The Meter is not a verdict
Consensus summarizes what the literature says as a Yes, Possibly, or No bar. It is a map of the papers, not the answer key, and this lesson drills the difference.
Consensus Meter
Aggregates how relevant papers lean on a yes-or-no question into percentage bars. Use it for orientation on a new topic, never as the final word: it counts papers, not the strength of each one.
Study filters
Narrow results by study design, sample size, or recency, so meta-analyses and randomized controlled trials rise above small pilot studies. This is where a search turns into an evidence check.
Study Snapshots
Per-paper cards that surface population, sample size, method, and outcome without opening the PDF. Read a few snapshots and you know whether the Meter is built on strong or shaky studies.
What you will do in the lesson
You answer and try things, step by step.
- Turn a topic into a question. You take a vague topic and rewrite it as a question a study could actually answer. Consensus rewards precise questions and punishes fuzzy ones, visibly.
- Read the Meter properly. Given a Meter split, you say what it does and does not tell you. Seventy percent of papers leaning yes is not a seventy percent chance of being right.
- Filter to the strong stuff. Apply a study-design filter and watch the results, and sometimes the picture, shift. You practice noticing when weak studies were doing the heavy lifting.
- Open a Snapshot. Pick one paper and read its Snapshot: who was studied, how many, what was measured. Then you decide how much weight that single result deserves.
- Spot the unanswerable question. You identify a question that looks scientific but is really a values question, where no Meter should convince you. Knowing the tool's edges is part of using it.
Before you start
Where do Consensus results come from?
From peer-reviewed academic literature, drawing on a corpus of over 200 million papers. Nothing in the results comes from blogs, news sites, or forums, which is the entire point.
What does the Meter percentage actually mean?
It is the share of relevant papers whose findings lean each way, not a probability that the claim is true. Ten small studies saying yes can be outweighed by one large trial saying no, which is why the lesson pairs the Meter with filters and Snapshots.
Do I need to install anything for this lesson?
No. Everything runs right here in the page, without an account. Consensus itself has a free way to try it when you want to run your own questions.
How is this different from Google Scholar?
Scholar hands you a ranked list of documents and leaves the synthesis to you. Consensus starts from your question, classifies how each paper answers it, and summarizes the lean. You still open the papers, but you arrive knowing which ones matter.
