The discussion section of a thesis is the chapter where you explain what your results actually mean. It does not repeat the numbers from the results, it interprets them: it compares your findings with the existing literature, explains agreements and contradictions, acknowledges the limitations of your study, and from all of this draws recommendations. Put simply, the results say „what I found", while the discussion says „what it means and why it matters".
What the discussion is and what it is for
The discussion is where data turns into knowledge. While the results show in raw form what your measurements and analysis produced, the discussion places those facts into the wider context of your field. It answers the question every reader and every examiner will ask: „Fine, so what follows from this?"
A good discussion does several jobs at once. It returns to the research questions or hypotheses from the introduction and shows how your findings answer them. It connects your own results with the theoretical part, so it is clear the thesis holds together. And finally, it honestly states where the boundaries of your conclusions lie, so that nobody reads them more strongly than the data allow.
The ability to interpret is precisely what separates independent scholarly work from a simple summary report. At the defense, the committee often does not assess so much which numbers you measured, but whether you understand them and can place them in context. A weak discussion can spoil even good research, whereas a thoughtful discussion can rescue even results that did not turn out as expected.
The difference between results, discussion and conclusion
These three sections are the ones students most often blur together, even though each has a different job. The easiest way to remember them is through the three different questions they answer.
| Section | Which question it answers | What it contains |
|---|---|---|
| Results | What did I find? | Bare facts, numbers, tables, charts, no evaluation |
| Discussion | What does it mean and why? | Interpretation, comparison with the literature, limitations, implications |
| Conclusion | Did I meet my aim? | Brief summary, statement about the aim, recommendations |
In the results you therefore do not evaluate, you only present. In the discussion, by contrast, you must not introduce new raw data that did not appear in the results. And the conclusion brings no new arguments or comparisons, it only closes off what the discussion analyzed. At some institutions the results and discussion are merged into a single „Results and Discussion" chapter, but even then keep in mind that presentation and interpretation are two distinct steps.
If you are unsure where a given piece of information belongs, it helps to look at the overall structure of a thesis, where the individual chapters fit together logically.
How to interpret your own results
Interpretation is the core of the whole discussion. It does not mean repeating what the reader already saw in a table, but explaining what a given result says about the problem you studied.
Move from the specific to the general. First pick a finding that matters, then put it into words, and only at the end explain what follows from it. Instead of the sentence „the mean value was higher in group A", write why it matters: what it suggests about the relationship between the variables, which research question it helps answer, and whether it confirms or refutes your hypothesis.
A useful tool is to ask three questions about every result:
- What does this result show? A brief interpretation, not a description of the number.
- Does it agree with what the literature claims? Agreement, partial agreement or contradiction.
- What are the implications? For theory, for practice or for further research.
At the same time, watch your scope. Interpret only what your data can genuinely support. If you studied a small sample in a single region, do not draw conclusions about the entire population. What is more, correlation is not causation, so if you found that two phenomena are related, it does not automatically mean that one causes the other.
Comparing your findings with the literature
A discussion should be a conversation, not a monologue. Set your results next to what other authors, the ones you mentioned in the theoretical part, found on the same topic. This is exactly where it shows whether you really read the literature or merely cited it.
When comparing, three situations can arise and each needs to be handled differently:
- Your findings confirm existing research. State which authors you agree with, and add value by placing the confirmation into your own context (a different sample, a different period, a different setting).
- Your findings partly differ. Name where you agree and where you do not, and try to explain what the difference may be caused by.
- Your findings contradict the literature. This is not a mistake, on the contrary it can be the most interesting part of the thesis. The important thing is to offer a possible explanation, not to pretend the contradiction does not exist.
Always refer to specific sources, not to „many authors" or „most studies". Citations in the discussion must be as precise as in the theoretical part and must match the methodology you used, so the reader can tell whether your results and those of others are comparable at all.
How to explain unexpected or contradictory results
Almost every piece of research yields something that surprises the author. A hypothesis is not confirmed, one indicator comes out the opposite of what you expected, or the data do not behave according to theory. This is no reason to panic or to cover anything up. Honest handling of unexpected results is precisely a mark of mature work.
The worst thing you can do is to pass over an inconvenient result in silence or „force" it toward the desired conclusion. The committee will detect such adjustments and it will damage the credibility of the whole thesis. It is better to admit the contradiction and offer possible causes.
Common explanations for unexpected findings include:
- the specifics of the sample (size, composition, selection of respondents),
- properties of the method or data-collection instrument you used,
- the influence of the time, place or context in which the research took place,
- different definitions of concepts than in the studies you compared with,
- random factors that could not be fully controlled.
If a hypothesis did not work out, do not forget that even a refuted hypothesis is a valid result. Science works in such a way that part of the assumptions are confirmed and part are not. The important thing is to interpret the result honestly and not to create conclusions the data do not support.
Limitations of the study and why you should acknowledge them
Limitations are the constraints that affected what and how you were able to find out. They include, for example, a small or unrepresentative sample, a short time period, limited availability of data, the chosen method, or subjectivity in qualitative assessment.
Many students are afraid to admit limitations because they feel it weakens their work. The opposite is true. Consciously named limitations show that you understand the methodology and that you view your own results critically. Work without a single limitation looks naive, because perfect research does not exist.
Limitations, however, need to be presented in a balanced way. Do not apologize endlessly and do not talk your own work down. The aim is to soberly define where the boundaries of your conclusions lie, and ideally to hint right away how they could be overcome in future research. This way the limitations naturally bridge into recommendations.
