Watson-Glaser Critical Thinking Appraisal Evaluating Causal Relationships 2 — Questions and Answers
Question 1: A newspaper headline reads: 'Coffee Drinkers Live Longer.' The supporting study controlled for smoking and exercise. What is still needed to establish causation?
- A randomized controlled trial to rule out remaining confounders. (Correct answer)
- A larger headline sample size.
- Longitudinal data on coffee brand preferences.
- Government regulation of coffee consumption studies.
Correct answer: A randomized controlled trial to rule out remaining confounders.
Observational studies, even when controlling for known confounders, cannot fully establish causation the way a randomized experiment can.
Question 2: Two variables, X and Y, have a correlation of 0.90. A critical thinker should conclude:
- X and Y are strongly related, but causation requires further evidence. (Correct answer)
- X causes Y with 90% certainty.
- Y causes X with 90% certainty.
- X and Y are identical measures.
Correct answer: X and Y are strongly related, but causation requires further evidence.
A high correlation indicates a strong association but does not identify which variable causes the other, if either does.
Question 3: A pharmaceutical company reports that patients who adhered to their drug regimen recovered faster. A critic points out this is likely due to the 'healthy adherer effect.' What does this mean?
- Patients who take medications as directed also follow other healthy behaviors that aid recovery. (Correct answer)
- Adherent patients received higher doses of the drug.
- The drug only works in patients who are already healthy.
- Non-adherent patients were excluded from the study.
Correct answer: Patients who take medications as directed also follow other healthy behaviors that aid recovery.
Adherent patients tend to lead healthier lifestyles overall, confounding the drug's apparent effect.
Question 4: During an economic recession, both unemployment and crime rates rise. A politician argues unemployment causes crime. What is the most precise critical thinking response?
- Both may be caused by the recession itself, and the relationship between them may not be directly causal. (Correct answer)
- Crime statistics are unreliable during recessions.
- Unemployment only causes property crime, not violent crime.
- The correlation holds only in urban areas.
Correct answer: Both may be caused by the recession itself, and the relationship between them may not be directly causal.
Economic downturns can simultaneously drive both metrics, making a direct unemployment-to-crime causal chain an oversimplification.
Question 5: A school implements a reading program. Test scores improve. To best determine if the program caused the improvement, researchers should:
- Compare results to a control group that did not receive the program. (Correct answer)
- Measure reading speed before and after only for participating students.
- Survey teachers about their impressions of student progress.
- Analyze test scores from the previous five years.
Correct answer: Compare results to a control group that did not receive the program.
Without a control group, it is impossible to rule out natural score improvements over time or other simultaneous changes.
Question 6: Which of the following best represents a necessary condition for establishing a causal relationship between A and B?
- A must precede B in time, they must be correlated, and alternative explanations must be ruled out. (Correct answer)
- A and B must occur simultaneously in every observed case.
- A must be the only factor present when B occurs.
- B must never occur without A occurring first.
Correct answer: A must precede B in time, they must be correlated, and alternative explanations must be ruled out.
The standard criteria for causation include temporal precedence, covariation, and elimination of competing explanations.
Watson-Glaser Critical Thinking Appraisal Evaluating Causal Relationships 2
A newspaper headline reads: 'Coffee Drinkers Live Longer.' The supporting study controlled for smoking and exercise.
What is still needed to establish causation?