Crime taxonomy · Words and context

Incorrect or unsubstantiated attribution

Crediting a change to a cause the chart cannot prove.

What it looks like

A title, caption or annotation says a policy, leader, product or event caused the change shown, but the chart cannot support it: the change started before the cause, there is no comparison group, or other obvious drivers are ignored. It also covers a cited source that could not have produced the data shown.

Example

The same data shown honestly: The full series shows the fall began years earlier, a similar city without the policy fell as much, and the note names the other drivers: cleaner buses and taxis, fewer diesel cars, more remote work. — **The fix.** The full series shows the fall began years earlier, a similar city without the policy fell as much, and the note names the other drivers: cleaner buses and taxis, fewer diesel cars, more remote work.

Why it is bad

A chart shows that something changed, not why. The caption quietly adds a causal claim the data never tested.
Long-running trends get claimed by whoever was in charge at the end of them.
A credible source name lends authority to a conclusion that source may never have drawn.
A typical example: maps of falling NO₂ in a city captioned as the result of bike lanes and car restrictions, ignoring that buses and taxis switched to electric or hybrid, cars moved away from diesel (the main NO₂ source), and post-pandemic remote work cut the number of trips.

How to fix it

Describe what changed, and claim a cause only when the evidence supports it.
Show the trend before the cause, so readers can see whether it changed course.
Add a comparison: a similar place, group or period without the policy or event.
Name the other drivers (technology, the economy, a pandemic, a new measurement method) and say how they were accounted for.
Quote what the cited source actually concluded, and link to it.

When it is not a crime. Causal titles are fine when the evidence is there, such as a controlled study, a clear break in the trend at the moment of the cause, or a comparison group that did not change.