Writing a dissertation is one of those projects that can seem manageable when you look at one chapter, but intimidating when you look at the whole document. You have to choose a worthwhile topic, find and evaluate research, decide how to investigate your question, analyse evidence, and turn everything into a clear academic argument.

The good news is that you do not need to figure everything out at once. A dissertation is easier to handle when you treat it as a series of smaller decisions rather than one enormous writing project.

In this guide, I’ll take you through the process from choosing your topic to submitting the finished dissertation, including the mistakes that can make the process harder than it needs to be.

1. Start by Understanding Your Dissertation Requirements

Before searching for sources or writing your introduction, find out exactly what your university expects.

Dissertations differ between institutions and subjects. A psychology dissertation, for example, may involve original empirical research, while a history dissertation may be based mainly on archival and secondary sources. A computer science dissertation might involve software development, experiments, data analysis, or a combination of these.

Your department's handbook should tell you about:

• Required word count
• Submission deadlines
• Referencing style
• Required chapters
• Formatting
• Ethical approval
• Research methods
• Assessment criteria
• Supervisor meetings
• Appendices and supporting material

Do not build your dissertation around a generic template you found online and assume it will satisfy your department. University guidance is more important than any general writing guide. For example, the London School of Economics notes that its suggested dissertation structure is a starting point and advises students to confirm the appropriate structure with their supervisor.

That is a useful principle throughout the project: your university's requirements come first.

2. Choose a Topic You Can Actually Research

A dissertation topic needs to be interesting, but interest alone isn't enough.

You also need to consider whether you can realistically research the topic with the time, resources, data, participants, and academic literature available to you.

Imagine that you want to study artificial intelligence in education. That subject is far too broad for most dissertations. You could narrow it to the use of generative AI for feedback on undergraduate writing, for instance.

The difference is important.

"Artificial intelligence in education" gives you hundreds of possible directions.

"The effect of generative AI feedback on undergraduate students' essay revision practices" gives you a much clearer starting point.

When narrowing your topic, think about:

• Who or what you want to study
• Where the research will take place
• The period you want to examine
• The particular problem you want to investigate
• The variables or concepts involved
• The evidence you can realistically access

A good dissertation question should be manageable and answerable. Research guidance from the Chinese University of Hong Kong similarly emphasises that research questions should define the scope of the work and be specific enough to investigate with available evidence.

3. Turn Your Topic Into a Research Question

Once you have a topic, turn it into a question.

This is where your dissertation starts becoming a research project rather than simply an essay about a subject.

Depending on your discipline, you might ask:

• What factors influence a particular outcome?
• How does one phenomenon affect another?
• Why does a particular process occur?
• To what extent is one factor associated with another?
• How do people experience or understand a particular issue?

Suppose your topic is social media and university students.

Instead of asking:

"Is social media bad for students?"

you could ask something more researchable:

"How does daily social media use influence study habits among undergraduate students?"

You can then develop objectives that support the question.

For example:

1. To examine patterns of social media use among undergraduate students.
2. To identify common study habits within the sample.
3. To investigate whether reported social media use is associated with particular study behaviours.

Your research question should remain visible throughout the dissertation. Ideally, your literature review, methodology, analysis, discussion, and conclusion should all connect back to it.

4. Research the Existing Literature

Now comes the reading.

This stage can become chaotic if you search randomly and save every interesting article you find. A better approach is to search deliberately using the main concepts in your research question.

Start with academic databases, peer-reviewed journals, books, government publications, professional organisations, and other authoritative sources appropriate to your field.

As you read, don't just ask, "What does this article say?"

Ask:

• What question did the researchers investigate?
• What evidence did they use?
• Which method did they choose?
• What did they find?
• What were the limitations?
• Do other researchers agree?
• How relevant is the study to my research question?

Swansea University's guidance recommends evaluating the evidence, methodology, strengths, weaknesses, and relevance of individual studies rather than simply describing them.

I also recommend keeping a research spreadsheet from the beginning. Record the citation, research question, methodology, findings, limitations, and the specific reason the source matters to your project.

It is much easier to write a literature review when you know why you saved each source.

5. Write a Literature Review That Does More Than Summarise

A literature review is not supposed to be a collection of mini book reports.

Your job is to show how existing research fits together.

For example, imagine you have found ten studies about AI-assisted academic writing. Five report potential benefits, three raise concerns about student dependence, and two find mixed results.

Instead of writing:

Smith (2023) found X.
Jones (2024) found Y.
Ahmed (2025) found Z.

you can organise the discussion around the underlying debate.

What does existing research agree on?

Identify areas where findings are reasonably consistent.

Where does the evidence conflict?

Discuss differences in findings, populations, research designs, or theoretical interpretations.

What remains uncertain?

This is particularly important because uncertainty can lead you toward your research gap.

What are the weaknesses in previous studies?

Look at sample sizes, measurement choices, research settings, assumptions, or methodological limitations.

