Evidence and Proof
People make claims constantly: this diet cures cancer, this politician lied, this product works better than competitors. Some of those claims are true, some are false, and many are somewhere in between. What separates a well-supported claim from an empty assertion is evidence — observable, verifiable information that gives us reason to think the claim is accurate. Learning to ask 'what is the evidence?' and then evaluate the quality of whatever answer you receive is one of the sharpest tools in a critical thinker's kit.
A claim without evidence is just an opinion. Opinions are fine in the right context — you are entitled to prefer one movie over another based on taste alone. But when a claim is about facts in the world — what happened, what causes what, which treatment works — opinions need to be backed by evidence. Otherwise, any two people can assert opposite things and there is no way to resolve the disagreement.
The Evidence Hierarchy
Not all evidence is equally strong. Scientists, doctors, lawyers, and historians all recognize that some forms of evidence are more reliable than others. At the top of the hierarchy is experimental evidence collected under controlled conditions — the kind produced by randomized controlled trials in medicine, or replicated laboratory experiments in science. Below that comes observational evidence: careful measurement of what happens in the real world, without artificial control. Lower still are expert consensus statements, case studies, and individual expert opinion. At the bottom of the pile are anecdotes — personal stories — and testimonials.
An anecdote is a personal story: 'My grandfather smoked every day and lived to 95.' Anecdotes are vivid and emotionally compelling, but they are weak evidence because a single case cannot represent the full range of outcomes. The grandfather who smoked and thrived is memorable; the thousands who smoked and died early are invisible. Human memory selectively preserves dramatic stories, which is exactly why statistical data across large samples is more reliable.
The gold standard for evidence about cause and effect is the randomized controlled trial, or RCT. In an RCT, researchers randomly assign participants to two groups: one receives the treatment being tested, the other receives a placebo or a comparison treatment. Because assignment is random, the two groups should be similar in all other respects. Any difference in outcomes can therefore be attributed to the treatment, not to pre-existing differences between participants. RCTs are expensive and sometimes impossible to run ethically, which is why other forms of evidence matter too.
Recognizing Weak Evidence in Practice
In everyday life, people rarely cite randomized trials. Instead, you will encounter several common patterns of weak evidence that are worth recognizing. The celebrity endorsement: a famous actor claims a product changed their life. Their fame gives them no expertise on the product's effects. The cherry-picked statistic: a company reports that 8 out of 10 dentists recommend their toothpaste — without revealing that only 10 dentists were surveyed, and two declined to participate. The appeal to authority on the wrong topic: a Nobel Prize winner in chemistry opines on economic policy; their prize proves expertise in chemistry, not economics.
Screenshots and out-of-context quotes are especially deceptive. A screenshot can be fabricated entirely, or it can be real but cropped to remove context that would change its meaning completely. Before treating a screenshot as evidence, always try to find the original source and read the full context. The same applies to statistics — a statistic without context can be technically true and deeply misleading at the same time.
Match each type of evidence to its key strength or weakness.
Terms
Definitions
Drag terms onto their definitions, or click a term then click a definition to match.
Burden of Proof
A key principle in logic and law is that the burden of proof rests on the person making the claim, not on those who are skeptical. If someone claims that a new supplement prevents memory loss, they must provide evidence. The skeptic does not need to prove it does not work; the absence of proven benefit is itself informative. This asymmetry matters enormously. It prevents you from being endlessly put on the defensive about things that have no evidence: 'Prove that it does not work!' is not a valid response to a lack of supporting evidence.
Absence of evidence means no one has found proof yet — this is weak but real information. Evidence of absence means researchers looked carefully and found nothing — this is stronger, and gives good grounds for skepticism. Confusing these two ideas leads people to say 'you cannot prove it does not work' as if that settles the question in favor of the claim. It does not.
A health website claims that 'thousands of customers report feeling more energetic after using our supplement.' Why is this weak evidence that the supplement actually causes increased energy?
A friend says, 'You cannot prove that aromatherapy does not cure headaches, so you cannot say it does not work.' What is wrong with this argument?
Evidence Quality Audit
- Step 1: Find three different claims online — one from a news article, one from a social media post, and one from a company website.
- Step 2: For each claim, identify what evidence (if any) is provided to support it.
- Step 3: Rate the evidence strength for each on a scale: Strong (controlled study or expert consensus), Moderate (observational data or multiple expert opinions), Weak (anecdote, testimonial, or no citation), None (bare assertion).
- Step 4: For the weakest-evidence claim, write two sentences describing what kind of evidence would actually be convincing.
- Step 5: Reflect: Were you surprised by how little evidence was provided for claims that seemed confident and authoritative?