AP CSP Day 43: Correlation vs Causation | Cycle 2
Share
Confounding variables are hidden factors that causally affect both variables in a correlation, creating the appearance of a direct relationship that does not exist. For example, ice cream sales and drowning rates both increase in summer, but ice cream does not cause drowning; hot weather is the confounding variable. AP CSP Cycle 2 correlation questions present realistic data scenarios and ask students to identify the most plausible confounding variable or to explain why the stated causal conclusion is not supported. Proposing an alternative explanation that involves a confounding variable is a strong exam response strategy.
📚 Study the Concept First (Optional) Click to expand ▼
Confounding Variables: Harder Analysis
What Is a Confounding Variable?
A confounding variable is a third variable that causes changes in both variables being studied, creating the appearance of a direct relationship between them. Identifying confounders requires domain knowledge and creative thinking about alternative explanations.
Ruling Out Confounders
The only way to rule out confounders is a randomized controlled experiment where all other variables are held constant. Observational data, no matter how large the dataset, cannot definitively rule out unmeasured confounders.
Practice Question
A study finds that students who eat breakfast daily score higher on standardized tests than students who skip breakfast. A newspaper headline states: "Eating Breakfast Boosts Test Scores!" Which of the following best evaluates this headline?
The study found a correlation (breakfast eating is associated with higher scores) but did not establish causation. A confounding variable like household income could independently influence both: higher-income families may be more likely to eat breakfast AND have access to educational resources that improve test scores.
A) The study shows correlation, not a controlled experiment proving causation. Even if breakfast helps cognition, the study design does not isolate that effect. B) Breakfast may have benefits, but the issue is that this particular study does not prove causation. D) Correlation studies do not prove causal links regardless of data quantity.
Students accept headlines at face value without evaluating whether the underlying study design (observational vs. experimental) supports causal claims. Observational studies show correlation, not causation.
Headlines often overstate study findings. When a question describes an observational study (no controlled experiment), the correct answer will almost always reference correlation, not causation.
Keep Practicing!
Consistent daily practice is the key to AP CSP success.
AP CSP Resources Get 1-on-1 Help