This is part five of a 12 week series where, on Mondays, I plan to explore the tactics of verbal cheap shot artists – people who can’t, or won’t use valid arguments to present their case, but instead resort to verbal cheap shots. To make sure you don’t miss a single article, be sure and sign up for my RSS feed or subscribe by email (both are free!) so you can get a new article each day! Check out my other articles in the Verbal Cheap Shots category.
Correlation and causation are commonly held fallacies. That correlation implies causation is a logical fallacy where two events occurring together are claimed to be cause and effect. The formal Latin phrase is cum hoc ergo propter hoc (“with this, therefore because of this”) and can be structured like this:
A is related to B
B is also related to C
Therefore, A causes C
My financial adviser sells only loaded funds.
Loaded mutual funds produce superior returns.
My financial adviser will produce superior returns.
Police officers eat lots of fast food.
Police officers have a high divorce rate.
Therefore, fast food cause divorce.
This arguments are examples of a false categorical syllogism. It ignores the possibility that the correlation is coincidence. If fast food and divorce were strongly correlated across cultures and large populations, and double blind studies could provide proof, it probably isn’t a coincidence. In that case, the fallacy may ignore the possibility that there is a common cause of fast food consumption and divorce.
Another important factor we should consider is the presence or absence of a known mechanism which may explain how one event causes the other. Could the consumption of too much fast food indicate a person’s disregard for their health which could stem from a poor self image? Could a poor self image be a contributor to divorce?
Ice-cream sales are strongly correlated with crime rates.
Therefore, ice-cream causes crime.
Could the explanation be that high temperatures increase crime rates (presumably by making people irritable) as well as ice-cream sales? Clearly, we can’t conclude that ice-cream causes criminal tendencies (or that criminals prefer ice-cream to other treats). Our natural tendency is to expect the correlation to point us towards a causal structure and though correlations may imply some sort of causal story, the real cause could be something more complicated or entirely different. Perhaps committing a crime make you crave ice cream? Nah. That doesn’t make sense either.
Sometimes the direction of the causation is wrong and should be reversed.
Gun ownership is correlated with crime.
Therefore, gun ownership leads to crime.
The facts could easily be the other way round: an increase in crime could lead to more gun ownership by concerned citizens.
Just because items happen together, or even occur in a linear or chronological pattern, doesn’t mean they’re causal.
Don’t fall for the correlation implies causation fallacy!