Planning a Capstone Project You Can Finish

A finishable capstone project is not the biggest project you can imagine.

It is the strongest project you can complete, revise, and submit within the time, access, evidence, and methodological constraints you actually have.

That distinction matters because students often begin capstone planning with an intellectual question:

What would be interesting to research?

I recommend adding a second question immediately:

What version of this research can I complete responsibly by the deadline?

The two questions have to be answered together.

A promising capstone topic can still become an unmanageable project if it requires:

• more participants than you can recruit;
• data you cannot access;
• a method you have not learned;
• ethics approval you have not budgeted time for;
• a literature base too large to review adequately;
• a comparison across too many populations or locations;
• several research questions that each deserve their own study;
• extensive supervisor feedback too close to submission;
• a final product much larger than the assignment requires.

So my basic planning formula is:

Research question + evidence + method + access + time + required deliverable = project scope

If one of those variables changes, the scope may need to change too.

Purdue OWL’s current research-planning guidance makes a similar point when it asks researchers to test whether a question is narrow enough for the proposed document and then identify what data and methods would actually be required to answer it: https://owl.purdue.edu/owl/graduate_writing/introduction_to_writing/documents/getting-started-with-grad-writing/articulating-a-research-agenda.pdf

George Mason University’s Writing Center likewise describes an effective research question as focused enough to be answered thoroughly within the space available: https://writingcenter.gmu.edu/writing-resources/research-based-writing/how-to-write-a-research-question

For a capstone, I would expand “space available” to include the entire project environment:

word count, weeks, skills, approvals, access, supervision, data, and revision time.

That is what this guide will help you plan.

Start With the Deadline, Not the Topic

Most students naturally plan forward.

The problem is that long projects contain dependencies.

You cannot analyse data you do not yet have.

You may not be able to collect data before receiving approval.

Your advisor cannot give useful feedback on a chapter you have not drafted.

You cannot make final revisions if your “final draft” is completed the night before submission.

So I prefer backward planning.

Write the final deadline at the bottom of the page.

Then ask:

What must be complete immediately before this?

Then:

What must be complete before that?

Continue until you reach today.

A simplified sequence might look like:

Final submission ← formatting and final proofread ← final revisions ← supervisor feedback ← complete draft ← analysis and findings ← data collection ← approval/access/recruitment ← method finalized ← literature review and research question ← preliminary research ← topic exploration

Cornell Graduate School explicitly recommends understanding the sequence of draft submission, examination, revisions, approvals, and final submission in order to establish a realistic timeline: https://gradschool.cornell.edu/academic-progress/degrees-fields/research-degree-requirements/thesis-dissertation/writing-your-thesis-dissertation/understanding-deadlines-and-requirements/

Your capstone may not require a thesis defense or graduate-school submission, but the planning principle is transferable:

The due date is the end of a chain, not the only date that matters.

Identify the Real Deliverable Before You Design the Research

“Capstone project” can mean very different things.

Before developing the research question, determine what your program actually expects.

Your deliverable might be:

• a research paper;
• an empirical study;
• a systematic or structured literature review;
• a policy analysis;
• a case study;
• a program evaluation;
• a business plan;
• a design project;
• an engineering prototype;
• an educational intervention;
• a clinical practice project;
• a portfolio;
• a report plus presentation;
• a report plus poster;
• a thesis-style document.

Do not assume the most complex option is the most impressive.

The assessment criteria matter more.

Read:

• the assignment brief;
• rubric;
• capstone handbook;
• departmental guidance;
• ethics requirements;
• formatting rules;
• presentation requirements;
• examples of previous projects if your program provides them.

Then write a one-sentence deliverable statement.

For example:

By May 1, I must submit a 7,500-word evidence-based policy analysis evaluating two municipal approaches to reducing heat risk among older adults, using published research and public data.

That sentence already constrains the project.

You know:

• what you are producing;
• approximate length;
• type of analysis;
• number of policies;
• population;
• evidence source;
• deadline.

This is much more useful than:

My capstone is about climate change and public health.

The Research Question Is a Scope Control Tool

Students often treat the research question as something they write for the introduction.

I treat it as one of the main project-management tools.

A good research question tells you:

• what belongs in the project;
• what does not;
• what evidence you need;
• which methods are relevant;
• what literature you should prioritize;
• how large the analysis will become.

George Mason recommends that a research question be clear, focused, concise, complex, and arguable: https://writingcenter.gmu.edu/writing-resources/research-based-writing/how-to-write-a-research-question

For a capstone, I add one more criterion:

Feasible.

A question can be academically excellent and still be inappropriate for your project.

Test the Question Against Five Constraints

Before committing to a question, test it against:

1. Time
2. Evidence
3. Access
4. Method
5. Deliverable size

Consider this question:

How does social media affect mental health among university students?

It is broad in several ways.

