Human biases in judging threats and costs

December 29, 2008 9:38 PM

What are some examples of statistics or scientific findings that are in opposition to people's intuitions and conventional wisdom?

I am looking for examples of statistics or scientific/economic findings that are at odds with most people's intuitions, especially those that show how much we ignore the huge things that {kill us, cost us money, reduce our quality of life} and focus on the smaller ones. The most common one I hear is that despite our great fear of flying, we are [n] times more likely to die in a car accident than in a plane crash. What are some other examples like this? I am especially interested in examples of popular biases that may end up having an effect on public policy.

posted by lunchbox to science & nature (42 answers total)

Your example is, perhaps, a counter-example to what you're thinking. It's true that more people die in car crashes than in plane crashes, but it's also true that there's a lot more car travel. To really evaluate differences in danger levels you'd probably need to compare them based on deaths per million passenger miles -- or deaths per thousand passenger hours.

Based on deaths per unit passenger miles, it does turn out that air travel is safer than car travel, but not as much safer as I suspect you think based on comparisons of the raw death rates.
posted by Chocolate Pickle at 9:44 PM on December 29, 2008 [1 favorite]

Chocolate Pickle: the studies I read did indeed control for those variables.
posted by lunchbox at 9:49 PM on December 29, 2008

There's a legend about the early days of operations research during WWII: the RAF had been adding armor to the most heavily damaged areas of returning bombers. The OR team realized that this was wrong, since the sample was biased--they were only looking at the surviving planes. The areas that were relatively undamaged in these planes were the ones that needed more armor, since the bombers damaged in those areas weren't returning.

Also, you've probably heard of it already, but there's always the Monty Hall problem.
posted by equalpants at 10:02 PM on December 29, 2008

Whoops, thought of another one: Arrow's theorem shows that it is impossible to design a voting system that satisfies all of several desirable democratic properties.
posted by equalpants at 10:05 PM on December 29, 2008

Here's a relevant current post on the blue.
posted by zoinks at 10:11 PM on December 29, 2008

Overall crime rates in the United States are at or near historic lows for most places. People simply won't believe you when you tell them there is substantially less per capita crime (including sex offenses) than 20 years ago.
posted by Crotalus at 10:16 PM on December 29, 2008 [1 favorite]

...that the full moon does not really affect peoples' behaviour, it's simply confirmation bias?
posted by Savannah at 10:16 PM on December 29, 2008

In economics, the canonical example of a counter-intuitive principle is comparative advantage: the idea that, for example, two countries will both be better off from trading in two goods or services even if one of the countries is better at producing both of those goods or services. Paul Krugman has written in detail about why non-economists — including many politicians, as well as people who encourage you to “buy local” for economic reasons — often fail to understand or accept this idea.
posted by mpt at 10:17 PM on December 29, 2008 [3 favorites]

I think people tend to focus inordinately on the energy consumed by home lighting, when in fact it's a relatively small proportion of household energy consumption. Estimates I've seen to vary between 10 and 30 percent of overall energy consumption, while heating seems to account for 40-60 percent. Mitigating space heating costs by using more granular systems (like radiant,) closing off unused rooms or NOT BUILDING ENORMOUS USELESS VAULTED CEILINGS are all good ways to save energy without sacrificing comfort or filling your home with CFLs that give inferior light and contain mercury vapor and a wasteful disposable ballast circuit (not that I think CFL's are evil, I do think they're a net gain but they're not nearly the godsend they're sold as.)

Actually, the whole "Green" movement/industry is rife with examples of things that have relatively little impact but are given undue importance because they require little effort, are highly visible to others and/or can be easily turned into profitable products.
posted by contraption at 10:17 PM on December 29, 2008 [5 favorites]

You might want to look for the book "How We Know What Isn't So". It's an easy/entertaining read about statistics vs human intuition.
posted by mandymanwasregistered at 10:17 PM on December 29, 2008 [1 favorite]

The Monte Carlo fallacy: Vegas is bankrolled by people's gut belief that the law of large numbers means "large enough to cover my time at the table." The illuminated list of previous winning roulette numbers is there for a reason. (In a similar vein, the odds of winning the lottery.)
posted by holgate at 10:20 PM on December 29, 2008

Interestingly, countries where bicycle helmets are commonly worn (for example the US) have higher cycling related fatalities than countries where they are not (Scandinavia).

