Correlation vs Causation Explained With Examples
By William DavisJanuary 20, 2026
Quick Answer
Correlation means two things happen together. Causation means one thing makes the other happen.Just because two variables move together doesn't mean one causes the other. Ice cream sales and drowning deaths both increase in summer, but ice cream doesn't cause drowning. The hidden factor is hot weather.
You've probably heard the phrase: "Correlation does not imply causation." But what does it actually mean? And why do smart people get this wrong all the time?
Confusing correlation with causation is one of the most common mistakes in data analysis, news reporting, and everyday thinking. It leads to bad business decisions, misleading headlines, and wasted money on things that don't work.
By the end of this guide, you'll understand the difference clearly, know how to spot false claims, and learn how scientists actually prove causation.
What Is Correlation?
Correlation is a statistical measure that describes the size and direction of a relationship between two variables. When one variable changes, the other tends to change too.
Positive Correlation
Both move in the same direction. Height and weight: taller people tend to weigh more.
Negative Correlation
Variables move in opposite directions. Car speed and travel time: faster speed means less time.
Zero Correlation
No relationship. Shoe size and intelligence: shoe size tells you nothing about IQ.
Correlation is measured by the correlation coefficient (r), which ranges from -1 to +1. A value of +1 means perfect positive correlation, -1 means perfect negative correlation, and 0 means no correlation at all.
Correlation Patterns & The Confounding Variable Problem
Visualizing correlation coefficients (r) and why correlation does not imply direct causation
As X increases, Y consistently increases (e.g., Study Hours & Exam Score)
As X increases, Y consistently decreases (e.g., Car Speed & Travel Duration)
No linear relationship detected between variables (e.g., Shoe Size & IQ)
The Spurious Correlation Trap: Hidden Confounder Example
Variable X
Ice Cream Sales
Variable Y
Drowning Incidents
The True Confounding Variable (Z): Hot Summer Weather. Higher temperatures drive both higher ice cream consumption AND more people swimming at beaches/pools. Eating ice cream does not cause drownings!
What Is Causation?
Causation (also called cause and effect) means that a change in one variable directly produces a change in another variable. Variable A makes Variable B happen.
Clear examples of causation:
- • Flipping a light switch causes the light to turn on
- • Stepping on the brake causes a car to slow down
- • Adding heat to water causes it to boil
- • Lack of water causes plants to wither and die
The Three Requirements to Prove Causation
To claim that A causes B, you must meet all three of these conditions:
1. Correlation
A and B must be correlated. If there's no relationship, there can't be a causal one.
2. Temporal Precedence (Time Order)
The cause must happen before the effect. A must occur before B.
3. Non-Spuriousness (No Confounders)
You must rule out alternative explanations. No third variable (C) can be causing both A and B.
Example 1: Shoe Size and Reading Ability
Studies show a strong positive correlation between shoe size and reading test scores in children. Kids with bigger feet score higher on reading tests. Does this mean we should stretch children's feet to improve their reading?
Analysis:
The correlation: r ≈ 0.8 between shoe size and reading scores
The hidden variable: Age
What's really happening: Older children have bigger feet AND better reading skills. Age causes both. Shoe size doesn't cause reading ability.
Conclusion: This is a spurious correlation caused by a confounding variable (age).
Example 2: Exercise and Heart Disease
Research shows people who exercise regularly have lower rates of heart disease. Is this correlation or causation? Let's apply our three requirements.
Testing for Causation:
1. Correlation exists? Yes (r ≈ -0.4 between exercise frequency and heart disease risk).
2. Time order correct? Yes (longitudinal studies tracked cohorts over 10+ years).
3. Alternative explanations ruled out? Mostly. Controlled for diet, smoking, genetics, and biological mechanisms.
Conclusion: Strong evidence for causation. Exercise does reduce heart disease risk.
4 Common Mistakes When Interpreting Data
Mistake 1: Assuming correlation proves causation
Countries that consume more chocolate win more Nobel Prizes. That doesn't mean chocolate makes people smarter. Wealthier countries have both more chocolate and better education systems.
Mistake 2: Reversing cause and effect
Successful people wake up early. Does waking early cause success, or do successful people have jobs that require early mornings? The direction might be backwards.
Mistake 3: Ignoring confounding variables
People who drink wine moderately live longer than non-drinkers. But moderate wine drinkers also tend to have higher incomes, better healthcare access, and healthier overall lifestyles.
Mistake 4: Cherry-picking time periods
The number of films Nicolas Cage appeared in correlates with swimming pool drownings (r = 0.67). This is pure coincidence from selecting specific years. Given enough variables, random correlations will appear.
How Scientists Establish Causation
If correlation alone isn't enough, how do researchers prove causation?
Randomized Controlled Trials (RCTs)
The gold standard. Randomly assign participants to treatment or control groups. If the treatment group shows different outcomes, the treatment likely caused the difference. Random assignment eliminates confounding variables.
Natural Experiments
Sometimes nature or policy creates random-like conditions. For example, comparing health outcomes in areas where a new policy was implemented versus areas where it wasn't can reveal causal effects.
Hill's Criteria
When experiments aren't possible, epidemiologists use criteria like: strength of association, consistency across studies, specificity, temporal relationship, biological gradient, plausibility, coherence, and analogy.
Frequently Asked Questions
Can correlation ever prove causation?
Correlation alone cannot prove causation. However, strong correlation combined with controlled experiments, a clear mechanism, and elimination of alternative explanations can establish causation. The correlation is necessary but not sufficient evidence.
What is a spurious correlation?
A spurious correlation is a statistical relationship between two variables that appears meaningful but is actually caused by a third factor or pure coincidence. The divorce rate in Maine correlates with margarine consumption, but neither causes the other.
Why do people confuse correlation with causation?
Our brains naturally look for patterns and explanations. When we see two things happening together, we instinctively assume one causes the other. Media headlines also often imply causation from correlational studies to make stories more interesting.
Conclusion
Understanding correlation vs causation protects you from drawing wrong conclusions. Here's what to remember:
- Correlation means two variables move together. Causation means one makes the other happen
- Proving causation requires correlation, correct time order, and ruling out alternative explanations
- Always look for confounding variables that might explain both factors
- Randomized controlled trials are the best way to establish causation
