A suggested workflow for interpreting the findings. This article maps is aimed at developing the Discussion arguments for your main Discussion (see also: IMRaD sections)
Step 1: List your research questions
- List each result that’s relevant to that question
Define your questions clearly
If you cannot list specific research questions (not vague topics), stop here. You need to return to Asking questions and write a question first. You cannot interpret data without a specific question to answer.
Step 2: Brainstorming interpretations
This is a creative stage. There are no wrong answers (that’s for step 4).
- For each result, answer this question
If results x is true not just for your study population, but for all … (e.g. tidal flats worldwide, patients with this conditions, firms in this sector), how does this result answer the research question?
- Rule: Use only your results. No citations, no ‘Smith et al. found… (that’s for step 3)
Define your questions clearly (yes, I'm saying this again)
If you find yourself writing ‘the results suggest an effect,’ return to your research question—did you ask precisely enough to demand a specific answer?
Step 3: Support your interpretations
→ See also: Supporting with evidence
For each interpretation from step 2:
- List literature that supports this interpretation (e.g. in a table).
Step 4: Conflicting evidence
For each interpretation from step 2, list statements that don’t support this interpretation:
- Contradictory results
- Conflicting literature
- Assumptions you made about the data
- Potential study limitations
Step 5: Verify your interpretations are reasonable
Ultimately, interpreting the results is a subjective activity (see also Know the Product).
You will need more support for your conclusions if for example:
- Your conclusion moves far away from the original data (i.e. borders on being speculative)
- You have sceptical reviewers
- You make a lot of assumptions about the data
You will likely have blind spots, so the best way to check your conclusions is to share them with your supervisors and/or co-authors.
Tip
You don’t have to do these steps in sequential order. If you find that while you’re brainstorming, you keep coming up with reasons why your conclusion is wrong, you can already list those reasons in step 4.