Statistically, more lives would be saved if pedestrians or car users wore helmets, yet this is never even considered!

The result of lots of helmets is a scary brand attached to cycling, and lower overall rates of people using bikes.
posted by stepheno at 10:47 PM on December 29, 2008 [1 favorite]

I should link to cyclehelmets.org for more information on this counterintuitive example.
posted by stepheno at 10:48 PM on December 29, 2008

You might like Nick Bostrom's TED talk Humanity's Biggest Problems Aren't What You Think They Are. He argues that we don't spend enough resources on mitigating existential risk and increasing human potential. More philosophical than empirical; he seems to be saying that how we define risk is constrained by certain assumptions (e.g. that people must naturally die, or that humanity will exist indefinitely), and that if we question those assumptions then other problems spring into view.
posted by ads at 10:50 PM on December 29, 2008

Also, given certain conditions groups are consistently more accurate at making predictions or deducing answers than the smartest/most experienced/qualified individuals within them, according to James Suowiecki's The Wisdom of Crowds. Leads to some counter-intuitive results. The book opens with a story about Francis Galton's finding that the weight of an ox at a county fair was more accurately determined by taking the mean of the guesses of all attendees than by asking cattle experts alone.
posted by ads at 11:03 PM on December 29, 2008 [1 favorite]

The intuitive view of basic Newtonian physics held by many people is deeply flawed.

Basically, people don't have an intuitive grasp of forces, their balancing, or the resultant motion.

The one that many, many people apparently get wrong is this situation: you and a friend are both sitting in rolling office chairs a couple feet from each other. You weigh 180 pounds; your friend weighs 120 pounds. You put your feet up on his chair and push against him. Who moves? How much?

The answer is (as can easily be demonstrated) that you both move. He moves farther, but you both move. Most people apparently answer that only your friend moves, since you weigh more.
posted by Netzapper at 12:06 AM on December 30, 2008

The birthday paradox.

Also, an airplane can take off from a conveyor belt ....
posted by Rumple at 12:54 AM on December 30, 2008

A serious, scientific argument can be made that the conventional wisdom that recycling is good is actually wrong. For many materials, the high transportation costs and energy- or labor-intensive sorting processes required make recycling more wasteful than using virgin materials. This isn't saying that all recycling is bad--just that some is. (It makes sense to recycle aluminum cans, for example, but probably not paper products.)
posted by kprincehouse at 1:10 AM on December 30, 2008

Netzapper: Unless the office chairs have crappy wheels, in which case he stays still and I go sideways because of a stuck caster.

People are almost never able to intuitively grasp anything involving exponentials, at least not without a lot of work. Even polynomials above quadratic are pretty iffy. I think intuition generally tries to handle things by patching together a bunch of linear fits. It works pretty well most of the time.

People still often assume heavy objects fall proportionately faster, at least until they consider it a bit. I consider this a triumph of science education, actually, in that people really do seem to override their incorrect assumption when they remember Galileo.
posted by hattifattener at 1:13 AM on December 30, 2008

Compare the number of deaths in the United States as a result of terrorism this decade to the number of deaths from automotive travel. Which is more heavily reported?
posted by rodgerd at 1:16 AM on December 30, 2008

That hot water freezes faster than cold water?
posted by jan murray at 1:34 AM on December 30, 2008

Maybe you've already read it, but Levitt's Freakonomics is a cornucopia of the counterintuitive trivia that you're looking for. Entertaining, surprising, great for conversation, and ultimately frivolous.
posted by randomstriker at 2:56 AM on December 30, 2008

