Watson-Glaser Critical Thinking Appraisal Interpreting Evidence and Data Questions and Answers 2 — Questions and Answers
Question 1: A study reports: 'Students who slept 8+ hours scored 15% higher on tests than those who slept fewer than 6 hours.' A teacher concludes: 'Students in my class who sleep more will score higher on my next test.' What is the best interpretation of this evidence?
- The conclusion is fully supported — the study proves sleep improves test scores.
- The study suggests a correlation between sleep and performance; the teacher's conclusion is plausible but requires caution about individual variation and confounders. (Correct answer)
- The conclusion is invalid because the study was not conducted in the teacher's class.
- Sleep has no effect on test performance — the correlation is coincidental.
Correct answer: The study suggests a correlation between sleep and performance; the teacher's conclusion is plausible but requires caution about individual variation and confounders.
Research establishes a likely pattern, but applying population-level findings to a specific class requires recognizing individual variation, the study's conditions, and potential confounders.
Interpreting evidence correctly on the WGCTA requires understanding what a study establishes and what it doesn't. A 15% performance gap between sleep groups suggests a real association, but population-level correlations don't guarantee individual outcomes. Confounders could include: high-achieving students tend to have better time management (enabling both more sleep and better study). Applying research to a new context requires considering whether the study population matches the class, whether confounders apply, and whether individual variation limits predictability. Option A overextends the finding to certainty. Option C incorrectly dismisses valid generalization from research. Option D denies well-established evidence on sleep and cognition.
Question 2: A bar chart shows that city A has the highest number of reported thefts among five cities. Which interpretation is most accurate?
- City A is the most dangerous city of the five.
- City A has the highest theft reporting rate, which may reflect high crime, high reporting behavior, or both. (Correct answer)
- City A's police force is the least effective among the five cities.
- Residents of city A are more dishonest than residents of other cities.
Correct answer: City A has the highest theft reporting rate, which may reflect high crime, high reporting behavior, or both.
Reported theft numbers reflect both actual crime and reporting rates — high reports could indicate more crime, better policing that encourages reporting, or higher civic engagement in reporting.
Official crime statistics measure reported crimes, not all crimes. City A's high reported theft count could result from: genuinely higher crime rates; higher rates of victim reporting (due to better police-community relations, insurance requirements, or civic culture); more active policing that detects more crimes. 'Most dangerous' (Option A) implies all crime types and actual rates, not reported theft counts. Option C infers police effectiveness from crime counts — the opposite inference is equally plausible (effective police record more). Option D makes an unfounded character judgment about residents. Option B correctly captures the dual nature of crime statistics.
Question 3: Data shows that country X has an average life expectancy of 82 years, while country Y has 69 years. Which conclusion is best supported by this data alone?
- Country X has a better healthcare system than country Y.
- People in country X live longer on average than those in country Y. (Correct answer)
- Country Y has a lower quality of life than country X.
- Country X has lower poverty rates than country Y.
Correct answer: People in country X live longer on average than those in country Y.
Life expectancy data directly supports the statement about relative average lifespans — the other conclusions require additional data points beyond life expectancy alone.
Life expectancy is a measure of average lifespan — Option B directly restates what the data shows. Options A, C, and D all make inferences about underlying causes or related conditions. Healthcare quality is one factor among many (diet, lifestyle, environmental factors, genetics, inequality) that determines life expectancy. Quality of life involves subjective wellbeing, income, social connections, and security — not reducible to lifespan alone. Poverty rates are separate from life expectancy. On the WGCTA, the correct interpretation of evidence confines conclusions to what the data directly measures, not what might explain the data.
Question 4: A company's employee satisfaction survey shows that 73% of respondents rate their job satisfaction as 'high' or 'very high.' HR concludes that the company has a strong workplace culture. What caveat is most important?
- 73% is not a high enough percentage to conclude a strong culture.
- Survey response rates and self-selection bias may mean the 73% overstates actual satisfaction. (Correct answer)
- Employee satisfaction has no relationship to workplace culture.
- The survey should have used a scale of 1-10 instead of categorical responses.
Correct answer: Survey response rates and self-selection bias may mean the 73% overstates actual satisfaction.
If dissatisfied employees were less likely to complete the survey, the 73% figure may reflect the responses of more satisfied employees, overestimating the overall satisfaction level.
Self-selection bias in voluntary surveys is a major data quality concern. Employees who are highly satisfied or highly dissatisfied have stronger motivations to respond; those with moderate experiences may not complete the survey. More critically, dissatisfied employees may fear negative consequences or simply disengage from voluntary HR surveys. This can inflate satisfaction scores. The response rate — if 73% came from only 30% of employees — would be far less representative than if it came from 80%. Option A arbitrarily challenges the percentage without reasoning. Option C incorrectly denies a well-established connection. Option D focuses on survey design rather than the interpretation of the result.
Question 5: A health report states that a country's obesity rate increased from 18% to 24% over a decade. Which interpretation is best supported?
- The government's health policies completely failed to address obesity.
- More people are obese now than a decade ago, representing a 6-percentage-point increase. (Correct answer)
- The obesity increase was caused by increased fast food consumption.
- 24% obesity rate means the majority of the population is obese.
Correct answer: More people are obese now than a decade ago, representing a 6-percentage-point increase.
The data directly shows a 6-percentage-point increase in obesity prevalence — the other options introduce causes, policy judgments, or mathematical errors not supported by the data.
Option B correctly reads the data: 24% - 18% = 6 percentage points of increase, which means more people were classified as obese at the later time point. Option A introduces a policy evaluation ('failed') requiring context about what policies existed and whether any of the 6-point increase was mitigated. Option C provides a causal explanation (fast food) not present in the data. Option D misreads the percentage: 24% means fewer than one in four people are obese, not a majority. On the WGCTA, accurate data interpretation requires mathematical precision and confining conclusions to what the numbers directly state.
Question 6: A table shows that Region A has a teacher-to-student ratio of 1:15 while Region B has 1:28. Which conclusion is best supported?
- Students in Region A receive higher quality education than those in Region B.
- Region A has more teachers per student than Region B. (Correct answer)
- Region B's schools are underfunded compared to Region A.
- Students in Region B will perform worse academically than those in Region A.
Correct answer: Region A has more teachers per student than Region B.
The ratio directly indicates how many students each teacher serves — Option B restates this factual comparison without inferring outcomes or causes.
The teacher-to-student ratio is a resource-allocation metric. Option B simply restates what the ratio directly measures — Region A has one teacher for every 15 students, Region B one for every 28. Options A and D infer quality and academic performance, which require additional evidence (class structure, curriculum quality, teacher training, student demographics). Option C infers funding levels — lower ratios could result from smaller class sizes by design rather than inadequate funding. On WGCTA, the factual interpretation stays within the data boundaries: the ratio measures a staffing ratio, not education quality, academic outcomes, or funding levels.