“Social media” could mean:

• TikTok;
• Instagram;
• Reddit;
• messaging;
• passive browsing;
• active posting;
• total screen time;
• appearance-focused use;
• social comparison;
• algorithmic recommendations.

“Mental health” could mean:

• anxiety;
• depression;
• loneliness;
• body dissatisfaction;
• stress;
• sleep;
• well-being.

“University students” could mean:

• undergraduates;
• graduate students;
• international students;
• first-year students;
• students at one institution;
• students across several countries.

The question is not difficult because it lacks potential material.

It is difficult because it contains too much.

A narrower version might be:

How is appearance-focused Instagram use associated with body dissatisfaction among female undergraduate students aged 18–24?

Whether that question is appropriate depends on the project.

If you are doing a literature review, it may be workable.

If you plan to recruit participants and collect new data in eight weeks, you need to test feasibility much more carefully.

Narrow by Variable, Population, Place, Period, or Comparison

When a question is too broad, I usually narrow it along one or more dimensions.

Variable

Population

Place

Time Period

Narrower: Hybrid-work productivity studies published after 2020

Comparison

Outcome

Evidence Type

You do not need to narrow along every dimension.

The goal is to reach a question that your available evidence and method can actually answer.

Do Preliminary Research Before Freezing the Question

There is a common planning mistake:

Choose the exact question first. Then search for evidence.

Sometimes that works.

Sometimes you discover three weeks later that the evidence you need barely exists.

I recommend doing a preliminary research pass first.

Purdue’s “Creating a Research Space” framework encourages researchers to examine the existing scholarship, identify a territory or topic, notice a gap or unresolved question, and then establish the research niche: https://owl.purdue.edu/owl/graduate_writing/introduction_to_writing/documents/getting-started-with-grad-writing/cars-model-activity.pdf

George Mason similarly recommends preliminary research before finalizing a research question: https://writingcenter.gmu.edu/writing-resources/research-based-writing/how-to-write-a-research-question

During the preliminary search, you are not trying to read everything.

You are testing the project.

Ask:

• Is there enough scholarly literature?
• Is there too much literature?
• Which terms do researchers actually use?
• What methods dominate the field?
• Are recent reviews available?
• Is the question already answered more directly than I expected?
• Are the populations I care about represented?
• Can I access the key studies?
• Does the evidence support the comparison I planned?

This early search can save enormous amounts of time.

Make a Feasibility Map Before You Write the Proposal

I recommend putting the project on one page.

Use six boxes.

1. Research Question

Write one primary question.

If you have five primary questions, you may have five projects.

2. Evidence

What evidence will answer the question?

Examples:

• peer-reviewed articles;
• public datasets;
• interviews;
• survey responses;
• archival documents;
• policy documents;
• financial records;
• observational data;
• experiments;
• case materials;
• technical measurements.

3. Method

What will you do with that evidence?

Examples:

• thematic analysis;
• statistical analysis;
• comparative analysis;
• content analysis;
• literature synthesis;
• policy evaluation;
• case-study analysis;
• design testing.

4. Access

What do you need permission or practical access to obtain?

Examples:

• participants;
• organization records;
• software;
• laboratory equipment;
• proprietary databases;
• a research site;
• supervisor approval;
• institutional ethics review.

5. Deliverables

What must you submit?

Include:

• written report;
• appendices;
• presentation;
• poster;
• prototype;
• code;
• dataset;
• reflection;
• executive summary.

6. Deadline

How many usable research weeks remain?

Not calendar weeks.

Usable weeks.

If you have 12 weeks but:

• one is spring break;
• two contain final exams;
• one includes another major assignment;
• your supervisor needs one week for feedback;

you do not have 12 equivalent research weeks.

Planning should reflect the real calendar.

Use a “Minimum Viable Capstone”

I use this phrase deliberately.

Your minimum viable capstone is the smallest version of the project that:

• answers a meaningful question;
• meets the rubric;
• uses an appropriate method;
• demonstrates academic rigor;
• can be completed by the deadline.

This is not the weak version.

It is the core version.

Suppose your original plan is:

Interview 40 students at four universities and compare how first-generation status affects attitudes toward AI-assisted writing.

Your minimum viable version might be:

Conduct 12–15 interviews at one institution and examine how first-generation students describe the role of AI tools in drafting and revision.

Or, if participant recruitment creates too much risk:

Conduct a structured literature review of current research on first-generation students, academic writing support, and generative-AI use, then identify gaps for future empirical research.

Whether either option is academically appropriate depends on the program.

The important planning move is that you know what can be removed without destroying the project’s intellectual purpose.

Separate the Core Project From the “Would Be Nice” Project

Create two lists.

CORE

Things required to answer the question.

OPTIONAL

Things that could make the project richer if time permits.

For example:

Core:

• one primary population;
• one main research question;
• literature review;
• one data source;
• one analytical method;
• final written report.

