Transcript: Science starts by looking for patterns in data. Therefore it’s important to understand the distinction between causation and correlation. Scientists believe in causation, the general idea that events have causes. However science starts by looking for patterns in observational data. Typically two quantities may be plotted on a graph against each other. If there’s a correlation, science tries to look for a cause. However it’s not always possible to find a cause, or it’s not correct to infer a cause. For example, it took 30 years of research before the government was sufficiently convinced of the correlation and the causation of smoking and cancer rates to put health warnings on all packets of cigarettes. So we must be careful of the distinction between two quantities that are correlated and whether one causes the other. Sometimes there may be an underlying variable or third quantity that relates to the causation. In astronomy we plot the Hertzsprung-Russell diagram where the luminosity and the effective temperature of main sequence stars are tightly correlated. However the underlying variable in this case is mass, a quantity not plotted at all. So scientists must be very careful not to make the jump from causation to correlation without a justified physical theory that makes predictions that can be confirmed.