Show Mobile Navigation
Misconceptions

10 Dumb Lies That Tricked People With Numbers

Adam Rowe


Statistics, percentages, and numbers of any kind hold a strange power: People believe them. State an opinion, and no one cares. State a fact, and it matters. As a result, people tend to apply statistics to anything, no matter how ridiculous.

One analyst even went so far as to compose a formula for the perfect butt:

butt-formula-gray

Sadly, it’s calibrated for the female derriere, so men’s butt attractiveness remains an unquantifiable mystery.

Granted, no one took the butt formula seriously. But historically, plenty of mass lies have worked because of the distorted math supporting them.

10 Death By Coconut

10-falling0coconut_000003976655_Small

Perhaps you’ve heard the urban legend that 15 times as many people die from coconuts falling on them as from shark attacks. Maybe you didn’t believe it at first, but it’s commonly backed up by statistics. At least, it appears to be if you don’t understand how data extrapolation works.

Coconuts don’t kill nearly as many people as sharks, but they kill a significant percentage of people in one location: the Provincial Hospital in Papua, New Guinea. A 1984 research paper published by Dr. Peter Barss in the Journal of Trauma started the whole coconut myth when Barss noted that 2.5 percent of the admissions to his hospital over a four-year period had been caused by falling coconuts.

The problem? In May 2002, respected shark researcher George Burgess used an extrapolation of that data to cover the rest of the world—which would only make sense if every other location had the same number of palm trees as New Guinea. Nevertheless, the claim stuck. Burgess is often cited as a definitive source of this misleading “fact” that falling coconuts kill 15 times as many people as sharks.

What does Dr. Barss think of the general public’s response? He’s not amused. “When you’re treating these injuries daily, it’s not funny at all,” he told the Canadian Medical Association. He’s right, too. The average coconut is 2.0 kilograms (4.5 lb) and falls 25 meters (80 ft), resulting in an impact force of up to 1,000 kilograms (2,200 lb).


9 Blonds Are Becoming Extinct

9-natural-blond_000066229801_Small

People with naturally blond hair will die out within 200 years. According to a factoid published by the BBC in 2002, blonds “are an endangered species and will become extinct by 2202.” Many other publications made the same claim, including The Sunday Times. They stated that “the last natural blond is likely to be born in Finland during 2202.”

But it was a complete hoax, as the World Health Organization (WHO) was forced to explain. The dry humor of the WHO clarification is tough to miss:

In response to recent media reports . . . WHO wishes to clarify that it has never conducted research on this subject. Nor, to the best of its knowledge, has WHO issued a report predicting that natural blondes are likely to be extinct by 2202. [ . . . ] We have no opinion on the future existence of blondes.

8 Statistics Can Create The ‘Most Wanted Painting Ever’

8-blue-painting_000062242598_Small

In 1994, polling firm Marttila & Kiley collected a swath of statistics in the US to determine what qualities Americans found most desirable in paintings. Around 66 percent preferred “soft curves” over “sharp angles.” Approximately 33 percent wanted to see a fall painting while just 15 percent wanted to see winter. A whopping 88 percent preferred an outdoor scene over an indoor one.

The survey was undertaken for two Russian-American painters, Vitaly Komar and Alex Melamid. They went on to craft the most wanted—and the least wanted—paintings for Russia, China, Italy, and 10 other countries.

The painters didn’t publish an account of their interpretation of the data. But some of the poll data generated confusing results. For example, the color blue was the winner of both the “favorite color” poll and the “second favorite color” poll at 44 percent and 19 percent of American voters, respectively. This potentially complicates the question of what percentage of blue to put in the painting.

As Komar and Melamid have not yet reached international stardom, their reliance on this data does not appear to have justified their claims. As artist Jonathan Keats commented, “If this is America’s most wanted painting, just imagine what laws the president and Congress think you want passed.”



7 Women Will Run Faster Than Men By 2156

7-woman-outpace-man_000059569948_Small

In 2004, the peer-reviewed, well-respected journal Nature published an article asking, “Will women outpace men in 2156?” The journal’s study relied on data from Olympics-level, 100-yard dashes dating back to 1900. The University of Oxford’s Andrew Tatem had originally analyzed the numbers.