Two classic articles on this topic are Tversky and Kahneman, Judgment Under Uncertainty: Heuristics and Biases, and The framing of decisions and the psychology of choice. The first discusses several general cognitive biases that affect the way people judge the likelihood of certain outcomes (such as the "law of small numbers" cited above); the second discusses how the exact same situation, framed in different ways, may lead people to make contradictory decisions (for example, buying a $20 scarf on sale for $15, but not a $100 shirt on sale for $95, even though both save you the same absolute amount of money).
posted by googly at 3:38 AM on December 30, 2008

Eating fat doesn't make you fat. [1 2]
posted by jbickers at 4:29 AM on December 30, 2008 [1 favorite]

That you are about as likely to be killed by an asteroid impact as you are to be killed in an aviation accident.
posted by mr. strange at 5:40 AM on December 30, 2008

People's fear of an event is linked to the severity of the event more than the likelihood of its occurrence, so fatal events are a rich source of this phenomenon. Check the odds of dying page and look at the how unlikely the objects of common phobias are to harm you.

My favourite example is the AskMe Botulism Police. Check any food storage question on AskMe, and you'll find someone who raises the spectre of botulism and tells you not to do it. This, despite the fact that the chances of getting botulism are on par with being legally executed, dying from a wasp sting, or being struck by lightning.
posted by Jakey at 6:17 AM on December 30, 2008

People's fear of an event is linked to the severity of the event more than the likelihood of its occurrence, so fatal events are a rich source of this phenomenon. Check the odds of dying page and look at the how unlikely the objects of common phobias are to harm you.

My favourite example is the AskMe Botulism Police. Check any food storage question on AskMe, and you'll find someone who raises the spectre of botulism and tells you not to do it. This, despite the fact that the chances of getting botulism are on par with being legally executed, dying from a wasp sting, or being struck by lightning.

Of course, the chances of getting botulism would be higher than they currently are were everyone flippant about the chances of getting botulism.
posted by Flunkie at 6:26 AM on December 30, 2008

Statistically, more lives would be saved if pedestrians or car users wore helmets, yet this is never even considered!

I've thought about wearing a helmet while driving. When I get to the point where I really don't care what anyone thinks of me, I'll do it.
posted by Joe Beese at 7:02 AM on December 30, 2008

Most general scientific findings go against traditional "intuitive" thought. The planet you are standing on is spinning and its round. Its not in the center of anything. It floats in a freezing void. Time is not an absolute, its completely relative. Solid objects contain more empty space than matter. Music, harmony, etc exist solely in the brain. Objects dont have a color, its the light they reflect being interpreted by your eyes and brain that gives them color. There are no such things as psychics, ghosts, seeing the future, etc. There are no creators and biodiversiy exists because of evolution. Germs are the cause of most disease. Homepathy cannot help with serious illness.

Humans have a strong inability to understand costs while gambling. Gambling often is fueled by "magical thinking" which makes people believe that they can influence the outcome or can feel when luck is upon them. Odds at a typical casino are so terrible that its never worthwhile to go to one. The lottery is another example of this.

Also, there was a recent study that showed that incompetent people had a very unrealistic view of their skills while competent people had a more realistic view of their skills. This makes incompetent people much worse at assessing risks and their own ability to handle them.
posted by damn dirty ape at 7:12 AM on December 30, 2008

10 environmental heresies.
posted by Jaltcoh at 7:32 AM on December 30, 2008

If you're a man, you have a greater chance than a woman of being the victim of a crime.

Empirical research on "Megan's Laws" challenges assumptions of people on both sides of the debate.

The financial cost of the Iraq war has been relatively tiny in the grand scheme of things.

People love to say the American criminal justice system isn't working, conveniently ignoring the fact that crime is falling. (You can be sure that if crime were rising, we'd be hearing a lot more about it from the left.)

In addition to Freakonomics, other books you'd enjoy are A Mathematician Reads the Newspaper (John Allen Paulos) and Stumbling on Happiness (Daniel Gilbert).
posted by Jaltcoh at 7:41 AM on December 30, 2008

Jakey's examples are great. Every now and then, somebody wants to do something boneheaded - the low incidence of botulism doesn't mean you should half-ass your canning. (Oh, what Flukey said.) But mostly, people throw it out at the drop of a hat.

