This Yale lecture comes from the course Understanding Medical Research: Your Facebook Friend is Wrong, a class built to help nonspecialists read health studies critically. The topic here is effect modification, the statistical phenomenon where a treatment or exposure's impact on an outcome differs depending on a third variable, such as age, sex, or an underlying condition. The instructor works through how researchers detect this pattern in data, why it differs from confounding, and why ignoring it can produce misleading headlines about a drug or intervention that actually helps one subgroup while doing little or nothing for another. Concrete examples ground the statistics, showing how a single overall result can hide very different effects in different populations. At fourteen minutes, it functions as a focused unit within the larger course rather than a freestanding survey, aimed at viewers who want to understand why the same study can be reported as both a breakthrough and a disappointment.