Start here · Essential guide
How to Tell a Correlation From a Cause in Trend Stories
A practical set of questions for separating a pattern from a cause when a trend headline says one thing leads to another, and why the gap matters.
Evidence and Method
Correlation, Causes and Competing Explanations
Reasoning tools for judging why a trend happened.
Confounding Variables Explained With Everyday Examples
How a hidden third factor can make two unrelated trends look connected, explained through ordinary examples and simple ways to spot and reduce the problem.
How Researchers Use Natural Experiments
How sudden policy or platform changes let researchers compare before and after without running a trial, and the assumptions such comparisons depend on.
Reverse Causation and When the Arrow Points the Other Way
How an apparent effect can actually be the driver, with examples about habits and mood and the questions that help spot reversed arrows in trend claims.
Selection Effects and Why Early Adopters Mislead
Why the people who join a trend first tend to differ from everyone else, and how that can distort conclusions about who a trend is for and what it does.
Survivorship Bias in Success Stories About Trends
Why stories about winners hide the many that failed, how this skews our view of trends and advice, and the questions that help reveal what is missing.
The Ecological Fallacy and Group Patterns Versus Individuals
Why a pattern across regions or groups does not describe any one person inside them, with examples and questions for reading group-level trend data.
Why Two Reasonable Explanations Can Fit the Same Trend
How competing accounts of one trend can both match the evidence, and the kinds of data, timing and comparisons that might tell the explanations apart.
Related topics