You often hear people say “correlation does not imply causation” as a way of cautioning you from making unwarranted conclusions from scientific data. However, science does at least sometimes give us knowledge of causal relationships in the world. For instance, we know that smoking causes cancer, that increased levels of greenhouse gases in the atmosphere causes the global temperature to rise, and that interrupting the electrical circuit connecting my switch and my lightbulb causes the light to go out. But all an experiment or observational study provides us with is a data set that may include correlations between the values of the independent variable and dependent variable. So, if experiments and observational studies give us correlations between variables, and correlation does not imply causation, how can we learn about causes through science? This video tackles this complicated question. The video is aligned with Chapter 11 of the book Recipes for Science: An Introduction to Scientific Methods and Reasoning, by Angela Potochnik, Matteo Colombo, and Cory Wright.