Wisconsin–Madison's kinesiology handbook calls the dissertation proposal a formal checkpoint to make sure the committee agrees on the research plan (UW–Madison, Kinesiology Graduate Program Handbook 2026–2027).

So the quotable rule of this guide is: spend about a third of the talk on the method and about a quarter on ethics and feasibility; the committee approves plans, not topics. That is our editorial guidance, built from the handbooks above, not a published standard. A proposal talk that spends five minutes on the literature and one minute on the timeline is defending the wrong thing.

Proposal defense vs final defense vs module proposal vs grant pitch

A proposal defense decides whether a plan may go ahead, a final thesis defense decides whether finished research earns the degree, a module proposal is graded coursework, and a grant pitch decides whether money is released. The slides change because the decision changes: a proposal talk has no results to show, so the method, the risks and the plan carry the weight that results carry later.

Two practical consequences follow. A committee that has read your proposal does not need the literature retold; it needs to see that you can defend your choices. A module audience usually has not read anything, so a module proposal spends a little more of its first two minutes on context. For a short academic talk about finished results, the conventions in our guide to turning a research paper into a conference presentation apply instead.

How long is a proposal defense? What three handbooks say

A proposal presentation usually runs 15 to 30 minutes and is followed by a longer question session, but the only length that counts is the one in your programme's handbook. The three handbooks we opened on 16 October 2026 range from about 20 minutes of talk to a session of about two hours, and one of them leaves the length to the student and advisor.

The MIT–WHOI figures come from the joint program's handbook extracts, which quote each committee's own rules side by side, so they show how much the format varies even inside one programme. The Rutgers figures come from its dissertation proposal defense checklist, which also says the defense can go ahead with three of the four committee members present. Before you plan a single slide, find four facts in your own handbook: the talk length, the question time, when the written proposal is due, and the possible outcomes.

Two rules hold whatever the slot. Rehearse to about 90% of the limit, because a live run is almost always slower. And when the question session is two or three times longer than the talk, as in the JCBO example, your backup slides matter as much as the main deck.

Timing by slot: 10, 15 and 20 minutes

In a 15-minute proposal slot, spend 2 minutes on the problem, gap and question, 2 on the literature, 1.5 on aims and framework, 5 on the method, 1 on ethics, 2.5 on the timeline, budget and risks, and 1 on the contribution. The table scales the same proportions to 10 and 20 minutes and shows the slide count for each block at 15 minutes.

The method share is 3.5 of 10 minutes (35%), 5 of 15 (33.3%) and 6.5 of 20 (32.5%): about a third in every slot. Ethics plus feasibility is 2.5 of 10 (25%), 3.5 of 15 (23.3%) and 5 of 20 (25%): about a quarter. The literature never gets more than 2.5 minutes, because a committee that has read your proposal will ask about the literature if it wants more.

For a 20-minute slot, use 14 slides: split the sample slide into "sample and setting" and "sample size", and give risks and plan B a slide of their own. For a 10-minute slot, use 10: fold the framework into the questions slide and data collection into the analysis slide. The general pacing rules, including when to go faster than a slide a minute, are in our slides-per-minute guide. Write the cumulative time at the end of each block on your cue cards (at 15 minutes: 2:00, 4:00, 5:30, 10:30, 11:30, 14:00, 15:00) and check the clock at 10:30. If you are more than 30 seconds behind there, shorten the contribution, never the feasibility slide.

The 12-slide research proposal outline, slide by slide

A 15-minute research proposal presentation fits 12 slides: title and question, problem and gap, literature, aims and research questions, conceptual framework, design, sample and setting, data collection, analysis plan, ethics, timeline and feasibility, and expected contribution. Each slide answers one thing the committee needs to approve, and each title states that answer as a sentence.

Three details make the outline work. Slide 1 carries the question, not just the topic, so the committee knows what you are asking from the first second. Slide 7 puts the sample-size formula on the slide, because "is the sample big enough?" is the most predictable question in any quantitative proposal. And slide 12 ends with an explicit request, which turns the close into the decision the meeting exists to make. Keep slide 12 on screen during questions; our guide on how to end a presentation explains why a bare "Thank you" wastes that space.