A strong literature review gradually makes a case for your own research. Both Swansea University and Edge Hill University emphasise that a literature review should identify themes, debates, limitations, and gaps rather than attempt to describe every source available.

6. Find the Research Gap

At some point, your reading should lead to the question:

What still needs to be investigated?

The answer is your research gap.

A gap doesn't necessarily mean that nobody has ever researched your subject. You might find that previous studies have focused on a different country, age group, industry, dataset, time period, or theoretical approach.

For example, perhaps most research on a particular technology has focused on postgraduate students, while very little has examined first-year undergraduates.

That difference could provide a reasonable basis for a new study.

Your introduction should eventually make this progression clear:

Existing knowledge → unresolved problem → research gap → research question → your study

That logical chain gives the dissertation a purpose.

7. Choose Your Research Methodology

Once you know what you want to find out, decide how you are going to find it out.

Your methodology needs to match your research question.

Qualitative research

Qualitative methods are often appropriate when you want to explore experiences, opinions, meanings, motivations, or social processes.

You might use:

• Interviews
• Focus groups
• Observations
• Case studies
• Document analysis
• Thematic analysis

Quantitative research

Quantitative research is useful when your project involves numerical measurement, statistical relationships, comparisons, or hypothesis testing.

Possible methods include:

• Surveys
• Experiments
• Existing datasets
• Statistical modelling
• Correlation or regression analysis

Mixed methods

A mixed-methods project combines qualitative and quantitative evidence where doing so provides a better answer to the research problem.

Whatever approach you choose, explain why it is appropriate.

Your methodology chapter should normally cover your research design, participants or data sources, sampling, data collection, analytical procedures, and ethical considerations. It should also explain the reasoning behind important methodological choices. Monash University's thesis guidance describes the methods chapter as a place to explain and justify how the research was conducted.

8. Check Ethical Requirements Before Collecting Data

If your dissertation involves people, personal information, sensitive data, experiments, or other potentially ethical concerns, find out whether you need formal approval.

Do this before collecting the data.

Depending on your institution and research design, you may need to provide information about participant consent, confidentiality, data storage, potential risks, recruitment, and withdrawal procedures.

Leaving ethics until the last minute can delay an entire dissertation, so treat it as part of your research planning rather than an administrative detail.

9. Build the Dissertation Structure

A typical empirical dissertation might contain:

1. Abstract
2. Introduction
3. Literature review
4. Methodology
5. Results or findings
6. Discussion
7. Conclusion
8. References
9. Appendices

However, this isn't a universal formula. Some projects combine chapters, while others require additional theoretical, contextual, or reflective sections.

The University of Portsmouth's current guidance for primary research dissertations follows a similar progression from introduction and literature review through methodology, findings, conclusion, references, and appendices.

The important thing is that every chapter has a clear purpose.

Introduction

Explain what you are studying, why it matters, the research problem, your question or objectives, and how the dissertation is organised.

Literature review

Show what previous researchers have discovered, where the debates are, and what remains unresolved.

Methodology

Explain how you conducted the research and why you selected those methods.

Findings

Present your evidence in a logical order.

Discussion

Explain what the findings mean and how they relate to existing research.

Conclusion

Answer the research question and explain the significance and limitations of your work.

10. Make a Dissertation Schedule

Don't write "finish dissertation" on your calendar and hope motivation takes care of the rest.

Break the project into deadlines.

For example:

Your actual timetable will depend on your programme.

Give yourself some spare time too. Research rarely goes exactly according to plan. An interview may be cancelled. A dataset may need cleaning. A supervisor may suggest substantial changes. You may discover halfway through the project that your original approach needs adjusting.

A little breathing room can make those problems manageable.

11. Start Writing Before You Feel Ready

One of the biggest mistakes is waiting until all the research is finished before writing anything.

Writing is part of the thinking process.

As you read, write notes in your own words. Create outlines for chapters. Draft paragraphs when you have enough evidence to support them.

You don't necessarily have to begin with Chapter 1.

Some students find the methodology easier to draft once the research procedures are complete. Others prefer starting with the literature review. Your abstract can usually wait until the end because you will know much more about the dissertation then.

The University of Tennessee's dissertation guide explicitly notes that students can draft the introduction, conclusion, and abstract at different points and may find it easier to complete some of these sections after producing a full draft.

The important thing is to keep moving.

Instead of thinking, "I need to write 15,000 words," give yourself a smaller task:

"Today I am going to write 500 useful words about my sampling strategy."

That is much easier to start.

12. Analyse the Evidence Honestly

When your data are ready, return to your research question and analysis plan.

Don't search through the data simply looking for a result that supports what you expected.

If the evidence contradicts your original assumption, that's not automatically a problem. Your responsibility is to report what the research shows and explain the possible reasons for the outcome.

For quantitative projects, make sure the statistical techniques you use are appropriate for the data and research design.

For qualitative projects, explain how you moved from raw material to codes, categories, themes, or interpretations.

Keep a record of important analytical decisions. If someone reading your dissertation asks, "Why did you do this?", you should be able to explain the reasoning.