Optional:

• second population;
• additional survey;
• extra interviews;
• geographic comparison;
• supplementary visualization;
• follow-up analysis.

Do not make optional features prerequisites for completion.

This is how projects become impossible to finish.

The project should work even if every optional component disappears.

One Strong Research Question Is Usually Better Than Five Weak Ones

Long projects tempt students to expand.

Because the capstone feels important, the question grows:

What are the causes, effects, perceptions, ethical implications, policy responses, and future trends related to X?

That is not depth.

It is accumulation.

Each research question creates work.

A new question may require:

• new literature;
• new variables;
• new data;
• another method;
• another results section;
• another discussion section.

Purdue’s research-agenda guidance explicitly suggests asking whether a question is narrow enough for the proposed document and whether an oversized question could instead be divided into multiple research projects: https://owl.purdue.edu/owl/graduate_writing/introduction_to_writing/documents/getting-started-with-grad-writing/articulating-a-research-agenda.pdf

That is good capstone advice.

A strong project often has:

one primary question;

and, if necessary,

two or three tightly connected subquestions.

Not seven independent questions sharing the same topic.

Your Method Should Follow the Question

Students sometimes choose a method because it sounds impressive.

“I want to do interviews.”

“I want to run a survey.”

“I want to use machine learning.”

“I want mixed methods.”

Then they try to invent a question that justifies the method.

Reverse that order.

Ask:

What evidence would allow me to answer this question responsibly?

Then:

What method is appropriate for that evidence?

Example:

Question: How do graduating nursing students describe the role of simulation feedback in building clinical confidence?

The topic is similar.

The research design is not.

Method Complexity Has a Time Cost

Every method carries hidden work.

Interviews do not mean:

conduct interviews → write findings.

They may require:

• protocol design;
• ethics review;
• recruitment;
• scheduling;
• consent;
• recording;
• transcription;
• anonymization;
• coding;
• analysis;
• quotation selection;
• secure data management.

Surveys may require:

• item selection;
• instrument design;
• validation considerations;
• pilot testing;
• recruitment;
• response monitoring;
• data cleaning;
• statistical analysis.

Secondary datasets may require:

• data access;
• documentation review;
• cleaning;
• missing-data decisions;
• variable construction;
• software skills.

Literature reviews may require:

• database selection;
• search terms;
• inclusion criteria;
• screening;
• source management;
• synthesis.

Do not estimate the project based only on the visible part of the method.

Estimate the whole workflow.

Treat Ethics Approval as a Dependency, Not an Administrative Detail

If your project may involve human participants, identifiable private information, or another form of regulated research, determine your institution’s requirements early.

Do not assume:

“My project is small, so I probably do not need approval.”

Your institution—not the student—should determine whether formal review, exemption, or another process applies.

For research covered by U.S. HHS human-subject regulations, the Office for Human Research Protections states that nonexempt human-subject research requires IRB approval before research begins: https://www.hhs.gov/ohrp/regulations-and-policy/guidance/faq/investigator-responsibilities/index.html

OHRP’s training materials also make clear that recruitment cannot begin before required approval: https://www.hhs.gov/ohrp/education-and-outreach/online-education/human-research-protection-training/lesson-4-irb-review-of-research/index.html

If you are outside the United States, the terminology and process may differ, but the planning principle remains:

Check the rules before scheduling data collection.

Ethics review can affect:

• timeline;
• recruitment language;
• consent procedures;
• data storage;
• participant eligibility;
• methodology;
• what information you are permitted to collect.

If approval is required, it belongs near the beginning of the project timeline.

Build the Timeline Around Dependencies

A project schedule should not simply list tasks.

It should show what depends on what.

For example:

You cannot finalize interview questions until the research question is stable.

You cannot recruit participants before required approval.

You cannot analyse interviews before enough data have been collected and processed.

You cannot write a final discussion before you know the results.

You cannot submit a final version before receiving and incorporating required feedback.

A practical dependency chain might be:

Question approved ↓ Method approved ↓ Ethics/access approved ↓ Recruitment opens ↓ Data collected ↓ Data cleaned/transcribed ↓ Analysis completed ↓ Findings drafted ↓ Discussion drafted ↓ Full draft sent for feedback ↓ Revision ↓ Formatting and proofing ↓ Submission

When one early stage moves, everything downstream may move too.

This is why early delays are disproportionately expensive.

Plan Backward From a Personal Deadline

If the official deadline is Friday at 11:59 p.m., do not make Friday at 10:30 p.m. your project plan.

Set an earlier personal completion date.

For example:

The exact dates will vary.

The principle will not:

Give the project room to go wrong.

