News coverage regularly presents a single statistic — a percentage, an average, a rate of change — as a clean, self-explanatory fact, when in reality nearly every reported statistic rests on a set of methodological choices that can meaningfully change what the number actually represents, and understanding a few of the most common pitfalls makes anyone a considerably more sophisticated consumer of data-driven news.
Average versus median: a genuinely common source of misleading headlines
The mean (average) and the median (the middle value in a data set) can diverge substantially when a data set includes extreme outliers, and income and wealth statistics are the most common real-world example: average income figures for a given population are frequently pulled significantly upward by a small number of extremely high earners, producing an “average” figure that doesn’t accurately represent the typical, median person’s actual experience — which is exactly why economists studying income distribution generally prefer median figures for representing a “typical” household, while average figures remain more common in headline reporting, sometimes because they produce a more dramatic number.
Correlation headlines routinely imply causation the underlying data can’t support
A statistically genuine correlation between two variables (rates of ice cream sales and drowning incidents both rising in summer, to use a classic teaching example) doesn’t establish that one causes the other — both may be driven by a shared underlying third factor (warmer weather, in that example) that the simple correlation alone can’t distinguish from direct causation. Legitimate causal claims require considerably more rigorous study design — controlled experiments, or careful statistical methods designed specifically to account for confounding variables — than a simple observed correlation provides on its own, a distinction that’s frequently lost or glossed over in headline-driven news coverage of a single new study.
The single most useful data-literacy habit for reading news statistics: before accepting a headline number, ask what exactly was measured, how, and over what population — the answer often changes the story significantly.
Sample size and selection matter more than most readers check
A statistic drawn from a small or non-representative sample can produce a headline-worthy number that doesn’t actually generalize reliably to the broader population it’s being presented as describing — survey and study methodology sections routinely disclose these limitations, but they rarely make it into news headlines or summary coverage, which is exactly why reading past the headline into a study’s actual methodology, when the stakes of a claim matter enough to warrant it, remains one of the most reliable ways to evaluate whether a widely shared statistic is genuinely as solid as it’s being presented.