cum hoc ergo propter hoc (or, correlation is not causation. I'm using latin to make myself sound intellectual) is probably my favourite logical fallacy.
In its simplest form it can be stated in the following way:
This logical fallacy is probably underlined most explicitly with the following argument (courtesy of the church of the flying spaghetti monster). Consider the proposition that global warming, earthquakes, hurricanes, and other natural disasters are a direct effect of the shrinking numbers of pirates since the 1800s. Although at first glance the proposition is ridiculous look at what happens if you plot them against each other:
This should do the job of pretty comprehensively showing us that there can be no conclusion made regarding the existence or the direction of a cause and effect relationship only from the fact that A is correlated with B. In fact we have to worry about a whole slew of possibilities: A causes B; B causes A; C causes A and B; A causes B AND B causes A; the relationship is so complex that it may be labelled as coincidental.
It is this last category that the pirates/global warming connection falls under. This whole post arose from a conversation I had with Gem* tonight about the widely reported link between smoking marijuana and bringing out schizophrenia in those that are otherwise vulnerable. I think it is brilliant and subtle example of the difference between correlation and causation. With a bit of thought it is evident that there are in fact (at least) three explanations for an observed correlation between smoking marijuana and developing schizophrenia:
The correlations found in the study are not as easily interpreted as many people (and newspapers) originally told us. I certainly remember the "weed will make you go mental" message being repeated by various educators when I was young. I'm sure now that if I were to phone the national drugs helpline that I would hear the same advice (as a matter of fact I may look it up and ask them at some point).
It is one of the most gloriously messy things about doing science that there is rarely an obvious A means B connection. The world is a messy, chaotic, unpredictable and beautiful place. I wouldn't have it any other way, but I do wish that people would put a little bit of thought in before stating that the latest study demonstrating a correlation means that A caused B.
*for anybody that doesn't know, Gem is a mental health nurse
In its simplest form it can be stated in the following way:
- I observe a correlation between event A and event B
- A therefore causes B
This logical fallacy is probably underlined most explicitly with the following argument (courtesy of the church of the flying spaghetti monster). Consider the proposition that global warming, earthquakes, hurricanes, and other natural disasters are a direct effect of the shrinking numbers of pirates since the 1800s. Although at first glance the proposition is ridiculous look at what happens if you plot them against each other:
This should do the job of pretty comprehensively showing us that there can be no conclusion made regarding the existence or the direction of a cause and effect relationship only from the fact that A is correlated with B. In fact we have to worry about a whole slew of possibilities: A causes B; B causes A; C causes A and B; A causes B AND B causes A; the relationship is so complex that it may be labelled as coincidental.
It is this last category that the pirates/global warming connection falls under. This whole post arose from a conversation I had with Gem* tonight about the widely reported link between smoking marijuana and bringing out schizophrenia in those that are otherwise vulnerable. I think it is brilliant and subtle example of the difference between correlation and causation. With a bit of thought it is evident that there are in fact (at least) three explanations for an observed correlation between smoking marijuana and developing schizophrenia:
- Smoking marijuana causes schizophrenia (A causes B)
- People with a predisposition to schizophrenia are also those that are likely to enjoy smoking marijuana (B causes A)
- A hidden third factor, poverty, increases the chances of both smoking marijuana and being diagnosed with schizophrenia (C causes A and B)
The correlations found in the study are not as easily interpreted as many people (and newspapers) originally told us. I certainly remember the "weed will make you go mental" message being repeated by various educators when I was young. I'm sure now that if I were to phone the national drugs helpline that I would hear the same advice (as a matter of fact I may look it up and ask them at some point).
It is one of the most gloriously messy things about doing science that there is rarely an obvious A means B connection. The world is a messy, chaotic, unpredictable and beautiful place. I wouldn't have it any other way, but I do wish that people would put a little bit of thought in before stating that the latest study demonstrating a correlation means that A caused B.
*for anybody that doesn't know, Gem is a mental health nurse