How recommendations grow out of the discussion
Recommendations are the logical culmination of the discussion. They should not fall from the sky but follow directly from what you found and what it means. If the discussion showed that some factor has an effect, a recommendation says what to do about it in practice or what still needs to be examined.
Distinguish two types of recommendation. Recommendations for practice say how to use your findings in a real-world setting: in a company, an institution, in teaching, in policy. Recommendations for further research point to which questions remained open and where the next authors should look, often precisely where you ran into limitations.
A good recommendation is specific and tied to a finding. Instead of the generic „it is recommended to pay more attention to the topic", write what exactly, for whom and why. You then summarize part of the recommendations in the conclusion as well, but while the discussion explains and justifies them, the conclusion only states them briefly.
The structure of the discussion step by step
The discussion has no single mandatory template, but a proven logic runs from your own findings through the literature and limitations to the recommendations. Such an order leads the reader naturally.
- A brief reminder of the key findings. One or two sentences that recall what mattered most, without repeating the entire table.
- Interpretation of the results. What the findings mean in relation to the research questions and hypotheses.
- Comparison with the literature. Agreements, differences and possible causes of contradictions.
- Explanation of unexpected results. Honest handling of what did not turn out as expected.
- Limitations of the study. A sober definition of the boundaries of validity.
- Recommendations. For practice and for further research, growing out of the findings and limitations.
Keep the same order of topics in which you presented the results. If in the results you proceeded by research question, stick to that in the discussion too. The reader can then easily match each interpretation to the corresponding finding.
Sample phrasings for the discussion
Treat these sentences as a skeleton you fill with your own content, not as text to copy. The committee will recognize borrowed phrases and every thesis goes through an originality check. They serve to keep you from getting stuck on a blank page.
Interpreting a finding: „The values obtained suggest that [variable] is related to [phenomenon], which corresponds to research question number [X]."
Agreement with the literature: „This result is consistent with the findings of [Surname, year], who observed a similar relationship in a different setting."
Contradiction with the literature: „Unlike [Surname, year], our data did not confirm this relationship. The difference may be related to a different sample composition and a different period of data collection."
Explaining an unexpected result: „The hypothesis was not confirmed. A possible explanation is the limited sample size, which did not allow a weaker effect to be captured."
Acknowledging a limitation: „The study is limited by the size of the sample and its focus on a single region, so the findings cannot be generalized to the whole population."
A recommendation: „Based on the findings, we recommend that [a specific party] introduce [a specific measure]. Further research should focus on [an open question]."
Notice that each sentence contains a place to fill in and calls for specificity. It is precisely specificity that separates a strong discussion from empty waffle.
Common mistakes in the discussion
Only describing results without interpreting them. The most common mistake. The author repeats the numbers from the previous chapter in different words but does not say what they mean. A discussion that could be moved wholesale into the results is not a discussion.
No link to theory. The results stand alone and are nowhere compared with the literature from the theoretical part. The work then looks as if the author did not know the field at all.
Claims the data cannot support. Generalizing from a small sample, mistaking correlation for causation, or conclusions that go far beyond what was measured.
Hidden or sugar-coated unexpected results. An attempt to „force" the results toward the desired conclusion. An honest analysis of a contradiction is always more valuable than a pretended agreement.
Missing or merely formal limitations. Either the limitations are not there at all, or there is one general sentence with no link to the actual research. Limitations should grow out of your methodology.
Recommendations with no link to the findings. Generic advice that would fit any thesis. A recommendation should follow from a specific result, otherwise it feels like filler.
Frequently asked questions
What is the difference between results and discussion?
The results present bare findings, numbers, tables and charts, with no evaluation. The discussion interprets those findings: it explains what they mean, compares them with the literature, acknowledges limitations and draws implications. So in the results you do not evaluate, and in the discussion you do not introduce new raw data.
Can the discussion be combined with the results into one chapter?
Yes, many institutions and many fields use a combined „Results and Discussion" chapter. Even then, however, keep in mind that presenting facts and interpreting them are two distinct steps. Always follow your institution's guidelines, which take precedence over general advice.
What should I do if my hypothesis was not confirmed?
Nothing bad has happened, a refuted hypothesis is a valid result. Acknowledge the non-confirmation honestly, interpret it and offer a possible explanation, such as the sample size, the method used or the specifics of the context. The important thing is not to create conclusions the data do not support.
How do I acknowledge limitations in the discussion without weakening my thesis?
Present the limitations soberly and in a balanced way. Name where the boundaries of validity of your conclusions lie, but do not apologize endlessly and do not talk your own work down. Ideally, hint right away how the limitation could be overcome in future research. Consciously named limitations come across as competent, not weak.
Do recommendations belong in the discussion or in the conclusion?
In both, but at a different depth. The discussion develops and justifies the recommendations on the basis of the findings and limitations. The conclusion only states them briefly, with no new arguments or comparisons. Recommendations should always follow from specific results, not from an impression.
How long should the discussion be?
There is no binding length and it varies by type of work, field and institution. In a bachelor's thesis the discussion tends to be shorter, in a master's thesis usually more extensive, because it works with more substantial research. More important than the number of pages is the depth of interpretation and the connection to the literature.
The discussion is where it shows whether you can truly work with your research or merely collect numbers. If you interpret your results honestly, compare them with the literature, acknowledge the limitations and draw specific recommendations from them, you gain a chapter that holds the whole thesis together. If you need expert help with the discussion or with the whole thesis, take a look at our services or write to us through a no-obligation order and we will advise you on exactly what your work needs.