Qualitative and mixed-methods proposals keep the same 12 slides with different contents: slide 7 shows the sampling strategy and the reasoning for the number of participants (for example, purposive sampling until no new themes appear, with a planned range), slide 8 the interview or observation guide, and slide 9 the analytic approach, such as thematic analysis with a named framework and how coding will be checked.

The gap slide and the research questions

A gap slide states, in one sentence the committee could repeat, what is not yet known and why that matters; the research questions slide then turns the gap into two or three questions your design can answer. If the committee cannot repeat your gap after slide 2, nothing later in the talk will feel necessary.

Write the gap with this frame: "[What is known] is well established for [population or setting], but [what is missing] for [your population or setting], so [the consequence]." Then check that each research question answers part of the gap and that the design on slide 6 can answer each question. Questions, hypotheses and objectives are different things, and committees notice when a slide mixes them.

Two or three questions is the right number for a master's proposal; a PhD proposal may carry three studies with a question each. More than four usually means the scope is too wide for the timeline, which is exactly what slide 11 will expose.

The literature slide: three themes and the gap

The literature slide in a proposal presentation shows three themes from the existing research, two or three key sources for each, and the gap they leave, all on one slide. It is not a list of 20 citations; it is the argument that leads to your research question, and the full review stays in the written proposal.

Build it as three columns or three rows. Each theme gets a one-line finding in plain words ("Ward nurses score well on when to clean hands, less well on how long"), the two or three studies behind it as short citations, and a mark showing whether outpatient settings were included. The gap appears under the three themes as one sentence, the same sentence as slide 2. If a theme does not lead to the gap, it belongs in the written proposal, not on the slide.

Cite on the slide the way your discipline cites in print, in small type, and keep the full references for a backup slide or the handout. When AI helps you draft any slide, check every citation against the original paper before you present it; our guide to AI presentations with research citations explains how to keep sources traceable.

Method slides: design, sample, instruments and analysis

The method section of a proposal presentation needs four slides, one each for the design, the sample and setting, the data collection and the analysis plan, and each must say why this choice fits the research questions. Committees approve a method when they can see the alternatives you considered and why you rejected them.

• Design. Name it in the discipline's terms (cross-sectional survey, randomised trial, case study, quasi-experiment, ethnography) and add one line on the rejected alternative: "An intervention study would answer question 2 more strongly, but needs two years and a control clinic."
• Sample and setting. The population, the sampling frame (a staff list, a register, a set of archives), the sampling method and the size with its formula, plus the named person or body who has agreed to give access.
• Data collection. The instrument, where it comes from and how it was validated, any adaptation, the pilot, the mode (paper, online, interview) and the dates.
• Analysis plan. One row per research question: the variable, the statistic or analytic approach, and what result would answer the question. A table works better than bullets here.

Draw the design as a simple flow rather than a paragraph: population, sampling, data collection, analysis, output. A diagram lets the committee point at a box when it asks a question, which keeps the discussion on your plan. For charts of pilot data or expected distributions, the rules in our guide on how to present data apply.

The sample-size slide: formula, inputs, correction and non-response

A sample-size slide shows the formula, every input with its source, the result, any correction for a small population, and how many people you must invite to reach that number after non-response. For a survey estimating a proportion, Cochran's formula with 95% confidence, a margin of error of 5% and p = 0.5 gives 385 responses before any correction.

The formula, as given in the University of Florida's extension guide to determining sample size, is n₀ = Z² × p × q ÷ e², where Z is 1.96 for 95% confidence, p is the expected proportion (0.5 when unknown, the most cautious choice), q = 1 − p, and e is the margin of error. With those inputs, 1.96² × 0.5 × 0.5 ÷ 0.05² = 3.8416 × 0.25 ÷ 0.0025 = 384.16, rounded up to 385. The same guide gives the correction for a small population, n = n₀ ÷ (1 + (n₀ − 1) ÷ N), where N is the population size, and notes that researchers often add 10% for people they cannot contact and often increase the sample by 30% to allow for non-response (UF/IFAS, Determining Sample Size, PEOD6).

