Empirical Dissertation: Structure, Appendix and Ethics

Two dissertations, the same survey, the same 180 responses, two very different marks. Most empirical dissertations lose marks not in the data but in the writing up. The cause sits almost always in the same place: finding and interpretation get mixed up.

An empirical dissertation is written by first building the theory and the research question, then setting out in a methodology chapter how the data were gathered and analysed, reporting the findings without judging them and only weighing them up in the discussion. Ethics approval comes before any of it. By the end of this guide you will have an outline you can put in front of your supervisor and an ethics timetable to match.

📌 The key points at a glance

• With no ethics approval Leeds caps the dissertation mark at 40%.
• Manchester Academic Phrasebank keeps commentary for the discussion section.
• QAA never uses dissertation for doctorates, only thesis.
• Leeds excludes appendices and bibliography from the word count.
• UK GDPR sits alongside the amended DPA 2018.

What separates an empirical dissertation from a literature review

An empirical dissertation tests its research question against data instead of answering it from the literature alone, and it sets out openly how those data came about and by which rules they were analysed.

Empirical does not have to mean that you send out questionnaires yourself. Leeds University Business School separates primary data, ‘information you collect yourself for your project’, from secondary data, ‘information that has already been collected by someone else, for example a government census or an academic report’. Re-analysing an existing dataset is empirical work too, and Leeds is explicit that ethics approval is required for both kinds of dissertation.

Both forms are equally valid, they simply answer different kinds of question. A question about the state of a debate needs no fresh data collection, a question about how a particular group behaves does. If you are still weighing the two up, the overview of empirical research helps you decide.

Structure of an empirical dissertation: a worked outline

The structure of an empirical dissertation follows the same split almost everywhere in the UK: a theoretical part that defines terms, orders the field and arrives at the research question, then an empirical part with methodology, findings and discussion.

No single outline is prescribed nationally. What binds you is your own module handbook and the marking criteria of your school, and the differences between subjects are large. The order below covers the standard case in the social sciences and business.

1. Introduction: why the topic matters, what you want to find out.
2. Literature review: models and state of research, ordered not listed.
3. Research question and hypotheses: derived from the literature, not fitted afterwards.
4. Methodology: design, instrument, sample, fieldwork dates, analysis and ethics approval.
5. Findings: results ordered by hypothesis or sub-question, reported without judgement.
6. Discussion: your findings set against the literature, plus the limits.
7. Conclusion: the answer to the research question and what remains open.
8. References and appendices: bibliography, questionnaire, consent materials and approval documents.

An outline is a template, not a rule. If your topic calls for it, you can cut findings and discussion differently, but agree that with your supervisor first. How you get from a topic to finished fieldwork is set out in the guide to empirical research in final papers.

Dissertation or thesis: what the two words mean in the UK

In the UK, dissertation normally refers to the undergraduate or taught master’s project, while thesis is reserved for the doctorate. The QAA Doctoral Degree Characteristics Statement of February 2020 never uses the word dissertation at all: candidates are assessed on their ‘thesis, portfolio, artefact or composition’ and defend it at an oral examination, the ‘viva’ or ‘viva voce’.

At master’s level the line blurs. For research master’s degrees the QAA Master’s Degree Characteristics Statement of the same date says assessment ‘usually requires a dissertation or thesis, or other output’ and that ‘the thesis is commonly defended in an oral examination’. Which word your school uses is the one that counts.

How much theory an empirical dissertation needs

No UK-wide figure sets how much of your text should be theory. The QAA Master’s Degree Characteristics Statement puts its number on course time, not page count, and hedges it twice: for specialised or advanced study master’s degrees ‘Frequently, at least a third of the course is devoted to a research project’, and for an MRes ‘at least half of the course is commonly devoted to one or more research projects’. Read that as effort, not a page ratio.

IMRaD is a journal convention, not a UK requirement

IMRaD stands for Introduction, Methods, Results and Discussion, and it orders journal articles rather than student work. The QAA leaves the substance to the institution: ‘Higher education providers determine the content of a master’s degree in terms of the knowledge and understanding, expertise and skills that the student is intended to acquire.’ The University of Manchester Academic Phrasebank adds that it is ‘not uncommon for the Results section to be combined with the Discussion section’.

