Actually, the most common error in understanding scientific or statistical information is conceptual rather than arithmetic. It’s the confusion between correlation and causation. Here is a real-life example in which educators (who should know better) made a colossal blunder.
A few years ago, a school board discovered that there was a strong correlation between the size of students’ feet and their reading scores. Measurements of foot length were taken as predictors of future success until someone discovered that the data collector neglected to adjust foot size for age. Of course the students who were older had (on average) bigger feet and also higher reading scores. For children, both foot size and reading scores are correlated with age, however, the larger foot size does not “cause” the higher reading scores. This example shows that drawing correct inferences from data is no mean feat.
Following the discovery of the negative correlation between moderate wine consumption and the risk of mortality through heart disease, researchers began to search for a causal link. In a 2006 meta-analysis of 4235 studies on the topic, the authors, Paul Ronksley et al. stated:
Corrao et al. reported a J-shaped relation between alcohol intake and coronary heart disease, whereas the review by Maclure described this relation as L-shaped because he did not observe any increase in coronary heart disease risk associated with higher alcohol consumption. Our up-dated meta-analysis supports the latter association for coronary heart disease, with a 25–35% risk reduction for light to moderate drinking that also is present with heavier drinking. … Furthermore, the association of alcohol consumption is complex and differs by stroke subtype, with a slightly lower risk of ischaemic [cardiac] stroke but higher risk of haemorrhagic stroke.
In addressing causation, the authors of that meta-analysis stated:
The association between alcohol consumption and decreased cardiovascular risk is not in question, as additional research has not changed this conclusion. Clearly, observational studies cannot establish causation.
Whether moderate consumption of red wine has positive health effects continues to be hotly debated, distinguishing between correlation and causation continues to be a challenge to researchers. It’s frustrating enough to drive people to drink.