Cornell’s thesis planning guidance builds explicit time between complete drafts, exams, revisions, approvals, and final submission: https://gradschool.cornell.edu/academic-progress/degrees-fields/research-degree-requirements/thesis-dissertation/writing-your-thesis-dissertation/understanding-deadlines-and-requirements/

Duke similarly requires major milestones before final thesis or dissertation submission, including initial submission, advisor review, defense scheduling, and final filing: https://gradschool.duke.edu/academics/milestone-and-graduation-information-and-deadlines/

Your undergraduate or professional capstone may have a simpler process.

It still needs revision space.

Add Buffer Where Uncertainty Is Highest

Not every task deserves the same buffer.

High-uncertainty tasks include:

• ethics approval;
• external permissions;
• participant recruitment;
• interviews;
• equipment access;
• data requests;
• supervisor feedback;
• software problems;
• complicated analysis.

Low-uncertainty tasks include:

• formatting references;
• creating a title page;
• checking heading levels.

Do not distribute your spare time evenly.

Place more buffer around the parts you control least.

For example:

If you need 15 interviews, do not plan to complete Interview 15 the day before analysis starts.

If you need an organization to send a dataset, do not assume it will arrive the same day you request it.

If you need your supervisor to read 7,000 words, do not send the draft 24 hours before you need comments.

Uncertainty should influence the schedule.

Plan by Milestones, Not by “Work on Capstone”

“Work on capstone” is not a useful task.

It gives you no definition of done.

Use milestones.

Examples:

A milestone should produce something visible.

That makes it easier to track progress and detect slippage early.

Use Weekly Outputs

For a semester-long project, I recommend deciding what should exist at the end of every week.

Example:

This is only an illustration.

Your method may require a completely different schedule.

The useful feature is not the exact week.

It is that each week has an output.

Make the Literature Review Serve the Project

Capstone students can lose weeks “reading around the topic.”

Reading is necessary.

Unbounded reading is not a plan.

George Mason describes the literature review as a map of the scholarly conversation that establishes what is known, identifies relationships among sources, and helps clarify what the project can contribute: https://writingcenter.gmu.edu/writing-resources/research-based-writing/writing-a-literature-review

Use that function to control your reading.

Your literature review should help answer:

• What is already known?
• Where do scholars agree?
• Where do they disagree?
• What methods have been used?
• What population or context is understudied?
• What definitions matter?
• What gap or problem justifies my project?
• What evidence will I need to interpret my findings?

Do not collect sources simply because they mention your topic.

A source earns a place because it helps build the research problem, method, analysis, or interpretation.

Set a Literature Search Stop Rule

You may never reach a point where there is literally nothing else to read.

So define what “enough preliminary research” means.

A practical stop rule might be:

I can move from exploration to focused review when:

• I can define the key terms;
• I can identify the major research strands;
• I have found several recent high-quality studies or reviews;
• I understand which methods are commonly used;
• I can explain the unresolved question my project addresses;
• new searches are increasingly returning studies I have already seen.

This does not mean the literature review is finished.

It means you know enough to stop browsing broadly and start researching deliberately.

Build a Literature Matrix Early

For each important source, record fields such as:

Author/year Research question Population/sample Method Key finding Limitations How it relates to my question Where I might use it

This prevents a common capstone problem:

You read 40 papers and remember that one of them contained an important finding, but you no longer know which one.

The matrix also reveals patterns.

You may notice:

• most studies use one population;
• results differ by method;
• a key concept is defined inconsistently;
• older studies dominate;
• several authors cite the same foundational paper.

Those patterns can shape the project itself.

Decide What You Will Not Research

A scope statement should contain exclusions.

For example:

This project examines undergraduate students rather than graduate students.

It focuses on peer-reviewed English-language research published from 2020 onward.

It evaluates two policy approaches rather than attempting a national comparison.

It considers academic-writing applications of generative AI rather than programming or image generation.

It uses publicly available secondary data and does not involve participant recruitment.

These exclusions protect the project.

They tell you what to ignore when interesting but irrelevant material appears.

A strong capstone is partly defined by what it refuses to become.

Write a One-Paragraph Scope Statement

Before deep research begins, write something like:

This project examines [specific problem] among/in [population/context]. It focuses on [variables, cases, texts, policies, or evidence] during [period, if relevant]. The project will use [method/data source] to answer [research question]. It will not attempt to address [major exclusions]. The final deliverable will be [format/length], with analysis completed by [internal deadline] to allow time for feedback and revision.

If you cannot write this paragraph clearly, your project may not yet be scoped.

Test the Project With a “What If?” Audit

Ask what happens if something goes wrong.

What if recruitment is slower than expected?

What if one interviewee cancels?

What if your dataset is incomplete?

What if one database does not provide full text?

What if your initial statistical test is inappropriate?

What if your advisor asks you to narrow the question?

What if ethics approval takes longer than expected?

What if the organization you planned to study withdraws access?

A project is more resilient when you already know the fallback.

For example:

You should not invent a fallback that violates the approved method or assignment.