Our reading tip

Clark, Tom; Foster, Liam; Sloan, Luke; Brookfield, Charlotte and Bryman, Alan (2026): Bryman’s Social Research Methods. Seventh edition, Oxford University Press, 736 pages, £54.99. Chapter 26 covers writing up and chapter 6 covers ethics, which is the part most UK dissertation handbooks assume you already know.

The empirical shape with theory, method, results and discussion is only one pattern among several. Which others exist, what order the formal sections follow and how deep to number your chapters is covered in the guide to dissertation structure.

The methodology chapter: what has to be in it

The methodology chapter of an empirical dissertation serves one purpose: a stranger should be able to repeat your fieldwork from it without asking you a single question. Everything needed for that belongs in, everything else does not.

The most common fault in this chapter is not omission but the wrong kind of writing. Many dissertations spend two pages explaining what an online survey is and not one sentence justifying why it suited this research question. Textbooks can explain. Only you can justify.

What belongs in the methodology chapter

Eight items make a methodology chapter reproducible, and most module handbooks ask for all of them. Work through the list before you hand the chapter to your supervisor, because a missing fieldwork date or a missing ethics reference is the kind of gap a marker spots in seconds and cannot award marks around.

• Research design and the reason it fits the research question.
• The instrument and its origin: your own or adapted.
• Operationalisation, so which question measures which construct.
• Sample, recruitment route and the number of usable responses.
• Fieldwork dates, given as calendar dates.
• Pilot: how many people, and what you changed as a result.
• Analysis techniques and the software used.
• Ethics approval with its reference, date and review route.

One item gets skipped again and again: the pilot. Letting two or three people from your target group fill in the questionnaire costs an afternoon and catches the questions nobody reads the way you meant them. Which decisions come before it is set out in the guide to research design.

Ethics approval comes before the first response

Ethics approval is the one step you cannot take retrospectively. Leeds University Business School calls it ‘the single, non-negotiable step you must complete before collecting any data’ and applies a sliding scale: 5 marks off if the approval is not submitted with the dissertation, 10 if it was granted after collection began, 15 if both apply.

Without any approval at all, Leeds caps the written dissertation at 40%, which is the bare pass. The University of Manchester lists five possible routes of ethical review, from division or school review to Health Research Authority approval, and requires review whenever research involves person-identifiable or special category data, recordings of participants or vulnerable groups. Your own school decides which route applies to you.

💡 Tip

Leeds asks for the ethics application at least two weeks before collection starts, because supervisors often send plans back for changes. Build those two weeks into your timetable together with the participant information sheet and the consent wording, and write the methodology chapter while you collect rather than afterwards.

Reporting findings without interpreting them yet

The findings chapter reports what you measured, not what you make of it. The split is the clearest fracture line in student dissertations and also the easiest one to avoid.

The University of Manchester Academic Phrasebank draws the line as precisely as it can be drawn: in a results section writers ‘comment on the significant data’ shown in tables and figures, while ‘more elaborate commentary on the results is normally restricted to the Discussion section’. Describing a number is allowed. Judging it is not.

Two orders work for the chapter: hypothesis by hypothesis, or along the sub-questions of your research question. Which one you pick is a matter of taste, mixing the two is not an option. Which analysis suits your data is covered in the guide to data analysis methods.

The discussion: placing, comparing and admitting limits

The discussion answers what your findings mean and how far they carry. The discussion is the chapter where numbers turn into a statement, and the only one in which you are allowed to judge.

The University of Manchester Academic Phrasebank describes discussion sections as ‘probably the most complex sections in terms of their elements’ and notes that they are usually built as a series of discussion cycles around each result. Four moves carry a good discussion, normally in this order.

1. Restate. The central findings in a few sentences, without the tables.
2. Connect. Does your result confirm, contradict or refine previous research?
3. Limit. Name your limitations and weigh them, rather than listing.
4. Look ahead. The question your finding raises for a follow-up study.

Limitations often decide the impression a dissertation leaves. A small, non-random sample is no disgrace as long as you name it and say what your claim is therefore limited to. Which standards apply here is set out in the guide to quality criteria.

⚠️ Careful

Limitations are a quality signal, not a confession. The mistake is not the small sample, the mistake is the sentence ‘the results are representative of UK students’ underneath it. Write instead who your finding holds for and who it does not.

What goes in the appendices and what does not

The appendices of an empirical dissertation hold everything a marker needs in order to judge your fieldwork but that would break the flow of the main text. First in line is the instrument itself, your questionnaire or interview schedule.

