Confounding Variables Examples in Statistics
Start with the recap, study the fully worked examples, then use the practice problems to check your understanding of Confounding Variables.
This page combines explanation, solved examples, and follow-up practice so you can move from recognition to confident problem-solving in Statistics.
Concept Recap
A confounding variable is a third variable that influences both the independent variable and the dependent variable simultaneously, creating a spurious association between them that can be mistaken for a direct causal relationship. Confounders are a major threat to the internal validity of observational studies.
Ice cream sales and drowning deaths correlate. Confounding variable: hot weather. It causes both! Without recognizing confounders, you'd wrongly blame ice cream for drowning.
Read the full concept explanation →How to Use These Examples
- Read the first worked example with the solution open so the structure is clear.
- Try the practice problems before revealing each solution.
- Use the related concepts and background knowledge badges if you feel stuck.
What to Focus On
Core idea: Confounding Variables checks whether the study design supports a fair comparison before interpreting the outcome.
Common stuck point: Students often know a procedure related to confounding variables but skip the recognition step: Did the study use a design feature that makes the groups comparable before the outcome is measured? That leads to a calculation or graph that looks reasonable but answers a different question.
Sense of Study hint: Ask: Did the study use a design feature that makes the groups comparable before the outcome is measured?
Common Mistakes to Watch For
Before you work through the examples, skim the mistake guide so you know which shortcuts and sign errors to avoid.
Worked Examples
Example 1
mediumA survey claims 'people who eat breakfast earn more.' Identify a confounder and explain how it could fully explain the link.
Answer
First step
1
Steady job holders likely eat regular breakfasts.
See the full worked solution + why-it-works coaching
SetupKey insightWhy it worksCommon pitfallConnection
Unlock answer keys One Family plan — every worked solution, all subjects
Example 2
mediumA garden experiment puts all the fertilized plants in the sunny corner. Show why the corner is a confounder and how to fix the design.
Example 3
mediumA new website redesign launches simultaneously with a holiday sale. Sales rise. Explain why the holiday is a confounder for evaluating the redesign and one experimental design that avoids this.
Example 4
hardA research team finds patients given a new surgery die less often than patients given the old surgery. Discover that the new surgery is offered only at large urban hospitals. Identify the confounder and propose a study that would resolve it.
Example 5
hardIn a vitamin study, supplement-takers report fewer colds (no assignment). After matching on age, sex, and exercise, the effect shrinks but persists. Can we conclude vitamins cause fewer colds? Explain.
Example 6
hardA municipal data set shows that as the number of pools per neighborhood grows, so does the rate of skin cancer. Identify the confounder and explain why building fewer pools would not lower cancer.
Example 7
challengeA famous study claimed hormone replacement therapy lowered heart disease in women based on observational data, but the later randomized Women's Health Initiative trial found the opposite. Explain how confounding produced the misleading observational result.
Example 8
easyA study finds that children who have more books at home get higher test scores. A researcher concludes that buying more books will raise test scores. Identify the confounding variable.
Example 9
mediumHospitals A and B both perform a surgery. Hospital A has a 90% survival rate, Hospital B has 95%. However, Hospital A takes more high-risk patients. When only high-risk patients are compared, Hospital A has a higher survival rate. Explain this paradox.
Practice Problems
Try these problems on your own first, then open the solution to compare your method.
Example 1
easyIce cream sales and drowning deaths rise together. What hidden variable likely causes both?
Example 2
easyA confounding variable must be associated with what two things?
Example 3
easyTowns with more firefighters have more fire damage. What confounder explains this without firefighters causing damage?
Example 4
easyWhich study design best removes confounding variables: observational survey or randomized experiment?
Example 5
easyChildren with bigger feet read better. The confounder is most likely what?
Example 6
easyA confounder creates what kind of association between two variables that aren't directly causally linked?
Example 7
easyConfounding variables are described as a major threat to what kind of validity in observational studies?
