In this program, students learn that one of the most common errors found in the news media regarding scientific and health-related studies is the confusion between correlation and causation.
In some cases it seems obvious that one action will cause another, whereas at other times it is not as clear. Many studies are designed to test a correlation using statistical tools to measure two variables over time. As Daniel Gilbert points out, a perfect correlation occurs when both variables change by a fixed amount. Correlations, tracked by using the correlation coefficient (represented by lowercase r), can be positive or negative and vary in strength. A strong positive correlation occurs when both variables increase equally, whereas in a negative correlation, one unit increases while the other decreases.
Establishing causality can be more challenging, so researchers conduct controlled studies using two comparable and randomly-chosen groups. Each group performs a task under different circumstances than that of the other group. The outcomes are then compared to determine whether the different outcomes are caused by having different experiences. The narrator then guides the viewer through an experiment on exercise and memory performance, demonstrating how sometimes the presence of a third variable can cause differences to occur.
Some variables are impossible and unethical to control for, such as asking a subject to smoke cigarettes. Therefore, psychologists and other scientists must adhere to a strict code of ethics when conducting research.