Confounding occurs when an extraneous variable correlates with both the dependent and independent variables, potentially leading to a false assumption about their relationship. It is crucial to identify and control for confounders to ensure the validity of causal inferences in research studies.