Which of the following best distinguishes a lurking variable from a confounding variable in observational studies?
- AA confounding variable is tied to both explanatory and response and was measured; a lurking variable was not.Correct
- BA lurking variable is measurable, while a confounding variable is unmeasurable.
- CA lurking variable causes the response directly, while a confounding variable only indirectly affects the response.
- DConfounding variables are only relevant in experiments, while lurking variables are only relevant in observational studies.
Why A is right
While definitions vary slightly by source, in AP Statistics: a lurking variable is one that influences the observed relationship between the explanatory and response variables but was not included in the study (not measured). A confounding variable is similar but is sometimes defined as one that is associated with the explanatory variable and causally related to the response. The key distinction often used is that a confounding variable is accounted for (measured) in the study, while a lurking variable is not.
Why the other 3 choices miss
- B
- Incorrect because lurking variables may be measurable in principle — they simply weren't measured in the study. The distinction is not about measurability but about whether the variable was included in the analysis.
- C
- Incorrect because both lurking and confounding variables can have direct or indirect effects on the response. The distinction is about whether they were included in the study, not about the causal pathway.
- D
- Incorrect because both lurking and confounding variables are relevant concerns in observational studies. Confounding can occur in experiments too if the experiment is not well-controlled, though randomization mitigates it.
A regression of exam score on hours of sleep produces the equation y-hat = 40 + 5x with a 95% confidence interval for the slope of (2.8, 7.2). A student claims that this proves sleeping more hours causes higher exam scores. This claim is not justified because
- Athe confidence interval is too wide to draw any conclusions about the true population slope
- Bobservational data show association, not causation; lurking variables such as study habits may explain bothCorrect
- Cthe confidence interval does not contain zero, so the data establish that more sleep causes higher exam scores
- Dthe y-intercept of 40 is too high for the model to be valid
Why B is right
Regression analysis of observational data can establish a statistically significant association but cannot prove causation. Lurking variables (study habits, overall health, motivation, stress levels) may be associated with both sleep duration and exam performance. Without a randomized experiment controlling for confounders, the causal claim is not supported.
Why the other 3 choices miss
- A
- Incorrect. The confidence interval (2.8, 7.2) does not contain zero, indicating a statistically significant positive relationship. Width alone does not invalidate conclusions about the population slope.
- C
- Incorrect. A significant positive slope demonstrates association, not causation. Statistical significance answers 'Is there a relationship?' not 'Does one variable cause the other?' It cannot establish that more sleep causes higher scores.
- D
- Incorrect. The y-intercept of 40 represents the predicted exam score for zero hours of sleep, which is an extrapolation. This does not invalidate the model within the data range or affect the causal argument.
A student computes the correlation between shoe size and IQ score for a sample of 200 adults. The student finds r = 0.72 and concludes that having larger feet causes people to have higher IQs. Which of the following identifies the flaw in this reasoning?
- ACorrelation does not imply causation; a lurking variable such as age or overall body size may explain both.Correct
- BA correlation of 0.72 is too weak to draw any conclusions about the relationship.
- CThe correlation should have been computed using Spearman's rank method for body measurements.
- DThe sample size of 200 is too small to compute a valid correlation.
Why A is right
Even a strong correlation does not establish causation. In this case, a lurking variable such as age (children have smaller feet and lower measured IQ) or overall body size could explain both variables simultaneously. The correlation captures a statistical association, not a causal mechanism.
Why the other 3 choices miss
- B
- Incorrect because r = 0.72 is actually considered a moderately strong correlation. The problem is not the weakness of the correlation but the invalid causal inference.
- C
- Incorrect because Pearson correlation is appropriate for quantitative data. There is no requirement to use Spearman's rank correlation for physical measurements.
- D
- Incorrect because n = 200 is actually a reasonably large sample for computing correlation. Sample size affects the precision and significance of r, but 200 is sufficient.