Client draft
Profile
✓ Active — content personalised to your coaching profile
Campaign
💡 Tip: Tick the checkbox on any study card to select it, then click ✨ Create content → in the purple bar to generate a combined email, newsletter or post covering all selected studies at once — great for building themed content around a topic.
🔑 What do the badges mean?
Study type — how much evidential weight it carries
◆ Meta-analysisStrongest — when it pooled trials. Combines the numbers from many studies into one result, so it reflects the whole evidence base rather than a single study. Check what it pooled: a review that combined observational studies inherits their limits and shows a pattern, not a cause.
▲ Systematic reviewVery strong. Searches and appraises all the studies on a question to a set method, without necessarily pooling the numbers.
● RCTStrong. Participants randomly assigned to a treatment or a control, which is what allows a study to show cause and effect.
◐ Non-randomised trialModerate. Something was deliberately changed and outcomes measured before and after — but participants chose to take part rather than being randomly assigned. The change is real; how much of it the intervention caused is not settled.
CohortSupporting. Follows a large group over time without assigning anything. Used where no trial exists — often because a trial would be impractical or unethical.
■ GuidelineNot a study — a recommendation. A clinical guideline or expert commission report. An expert panel reviewed the evidence and agreed what should be done; it ran no experiment of its own, and any figure in it comes from research the panel reviewed or a model it built. Part of what a guideline says is trial evidence and part is expert consensus, and the good ones say which is which. Read it as current best practice, not as a finding — and remember guidance is periodically updated and superseded.
Cross-sectionalWeakest. A snapshot of a population at a single moment. It can show that two things occur together, but not which came first — so treat it as a starting point for a question, never as an answer.
Flags — read these before you build content on a study
≈ ObservationalShows an association, not causation. The two things move together, but the study cannot prove one caused the other. Say "linked to", never "causes".
≈ Not randomisedSomething was changed and measured, so you can describe what happened. But those taking part chose to, so part of the change may be down to who volunteered. Don't say the intervention caused it, and don't apply it to people who didn't opt in.
≈ Observational · pooledA meta-analysis or review that combined observational studies. Pooling makes the estimate more precise, but it does not make it causal — twelve cohort studies added together are still twelve cohort studies. Say "linked to", never "causes".
≈ Partly observationalA review that pooled both trials and observational studies. You can describe what the trials found, but the headline pooled figure isn't proof of cause, because not every study behind it tested a change.
⚠ Lower-tier journalPublished in a broad-scope open-access journal with less selective peer review. Sound enough to include, but cross-reference before relying on it alone.
★ LandmarkAn older study kept deliberately because it remains the best available evidence on its topic. Age here is not a weakness — nothing better has replaced it.
Every study also shows its publication year. Where a DOI is available, a ↗ View study link takes you to the original paper.