The correction matters most in exactly the studies students propose: one hospital, four clinics, a school district, a company. In the worked example below, skipping it would have made the plan look infeasible. A first draft took 385, added a 15% buffer (385 ÷ 0.85 = 452.9, so 453 invitations), and found that 453 of 520 nurses, 87.1% of everyone, would have to be invited. The corrected calculation needs 222 responses and 318 invitations at a 70% response rate, 61.2% of the nurses.

Show the inputs and their sources on the slide in small type: the confidence level, the margin of error, the value of p and why, the population size from the staff list, and the expected response rate from a named earlier survey. Two extra lines answer the follow-up questions. First, the precision check: with 222 of 520, the margin of error is 1.96 × √(0.25 ÷ 222 × 298 ÷ 519) = ±5.0%, so the target holds. Second, what the size was calculated for: if a second question compares groups, say whether the sample was also sized for that comparison or whether that question is exploratory and will be reported with confidence intervals.

Trials, experiments and comparisons between groups use a power calculation instead (the effect size you expect, the significance level, the power, usually 80% or 90%, and the resulting number per group). The slide pattern is the same: formula or software, inputs with sources, result, adjustment for drop-out, invitations.

The ethics and data protection slide

The ethics slide names the approvals the study needs and when they will be obtained, how participants will give informed consent, how their identity and data will be protected, and when data will be deleted. Studies with human participants normally need approval from a research ethics committee or institutional review board (IRB) before any data collection starts, so the approval also belongs on the timeline.

In the United States, research with human participants at institutions that follow the federal rules falls under the HHS regulations at 45 CFR 46. HHS describes five subparts: subpart A, the Common Rule, revised in the "2018 Requirements"; subparts B, C and D with additional protections for pregnant women and fetuses, prisoners and children; and subpart E on IRB registration (HHS Office for Human Research Protections, 45 CFR 46). In the United Kingdom, studies involving the NHS may need review by an NHS research ethics committee, which, in the Health Research Authority's words, gives "an opinion about whether the research is ethical"; the HRA provides a decision tool for checking whether a study needs one (Health Research Authority, Research Ethics Committee review). Your university's own ethics office tells you which route applies to your study.

Keep the slide to five lines and put the full consent form in the written proposal. The question committees ask most often here is about pressure: if your participants work for, study with or are treated by the people helping you recruit, say how you separate recruitment from that relationship.

Feasibility: timeline, access, budget, risks and plan B

A feasibility slide shows that the study can be finished on time and on budget: the months for each phase, the named access you already have, the cost line by line, and the two or three risks most likely to delay you, each with its plan B. Committees approve proposals with risks; they send back proposals whose risks the student has not noticed.

Draw the timeline as phases in months, not as a list of dates, and put the ethics approval on it as a phase with its own buffer, because the review can take longer than the form suggests and no data collection can start before it ends. Name the access: "The clinic network's nursing director has agreed in writing to share the staff list" is evidence; "the clinics are supportive" is not.

Three risks are enough. Each needs a trigger you can observe, not a feeling, so that the committee sees you will know when to switch. If your study depends on one gatekeeper, one machine or one dataset, that dependency is your first risk.

The 12 questions committees ask, and the slide that answers each

Proposal committees ask predictable questions about significance, novelty, design, sample, access, instruments, analysis, ethics, time and risk, and every one of them should be answered by a slide you can return to in a click. Map each question to its slide before you rehearse, and the question session becomes a tour of your own plan.

Answer each question in three parts: the direct answer in one sentence, the evidence or slide behind it, and what it means for the plan. For question 5 in the worked example: "Yes: 222 responses give a 5% margin of error for 520 nurses. The calculation is on slide 7. If returns fall below half the target by week 3, I invite all 520." That takes about 15 seconds. Committees also ask questions to test the plan rather than the student, and "how would you improve this?" is an invitation, not an attack: the JCBO handbook says the defense will likely include discussion of how the research plan might be improved. Write down every suggestion during the session.

