Numbers Don't Lie — But Headlines Can
Every day, news articles cite statistics to make their stories more authoritative. "Risk doubles," "nine out of ten dentists agree," "unemployment falls to record lows." These claims sound compelling — but without context, they can paint a very distorted picture.
Becoming a smarter reader of statistics doesn't require a math degree. It requires knowing what questions to ask.
Key Questions to Ask About Any Statistic
- What is the sample size? A study of 12 people tells you far less than one of 12,000. Small samples produce unreliable results.
- Who conducted the study? Research funded by an interested party — a pharmaceutical company, a food brand, a political group — may have conflicts of interest baked in.
- What was the comparison group? "Better than average" means nothing if the average isn't defined.
- Is this correlation or causation? Two things happening together doesn't mean one causes the other.
- What's the actual number behind the percentage? A "100% increase in cases" could mean going from 1 case to 2.
Absolute vs. Relative Risk: A Critical Distinction
This is one of the most commonly misunderstood statistical concepts in health reporting.
- Relative risk: "Drug X reduces heart attack risk by 50%."
- Absolute risk: "Drug X reduces heart attack risk from 2% to 1%."
Both statements describe the exact same drug. The relative version (50%) sounds dramatic. The absolute version (a 1 percentage point reduction) gives you the full picture. Always look for the absolute numbers.
Averages Can Mislead
When a statistic uses "average," it's worth asking: which kind?
| Measure | What It Means | When It's Misleading |
|---|---|---|
| Mean | Sum divided by count | Skewed by extreme outliers |
| Median | Middle value in a sorted list | Hides the spread of data |
| Mode | Most frequently occurring value | Can be far from typical values |
When a headline says "average salary is $75,000," a median would better represent most workers if a handful of very high earners are pulling the mean upward.
Watch Out for Cherry-Picked Timeframes
Graphs are powerful tools — and powerful tools for distortion. A stock, a temperature trend, or a crime rate can look dramatically different depending on where you start and end the timeline. Always ask: why was this period chosen? What does the longer-term trend look like?
The "Statistical Significance" Caveat
A result described as "statistically significant" simply means it's unlikely to have occurred by chance — it does not mean it's large, important, or practically meaningful. A medication could produce a statistically significant improvement that is clinically trivial in real life.
Build Your Statistical Intuition
The goal isn't skepticism for its own sake — it's informed reading. A few habits that help:
- Look for the original source, not just the headline.
- Seek out the methodology section of any cited study.
- Compare multiple outlets covering the same data.
- Ask yourself: does this finding match the real world as I experience it?
Statistics are invaluable tools for understanding the world. Learning to read them critically puts you in control of what you believe — and why.