The point is to identify the project's critical failure points early.

Know Your Critical Path

In project management, the critical path is the sequence of tasks that directly determines the finish date.

For a capstone, your critical path may be:

ethics approval → recruitment → interviews → transcription → coding → findings → discussion → full draft → revision

If recruitment starts two weeks late, the project may finish two weeks late unless something else changes.

Formatting references is probably not on the critical path.

This distinction matters because students sometimes spend large amounts of time optimizing low-impact tasks while the critical work remains blocked.

If your critical path is at risk, reduce scope or solve the dependency.

Do not compensate by making the title page prettier.

Your Supervisor Is Part of the Timeline

Advisor or supervisor feedback is not an emergency service that appears when your draft is finished.

Agree on expectations early.

Cornell’s advising guidance recommends establishing a meeting schedule, preparing materials in advance, maintaining a written plan and timeline, and discussing whether an advisor prefers complete drafts or sections: https://gradschool.cornell.edu/academic-progress/opportunities-resources-support/advising-guide-for-research-students-2025/

Ask early:

• How often should we meet?
• What should I send before meetings?
• Do you prefer sections or full drafts?
• How much turnaround time should I expect?
• Which decisions require your approval?
• When is the latest useful date for sending a full draft?
• Are there periods when you will be unavailable?

These answers affect your project plan.

If your supervisor needs seven days to return feedback, your draft deadline is not the final submission date minus one day.

Use Meetings for Decisions, Not Status Reports

Do not spend half of a 20-minute meeting explaining what you did.

Send a concise update beforehand when appropriate.

Then bring decisions.

For example:

I found that the literature uses three different definitions of student engagement. I think Definition B fits my question because X. Do you agree?

My recruitment is slower than expected. I can either extend recruitment one week or reduce the sample target and deepen the interviews. Which is more defensible?

The original question requires data I cannot access. I have narrowed it from X to Y. Does the revised question still meet the capstone objective?

A supervisor can help more when the problem is visible.

Start Writing Before the Research Is “Finished”

Long projects become dangerous when students divide the semester into:

research phase;

then writing phase.

Writing is part of research.

Cornell Graduate School specifically advises students not to postpone writing while waiting to read or analyse everything, and recommends starting small with outlines and sections: https://gradschool.cornell.edu/career-and-professional-development/pathways-to-success/build-your-skills/tips-and-takeaways/research-and-writing-tips/

You can often draft early versions of:

• background;
• definitions;
• research problem;
• literature review sections;
• methodology;
• search strategy;
• limitations of the planned design.

Writing early reveals gaps.

You may discover:

• the question is still unclear;
• two concepts overlap;
• a section has no evidence;
• the method does not fully answer the question;
• the literature supports a narrower claim.

Those are useful discoveries in Week 5.

They are painful discoveries in the final week.

Build a Skeleton Draft Early

Purdue’s roadmap guidance recommends creating short explanations of the project as a whole and using outlines to clarify what each section needs to do: https://owl.purdue.edu/owl/graduate_writing/documents/creating-a-roadmap.pdf

I recommend creating the complete document structure before you have complete content.

For example:

Title

Abstract [write last]

Introduction

• problem
• significance
• research question
• scope

Literature Review

• theme 1
• theme 2
• theme 3
• gap

Methods

• design
• evidence/data
• inclusion/recruitment
• analysis
• ethics

Findings

• finding 1
• finding 2
• finding 3

Discussion

• answer to research question
• connection to literature
• alternative explanations
• limitations
• implications

Conclusion

References

Appendices

Now the capstone is no longer one enormous blank document.

It is a set of smaller deliverables.

Do Not Give Every Section Equal Time

Some sections are project bottlenecks.

For an empirical study, analysis may require more time than the introduction.

For a policy capstone, assembling and evaluating the evidence may dominate.

For a literature review, search, screening, and synthesis may take more time than final prose.

For a design project, testing and iteration may consume the schedule.

Allocate time according to difficulty and dependency, not expected word count.

A 500-word methods section may represent weeks of planning.

A 1,000-word introduction may be drafted in two days.

Words are not a reliable measure of project effort.

Use a Scope-Cut Ladder When You Fall Behind

If the project slips, do not immediately assume the solution is to work all night for three weeks.

Cut complexity deliberately.

A scope-cut ladder might be:

The later you make these changes, the more expensive they become.

That is why progress review matters.

Hold a Scope Review at the One-Third Point

When roughly one-third of your available project time has passed, ask:

• Is the research question still answerable?
• Do I have access to the needed evidence?
• Are approvals complete?
• Is the literature manageable?
• Is the method working?
• Am I on the critical path?
• Has any optional work accidentally become mandatory?
• Do I have enough time left for writing and revision?

If the answer to several questions is no, narrow the project then.

Do not wait until 80% of the semester has passed.

