Correlation vs Causation — Practice

Practice bank for UCAT Decision Making questions testing your ability to distinguish correlation from causation.

---

Quick-Reference: Key Concepts

Definitions

• Correlation: Two variables are associated — when one changes, the other tends to change too
• Causation: One variable directly causes a change in the other
• Correlation does NOT imply causation

Why Correlation ≠ Causation

Three alternative explanations for an observed correlation:

Explanation	Example
Reverse causation	"Hospitals with more doctors have higher death rates" — sicker patients go to larger hospitals, not the other way round
Confounding variable	"Ice cream sales and drownings are correlated" — both are caused by hot weather (the confounder)
Coincidence	Two unrelated trends happen to move together by chance

When CAN We Infer Causation?

The gold standard is a randomised controlled trial (RCT):

1. Random allocation eliminates confounders
2. Control group isolates the effect of the intervention
3. Blinding reduces bias

Bradford Hill Criteria (Simplified for UCAT)

• Strength of association (larger effect = more suggestive)
• Consistency (replicated across studies)
• Temporality (cause must precede effect)
• Dose-response (more exposure = more effect)
• Plausibility (a biological mechanism exists)

---

Strategy

1. When a question describes an association, ask: "Could this be explained by a confounder or reverse causation?"
2. Look for the study design — only RCTs can establish causation
3. Observational studies can show association but not causation
4. Target: 45–60 seconds

---

Practice

Complete the 10 assessment questions.