Watson-Glaser Critical Thinking Appraisal Interpreting Evidence and Data 5 — Questions and Answers
Question 1: A hospital advertises a 95% survival rate for its cardiac surgery patients. A competing hospital advertises 90%. Which piece of information is most critical before concluding the first hospital is better?
- The hospital's founding year
- Whether the case mix (severity of patients) differs between the hospitals (Correct answer)
- The hospitals' locations
- Which hospital charges higher fees
Correct answer: Whether the case mix (severity of patients) differs between the hospitals
A hospital that treats more high-risk patients may have a lower survival rate even if it performs equally well or better.
Question 2: A longitudinal study tracks 500 smokers and 500 non-smokers for 20 years and finds smokers have significantly higher rates of lung disease. This study design best supports which type of conclusion?
- Smoking cannot affect lung health
- A strong causal link is plausible, though experimental proof requires a randomized trial (Correct answer)
- The finding is invalid because it took 20 years
- Only 500 subjects means the data is unreliable
Correct answer: A strong causal link is plausible, though experimental proof requires a randomized trial
Longitudinal observational studies can strongly suggest causation but cannot definitively prove it the way a randomized controlled trial can.
Question 3: Sales data shows that umbrella purchases spike every April. An analyst claims umbrellas are suddenly more popular in spring. What alternative explanation should be considered?
- People forget they own umbrellas
- Seasonal rainfall in April drives umbrella purchases rather than a change in popularity (Correct answer)
- April has more shopping days than other months
- The data only covers one year and must be wrong
Correct answer: Seasonal rainfall in April drives umbrella purchases rather than a change in popularity
Seasonal weather patterns, not a shift in consumer preferences, most plausibly explain the recurring April spike.
Question 4: A report presents the median household income as $65,000 and the mean as $95,000 in the same dataset. What does this discrepancy most likely indicate?
- One of the figures must be an error
- A small number of very high earners are pulling the mean above the median (Correct answer)
- Median is always higher than mean for income data
- The dataset contains no low-income households
Correct answer: A small number of very high earners are pulling the mean above the median
When the mean substantially exceeds the median, it typically signals a right-skewed distribution caused by high-value outliers.
Question 5: An experiment finds that a new fertilizer increases crop yield by 15% in greenhouse conditions. A farmer plans to apply it immediately at scale. What interpretive caution is most important?
- Greenhouse experiments are always fraudulent
- Greenhouse results may not generalize to outdoor field conditions with different soil, weather, and pests (Correct answer)
- A 15% gain is too small to be useful
- The farmer should wait 10 years for confirmation
Correct answer: Greenhouse results may not generalize to outdoor field conditions with different soil, weather, and pests
Controlled laboratory or greenhouse conditions often differ substantially from real-world environments, limiting the direct applicability of findings.
Question 6: A dataset shows that countries with higher chocolate consumption win more Nobel Prizes per capita. The best interpretation is:
- Chocolate boosts cognitive performance needed for Nobel-level work
- Both variables correlate with a third factor, such as national wealth or research investment (Correct answer)
- Nobel Prize committees favor chocolate-eating nations
- The data proves chocolate should be subsidized for scientists
Correct answer: Both variables correlate with a third factor, such as national wealth or research investment
Wealthier nations tend to both consume more chocolate and invest more in scientific research, making wealth the likely confounding variable.
Question 7: A graph shows that every year a researcher has published a paper, the stock market has risen. The researcher claims their publications predict market gains. This reasoning is flawed because:
- Academic papers should not discuss finance
- Coincidental co-occurrence of two independent events does not establish prediction or causation (Correct answer)
- The stock market cannot be graphed over time
- The researcher should publish more papers to raise the market
Correct answer: Coincidental co-occurrence of two independent events does not establish prediction or causation
Two variables moving together over a small number of cases can easily be coincidental, especially without a plausible mechanism.
Watson-Glaser Critical Thinking Appraisal Interpreting Evidence and Data 5
A hospital advertises a 95% survival rate for its cardiac surgery patients.
A competing hospital advertises 90%.
Which piece of information is most critical before concluding the first hospital is better?