No UK-wide rule governs appendices, and schools differ in the detail. The dissertation format regulations at Leeds keep appendices, the bibliography and acknowledgements out of the word count, while the abstract, contents page, tables and footnotes inside the main text all count.

Moving a load-bearing argument into an appendix is what costs marks. Leeds puts it plainly: students ‘must not use appendices to present matters of substance which should be included in the main body of the text’, and ‘the text must remain understandable without reference to any appendices’. Every page is numbered, appendices included.

One more habit is worth building: export the questionnaire as a PDF before you send it out, filter logic and answer options included. Once the survey has closed, a printable version can be hard to recover, and the appendices are where the marker looks first.

What happens to your data after you submit

After submission a question arises that few guides answer: do you have to keep your raw data or delete them? Both duties exist, they come from different rule books and both can apply to the same dissertation.

The UK Research Integrity Office publishes the Code of Practice for Research, updated to version 3.5 in July 2025. The update aligns the code with the 2025 Concordat to Support Research Integrity, adds a section on artificial intelligence and ties data management to the FAIR principles, which point towards keeping data reusable rather than deleting them.

The Concordat itself is worth naming correctly. The 2025 edition is its third, following 2012 and 2019, and it is ‘hosted on behalf of the UK research community by the UK Committee on Research Integrity’. Care and respect for participants is one of its five principles.

From data protection comes the counter-pressure. The Information Commissioner’s Office states that ‘the GDPR is retained in domestic law as the UK GDPR’ and that it ‘sits alongside an amended version of the DPA 2018’. Personal data gathered for a student project fall under the same storage limitation principle as any other personal data.

The tension between keeping and deleting usually dissolves through anonymisation. Where names, email addresses and other identifiers were never stored in the first place, what remains at the end is a dataset that may be kept because it identifies nobody. Which settings that takes is covered in the guide to data protection in surveys. Your institution’s own policy is what binds you, and this article is not legal advice.

Common mistakes in an empirical dissertation

Five mistakes turn up in empirical dissertations again and again. Four of them are writing mistakes rather than research mistakes, and the fifth is a matter of timing that no amount of writing can repair.

The good news about the writing mistakes is that a single revision pass catches all four, and none of them needs new data or new reading. The fifth one needs a calendar.

Interpretation slips into the findings chapter

The commonest mistake grows out of helpfulness: you want to make it easier for the reader and put the meaning in straight away. The discussion then has nothing left to say and repeats itself, so both chapters look half finished. The test is simple. If a sentence in the findings chapter carries a judging word such as ‘encouraging’, ‘surprising’ or ‘unfortunately’, it belongs one chapter further on.

Data collection starts before ethics approval arrives

Collecting first and applying afterwards is the most expensive mistake in a UK dissertation, because it is the one a marker can see in the paperwork. Leeds takes 10 marks off a dissertation whose approval was granted after collection began, and 15 if the approval is also missing from the submission. Neither penalty can be argued away later.

The methodology chapter explains the method instead of justifying the choice

A definition of the online survey sits in every textbook and interests nobody in a methodology chapter. What is wanted is the reason this way of collecting data suited your research question better than an alternative you considered. One sentence often does it: interviews would have gone deeper, but the sample size you needed would have been out of reach.

Raw software output is pasted into the text

Output windows from statistical software are built for the person operating them, not for a reader. Build your own table with the figures that bear on your hypothesis, and give every table and figure a number and a title, as the University of Manchester Academic Phrasebank advises. The full output goes to the appendices.

Limitations are written as clichés

‘The sample was small’ is an observation, not a limitation. A limitation names the claim that no longer holds: with 84 participants from a single programme you cannot test differences between subject areas, though you can describe the tendency inside that programme. Drawing that distinction is what shows methodological understanding.

Conclusion

An empirical dissertation does not win on spectacular data, it wins because every decision is documented and can be followed. Anyone who keeps method, finding and interpretation apart makes it easy for a marker to see the route that was taken. Behind all the formal requirements sits that single demand.

In the findings chapter you report what happened. In the discussion you say what it means. That line is half the mark.

Where to go next

• Still need to build the survey for your dissertation? Questionnaire design
• Unsure about your hypotheses? How to write a hypothesis
• Looking for the route from idea to analysis? Empirical research process

Still to run the fieldwork for your dissertation?

With empirio.ai, an online survey tool from Germany, you build the questionnaire, collect the responses anonymously and export the dataset at the end for the appendices of your dissertation.

Create a survey for your dissertation