How to write a discussion chapter

The results chapter answers "what did I find?"; the discussion answers "what does it mean and why does it matter?" This is usually where the committee judges your contribution, and it is also the chapter students find hardest, because knowing your results and interpreting them are different skills.

In some fields and journals, results and discussion are combined into a single chapter. Check your graduate school’s formatting guide or the target journal’s guidelines for chapter names and order; the structure below works in either layout.

The structure of a discussion chapter

1. State the answer up front

In the first paragraph, remind the reader of the research question and summarise the answer your findings give in a sentence or two. This is a synthesis, not a list of results: instead of "two of the three hypotheses were supported", write something like "the results show that X is associated with Y, but that the association weakens under condition Z".

2. Compare your findings with the literature

For each main finding, answer: Does it agree with earlier studies? If so, what does it add? If not, why might that be? When explaining a difference, look at differences in sample, instrument, context or design. Most of the studies you cite here should already have been introduced in the literature review; the discussion is where you close the conversation you opened at the start. The literature review guide can help you build that link.

3. Explain unexpected results

A result that does not support your hypothesis, or that surprises you, is not a weakness of the thesis; glossing over it is. List the possible explanations, discuss how well each fits your data and say which seems more likely, with your reasons. If there is no clear explanation, say so openly. Committees tend to ask about exactly these results.

4. Acknowledge limitations

Writing limitations is not about shrinking your findings but about showing honestly how far they can be trusted. For each limitation, state how it affects interpretation: rather than "the sample came from a single city", continue with "so generalising the findings to other regions requires caution". The limitations here should be consistent with those in your methods chapter; more in the methods chapter guide.

5. Set out the implications

Theoretical implications describe how your findings support, limit or extend an existing theory. Practical implications say who could do what differently. Write both only to the extent your findings actually support them; drawing policy prescriptions from a cross-sectional study will catch the committee’s attention.

6. Make concrete suggestions for future research

"More research is needed on this topic" is not a suggestion. Write concrete questions that arise from your limitations and unexpected results: with which sample, which design, which added variable. A good suggestion shows the reader the open question your thesis leaves behind.

The most common mistakes

Retelling the results

Listing the numbers from your tables again in the discussion is the most common mistake. If you need to refer to a finding, mention it briefly and move straight to its interpretation. The reader has already seen the numbers in the results chapter.

Claims the data cannot carry

Interpreting a correlational analysis with words like "affects" or "causes", generalising from a small or unusual sample, and confusing statistical significance with practical importance are the points committees object to most. Choose your verbs to match the strength of the finding: "shows", "indicates" and "suggests" are different degrees.

New results in the discussion

Presenting an analysis or result for the first time in the discussion, when it does not appear in the results chapter, confuses the reader and breaks the structure. If an interpretation depends on a new analysis, move that analysis into the results.

Ignoring contradicting literature

Citing only studies that agree with your findings is quickly noticed by a committee member who knows the field, and leads to the question "this study found the opposite; how do you explain that?" Raise contradicting results yourself first and discuss the likely reasons for the difference.

Drifting from the question

As the discussion goes on, it is easy to drift into topics that are interesting but unrelated to your research question. At the end of each paragraph, ask "what does this add to the answer to my question?"

A quick check before submission

• Does the first paragraph give the answer to the research question?
• Has every main finding been compared with at least one earlier study?
• Are results that do not support the hypothesis explained?
• Do the verbs match the strength of the findings?
• Is there any result in the discussion that is not in the results chapter?
• Is each limitation linked to the interpretation?

The discussion is one of the chapters that draws the most questions at a defense; we have collected the ones to expect in the thesis defense questions guide.

tezkeys reads your discussion together with your results and literature: it marks places where findings are interpreted more strongly than the data allow, passages that drift from the question and inconsistencies between numbers in tables and in the text, all on your own text; it does not rewrite the text for you. The report is AI-generated, can be wrong and does not replace your supervisor. See what a report looks like in the sample report.