13. Keep Findings and Discussion Separate

This distinction can make a surprisingly big difference to the quality of a dissertation.

The findings section primarily answers:

What did I find?

The discussion asks:

What do those findings mean?

Imagine your research shows that students who reported higher levels of social media use also reported shorter study sessions.

In the findings chapter, you would present the evidence.

In the discussion, you might consider why the relationship appeared, whether previous studies reported something similar, whether there are alternative explanations, and what the finding means for your research question.

Current university guidance makes the same distinction: findings should present the evidence clearly, while discussion sections interpret the significance of those findings in relation to the wider research.

Keeping those purposes separate helps prevent the dissertation from becoming repetitive.

14. Write the Conclusion Around Your Research Question

Your conclusion is not simply a shorter version of every previous chapter.

Its central job is to bring the research back to the question you started with.

A useful conclusion can cover:

• The answer to the research question
• The most important findings
• What the study contributes
• Important limitations
• Practical or theoretical implications
• Suggestions for future research

Don't introduce major new evidence at this point.

A good test is to ask yourself:

"If someone read only my research question and conclusion, would they understand what I discovered?"

If the answer is no, the conclusion probably needs revision.

Monash University and the University of Tennessee both emphasise that the conclusion should provide a clear response to the research purpose and explain the wider significance of the findings.

15. Reference Everything Correctly

Good research can be undermined by careless referencing.

Every source that contributes an idea, piece of evidence, quotation, dataset, image, or argument should be acknowledged according to your university's required style.

Use a reference manager if it helps you stay organised. But don't assume automatically generated references are always correct. Check the details yourself.

Pay particular attention to:

• Author names
• Publication dates
• Titles
• Journal information
• Page numbers
• DOI information
• In-text citations
• Reference-list entries

Good referencing isn't just about avoiding plagiarism. It allows your reader to trace the evidence behind your claims.

16. Protect the Originality of Your Work

Your dissertation should represent your own research and thinking.

That means keeping careful notes about where information comes from and distinguishing your own interpretation from someone else's argument.

Don't copy a paragraph into your notes and assume you'll remember to rewrite it later. That is an easy way to accidentally reproduce someone else's wording.

Instead, record the source immediately and summarise the idea in your own understanding.

Be particularly careful with artificial intelligence tools. If your university has rules about AI-assisted writing, research, coding, translation, or editing, follow them. Some institutions permit limited uses while others place much stricter restrictions on assessed work.

The safest approach is to understand your institution's policy before using AI in the dissertation.

If your project involves programming, data analysis, or Python, you should also understand and be able to explain the code and analytical decisions you submit. When you need help understanding a programming concept, resources such as Python assignment help can be used as supplementary support, but your submitted work should remain consistent with your institution's academic-integrity rules.

17. Revise in Several Passes

Don't try to fix everything simultaneously.

I find it much easier to revise a dissertation in stages.

First pass: Argument

Read for logic.

Does each chapter contribute to the research question? Are the claims supported by evidence? Does the discussion actually follow from the findings?

Second pass: Structure

Look at headings, paragraph order, transitions, tables, figures, and repetition.

Ask whether the reader can easily follow the argument.

Third pass: Evidence and references

Check citations, quotations, statistics, tables, and references.

Make sure important claims have appropriate support.

Fourth pass: Language

Only after the bigger problems are fixed should you focus heavily on grammar, wording, spelling, and sentence-level style.

This approach saves time because there is little value in perfecting a paragraph that you later discover needs to be removed.

18. Do a Final Examiner-Style Check

Before submitting, put yourself in the position of the person marking the dissertation.

Ask:

• Is the research question immediately clear?
• Is the topic sufficiently focused?
• Does the literature review demonstrate critical thinking?
• Is the research gap convincing?
• Does the methodology fit the question?
• Are the findings presented clearly?
• Does the discussion interpret rather than simply repeat results?
• Does the conclusion answer the research question?
• Are limitations acknowledged honestly?
• Are references complete and consistent?
• Have I followed the formatting requirements?
• Have I checked the final word count?
• Are all tables, figures, and appendices correctly labelled?

Read the dissertation once without editing it. You're looking for the overall argument.

Then read it again specifically for errors.

This final distance can reveal problems you won't notice while constantly rewriting individual paragraphs.

Final Thoughts

A dissertation becomes much less intimidating once you stop treating it as one enormous writing assignment.

It is really a chain of connected tasks.

You choose a manageable problem, turn it into a research question, investigate the existing literature, identify what is missing, select an appropriate methodology, collect and analyse evidence, and then explain what your findings mean.

The writing connects all of those stages.

If you keep returning to your research question, question your evidence, acknowledge the limitations of your approach, and follow your university's requirements, you will have a much stronger foundation for the final document.

And don't worry if your first draft is messy. It is supposed to be.

A dissertation is rarely written perfectly the first time. The quality usually emerges through research, drafting, supervisor feedback, restructuring, and careful revision. The goal isn't to make every sentence impressive. The goal is to make the research clear, credible, and defensible.