Hold a Completion Review at the Two-Thirds Point

At roughly two-thirds of the timeline, the question changes.

You are no longer asking:

What else could make this project interesting?

You are asking:

What must be completed to produce the strongest defensible final version?

At this stage:

• freeze major scope expansion;
• finish the evidence you already committed to;
• prioritize analysis;
• draft missing sections;
• identify limitations honestly;
• schedule feedback;
• protect revision time.

Late-stage curiosity is dangerous.

Write future-research ideas down.

Do not automatically add them to the current project.

A Capstone Does Not Need to Solve the Whole Problem

Students often over-scope because they believe the project must produce a major solution.

A capstone can make a meaningful contribution by doing something much smaller.

For example:

• comparing two approaches clearly;
• applying an existing framework to a new case;
• synthesizing recent evidence;
• identifying a gap;
• evaluating one program;
• analysing one dataset;
• testing a prototype;
• documenting one process;
• examining one population;
• clarifying a practical decision.

Cornell describes a thesis or dissertation as scholarly work that substantiates a specific point of view through original research: https://gradschool.cornell.edu/academic-progress/degrees-fields/research-degree-requirements/thesis-dissertation/writing-your-thesis-dissertation/guide-to-writing-your-thesis-dissertation/

A capstone may have different expectations, especially at undergraduate or professional-master’s level.

But “specific” is still a useful word.

Depth usually creates more value than uncontrolled breadth.

Build Limitations Into the Design Instead of Discovering Them at the End

Every capstone has limits.

Examples:

• small sample;
• one institution;
• self-report measures;
• short time period;
• English-language sources only;
• public data only;
• inability to infer causation;
• limited generalizability;
• incomplete records.

Do not treat limitations as embarrassing defects you reveal in the final paragraph.

Use them during planning.

Example:

If you can recruit only one institution, then do not frame the research question as though you are representing all university students.

If you use cross-sectional survey data, do not promise to identify causal effects.

If you conduct a literature review only in English, acknowledge the potential language restriction.

Good scope turns limitations into boundaries.

Match the Claim to the Design

This is one of the best ways to keep the capstone rigorous.

Ask:

What type of conclusion can this method actually support?

Do not design one type of project and write the conclusion of another.

A narrow, accurate claim is stronger than a broad claim your design cannot support.

Keep an Assumptions and Decisions Log

Long projects involve dozens of decisions.

Write them down.

Examples:

This record helps when writing:

• methods;
• limitations;
• discussion;
• supervisor updates.

It also prevents you from repeatedly re-litigating decisions you already made.

Track Risks, Not Just Tasks

A simple capstone risk register can have four columns:

Risk Probability Impact Response

Example:

Low participant recruitment High probability High impact Begin recruitment early; identify backup channels; agree minimum defensible sample with supervisor.

Data access delayed Medium probability High impact Request access before proposal approval if allowed; identify public alternative dataset.

Literature too broad High probability Medium impact Narrow population and date range; define inclusion criteria.

Analysis software unfamiliar Medium probability High impact Complete training early; test method on sample data.

Supervisor unavailable near deadline Known High impact Move full-draft deadline earlier.

This may feel formal.

For a complex capstone, it is simply a way to stop foreseeable problems from becoming surprises.

A Sample 16-Week Capstone Timeline

Here is a general example for a semester-based project.

Weeks 1–2: Define the Project

• read capstone requirements;
• identify assessment criteria;
• explore possible topics;
• conduct preliminary searches;
• draft 2–3 research questions;
• discuss feasibility with supervisor.

Weeks 3–4: Lock the Scope

• refine the research question;
• define population/case/time period;
• identify evidence;
• choose method;
• create scope statement;
• identify ethics/access requirements;
• build backward timeline.

Weeks 5–6: Build the Research Base

• finalize search strategy;
• collect core literature;
• build literature matrix;
• draft early literature-review sections;
• submit ethics/access applications if required;
• prepare instruments or data plan.

Weeks 7–9: Collect or Screen Evidence

Depending on method:

• recruit participants;
• conduct interviews;
• collect survey data;
• acquire secondary data;
• screen literature;
• gather case documents;
• test prototype.

Weeks 10–11: Analyse

• clean data;
• transcribe/categorize;
• code;
• run analyses;
• synthesize sources;
• evaluate cases;
• document decisions.

Week 12: Draft Findings

Do not interpret everything yet.

First make clear what you found.

Week 13: Draft Discussion and Conclusion

• answer research question;
• connect findings to literature;
• discuss alternative explanations;
• identify limitations;
• explain implications.

Week 14: Full-Draft Review

• send to supervisor if feedback is available;
• check argument coherence;
• test section connectivity;
• confirm that every section serves the research question.

Week 15: Revise

Prioritize:

1. argument;
2. analysis;
3. missing evidence;
4. structure;
5. paragraph clarity;
6. style.

