Thesis, Hypothesis, Research Question: The Difference

Your proposal contains one sentence, and your supervisor asks: is that your argument or your hypothesis? The question sounds like hair-splitting. It decides whether you spend the next three months in the library or building a questionnaire.

The difference between a thesis and a hypothesis lies in what happens to the sentence next: a thesis is a contestable claim that you defend with evidence and reasoning, while a hypothesis states a relationship between at least two measurable variables and is tested against data you collect yourself. By the end of this article you will be able to place every sentence in your proposal into one of four categories.

📌 The key points at a glance

• A thesis is a contestable claim defended through argument.
• A hypothesis links at least two variables into a testable statement.
• Every hypothesis is a thesis, but not every thesis is a hypothesis.
• In the UK, thesis also names the doctoral document itself.
• An assumption is declared rather than tested.

Thesis or dissertation? Why the word itself causes trouble

In British higher education, thesis carries two separate meanings, and only one of them is the subject of this article: the document you submit, and the arguable claim you defend inside it. Most confusion about theses and hypotheses starts here rather than with the methodology.

The University of Oxford's Department for Continuing Education puts the institutional meaning plainly: “In the UK, a thesis is the term used for a piece of writing submitted for an undergraduate degree or a doctorate, and a dissertation is the final piece of writing submitted for a Master's degree. In the US, the use of the terms is inverted.” Anyone reading American study guides alongside British ones will meet both conventions in the same afternoon.

Throughout this article, thesis means the claim, not the document. That is the sense in which it sits alongside hypothesis and assumption.

Choosing the topic of your dissertation comes first, and a topic on its own claims nothing. “Paid work and degree outcomes” marks out a field; it does not yet say anything anyone could disagree with.

What is a research question?

A research question is your topic narrowed into a single sentence: it names precisely what the project answers and keeps every chapter pointing in the same direction. It appears in the introduction and is answered in the conclusion.

Every piece of academic writing needs it, from a first-year essay to a doctoral thesis. From “paid work and degree outcomes” you might arrive at: “Does the number of hours students spend in paid employment relate to their final degree classification?” How to get there step by step is set out in our guide to formulating a research question.

A question mark is not compulsory, despite the name. A statement of aim does the same job: “This dissertation examines whether hours of paid work relate to degree classification.” What counts is not the punctuation but whether the final chapter delivers an answer. Your module handbook and your supervisor set the binding requirements.

💡 Tip

Keep your research question written on a card beside your keyboard and test every finished paragraph against it. A paragraph that contributes nothing to the answer belongs in an appendix or in the bin. This is the single most effective cure for a dissertation that keeps growing.

The difference between a thesis and a hypothesis

A hypothesis is a particular kind of thesis. Both claim something contestable, but only the hypothesis names two variables and the direction of the relationship between them, and only the hypothesis can therefore be measured rather than merely debated.

The relationship is written in one of two patterns, either as an if-then statement or as a the-more-the-more statement. “The more hours students work in paid employment, the lower their final degree classification” is a hypothesis. “Student employment is an underestimated pressure on degree outcomes” is a thesis, because the sentence carries a judgement rather than a measurement rule.

The choice governs your whole working method. Defending a thesis means literature, argument and counter-argument. Testing a hypothesis means a questionnaire, data collection and statistical analysis. Both are perfectly respectable, but the decision belongs at the proposal stage rather than halfway through writing up.

Every hypothesis is a thesis. The reverse is not true.

The counter-claim test: how to spot a real thesis

Whether your sentence works as a thesis takes about a minute to establish. State the opposite and see whether anyone could reasonably argue for it. The Goethe University Frankfurt turns this into a criterion in its guidance on academic writing: if no sensible counter-claim exists, you have not written a thesis. “Paid work lowers degree outcomes” passes the test. “Some students have part-time jobs” does not.

Where theory fits between the two

A theory is not a single sentence but a structure of assumptions about how part of the world works, including which factors matter and how they act on one another. Hypotheses are derived from a theory, not the other way round. In everyday speech theory means almost the opposite, namely an untested guess, and that double meaning is where most of the confusion comes from.

Turning a body of literature into a testable statement is a technique in its own right, with its own criteria. Our guide to formulating hypotheses sets it out in full.

What is an assumption?

An assumption is something you take as given for the purposes of your study without testing it. It sets the boundaries within which your findings hold, which is exactly why it belongs on the page rather than in your head.

In the employment example, one such boundary would be: “Students are counted as working part-time if they are paid for at least ten hours in an average term-time week.” You do not test that threshold, you set it. It still shapes your result, because a five-hour threshold would produce a different sample.

Alongside these deliberate boundaries, every project carries unspoken assumptions built into the question itself. Asking why students in paid work achieve lower classifications already presumes that they do. Sentences of that kind belong in the limitations section, otherwise they remain a blind spot that an examiner will find before you do.

The word is used in two widths. Narrowly, an assumption is the untested premise. Broadly, it is the umbrella term under which theses and hypotheses both sit as special cases. For writing purposes, the narrow sense is the one that matters.

