What’s the Big Idea?
Will red wine really make you live longer? Does coffee strengthen brainpower or turn bones to dust? (Yes! No! Maybe.) For every scientific study proclaiming that something is good for you, there’s an equally plausible study that says it will kill you. The reason, according to Jason Fletcher of the Yale School of Public Health, is that flavor-of-the-week research is based on correlation, or an indicative relationship among data, rather than causality – a difference that is as statistically significant as the one between “probably” and “beyond a reasonable doubt.”
The gold standard for experiments is the randomized control trial (RTC), in which subjects are arbitrarily assigned to participate in either the test group or a control group. That standard is simple to uphold if you’re observing inanimate objects, but has presented a major challenge in the study of human beings. A researcher who wants to understand whether, for example, a toddler’s exposure to books affects his or her future chances of employment cannot ethically prevent a group of kids from learning to read.