Key information
- ✓Return to the purpose and research questions you wrote in the introduction.
- ✓Formulate conclusions based on the results, not your own expectations.
- ✓Indicate the practical significance and limitations of the study.
- ✓Do not introduce new data, sources or extensive definitions.
Where to start with the ending?
At the beginning, briefly recall the subject and purpose of the work, but do not copy sentences from the introduction. Then indicate what material was analyzed and how it allowed to achieve the assumed goal.
Before writing, prepare a table containing your research questions, corresponding results, and one-sentence conclusions. Such an outline helps to maintain completeness and avoid random order.
How to formulate conclusions?
The conclusion should be an answer from the data, not a re-description of a number or graph. Explain what pattern you observed and what it means in the context of the problem you are analyzing.
Be careful with language. Correlational research indicates a relationship, but does not always allow us to talk about impact and causality. The scope of the claim cannot be broader than the method and test used.
Limitations and practical importance
State the limitations factually: sample size or selection, method of measurement, data availability and time context. They are not a list of excuses, but information needed to properly evaluate applications.
If the results have practical applications, identify a specific audience and possible action. The recommendation must result from the analysis and not be a loose idea unrelated to the results.
What not to include at the end of your work?
Do not add new theories, data, or quotes that have not been previously discussed. The ending is not the place to describe the subsequent chapters in detail.
Avoid empty statements that the goal has been achieved. Show this through synthetic answers. Finally, check the consistency of the introduction and conclusion: each important assumption should receive its closure.
Begin the conclusion with the answer, not the history of writing
A thesis conclusion should return to the problem introduced at the beginning. Briefly recall the objective and immediately show what was established. Do not chronologically recount reading literature, preparing a survey and writing chapters. That describes the work process, not its outcome. Readers need a synthesis: what answer the analysis provides and within which boundaries it is justified.
A working table linking specific objectives to findings can help. For each objective, record one main finding and its limitation. If you cannot identify a result, check whether the objective was actually achieved. Do not fill the gap with a general statement about the topic's importance. It is better to describe partial achievement honestly than to claim complete success without evidence. This table helps draft the text; it need not appear in the conclusion.
A conclusion, summary and recommendation are different elements
A summary describes what the thesis contains, a conclusion explains what follows from the analysis, and a recommendation proposes action. The concluding chapter may use these elements, but should not confuse them. Saying that chapter three contained survey results is not a conclusion. A conclusion should address an identified relationship, difference, problem or assessment of a solution. It need not repeat every numerical value the reader has already seen.
A recommendation requires an additional step: explaining why the proposed action addresses the identified problem. If the study concerned a small group's opinions, a sector-wide recommendation may exceed the evidence. You can propose further investigation or a pilot instead of immediate implementation. In an engineering thesis, a recommendation may concern a specific design improvement arising from tests and assessment criteria. Every proposal should have a visible rationale.
An example of moving from an observation to a cautious conclusion
Educational example: interview participants describe difficulty finding current instructions. The observation is a recurring problem in the collected material. A conclusion might be that information availability needs consideration when assessing the process. A recommendation is to consider a shared place for publishing instructions and test its usefulness. This does not automatically establish that the solution will improve every organisational outcome or eliminate every communication difficulty.
Preserve that reasoning sequence in the conclusion. Show what was established, how you interpret it and what action could be considered. If data on the action's effects are missing, do not present effectiveness as a fact. This example describes neither a real client nor a completed study. Its purpose is to explain the difference between levels of inference. Your own conclusion should rest on material presented earlier in the thesis.
Limitations should shape the conclusions themselves
A limitation is not a compulsory sentence appended without connection to the analysis. Explain how sampling, measurement, source coverage or test conditions narrow interpretation. If a study covers one institution, the conclusion should primarily concern that context. If the data are self-reports, do not equate them with observed behaviour. Identifying a limitation should help readers interpret the result correctly.
Do not list every possible weakness of any conceivable study. Choose limitations that actually apply to the project and may affect assessment of its findings. Then suggest a specific direction for further work: another measurement approach, additional group comparisons, a longer observation period or testing the solution under different conditions. A proposal for further research should follow from an identified gap rather than a general call to explore the topic more deeply.
Final editing: remove new topics and unsupported generalisations
Read the conclusion and identify the supporting passage in the thesis for each claim. A new model, source or result probably requires earlier discussion. The conclusion is not the place to rescue a missing argument. Also check words such as proved, always, everyone and completely. In many degree projects, they are too strong for the method and material used.
Finally, compare the conclusion with the introduction and abstract. All three should describe the same objective and result while serving different purposes. Create smooth transitions between findings, limitations and the work's significance. Do not finish merely by stating that the topic matters. The last paragraph can clearly identify the thesis's main contribution to understanding the problem or developing a solution without exaggerating its reach.
Summary
A good ending shows that all parts of the work led to a common result. It should clearly answer questions, indicate the significance of the results, and honestly define the limits of conclusions. If, after reading this fragment, the reader understands what has been established, why it is important and what is still unknown, the work receives a logical and convincing conclusion.
Frequently asked questions
How many pages should the conclusion of the work have?
It depends on the volume and university guidelines. Most often, it is shorter than the introduction and takes several pages, but it must contain answers to all main questions.
Can the ending include footnotes?
They are usually not needed because the ending is based on the results of your own work. A footnote is only justified when information taken from a source appears.
Can it be said that the hypothesis was not confirmed?
Yes. The result and possible explanations should be factually indicated. Rejecting a hypothesis does not reduce the value of a properly conducted study.
Can I state in the conclusion that the objective was not fully achieved?
Yes. Explain what was established, what is missing and why. An honest account of partial achievement is better than an unsupported claim of success. Discuss with your supervisor beforehand whether the limitation requires a scope change or additional work.
Need support with your work?
Describe the topic, scope and deadline. We will review the materials, recommend the right support and prepare a clear quote.
