Selection bias gets its own session in this installment of Yale's course on reading medical research critically, part of a series aimed at teaching non-specialists to spot flawed claims before sharing them. The lecturer walks through how studies can reach skewed conclusions when the people or data included are not representative of the population a claim is supposedly about, covering examples such as volunteer bias, survivorship bias, and self-selection in health studies. The talk explains why a study's sample matters as much as its statistics, and how researchers try to correct for or at least disclose these distortions. Running about fifteen minutes, it stays focused on one concept rather than surveying the whole field, making it a compact companion piece to the course's broader argument that most viral health claims fail on methodology long before they fail on evidence.