According to the trend, women’s finish times are creeping up on the men’s times. Statistically projected into the future, this trend showed that women could be expected to be faster than male sprinters in the 2156 Olympics, give or take a statistical margin of error of 724 years. The margin of error didn’t appear in the article’s title.

Nevertheless, the data was misleading. According to physiologists, the study ignored the possibility of a natural cap on the trend caused by differences in testosterone and weight that favor men. One physiologist went so far as to call the analysis “flawed at a fundamental level.”

So what could have caused the trend? Perhaps health advances since the start of the 20th century have impacted athletes’ performances more than their genders. Even so, their genders will continue to have a measurable, if smaller, impact.

But common sense needs to temper this kind of analysis. Obviously, not all trends will continue indefinitely. If they did, female athletes would be running at the speed of light within another millennium.

6 50 Percent Of All Marriages End In Divorce

6-divorce_000073209423_Small

Divorce in the US peaked in the 1970s and has been declining ever since. Not in the eye of popular wisdom, though. Most naysayers cling to a single, official-sounding percentage to back up their belief in the moral (or at least matrimonial) decay of society: 50 percent of all marriages end in divorce.

According to New York Times reporter Tara Parker-Pope, the actual origin of this statistic is “murky.” In fact, she devoted an entire chapter of her marriage guide to debunking this myth.

Since the 1970s, it hasn’t been close to accurate because more women began going to college and delaying marriage. Many other factors may have also influenced divorce rates—from new divorce laws that artificially boosted the stats to shifting societal norms. Sadly, no definitive divorce statistic exists. But one thing is certain: The divorce rate is nowhere near 50 percent of married couples today.

The sticking power of this percentage is probably due to its status as a “political Swiss Army knife,” Parker-Pope says. Social conservatives believe that it represents the decline of traditional marriage while liberals may invoke it to bolster efforts to help single mothers.

5 Over 50 Percent Of Detroit’s Population Is Illiterate

5-illiteracy_000085321015_Small

In 2013, a controversial claim resurfaced: Over 50 percent of Detroit’s population is illiterate. This claim had also appeared in 2011 and was covered then by news outlets like CBS, Fox, and the Huffington Post. At that time, this conclusion was based on a “new” study that pulled the statistic from an earlier analysis, which used even older data.

The true source? A 1993 national survey that polled 26,000 people.

The data was already five years old when it was used in a 1998 analysis by the National Institute for Literacy. However, this poll never provided a verifiable picture of Detroit’s literacy.

The study charts Detroit’s functional literacy rate at 47 percent but with a major caveat that is rarely mentioned. Appendix A states that “there is no direct evidence available about the validity of the model’s predictions for the Congressional district or city/town/place census areas” and that “some caution is appropriate.” So the model may work for countries but hasn’t been proven to work for cities.

As provided by the National Center for Education Statistics, the most recent information about Detroit’s literacy is at the county level. According to 2003 data, those lacking “basic prose literacy skills” in Wayne County, Michigan, clocked in at 12 percent.

Data extrapolation is always risky. But the media furor about Detroit’s literacy reveals just how long one out-of-date statistic can linger. Ultimately, the conclusions drawn from this study say more about journalists’ problems with understanding statistics than about Detroit’s problem with illiteracy.



4 Women Are More Likely To Get Breast Cancer If They’ve Had An Abortion

4-breast-cancer_000014444202_Small

According to Joel Brind’s 1996 meta-analysis, a woman’s risk of breast cancer increases by 30 percent after an abortion. Under the 2003 Woman’s Right to Know Act, Texas abortion providers have to mention this connection to potential clients. However, the link is a widely disseminated bit of misleading data.

The flaw lies in the difference between correlation and causation. If the number of women with breast cancer appears to be correlated—even in a minor way—with the number of women who had abortions, that statistical relationship does not mean that abortion causes breast cancer.

There are mathematical and scientific reasons for this. But common sense also comes into play. In strongly pro-life regions, women may not admit to having an abortion when polled. That may be true even in regions without strong pro-life factions. Abortion is an intensely private matter. Why risk loss of privacy or negative repercussions by telling a statistician about it?

Furthermore, if statistics are drawn from a subset of women, such as those with breast cancer, their behavior may be different than that of all women taken together. A medical scare such as cancer may cause a woman to be more honest about having an abortion when asked about her medical history. Of course, there’s no way to know if this causation is true, either.

