Skip to content

Quiz: Causation and Study Design

Test your understanding of causation, confounding variables, study design, and Simpson's Paradox with these review questions.

1. What is the key difference between correlation and causation?

  1. Correlation involves numbers while causation involves words
  2. Correlation means variables move together; causation means one directly produces changes in the other
  3. Correlation is always positive; causation can be negative
  4. Correlation requires experiments; causation requires surveys
Show Answer

The correct answer is B. Correlation describes when two variables tend to move together (either in the same or opposite directions). Causation is a much stronger claim that changes in one variable directly produce changes in another. Establishing causation requires ruling out alternative explanations, which correlation alone cannot do.

Concept Tested: Correlation vs. Causation

2. A study finds that students who eat breakfast have higher grades. Parental involvement (ensuring breakfast AND helping with homework) might explain both. What is parental involvement in this scenario?

  1. The response variable
  2. The explanatory variable
  3. A confounding variable
  4. A dependent variable
Show Answer

The correct answer is C. A confounding variable is associated with both the explanatory variable (breakfast eating) and the response variable (grades). Parental involvement could cause both more breakfast eating and better grades, creating a misleading association between breakfast and grades. This makes it impossible to determine if breakfast itself affects performance.

Concept Tested: Confounding Variable

3. A researcher surveyed 500 adults about their coffee consumption and measured their blood pressure. This is an example of what type of study?

  1. A randomized experiment
  2. An observational study
  3. A clinical trial
  4. A controlled experiment
Show Answer

The correct answer is B. An observational study measures variables without imposing any treatment. The researcher simply observed existing coffee habits rather than randomly assigning people to drink or not drink coffee. Because there's no random assignment, confounding variables could explain any association found between coffee and blood pressure.

Concept Tested: Observational Study

4. What makes an experiment capable of establishing causation while an observational study cannot?

  1. Experiments use larger sample sizes
  2. Experiments randomly assign subjects to treatments, balancing confounding variables
  3. Experiments always have control groups
  4. Experiments measure more variables
Show Answer

The correct answer is B. Random assignment is the key feature that allows experiments to establish causation. By randomly assigning subjects to treatment or control groups, all confounding variables (known and unknown) are distributed approximately equally across groups. Any difference in outcomes can then be attributed to the treatment rather than pre-existing differences.

Concept Tested: Experiment and Random Assignment

5. Hospital A has an overall survival rate of 90%, while Hospital B has a rate of 80%. However, Hospital B has higher survival rates for both mild and severe cases when examined separately. What phenomenon is this?

  1. Regression to the mean
  2. Selection bias
  3. Simpson's Paradox
  4. The placebo effect
Show Answer

The correct answer is C. Simpson's Paradox occurs when a trend present in different groups reverses or disappears when the groups are combined. This happens because the hospitals treat different proportions of mild versus severe cases. Hospital A may look better overall simply because it takes easier cases, even though Hospital B performs better in each category.

Concept Tested: Simpson's Paradox

6. A variable that affects the relationship being studied but is not included in the analysis is called what?

  1. A dependent variable
  2. An independent variable
  3. A lurking variable
  4. A control variable
Show Answer

The correct answer is C. A lurking variable is not measured or included in the analysis but affects the relationship between the variables being studied. Lurking variables can create misleading associations by influencing both the explanatory and response variables from behind the scenes, leading researchers to incorrect conclusions.

Concept Tested: Lurking Variable

7. Researchers want to test whether a new medication reduces cholesterol. They randomly assign 100 patients to receive the medication and 100 to receive a placebo. What type of study is this?

  1. An observational study
  2. A case-control study
  3. A survey
  4. An experiment
Show Answer

The correct answer is D. This is an experiment because researchers deliberately impose a treatment (medication vs. placebo) and randomly assign subjects to groups. The random assignment helps ensure that groups are equivalent in all ways except the treatment, allowing causal conclusions if a difference is observed.

Concept Tested: Experiment

8. To determine if a variable is confounding, it must be associated with which of the following?

  1. Only the explanatory variable
  2. Only the response variable
  3. Both the explanatory and response variables
  4. Neither variable, but still affect the outcome
Show Answer

The correct answer is C. For a variable to be confounding, it must be associated with both the explanatory variable AND the response variable. This dual association creates a tangled web where the effects of the confounding variable and the explanatory variable cannot be separated, making causal claims impossible without controlling for the confounder.

Concept Tested: Identifying Confounding

9. A news article claims "Study proves that drinking green tea prevents cancer." The study surveyed 10,000 people about their tea consumption and tracked cancer diagnoses over 10 years. What is wrong with this headline?

  1. The sample size is too small
  2. Ten years is not long enough to study cancer
  3. An observational study cannot prove causation
  4. Green tea has never been studied before
Show Answer

The correct answer is C. This is an observational study since researchers observed existing tea-drinking habits rather than randomly assigning people to drink tea. Observational studies can only show association, not causation. People who drink green tea might also exercise more, eat healthier, or differ in other ways that actually prevent cancer. The word "proves" is inappropriate.

Concept Tested: Comparing Studies (Observational vs. Experiment)

10. In a randomized experiment, what is the purpose of the control group?

  1. To provide a baseline comparison for the treatment effect
  2. To ensure the experiment is ethical
  3. To increase the sample size
  4. To eliminate all confounding variables
Show Answer

The correct answer is A. The control group provides a baseline for comparison. By measuring outcomes in subjects who don't receive the treatment (or receive a placebo), researchers can determine whether any changes in the treatment group are actually due to the treatment rather than natural changes, placebo effects, or other factors affecting all subjects.

Concept Tested: Experiment Design