Week 16: Finish

• references;
• tables/figures;
• appendices;
• formatting;
• citation audit;
• proofreading;
• file checks;
• submission.

This timeline will not fit every method.

An ethics-heavy empirical project may need approval much earlier.

A design capstone may require more iteration.

A literature review may shift most of the middle weeks toward screening and synthesis.

Use the logic, not the exact week numbers.

A Before-and-After Scope Example

Suppose a student starts with:

I want to study how AI is changing higher education.

The topic is interesting.

The project is not yet researchable.

Version 1: Still Too Broad

How is generative AI affecting university teaching and student learning?

Problems:

• “university teaching” is huge;
• “student learning” is huge;
• effect can mean dozens of outcomes;
• no population;
• no course context;
• no evidence type.

Version 2: Narrower

How does generative AI affect feedback in undergraduate writing courses?

Better.

But “affect feedback” is still unclear.

Version 3: Capstone-Ready Literature Review

What does peer-reviewed research published from 2023–2026 report about the use of generative AI to provide formative feedback in undergraduate writing instruction?

Now we can identify:

That question is smaller.

It is also much easier to answer rigorously.

A Before-and-After Empirical Example

Original:

How does stress affect nursing students?

This could become dozens of studies.

Narrowed:

How do final-year nursing students describe the relationship between clinical-placement workload and perceived academic stress during their final semester?

Now the student has:

• one population;
• one stage of study;
• one context;
• one main phenomenon;
• a question suited to qualitative exploration.

Then feasibility still has to be tested:

Can participants be recruited?

Is approval required?

Can interviews be completed in time?

Is the sample manageable?

Can the student analyse qualitative data?

If not, narrow or redesign before committing.

When to Change the Project

Changing scope is not automatically failure.

Sometimes it is evidence of good research judgment.

Consider changing the project when:

• the needed data are unavailable;
• required approval cannot be obtained in time;
• recruitment is not working;
• the question is still too broad after preliminary research;
• the literature does not support the planned comparison;
• the method cannot answer the question;
• the analysis requires skills you cannot reasonably develop in time;
• supervisor feedback identifies a fundamental design problem;
• the remaining timeline leaves no realistic revision period.

The earlier you make the change, the more options you have.

When Not to Change the Project

Do not redesign the project every time you discover a new interesting paper.

Do not add a new population because one source mentions it.

Do not change methods because another method sounds more advanced.

Do not add a second research question because you have spare ideas.

A capstone needs intellectual flexibility.

It also needs scope discipline.

Keep a “future research” document.

Put interesting ideas there.

That gives them somewhere to go without expanding the current project.

My Capstone Feasibility Checklist

Before approving your own plan, ask:

REQUIREMENTS

[ ] Do I know exactly what I must submit?

[ ] Have I read the rubric and capstone handbook?

[ ] Do I know the final word count, format, presentation, and submission requirements?

QUESTION

[ ] Can I state one primary research question clearly?

[ ] Is it narrow enough for the project length?

[ ] Is it complex enough to require analysis?

[ ] Is it answerable with evidence I can realistically obtain?

EVIDENCE

[ ] Do I know what evidence will answer the question?

[ ] Have I confirmed that enough evidence exists?

[ ] Can I access the essential sources, participants, documents, or data?

METHOD

[ ] Does the method match the question?

[ ] Do I understand the work required by the method?

[ ] Can I complete the analysis with the skills and tools available?

ETHICS AND ACCESS

[ ] Have I checked whether ethics or institutional approval is required?

[ ] Have I identified permissions, data agreements, recruitment, or access dependencies?

[ ] Are these tasks scheduled before data collection?

SCOPE

[ ] Have I defined what the project will not cover?

[ ] Can I remove optional components without breaking the core project?

[ ] Do I have a minimum viable version?

TIME

[ ] Have I planned backward from the deadline?

[ ] Does my schedule include feedback and revision?

[ ] Have I added buffer around uncertain tasks?

[ ] Do I know my critical path?

WRITING

[ ] Will I begin drafting before every research task is complete?

[ ] Do I have a skeleton outline for the entire report?

[ ] Does each planned section serve the research question?

SUPERVISION

[ ] Have I agreed on feedback expectations with my supervisor?

[ ] Do I know when they need drafts?

[ ] Have I scheduled major decision points?

RISK

[ ] What are the three most likely reasons this project could be delayed?

[ ] Do I have a reasonable response to each?

If several answers are “not yet,” the project is not ready to scale up.

That is useful information.

Fix the plan while changes are still cheap.

Frequently Asked Questions

How Narrow Should a Capstone Topic Be?

Narrow enough that you can answer the research question thoroughly within the available time, word count, evidence, and method.

The right scope depends on the assignment.

A 5,000-word literature review and a year-long empirical thesis should not have the same research question.

How Many Research Questions Should a Capstone Have?