Thesis or hypothesis: what your project actually needs

What you formulate follows from your research design: projects without primary data carry an argument, quantitative projects carry hypotheses, and qualitative projects usually manage with the research question alone.

Use the words thesis and hypothesis only where your empirical research can actually deliver on them. Many undergraduate dissertations need nothing more than a precise research question. A decorative hypothesis promises an analysis that then never appears, and that gap is the first thing a marker notices.

Hypotheses are not compulsory, a point that rarely reaches student guides but is stated plainly at source. The Goethe University Frankfurt notes in its guidance that a dissertation or larger research project does not have to proceed by way of hypotheses at all. Hypotheses are described there as a customary feature of positivist research designs, and not obligatory even within those.

Discipline matters too, more than most handbooks admit. Psychology, economics and the social sciences work with hypotheses as a matter of course; history, literature and parts of politics work with arguments. The term slips most often in qualitative research, where propositions arrive at the end of a study rather than at the start.

Common mistakes with theses, hypotheses and assumptions

Four mistakes turn up in dissertations often enough to be worth checking before you submit your proposal. All four take five minutes to fix beforehand and a rewritten chapter afterwards.

Two concern the wording, two concern what a statistical test can actually deliver. The second pair surfaces late, in the results chapter, and is therefore the more expensive kind.

Mistake 1: the thesis is really a statement of fact

“The number of students in paid employment has risen” sounds like a thesis but is not one. Nobody can argue with it, they can only look it up, which is precisely why it fails the counter-claim test. A thesis begins where you interpret a figure, place it in context or give it a meaning that someone else would dispute.

Mistake 2: the hypothesis mixes up its two patterns

Two sentence patterns are available, and they get spliced together surprisingly often. Either write “if A, then B” or write “the more A, the more B”. Mixing them produces sentences that read as though a variable were being compared with a state. The rule of thumb is straightforward: states call for the if-then form, gradations call for the comparative form.

Mistake 3: a tested hypothesis becomes a proof

Data can support or reject a hypothesis; it cannot prove one. What gets tested is not your own supposition but the null hypothesis, the counter-assumption that no relationship exists. Write “the data support hypothesis 1” rather than “hypothesis 1 is proven”. With a thesis in the argumentative sense the vocabulary differs, and there the language of supporting and refuting is entirely correct.

Mistake 4: a relationship turns into a cause

“Paid work lowers degree classifications” claims a direction of effect that a one-off survey cannot carry. It is equally plausible that students struggling academically take on more paid hours. Write the claim as a relationship for as long as your design cannot support a causal statement, and say so openly. A genuine causal claim requires a causal research design.

All the terms in one worked example

The distinction settles fastest when the same project is run through every term in turn. The example below follows an undergraduate dissertation on paid work and degree outcomes.

• Topic: student employment and its consequences for academic performance
• Research question: Does the number of paid working hours relate to final degree classification?
• Thesis: student employment is an underestimated pressure on degree outcomes
• Hypothesis: the more hours students work in paid employment, the lower their final degree classification
• Statistical hypotheses:
  • Null hypothesis: there is no relationship between working hours and degree classification
  • Alternative hypothesis: there is a relationship between working hours and degree classification
• Assumption: students count as working part-time from ten paid hours in an average term-time week

The step to the statistical hypotheses reads back to front on first encounter, and it genuinely is. What gets tested is the null hypothesis rather than your own supposition. If the result is unlikely enough, the null hypothesis is rejected and the alternative hypothesis is provisionally retained.

How that chain fits into the shape of a whole dissertation, from proposal to discussion, is covered in our guide to writing an empirical dissertation.

Sources and further reading

The sources below carry the statements made in this article and double as citable references for your own methodology chapter.

• University of Oxford, Department for Continuing Education: Writing theses and dissertations. On the differing use of thesis and dissertation in the UK and the US.
• The University of Manchester: Academic Phrasebank, Introducing work. On stating the aims of a study and its research questions or hypotheses.
• Goethe University Frankfurt, Chair of International Institutions and Peace Processes: Guidance on formulating theses, hypotheses and assumptions. Source of the counter-claim test and of the hypothesis as a special case of the thesis. Published in German.

One caveat on all three: the Frankfurt guidance comes from political science, where argument-led work is the tradition. In quantitative disciplines the weight shifts towards hypotheses. Your own department's requirements are what bind you.

Conclusion

One question separates all four terms reliably: what happens to this sentence next? A research question gets answered, a thesis gets defended, a hypothesis gets tested, an assumption gets declared. Being able to say which applies to every sentence in your proposal is most of the groundwork done, and it makes the conversation with your supervisor considerably shorter.

Where to go next

• Want to see the whole process? The empirical research process: stages, steps and an example
• Planning a survey for your dissertation? Running a survey for your dissertation
• Unsure about academic style? Academic writing: style, tense and the first person

Hypotheses ready, data missing?

Once your hypotheses stand, you need responses. With empirio.ai, an online survey tool from Germany, you can build the matching questionnaire for free, share it as a link and watch the results being analysed as they arrive.

Create a survey for your dissertation