Prepare a backup slide for any question whose full answer does not fit its main slide: the detailed sample-size inputs, the questionnaire items, the full budget, the rejected designs. Number your slides so you can jump straight to one when asked. The thesis defense guide has a fuller Q&A rehearsal method that works for proposals too.

Preliminary results or a pilot: show them without overclaiming

Show preliminary results or a pilot in a proposal presentation as evidence that the plan works, not as findings: what you tested, on how many people or samples, what it told you about feasibility, and what you changed. A pilot of 20 participants proves the questionnaire takes 12 minutes; it does not prove the hypothesis.

• Say what it was for. "The pilot tested timing, wording and response rate", not "early results".
• Give n on the slide. Any percentage from a pilot gets its denominator next to it.
• Show the change it caused. "Two items were reworded after 6 of 20 nurses misread them" is the most persuasive pilot line there is.
• Keep effect estimates off the main slide. If you must show one, label it as a pilot estimate with its interval and put it in a backup slide.

The JCBO handbook asks students to discuss the proposed research "including any preliminary results", so a pilot slide is welcome where you have one. Where you do not, a sentence on how and when you will pilot the instrument does the same job.

Worked example: an MPH proposal, timed and calculated

The worked example below applies every table in this guide to one fictional proposal: an MPH student with a 15-minute slot and 30 minutes of questions, proposing a cross-sectional survey of hand-hygiene knowledge among nurses in four outpatient clinics. Every number is invented for teaching, and every calculation is shown so you can rerun it with your own inputs.

Population. 160 + 140 + 120 + 100 = 520 nurses, from the clinic network's staff list, which the nursing director has agreed in writing to share.

Sample size, first draft. Cochran: 1.96² × 0.5 × 0.5 ÷ 0.05² = 384.16, so 385. With a 15% buffer: 385 ÷ 0.85 = 452.9, so 453 invitations, or 453 ÷ 520 = 87.1% of all nurses. Dr Ortiz's first comment: that is nearly a census, and it ignores the population size.

Sample size, corrected. Finite population correction: 384.16 ÷ (1 + 383.16 ÷ 520) = 384.16 ÷ 1.7368 = 221.2, so 222 completed surveys. Precision check: 1.96 × √(0.25 ÷ 222 × 298 ÷ 519) = ±5.0%.

Invitations. The clinics' 2025 staff survey had a 70% response rate, so 222 ÷ 0.70 = 317.1, rounded up to 318 invitations, chosen at random from the staff list: 318 ÷ 520 = 61.2% of nurses.

Plan B. If fewer than 111 surveys (half of 222) are back by the end of week 3 of the 6-week collection window, invite the remaining 520 − 318 = 202 nurses. A census reaches the target as long as at least 222 ÷ 520 = 42.7% respond, well below the 70% expected.

Analysis. Question 1: the proportion with a score of 75% or more, with its 95% confidence interval. Question 2: a chi-square test of good knowledge by training in the last 2 years, then logistic regression adjusted for clinic and years of experience. The sample was sized for question 1; question 2 is reported with odds ratios and confidence intervals, and the slide says so.

Timeline (12 months). Literature and instrument adaptation, months 1–2 (2); ethics approval and a pilot with 20 nurses outside the sample, months 3–4 (2); data collection, months 5–7 (3); analysis, months 8–9 (2); writing and the final defense, months 10–12 (3). Total 2 + 2 + 3 + 2 + 3 = 12 months.

Budget. Printed questionnaires 318 × $0.40 = $127.20; pilot copies 20 × $0.40 = $8.00; a prize draw of three $100 vouchers = $300.00. Total $127.20 + $8.00 + $300.00 = $435.20. Plan B adds 202 × $0.40 = $80.80, for a maximum of $516.00.

Time split (15 minutes). Problem, gap and question 2; literature 2; aims, questions and framework 1.5; method 5; ethics 1; timeline, budget and risks 2.5; contribution 1. Total 15. Method share 5 ÷ 15 = 33.3%; ethics and feasibility 3.5 ÷ 15 = 23.3%.

