Cut through the hype by using AI and the right questions to get closer to what health research actually says.
Despite my training, I still get a lot of my health and mental health information from sources other than peer-reviewed scientific papers – YouTube channels I follow, news sites I admire, podcasts I enjoy.
The upside of these sources is that they often present information in more engaging, digestible, and human ways.
The downside? In the process of making things easier to understand (or more entertaining), the detail and nuance of the actual research can get lost.
In the worst cases, findings are misrepresented to serve someone’s agenda:
-
Exaggerated results to help sell a product
-
Overhyped claims to chase clicks or views
-
“Science-washing” poor advice by citing just one study
That’s why it’s worth being able to go back to the original source (the actual peer-reviewed study) and dig into what the researchers really found.
But reading original research can be hard, especially if it’s not in your area of expertise. For example, I am very comfortable with psychology research, but would struggle reading a detailed paper on nutritional science.
That’s where AI can help.
🤖 AI as a Research Companion
If you have the original study and a good set of questions to ask, an AI like ChatGPT can give you:
-
A clear explanation of the study’s purpose and results
-
The proposed mechanisms (how and why something works)
-
Critical insights on limitations, conflicts of interest, and alternative explanations
-
Realistic interpretations—not just what the study says, but what it doesn’t say
- Realistic expectations – what would happen if you changed your life on the basis of the research
I’ve been testing this out over the past few months (digesting a lot more original research) and have been genuinely impressed by how well AI can:
-
Translate jargon-heavy studies into plain language
-
Flag potential traps (e.g. correlation vs. causation)
-
Help non-experts become more research-literate
So I worked with ChatGPT to develop a set of questions that anyone can use to extract meaningful, balanced insights from a study.
You’ll find them below and in the attached document.
🧾 The Idea
Next time you see someone mention “a study” in an article or video, especially if it’s about a health claim, try this:
-
Find the original research (or as close as you can get).
-
Give the study to AI, along with the list of questions (cut and paste them into the AI prompt window)
-
Get a summary that cuts through the hype and brings you closer to what the study actually found.
The bigger picture goal here is building your health literacy, your ability to navigate a noisy world of health claims, and to make decisions that serve you better.
⚠️ Quick Note for Students in the Context of your Studies
This post is about using AI to build your own knowledge and protect yourself from misleading health advice. It’s not about using AI to write assignments. If you’re a student, always check your topic or assessment requirements before using AI tools for anything related to coursework.
👀 Why This Matters
I’m not here to demonise non-peer-reviewed health information. I work in that space. I write posts, create summaries, deliver talks.
But I also know there’s often a significant gap between how research is communicated in popular media and what it actually says. And sometimes, that gap can lead to bad decisions based on misunderstood science.
So if we want to take control of our health, it helps to be able to look under the hood, even just a little.
I hope these questions help.
I’ll keep refining them as I learn more. And I’ll update this post from time to time if the framework improves.
You can have the questions as a document if you’d like. Otherwise you can bookmark this post and cut and paste the questions straight from it.
🧠 AI Research Companion: 24 Questions to Understand and Use Health Research
Use these prompts with an AI chatbot and a research article to get a clear, realistic, and critically informed summary. Perfect for curious readers, health professionals, educators—or anyone trying to make sense of studies shared online.
🔍 Understand the Basics
- What was the study trying to find out, in plain language?
- Why does this topic matter now? What’s the real-world issue this study is responding to?
- What kind of study was this? (e.g. randomized trial, survey, lab experiment?)
- Who was in the study? Were the participants similar to me or the people I care about?
🛠️ Explore How It Works
- What exactly did they do? What were the interventions or conditions?
- How did they measure outcomes? Were the tools valid and meaningful?
- What were the key findings? Be specific, not just “it worked.”
- What were the size and direction of the effects? Were changes big enough to matter in real life?
⚙️ Dig Into the “Why” — Mechanisms & Explanations
- Do the authors explain why these results happened? What biological, psychological, or social processes are proposed?
- Are there other plausible explanations? Could something else explain the results?
🎯 Real-World Relevance
- Who might this help? And who might it not help?
- What actions could someone take based on this? Are those actions realistic and safe?
- What barriers or challenges might come up in trying this in real life?
⚖️ Stay Grounded — Limits, Biases, and Alternatives
- What limitations did the authors admit? Are there others they didn’t?
- Was this a small or preliminary study? Does the sample size affect how confident we should be?
- Are there any conflicts of interest, funding sources, or affiliations that might affect interpretation?
- What is the most logical counterargument to the study’s conclusions? Why might someone disagree?
🚨 Avoid Common Traps
- Is this being overhyped or oversimplified? (e.g. “X cures Y” headlines)
- Is correlation being confused with causation? Did the study actually test what it’s claiming?
- Do the findings apply to everyone? Or just a specific group (e.g. young people, mice, people with a certain condition)?
- What is the difference between statistical significance and real-life importance here?
🧭 Put It All Together
- How confident should I feel in acting on this study right now?
- What would I want to see in future studies to feel more certain?
- What would be sensible actions someone could take based on this study—without overreacting or ignoring uncertainty?
- If I had to explain this to a friend or colleague in 60 seconds, what would I say?