The weakest discussion sections retell the results in longer sentences. The reader has just read the results; what they need now is what those results mean.
The order that works
- The main finding, stated in one sentence.
- How it fits or conflicts with previous work, and why.
- A mechanism that could explain it.
- Alternative explanations and why you find them less convincing.
- Limitations, then implications.
Disagreement with previous work
- This is the most interesting part of most papers.
- Examine differences in sample, setting, measurement, design.
- Propose a testable explanation for the difference.
- Do not simply note that results differ and move on.
Alternative explanations
- Raise them before the reader does.
- Explain why your data make them less likely.
- A paper that anticipates objections reads as far more secure.
Implications
- Only claim what your design supports.
- State the conditions: for whom, in what circumstances.
- Practice recommendations beyond the evidence weaken the whole paper.
The closing sentence
- What the reader should take away.
- Not a summary of the paper.
- Specific, and proportionate to the evidence.
One thing worth remembering
Raise the alternative explanations yourself, before the reviewer does.
A paper that names the competing account and explains why the data disfavour it reads as work the author understands deeply — while the same paper with that paragraph missing invites exactly the objection it could have answered.
Câu hỏi thường gặp
What makes a discussion section weak?
Retelling the results in longer sentences — the reader has just read them and needs to know what they mean.
What order works?
Main finding in one sentence, how it fits or conflicts with previous work and why, a possible mechanism, alternative explanations, then limitations and implications.
How should you handle disagreement with previous work?
Examine differences in sample, setting, measurement and design, and propose a testable explanation rather than simply noting the results differ.
Why raise alternative explanations?
Because a paper that anticipates objections reads as far more secure than one that leaves the reader to think of them.
What should the closing sentence do?
State what the reader should take away — specific, proportionate to the evidence, and not a summary of the paper.