Most of my experience with this has been in the realm food safety. I worked for a public health association, and I'd get questions like - "My husband's best friend eats a chicken a day. Will he get the bird flu?" There's also something called the last meal bias - when somebody starts throwing up, they usually blame the last thing they ate, which is usually not what made them sick.

From a policy perspective, the starkest example I saw was the anthrax attacks in 2001. Five people died from from infection in those attacks. But people's reactions were out of hand. First responders were being called out to handle all kinds of random white powder that people found. State labs had to test the samples, which taxed their resources, and took time and equipment away from more routine testing. I know several state lab directors who felt that the time wasted in testing deluge of white powder caused by panic was far more deadly than the attacks themselves. "What if we'd caught that whooping cough outbreak earlier? What outbreaks slipped through the cracks while we were doing all this testing?"

This sort of fear makes for a lot of public health money focused on terrorism that could be better spent improving infrastructure.
posted by averyoldworld at 7:46 AM on December 30, 2008

I recently read about how up until the mid to late 20th century it was thought that as you age you should be less and less active in order to live longer. People thought it was the wear and tear of moving around that aged you and ultimately caused death.
posted by Octoparrot at 8:00 AM on December 30, 2008

My girlfriend thought it was really obvious, but it hadn't occured to me that miles-per-gallon and gallons-per-mile don't have a linear relationship until I read about it (i.e: going from 10mpg to 11mpg saves as much fuel as going from 33mpg to 50mpg.) More here.
posted by so_necessary at 8:53 AM on December 30, 2008

Also it's pretty standard that when polled on "satisfaction" regarding a public service, people are more satisfied personally than is suggested by their overall impression: so "How satisfied were you with your experience last time you required medical treatment?" gets better results than "How satisifed are you with the National Health Service?" And "Is there less crime on your street than five years ago?" gets better answers than "Is Britain safer than it was five years ago?" People seem to assume they just get freakily lucky with good doctors and safe streets existing in a world of waiting-room collapses and unprovoked stabbings.

Can't find a snappy link right now, but have a Google and hopefully you'll find something.

Also the book "Nudge" might suit you, it's about shaping public policy to take advantage of these cognitive biases (a list of them).

People are funny about money. Like if you lose your $20 concert ticket on the way to a concert you were only a little excited about, would you buy another $20 ticket when you get there? Maybe not. But if you didn't have a ticket yet and instead lost a $20 note on the way there, would you still go and buy a ticket when you get there? Probably more likely, even though it's essentially the same thing.

Anyway, Nudge and Freakonomics both have blogs you might enjoy.
posted by so_necessary at 9:15 AM on December 30, 2008

Nthing Freakonomics. Exactly what you are looking for. Especially the 'which is more dangerous' question, having a loaded gun in the house or a swimming pool out back?

"Blink" and "Outliers" by Gladwell are books which also speak to common incorrect bias.

And absolutely nail on what you are asking about is "Benford's Law."
posted by Kensational at 9:26 AM on December 30, 2008

An outline of the book Predictably Irrational.
posted by so_necessary at 10:22 AM on December 30, 2008

Sharp knives are safer than dull knives because they are less likely to catch and slip when cutting.

One would think that more massive stars last longer than smaller stars because they have more mass, but instead it's the opposite. Because they have more mass they burn that mass even more quickly and die out sooner than smaller, less massive stars.
posted by euphorb at 11:12 AM on December 30, 2008

Trebuchets on wheels throw further than trebuchets without.
posted by speedo at 12:00 PM on December 30, 2008 [1 favorite]

Evolution.
posted by exphysicist345 at 5:11 PM on December 30, 2008

Great answers, everyone! Thanks so much.
posted by lunchbox at 7:22 PM on December 30, 2008

Risk: The Science and Politics of Fear
posted by storybored at 8:48 PM on December 30, 2008