The request on slide 12. "I ask the committee to approve this plan so that I can submit it for ethics review in month 3." The full deck is the 12-slide outline above, with titles from its example column. The shorter 10-slide version built in NOXI folds the literature into the gap slide and the framework into the questions slide, and gives the sample size and the budget with its risks a slide each, so the sample-size and cost arithmetic above sits on two slides.

Here are two slides from this article's worked example, as NOXI built them, unedited.

The grant-pitch variant: five slides for a funding panel

A research grant pitch compresses the proposal into five slides: the problem, the aims, the approach, the team, and the budget with what the funder gets. A funding panel decides whether the money will produce the promised result, so feasibility and the team move to the centre, and the literature shrinks to one line on the problem slide.

Follow the funder's own form when it has one: its headings are the slide titles, in its order. A panel scoring 20 applications in an afternoon rewards the one that maps cleanly to its criteria.

After the defense: the three outcomes and the revision memo

A proposal defense usually ends with one of three outcomes: accept, accept with revisions, or revise and defend again. Rutgers' social work checklist names three votes, accept, accept with revisions or reject, taken while the student is out of the room, after which a rejected proposal is rewritten and defended again. The MIT–WHOI marine geology handbook asks the committee to agree a timetable when a plan must be modified.

The fastest way through "accept with revisions" is a revision memo: a one-page table with each committee comment, your response and where in the proposal you changed it. Send it to your supervisor within a week, while the discussion is fresh. If the outcome is to present again, the memo becomes the first backup slide of the second defense, so the committee sees each concern answered in order.

Using AI for a proposal presentation: check your programme's policy first

Whether you may use AI to build a proposal presentation depends on your university's and your programme's policy, and that policy wins over any tool's capabilities. Where it is allowed, use AI to turn your own written proposal into a first draft of the slides, then check every number, citation and claim against the proposal yourself and disclose the use if you are asked to.

The proposal document is the single source of truth: the deck should carry no number, source or design choice that is not in it. Our guide to turning a Word document into a PowerPoint with AI explains what to check after any conversion. The three prompts below assume your programme allows it; they work in NOXI or in any AI presentation tool. Replace the brackets and attach your proposal.

How to build the proposal deck in NOXI

NOXI turns a written research proposal into a structured deck, or builds it in your department's style from a template you upload, and you then check it line by line against the proposal. It is free, with no card and no watermark; uploading files needs a free account, and a full deck uses about 25 of the 1,000 monthly credits.

Frequently asked questions

How do you present a research proposal?

Present it as a plan the committee can approve: state the problem and the gap, summarise the literature in three themes, give two or three research questions, then spend about a third of the time on the design, sample, data collection and analysis, and about a quarter on ethics, the timeline, the budget and the risks. End with the expected contribution and an explicit request for approval.

How many slides should a research proposal presentation have?

Plan about 12 slides for a 15-minute slot, 10 for 10 minutes and 14 for 20 minutes. The 12 are title and question, problem and gap, literature, aims and questions, framework, design, sample and setting, data collection, analysis plan, ethics, timeline and feasibility, and contribution. Put detail such as the full budget or questionnaire items in numbered backup slides after the end.

How long is a proposal defense?

Your handbook sets it. Published examples include an overview of about 20 minutes followed by questions (Rutgers School of Social Work), a talk under 20 minutes followed by 40 to 60 minutes of questions (one MIT–WHOI Joint Program committee), and a 20 to 30-minute talk in a session of about two hours (another). Rehearse the talk to about 90% of the limit.

What questions are asked in a proposal defense?

Committees ask why the problem matters, what is new compared with the literature, whether the questions are answerable, why this design, whether the sample is big enough, whether you can get access, whether the instruments are valid, how you will analyse the data, what the ethical risks are, whether you can finish on time, what happens if recruitment fails, and what you will claim.

What is the difference between a proposal defense and a thesis defense?

