Episode number 11 of Series “Against all odds: inside statistics”.
Studies the idea that causation is only one of many possible explanations for strong statistical correlations. Others may be common responses to a third factor, a confounding factor or coincidence. We are shown the importance of lurking variables, and Simpson’s paradox. As an example of the way in which evidence for causation is gathered when experiments cannot be done, the video presents an historical documentary on smoking and lung cancer. Presented by Theresa M. Amabile.