But if women with breast cancer are more truthful about their medical histories, then women without cancer who had abortions would remain undocumented at higher rates. So we wouldn’t know exactly how many women who had abortions didn’t get breast cancer.

It’s also concerning that Brind appeared to draw his conclusion before he undertook his study. When Science News wouldn’t print his letter disagreeing with one of their articles on pregnancy and the risk of breast cancer, he may have decided to take action another way.

As quoted in the Christian publication World, Brind said, “I realize abortion increases the risk of breast cancer, and this journal [Science News] is covering it up. Somebody has to bring this out because a lot of women are being made sick and [will] die from this.”

Two years after Brind’s analysis, an unrelated study found “no excess risk of breast cancer” among those who had abortions, and a 2003 National Cancer Institute workshop further refuted the claim.

3 The Average Man Sleeps With 7 Women

3a-ladies-man_000030349326_Small

From 1999 to 2002, the National Center for Health Statistics collected data on the number of sex partners that US men and women have. According to the results of their study, the typical man sleeps with seven women while the typical woman sleeps with four men.

This is another case of misleading data due to one simple trick: straight-up lying.

In general, men want to be seen as having a “reasonably” extensive sexual past while women want to be seen as having a “reasonably” chaste sexual past. Unfortunately, their actual pasts might not line up, leaving them to lie about sex whenever they take surveys about the subject.

As a result, gender stereotypes about sex are self-perpetuating. To fit into society, enough people may lie about their sex partners to create a false statistic that depicts the nonexistent society that everyone’s trying to fit into.

To try to get more truthful responses, additional studies hooked up respondents to lie detector machines. These lie detectors weren’t always working, although the respondents didn’t know that. Faced with a possible check on their honesty, male and female participants reported equal numbers of sex partners.

2 The Existence Of The Bermuda Triangle

2-bermuda-triangle_000011927848_Small

The Bermuda Triangle is one of the most famous examples of a real-life mystery worthy of a conspiracy theory. Anytime a ship or plane vanishes within this area of the Atlantic Ocean, the popular theory returns that some unknown transportation-destroying force lies within the Bermuda Triangle.

According to some websites, this force swallowed 11 missing aircraft over a 22-year period. The secret to this frightening accumulation of data? A lack of context.

As reported by the BBC, a 2013 study noted that the Bermuda Triangle didn’t even make the list of the top 10 zones for shipping accidents or disappearances. Ships and planes go down everywhere, but any data outside the Bermuda Triangle is simply ignored by proponents of this theory.

In an interview with Playboy, Carl Sagan explained the fallacy this way: “Compared with other places in the world as well-traveled as that area of the Atlantic, do airplanes and ships go down more? The answer is no. Why is it always planes and ships that get lost? It’s because they can sink in water. If we started losing trains—if we had a Duluth Triangle in which trains began disappearing—that would be interesting.”

1 Joe McCarthy’s List

1-joe-mccarthy

Photo credit: United Press

On February 9, 1950, US Senator Joe McCarthy claimed that the US State Department was “thoroughly infested with communists.” His supposed proof: a list of 205 members of the Communist Party hidden in the government. He named no names but simply waved around a number.

But that number changed. Despite reports to the contrary, McCarthy claimed to have said “57” later. Then, on February 20, McCarthy split the difference and cited “81 loyalty risks.”

McCarthy didn’t make his numbers up wholesale. A 1947 investigation produced 108 case studies of State Department employees about whom questions of security or loyalty had been raised. By 1948, only 57 of them were still employed by the State Department.

We might assume that the investigation had cleared these 57 employees, or we might call them “card-carrying communists” as McCarthy did. His 81 number was drawn from the 108 case studies. Meanwhile, the 205 number came from an earlier 1946 report to Congress which confirmed that 79 of 284 potential security risks had been fired.

Put on the defensive, the US government scrambled to locate the source of McCarthy’s numbers. McCarthy was less concerned about that. He simply said, “I don’t answer accusations. I make them.”

The effects of his list were so strong that his influence is termed “McCarthyism,” a word that is now defined as the general practice of making unfair allegations. McCarthy also gives us one of the best examples of how numbers can be misused. He was able to lead a witch hunt simply by throwing out a few digits at random to a scared, polarized populace.

Adam Rowe is a science and technology writer with a love for science fiction. Don’t take his word for it: He runs a blog to prove it, Maddd Science.