Difference between correlation & causation

I have written on this blog before about "craftaphobia" - the fear of Freemasonry. A way of creating this anti-masonic sentiment is by mixing correlation and causation. This neatly describes the work of Stephen Knight and Martin Short, for example.

We are socialised into thinking about cause-and-effect through our upbringing. By using limited case studies such writers focus their readers in a limited area to try and lead their readers into thinking there is corruption between Masons. First of all the sample size is not representational. If there are 300,000 members and let's say a dozen instances involving 3 Masons that represent 0.0000012% of the population. (Arguably still too high, but that is a different issue).

Issues of correlation and causation are different. If red cars were involved in more accidents last year it does not necessarily mean that if you drive a red car you are more likely to have an accident. Other factors such as make, model, age, gender, weather are as likely to contribute. The same argument is true of "Masonic conspiracies" Just because 0.0000012% of masons were alleged to have conspired it does not necessarily follow that they did so because they were Masons. Likewise, it does not prove that other Masons are likely to do so.

Such is the sophistry of the anti-masonic literature.