A proposal defense happens before the research and decides whether your plan may go ahead, so it focuses on the gap, the method, ethics and feasibility. A thesis defense happens after the research and decides whether the finished thesis passes, so it focuses on results, their limits and the contribution. At UW–Madison Kinesiology, for example, the proposal committee usually continues as the final defense committee.

What should be included in a research proposal presentation?

Include the research question, the problem and gap, a one-slide literature summary, aims and questions or hypotheses, a conceptual framework, the design and why you chose it, the sample with its size calculation, data collection, the analysis plan, ethics and data protection, the timeline, budget and risks with a plan B, and the expected contribution with your request to the committee.

How do you present methodology in a presentation?

Use four slides: design, sample and setting, data collection, and analysis plan. On each, say why the choice fits the research questions and name the alternative you rejected. Draw the design as a flow from population to output, put the sample-size formula and inputs on the sample slide, and show the analysis plan as a table with one row per question.

How do you show a research timeline on a slide?

Show phases in months rather than a list of dates: for example literature 2 months, ethics approval and pilot 2, data collection 3, analysis 2 and writing 3, for 12 months in total. Put the ethics approval on the timeline as its own phase with a buffer, and add the two or three risks that could delay you, each with a trigger and a plan B.

Can I use AI to make my proposal presentation?

Only if your university's and programme's policy allows it, and the policy wins over any tool. Where it is allowed, let AI turn your own written proposal into a first draft of the slides, then check every number, citation and design choice against the proposal yourself, and disclose the use if you are asked. The deck should contain nothing that is not in your proposal.

Sources & methodology

How this guide was made: the presentation lengths, question times, submission deadlines and outcomes come from three programmes' own published rules (the MIT–WHOI Joint Program's handbook extracts, the Rutgers School of Social Work proposal defense checklist and the UW–Madison Kinesiology handbook), which are examples, not a standard; the sample-size formula, the finite population correction and the non-response guidance come from the University of Florida's extension guide; the ethics rules come from HHS's Office for Human Research Protections and the UK Health Research Authority. The timing table, the 12-slide outline, the question map, the risk table and the grant-pitch outline are editorial recommendations. Amira Haddad, Westbridge University, Harbourview Health and all their figures are fictional; every calculation was checked by hand. All sources were checked on 16 October 2026.

• [1] Proposal defense under 20 minutes followed by 40–60 minutes of committee questions (JCBO); a 20–30 minute presentation in a session of about 2 hours, a proposal of no more than 10 text pages submitted 3 weeks ahead, "not an examination", and the issues the discussion should cover (JCMG&G); defense about 2 weeks after submission with length agreed with the advisor (JCPO) — MIT–WHOI Joint Program, Thesis Proposal Defense/Presentation (handbook extracts).
• [2] An overview of about 20 minutes, a defense that is not public, three of four committee members required, and the votes accept, accept with revisions or reject — Rutgers School of Social Work, Dissertation Proposal Defense Checklist (2019).
• [3] Written proposal to the committee at least 7 days before the oral presentation, format set by the advisor, the proposal as a formal checkpoint for committee consensus, the proposal committee usually continuing as the final defense committee, and a formal assessment rubric — University of Wisconsin–Madison, Kinesiology Graduate Program Handbook 2026–2027: Dissertation Proposal.
• [4] Cochran's formula n₀ = Z²pq ÷ e² with the example of 385, the finite population correction n = n₀ ÷ (1 + (n₀ − 1) ÷ N), and the practice of adding 10% for non-contacts and 30% for non-response — Glenn D. Israel, Determining Sample Size, PEOD6, UF/IFAS (revised April 2009, reviewed June 2013).
• [5] The five subparts of 45 CFR 46, the Common Rule's 2018 Requirements, and research begun before 21 January 2019 — U.S. Department of Health and Human Services, Office for Human Research Protections, 45 CFR 46 (content last reviewed 11 February 2025).
• [6] NHS research ethics committees give an opinion on whether research is ethical, and the decision tool for whether a study needs one — Health Research Authority, Research Ethics Committee review.

Build your proposal deck in NOXI, free: upload your written proposal, match your department's template, and export an editable PowerPoint or PDF for the committee.