Representative Heuristic

A heuristic is a shortcut rule, or guide, by which one tries to organize one’s understanding of the world. The representativeness heuristic judges things by their resemblance to a category or prototype, invoking “the principle that members of a category should resemble a prototype” (Schick and Vaughn 2014, 33). This shortcut can sometimes support a useful, timely decision. If you see a violently barking dog foaming at the mouth, it would be rational to avoid it. You do not have to establish that it is rabid to recognize a reason for caution. But such reasoning can often lead us astray.

The harder‐to‐understand version of this mistake involves unjustifiably grouping similar things together – thinking that like goes with like. The easiest way to understand this mistake is to consider how it contributes to the conjunction fallacy (see Chapter 74). As a quick reminder of this fallacy, suppose a guy named Chris lives in Alabama, owns lots of guns, and loves to go hunting. Given that, which of these statements is more likely?

1. (A) Chris voted for Obama in 2012; or
2. (B) Chris voted for Obama in 2012 and is a member of the National Rifle Association (NRA).

Choice (B) may feel more representative of Chris, but (A) must be at least as likely. Now choice (A) is the right answer because the probability of two events being true together can never be greater than the probability of either one of them being true by itself. But what’s more relevant for our purposes is this question: Why does (B) seem to be the right answer to most people?

It’s because we tend to group things that are alike together. A prototype NRA member is imagined to be a southern gun‐owning hunter. Since (B) mentions the NRA, we tend to think that (B) “represents” Chris. And since that is not what we imagine when we think of an Obama voter, we think (A) does not represent Chris. This tendency to group similar things (in this case Chris and option (B)) together as a rule blinds us to the mathematical facts and leads us astray.

The other way the representative heuristic leads us astray is by making us apt to think that causes and their effects must resemble each other – that like causes like. This is most striking in the realm of alternative and “natural” medicine, where certain physical qualities of objects are thought to be transferable to those who ingest them. According to Schick and Vaughn (2014, 133), some Chinese ingest ground‐up bats as a treatment for vision problems because it is (incorrectly) assumed that bats have good vision, some Europeans treat asthma with fox lungs, and some Americans treat mental disorders with raw brains. Some even erroneously believe that eating genetically modified foods can genetically modify humans. In Vietnam, people grind up rhino horn as a male aphrodisiac. The logic literally is “since rhino horns are long and hard, ingesting them will make me long and hard.” Like causes like. Unfortunately, this fallacious logic is helping to drive the rhino to extinction.

The same demand for resemblance can encourage conspiracy theories. “Big events,” like presidential assassinations and terrorist attacks, “can’t have simple small explanations,” it is thought. No! Big events need big causes. JFK wasn’t shot by Oswald – it was the CIA working in conjunction with the Mafia and the Kremlin. The 9/11 attacks weren’t pulled off by 19 hijackers – it was really Bush and the US government. As (fictional) President Bush admitted to the boys on South Park:

Quite simple to pull off, really. All I had to do was have explosives planted in the base of the towers. Then on 9/11 we pretended like four planes were being hijacked, while we really just rerouted them to Pennsylvania, then flew two military jets into the World Trade Center filled with more explosives, [and] then shot down all the witnesses of Flight 93 with an F‐15 after blowing up the Pentagon with a Cruise missile. It was only the world’s most intricate and flawlessly executed plan ever, ever.

(Parker and Smith 2006)

The fallaciousness of such reasoning was demonstrated in a study by Patrick Leman (2003), when he had subjects read one of four different (fictional but convincing) newspaper accounts of a foreign president’s assassination. In one, the assassin was successful. In a second, he merely wounded the president. In another, the president was wounded but later died for another reason. In the last, the assassin missed completely. The subjects were much more likely to invoke a conspiracy to explain the first story than the others. But what rational reason could someone have for doing that? In the first story, the assassin’s gun was aimed correctly – in the others, his aim was slightly off. But if a conspiracy is not needed to explain how an assassin could manage to get a shot off that grazes or misses the target, why would one be needed to explain how the assassin’s aim was actually true? The difference between the stories is literally just a few centimeters, if even that.

Ever wonder why there was a movie made about the conspiracy theories surrounding JFK but you’ve likely never even heard of a Reagan assassination attempt conspiracy theory? It’s because of the representative heuristic. Mere attempts aren’t that big of a deal, so we aren’t apt to demand a big explanation. But such a demand is irrational. So the next time you are inclined to group similar things together or to link similar events and causes, think twice. Similarity can be a useful clue, but it does not establish category membership, probability, or causation.