Evaluating Statistical Claims: Observational Studies and Experiments Flashcards

A news headline reads: 'Students who eat lunch perform better in afternoon classes.' The data came from a study where students chose whether to eat lunch. Which of the following is most likely a confounding variable?

Answer: Overall health and family habits that affect both eating patterns and academic performance

Students from families that emphasize regular meals may also emphasize studying and routines, making family habits a confounding variable associated with both lunch-eating and academic performance.

A pharmaceutical company randomly assigns 1,000 patients with high blood pressure to either a new medication or a placebo. After 6 months, the medication group shows a significant reduction in blood pressure. The company claims the medication is effective for all patients with high blood pressure. What is the most valid concern?

Answer: The study sample may not represent all patients with high blood pressure, such as those with other conditions.

Random assignment supports causation within the sample, but the sample's characteristics (e.g., age, comorbidities) may limit how broadly the results apply.

Researchers observe that countries with more hospitals per capita also have longer life expectancy. They conclude that building more hospitals increases life expectancy. What is the most likely confounding variable?

Answer: National wealth, which increases both hospital capacity and access to food, education, and clean water

Wealthier nations can afford more hospitals and also have better nutrition, education, sanitation, and healthcare access — all of which increase life expectancy independently.

A social scientist conducts an experiment by randomly assigning rural communities to receive a new community center and comparing crime rates to communities that did not receive one. After 2 years, crime rates are lower in communities with the center. What is the scope of the valid causal conclusion?

Answer: In rural communities similar to those studied, community centers likely cause lower crime rates.

Random assignment supports causation; the causal conclusion applies to rural communities similar to those in the study, not to all communities globally.

A study surveys 800 teenagers about sleep habits and academic performance. Students who sleep 9 or more hours have higher GPA. A parent concludes that forcing teens to sleep 9 hours will improve their grades. Why is this conclusion flawed?

Answer: The association is from an observational study; factors like family structure and health may cause both more sleep and higher GPA.

Observational studies cannot establish that sleep causes higher GPA; structured home environments may promote both longer sleep and better academic habits.

Two researchers study the effect of tutoring on test scores. Researcher A randomly assigns students to tutoring or no tutoring. Researcher B surveys students who chose to get tutoring vs. those who didn't. Which researcher can draw causal conclusions, and why?

Answer: Researcher A, because random assignment controls for confounding variables.

Only Researcher A used random assignment, which balances confounders between groups and allows a causal conclusion; Researcher B's study is observational and subject to selection bias.

A report states: 'Among 500 randomly selected adults, those who meditate have lower cortisol levels.' A journalist writes: 'Meditation reduces cortisol.' What is wrong with the journalist's conclusion?

Answer: The study used a random sample but did not randomly assign participants to meditate, so causation cannot be concluded.

The study is observational — people who chose to meditate may differ in other ways (stress management, lifestyle) from those who don't, making the causal claim unjustified.