How to Write a Discussion Section for Your Thesis

Check your programme requirements first

Before writing, check:

• the graduation manual;
• the assessment criteria;
• the required chapter structure;
• instructions from your thesis supervisor;
• examples provided by your programme.

Check whether your programme expects:

• a separate discussion chapter;
• a combined conclusion and discussion;
• a combined results and discussion chapter;
• separate practical recommendations;
• future research suggestions;
• a reliability and validity section;
• theoretical and practical implications.

Use the terminology and structure required by your programme.

Some universities of applied sciences place greater emphasis on the practical meaning of the findings. Research universities may expect a more detailed theoretical discussion. This is not a fixed rule. Always follow the criteria for your own degree programme.

How to write your discussion in 7 steps

Step 1: Return to the research question

Read your research objective, main question and sub-questions again.

Ask:

• Which findings provide the clearest answer?
• Which findings require further explanation?
• Did the results answer every sub-question?
• Has the scope changed during the study?
• Which conclusions are supported by the data?

Your discussion should remain focused on the research you actually conducted.

Do not introduce a new question simply because an interesting result appeared.

Step 2: Summarise the central findings

Begin with a short overview of the most important findings.

Do not repeat the complete results chapter. Focus on the overall pattern.

For example:

The findings show that supervisor support, workload and career development all influence employees’ intention to remain with the organisation. Supervisor support appears to play a central role because it also affects how employees experience workload and development opportunities.

This opening gives the reader the main line of the findings.

Avoid listing every percentage, quotation or statistical result again.

Step 3: Interpret what the findings mean

Explain the meaning of the results within the context of your study.

Ask:

• What is the central pattern?
• Why might this pattern have occurred?
• Which explanation is supported by the data?
• Does the result apply to all participants?
• Did different groups respond differently?
• Does the result support the hypothesis or expectation?
• Which alternative explanations remain possible?

Make clear when you are presenting a possible explanation.

For example: One possible explanation is that employees experience supervisor support as a source of both practical help and recognition. However, the study did not test whether supervisor support directly causes higher retention.

The first sentence offers a plausible interpretation. The second prevents an unsupported causal claim.

Step 4: Compare the findings with literature

Connect your findings to the studies, theories and models discussed earlier in your thesis.

For each important finding:

1. state your result;
2. compare it with previous research;
3. explain whether it supports, contradicts or extends existing knowledge;
4. discuss a possible reason;
5. link the comparison back to your research question.

For example: The importance of supervisor support is consistent with earlier studies on employee retention. However, the present study found that support was especially important for remote employees. This may be because they have fewer informal opportunities to ask questions or build relationships with colleagues.

Do not cite sources only to show that you have read them. Explain the relationship between your findings and the literature.

Your results may:

• confirm earlier research;
• partly support it;
• contradict it;
• show that a relationship depends on context;
• apply a theory to a new group;
• reveal an aspect that earlier research overlooked.

Step 5: Discuss unexpected or conflicting results

Unexpected findings are not automatically mistakes.

They may show:

• that the original assumption was incomplete;
• that the context differs from earlier studies;
• that participants interpreted a concept differently;
• that another factor influenced the outcome;
• that the sample contained different subgroups;
• that the method affected the result.

First check whether the finding could result from:

• a coding or calculation error;
• missing data;
• an unclear question;
• a small subgroup;
• inconsistent data collection.

If the result remains valid, discuss it honestly.

For example: Contrary to expectations, employees with longer tenure reported a greater need for development opportunities. A possible explanation is that experienced employees saw fewer internal career options than newer employees.

Do not force unexpected findings to fit your original expectation.

Step 6: Explain implications and limitations

Discuss what the findings mean for theory, practice or both.

Theoretical implications

Depending on your study, explain whether the findings:

• support a theory;
• challenge an assumption;
• show that a relationship is context-dependent;
• extend an existing model;
• highlight a concept that requires further research.

Do not claim a major theoretical contribution when the study was small or exploratory.

Practical implications

For a practice-based graduation project, explain what the organisation can learn from the results.

For example:

• which needs are most important;
• which assumptions were incorrect;
• which groups require different approaches;
• which conditions affect a possible solution;
• which problem should be addressed first.

Keep practical implications separate from fully developed recommendations when your programme uses a separate recommendations chapter.

Limitations

Explain the limitations that affect the interpretation of the results.

For each important limitation, describe:

1. what the limitation was;
2. why it occurred;
3. which findings it may have affected;
4. what can still be concluded;
5. which claims need caution.

Do not simply list weaknesses.

Step 7: Identify future research

Suggest future research only when it follows logically from:

• an unanswered question;
• a limitation;
• an unexpected finding;
• a new pattern;
• a group not included in the study;
• a relationship that requires stronger testing.

Be specific. For example: Future research could examine whether the relationship between supervisor support and retention is also found in other departments and organisations.

Do not ask future researchers to repeat the complete study without explaining what they should change or add.

How to discuss reliability and validity

Your methodology explains which quality measures you used.

The discussion evaluates the remaining risks and their consequences.

You might discuss:

• whether the sample reflected the target group;
• whether participants interpreted questions consistently;
• whether a translated instrument may have changed meaning;
• whether researcher involvement affected interviews or coding;
• whether non-response created bias;
• whether the context limits generalisability;
• whether alternative explanations remain possible.

For example: All interviews followed the same topic guide and were recorded and transcribed. This improved consistency. However, the researcher also worked within the organisation, which may have influenced how openly participants discussed negative experiences.

Name only measures and limitations that actually apply to your study.

Discussion in qualitative research

A qualitative discussion often focuses on:

• the meaning of themes;
• differences between participants;
• contrasting cases;
• links between themes;
• comparison with earlier qualitative studies;
• context;
• credibility;
• transferability;
• researcher influence.

Do not treat the number of times a theme was mentioned as its only measure of importance.

A less common perspective may still be relevant when it reveals:

• a different experience;
• an exception;
• an unmet need;
• a condition under which the main pattern changes.

Use quotations in the results section to support themes. Use the discussion to explain what those themes mean.

Discussion in quantitative research

A quantitative discussion often focuses on:

• whether hypotheses were supported;
• direction and strength of relationships;
• unexpected group differences;
• effect sizes;
• practical relevance;
• comparison with earlier studies;
• alternative explanations;
• measurement quality;
• sampling and generalisability.

Statistical significance does not automatically mean that a finding is important in practice.

A non-significant result also deserves discussion when it relates directly to a research question or hypothesis.

Do not write that the null hypothesis has been “proven”. Explain only what the analysis supports.

Discussion in practice-based research

In a practice-based graduation project, the discussion connects research findings to the organisational problem.

Explain:

• which assumptions were confirmed;
• which assumptions were not confirmed;
• what the organisation now understands better;
• which target groups have different needs;
• which conditions influence a possible solution;
• what the findings mean for the central problem.

Do not move directly from one result to a preferred solution.

For example:

Employees reported that existing onboarding information was difficult to prioritise. This suggests that the problem is not only a lack of information, but also the timing and structure in which it is provided.

This implication can later support a recommendation. It is not yet the full recommendation itself.