As an editor, I often read discussions that are long winded, meandering and fail to make an impact.
They feel like an after-thought rather than a key part of the paper.
Here are some tips to make your discussion better. Full credit to @sanjaysaint who taught me this years ago!
First, what are the goals of the discussion? One possible list:
Summarize the key findings of your paper
Put your findings into context of what is known
Discuss strengths, limitations
Tell readers how reading this paper can help advance science
One way to achieve these goals is to think of each paragraph or two in your discussion as a block dedicated to a specific mission.
So, writing the discussion "by numbers" is a useful construct to follow.
Para 1: Summarize the key question/findings. Why is this an important topic, what were your main findings, why are they relevant.
Eg: “In this study, we found… We also found…… Our findings are important because….” In para 1, you're telling the reader why your study matters!
Para 2: Go over additional findings of importance.
This paragraph should focus on secondary outcomes, those from subgroups or other sensitivity analyses that add more richness to the main findings.
Eg: “In addition to our main findings, we also found that…..”
Para 3-5: Now that you’ve laid out your findings, put them into context.
Use these paras to show mastery over content. Define current state, cite key studies, describe what they've shown, what gaps remain and how do your data address these?
Para 4-5: Describe whether your findings are in line with others (if so, credit them) and if not, explain why not? Eg: “Our study support prior findings by researchers X et al … However, we also found important differences including… These differences may be explained by….”
Para 6-7: Limitations & strengths. Think like a reader! What may a reader think are key measurement, scientific, validity concerns? What did reviewers raise? Are there ways you have addressed these? Adding a sentence for how a limitation was managed is particularly impactful.
For example:
“This is an observational study and can only show associations, not causation. However, our use of method X provides greater rigor and confidence in our conclusion.”
Final paras: These are key and are often missing!
- What should future studies on this topic look like? What are the remaining questions? Does your study shed light on design or sample size or other technical issues for these studies?
- What are clinical, research or policy implications from your study? This is important and often missed! What should we do differently or be thinking about differently now that you have done this work?
-Finally, don't end with “More research is required.” More research will always be required!
Instead, close with the key take home message for your paper and what you want readers to remember.
EG: “Future studies that examine x,y,z accounting for a,b,c would be welcomed.”
Fin – pls add to this thread if you have additional ideas or thoughts about how to write an impactful discussion!
And look forward to seeing better discussions!
• • •
Missing some Tweet in this thread? You can try to
force a refresh
@sood_lonika asked recently about how to write good external letters for evaluation of promotion.
I write a LOT of these. I always begin by reviewing all the materials sent (CV, Exec Summary, NIH bio sketch, papers, personal statement, etc)
Then, here is how I approach them:
First para:
-Thank the [person] asking for the letter
-Establish your credentials for the review (most require you to be at rank or higher than the person being promoted)
-State you have reviewed relevant criteria
-Verify objectivity (no COI, "arms length" reviews, etc)
Second para:
-Briefly summarize the individuals background, credentials from their CV/Exec Summary/other material
-Remember, this is a colleague (could be you one day), so adopt a positive lens
-Aim to help your colleague succeed
@BMJ_Qual_Saf First, head over to improvePICC.com and click on “Implementation” in the top menu - this will take you to the Roadmap for Quality Improvement.