Description:
This is an error in causal reasoning that occurs when it is assumed that a correlation between two things must be a causal connection. Translated, the fallacy is called “with that, therefore because of that.” This fallacy has the following form:
Premise 1: There is a correlation between A and B (or As and Bs).
Conclusion: Therefore, A causes B (or As cause Bs).
This fallacy is like another classic causal fallacy, the Post Hoc Ergo Propter Hoc fallacy. The difference is that the Post Hoc fallacy occurs when it is inferred that A causes B merely because A occurs before B. In the Cum Hoc fallacy, the error involves assuming that correlation must entail causation.
Just because two things are correlated is not enough to justify inferring that there is a causal connection. In some cases, this is obvious. For example, no one would infer that cold weather is caused by people wearing jackets.
Not surprisingly, the fallacy is most likely to occur when it appears there might be a causal connection. For example, a person might find that there is a correlation between sleeping fully dressed and waking up with a headache and conclude that sleeping this way causes headaches. It would not be unreasonable to consider that clothes might have this effect, but mere correlation would not suffice to prove there is a connection.
The fallacy can even be committed when a causal connection holds between the two things. While it might seem odd, the key to the fallacy is not that there is, in fact, no causal connection between A and B. It is that adequate evidence has not been provided for the claim that A causes B. This is another example of the distinction between factual errors and bad reasoning.
This fallacy is often committed unintentionally due to a lack of caution in causal reasoning. Leaping to a causal conclusion is easier and faster than investigating the phenomenon. However, such leaps often land far from the truth of the matter.
The fallacy can also be intentionally committed. In these cases, the person inflicting the fallacy believes that there is no causal connection but uses correlation to try to persuade others that there is. This technique can be misused for a wide variety of nefarious purposes, ranging from generating clickbait “science” headlines to deceiving people about the efficacy of a medical product or procedure.
This fallacy is often committed or accepted when a person wants the correlation to be causation. Such cases can be considered a combination of Wishful Thinking and this fallacy. For example, someone who wants to believe that eating chocolate causes weight loss might be inclined to accept an attempt to use (possibly manufactured) correlation to “prove” causation.
If you are interested in strange correlations, Tyler Vigen maintains a collection of spurious correlations at tylervigen.com. As Vigen shows, correlation exists between such things as the divorce rate in my home state of Maine and the per capita consumption of margarine. Such non-causal correlations should be expected. If enough data is analyzed, numerous correlations between unrelated things will be found.
Defense: Because this fallacy is committed by drawing an unjustified causal conclusion, the key to avoiding it is careful investigation. While causes and effects do correlate, correlation is not causation. While a causal investigation will often begin with an investigation of correlation, it should not end there.
To avoid having this fallacy inflicted on you, the defense is to consider whether the claim is supported by anything beyond correlation. While statistical analysis goes way beyond the scope of this book, you should consider that there are numerous deceitful techniques to make it appear that a causal connection exists. But you also need to be careful about unwarranted skepticism about causal claims. People also reject well-supported causal claims because they do not want them to be true, which is often a case of Wishful Thinking.
Example#1
“You know what I’ve noticed? There is a correlation between when the President speaks on the economy and the Dow Jones. While it does not happen every single time, usually when he speaks the Jones dips. And the more he talks, the deeper the dip. If he wants to help the economy, he needs to stop talking about it. His speeches are bringing it down!”
Example #2
Sam: “After four years of college I’ve learned something important.”
Jane: “And what might that be, Socrates?”
Sam: “Sleeping in your clothes gives me a headache.”
Jane: “You’ve been sleeping in my clothes?”
Sam: “No, I mean the general thing. Well, I mean when I sleep in my clothes, I get a headache. I’m not sure why but sleeping with clothes on hurts my head. So that is why I started sleeping naked.”
Jane: “What does your roommate think of that?”
Sam: “He’s not happy. He calls me ‘junk man.’”
Jane: “So, do you no longer get headaches?”
Sam: “That is the odd part. I still do. But I’m sure the clothes cause headaches. Maybe I’m sleeping too close to them?”
Jane: “Yeah, I’m sure that is it.”
Example #3
Ashleigh: “I’ve decided I’m not eating ice cream before I go swimming.”
Nancy: “You know that isn’t true. The myth about eating before swimming, I mean.”
Ashleigh: “Oh, I know. But I heard the professor say in class that drowning deaths increase in proportion to the sale of ice cream. I’m not sure what he was talking about, but I’m sure that eating ice cream before swimming would be risky.”
Leave a Reply