Example 8
easyTo check if a variable is a confounder in 'coffee and heart disease,' which extra variable would you investigate first: shoe size or smoking?
Example 9
mediumA study finds people who take vitamins are healthier. Name a plausible confounder and explain how it could fully account for the association.
Example 10
mediumCities with more bookstores have higher average incomes. A claim says bookstores raise income. Identify the confounder and the correct causal direction it implies.
Example 11
mediumIn a fertilizer field study, the treated plots were also watered more. Explain how watering confounds the result and one design fix.
Example 12
mediumA study claims a job-training program raises wages, comparing graduates to non-applicants. What confounder threatens this, and how would random assignment fix it?
Example 13
mediumResearchers note coffee drinkers have more heart disease, but coffee drinkers also smoke more. After accounting for smoking, the coffee link vanishes. What does this reveal about smoking's role?
Example 14
mediumExplain why random assignment, not just measuring and adjusting for known confounders, is the stronger defense against confounding.
Example 15
mediumA report says students who eat breakfast get better grades, urging schools to serve breakfast. Identify a confounder and state what study would actually test the causal claim.
Example 16
challengeA study controls for income, age, and education yet still finds a treatment-outcome link. Can it now claim causation? Explain the residual confounding concern.
Example 17
challengeCountry-level data show higher chocolate consumption correlates with more Nobel laureates. Propose the most plausible confounder and explain why this is a textbook confounding (not causation) example.
Example 18
challengeA confounder C causes both X and Y. A naive analysis regresses Y on X and finds a strong effect. Explain in words why conditioning (stratifying) on C can make the X-Y association disappear, and what that implies about the original effect.
Example 19
mediumA study finds that students who use tutoring score higher. The tutored students also study more hours. Identify the confounder and how to control it.
Example 20
mediumResearchers observe that people who drink red wine have fewer heart attacks, but wine drinkers also tend to be wealthier with better healthcare. Identify the confounder and the safer conclusion.
Example 21
easySales of sunglasses and sunburns both rise in July. What is the most likely confounder?
Example 22
easyTowns with more churches also have more bars. Suggest a confounder.
Example 23
easyA study finds students with bigger backpacks score higher on math tests. Identify a likely confounder.
Example 24
easyTrue or false: a confounder must be measurable for it to bias an observational study.
Example 25
easyWhich is a confounder of the coffee-and-heart-disease link: shoe size or stress level?
Example 26
mediumIn a study, students who use a tutor have higher GPAs. Name two confounders that threaten the conclusion that tutoring raises GPA.
Example 27
mediumA claim says 'children who play music score higher on IQ tests.' Suggest a confounder and explain why a randomized study would be more convincing.
Example 28
mediumResearchers find runners have lower blood pressure than non-runners (no assignment). Name two confounders that could fully explain this.
Example 29
mediumIn a class study, students taught with a new textbook also had a more experienced teacher. Identify the confounder and a fix.
Example 30
mediumA study reports a 0.7 correlation between hours of TV watched and child obesity. Name two confounders and explain why correlation alone cannot establish causation.
Example 31
hardA health study claims 'people who eat fish twice a week live longer.' Suggest two confounders and one observational technique that partially adjusts for them.
Example 32
hardDefine 'Simpson's paradox' in one sentence and explain its connection to confounding.
Example 33
hardA workplace study finds people who attend yoga classes have lower stress. The HR analyst adjusts for income, age, and exercise. Why is the analyst still cautious about claiming yoga causes lower stress?
Example 34
mediumA study finds that people who own swimming pools are more likely to have skin cancer. Identify a possible confounding variable and explain how random assignment could address it.
Example 35
hardA company claims its energy drink improves athletic performance because athletes who drink it run faster. However, the athletes were not randomly assigned — those who drank the energy drink also trained harder. (a) Identify the confounding variable. (b) Explain what 'controlling for' a confounding variable means. (c) How could the study be redesigned?
Related Concepts
Background Knowledge
These ideas may be useful before you work through the harder examples.
correlation vs causationdata collection