One clear primary question is often enough.

Some projects need tightly connected subquestions, but every additional question increases the amount of literature, evidence, analysis, and writing required.

Do not add questions unless the project genuinely needs them.

When Should I Start Writing?

Early.

Once your research question and broad structure are stable, you can usually begin drafting background, literature-review material, methods, definitions, or project rationale.

Do not wait until every source has been read or every piece of data has been collected.

What If My Topic Has Too Much Research?

Narrow by:

• population;
• variable;
• outcome;
• place;
• time period;
• evidence type;
• theoretical perspective;
• comparison.

Then define explicit inclusion and exclusion criteria.

What If My Topic Has Almost No Research?

First check whether the search terminology is correct.

Then consider whether:

• the question is too new;
• the population is too narrow;
• the concept uses different terminology;
• adjacent research can inform the project;
• an empirical design is needed rather than a literature-only project.

Discuss the change with your supervisor before rebuilding the project.

Is a Literature Review Easier Than Collecting Original Data?

Not necessarily.

It removes some risks related to recruitment, approvals, and data collection, but a rigorous review still requires a clear question, systematic searching, selection decisions, evaluation, and synthesis.

Choose it because it matches the research problem and assignment—not because it sounds like the quickest option.

What If I Fall Behind?

First protect the core research question and required deliverables.

Then remove optional work.

Reduce unnecessary comparisons, subquestions, cases, variables, or supplementary analysis.

If the critical path is no longer realistic, discuss a scope change with your supervisor early.

Should I Plan to Finish on the Official Deadline?

No.

Plan to have a complete final-content version before the official submission deadline so that formatting, references, technical checks, and unexpected problems do not consume your final hours.

Is It Bad to Change My Research Question?

No.

Research questions often become more precise after preliminary reading, feasibility testing, or early evidence.

The important issue is whether the change produces a more coherent and defensible project.

Summary: Scope Is What Makes a Capstone Finishable

A capstone becomes manageable when the research and the calendar describe the same project.

Start with the deliverable.

Work backward from the real deadline.

Develop a question that is not only interesting but feasible.

Test it against:

• time;
• evidence;
• access;
• method;
• project length.

Then build the smallest rigorous version of the project first.

Keep optional work optional.

Identify dependencies.

Check ethics and access requirements early.

Start writing before the research feels complete.

Use visible weekly milestones.

Build feedback and revision into the schedule.

Review scope while there is still time to change it.

And remember that narrowing a project does not necessarily make it less ambitious.

A broad project can produce shallow conclusions because the writer has too little time to investigate any part properly.

A focused project gives you room to:

read carefully; analyse deeply; write clearly; revise seriously; and make claims your evidence can actually support.

That is the capstone I would rather see a student submit.

Not the largest project they could imagine.

The strongest project they could finish well.

Sources and Further Reading

Purdue OWL — Articulating a Research Agenda https://owl.purdue.edu/owl/graduate_writing/introduction_to_writing/documents/getting-started-with-grad-writing/articulating-a-research-agenda.pdf

Purdue OWL — Creating a Research Space https://owl.purdue.edu/owl/graduate_writing/introduction_to_writing/documents/getting-started-with-grad-writing/cars-model-activity.pdf

Purdue OWL — Creating a Roadmap https://owl.purdue.edu/owl/graduate_writing/documents/creating-a-roadmap.pdf

George Mason University Writing Center — How to Write a Research Question https://writingcenter.gmu.edu/writing-resources/research-based-writing/how-to-write-a-research-question

George Mason University Writing Center — Writing a Literature Review https://writingcenter.gmu.edu/writing-resources/research-based-writing/writing-a-literature-review

Cornell Graduate School — Understanding Deadlines and Requirements https://gradschool.cornell.edu/academic-progress/degrees-fields/research-degree-requirements/thesis-dissertation/writing-your-thesis-dissertation/understanding-deadlines-and-requirements/

Cornell Graduate School — Advising Guide for Research Students https://gradschool.cornell.edu/academic-progress/opportunities-resources-support/advising-guide-for-research-students-2025/

Cornell Graduate School — Research and Writing Tips https://gradschool.cornell.edu/career-and-professional-development/pathways-to-success/build-your-skills/tips-and-takeaways/research-and-writing-tips/

Duke Graduate School — Milestone and Graduation Information and Deadlines https://gradschool.duke.edu/academics/milestone-and-graduation-information-and-deadlines/

U.S. Department of Health and Human Services, Office for Human Research Protections — Investigator Responsibilities FAQs https://www.hhs.gov/ohrp/regulations-and-policy/guidance/faq/investigator-responsibilities/index.html

U.S. Department of Health and Human Services, Office for Human Research Protections — IRB Review of Research https://www.hhs.gov/ohrp/education-and-outreach/online-education/human-research-protection-training/lesson-4-irb-review-of-research/index.html