Correlation vs causation
Two numbers moving together (correlation) does not mean one causes the other. A third factor, the direction of cause or plain chance can explain it.
Formulamoved together ≠ one caused the other
Try it · 30 seconds · from Atterna’s practice
“Revenue grew 23% last week, driven by the AUTUMN20 promo.”
Which part can the data alone not confirm?
Preparing the practice…
How we work it through
- 1
List everything else that changed in the same period.
- 2
Look for a comparison group that did not get the change.
- 3
Write “coincided with” until you have that comparison.
On the practice dataset
The query we run on Meridian, Atterna’s practice marketplace, and what it returns.
SELECT week_id, SUM(spend_eur)
FROM meridian.marketing_spend
GROUP BY week_id ORDER BY week_id;Resultmarketing spend €3,200 · €3,350 · €10,700 — it tripled in the promo week too
The common trap
Writing “driven by the promo” in a summary when the data only shows the two happened at the same time.
When an AI answers
“Revenue grew 23% last week, driven by the AUTUMN20 promo.”
- What it didn’t know
- It did not know marketing spend tripled and new sign-ups rose the same week.
- Ask back
- “What evidence shows the promo caused it? What else changed that week?”
Read next
Examples use Meridian, Atterna’s fictional practice marketplace.