Intelligence tests have had many uses throughout their history--as tools to sort both schoolchildren and army recruits, and, most frighteningly, as methods to determine who is fit to pass on their genes. But as intelligence testing has made its mark on the educational and cultural landscape of the Western world, the concept of intelligence itself has remained murky. The idea that an exam can capture such an abstract characteristic has been questioned, but never rigorously tested--perhaps because conceiving of a method to measure the validity of a test that evaluates a trait no one fully understands is impossible. IQ tests have proved versatile, but are they legitimate? To what extent do intelligence tests actually measure intelligence?

Reporter and philosopher Walter Lippmann, who published a series of essays in the 1920s criticizing the Stanford-Binet test, wrote that it tests "an unanalyzed mixture of native capacity, acquired habits and stored-up knowledge, and no tester knows at any moment which factor he is testing. He is testing the complex result of a long and unknown history, and the assumption that his questions and his puzzles can in 50 minutes isolate abstract intelligence is, therefore, vanity."

Lippmann criticized the tests over 80 years ago, but already he recognized that people needed to approach their results with caution--advice now made more salient by a number of studies revealing interesting phenomena that validate his and other test skeptics' opinions.

As it turns out, a number of variables, none of which have to do with brainpower, can influence test scores.

For example, many researchers have discovered that people from minority racial groups often perform worse on intelligence tests than their white counterparts, despite a lack of evidence that they are actually less intelligent.

Another study has shown that intelligence is not fixed throughout one's lifetime: Teens' IQs changed by as much as 20 points over four years, raising questions about some modern IQ test uses. Many gifted programs, for example, use IQ tests to select student participants. But what does it mean if, over their school careers, students who fell below the cut-off grew more "gifted" than those in the program? Should they have been denied the enrichment opportunity, even though they later were revealed to be just as intellectually capable as the students who were allowed to enroll?

External cues that influence one's self-perception--such as reporting one's race or gender--also influence how one performs on intelligence tests. In a blog post for Scientific American, writer Maria Konnikova explains, "Asian women perform better on math tests when their Asian identity is made salient--and worse when their female identity is. White men perform worse on athletic tasks when they think performance is based on natural ability--and black men, when they are told it is based on athletic intelligence. In other words, how we think others see us influences how we subsequently perform."

If one's performance on IQ tests is subject to so many variables outside of natural ability, then how can such tests measure intelligence accurately? And does one's level of innate intelligence even matter? Is it correlated with success?

In one study, Robert Sternberg, a psychologist at Tufts University, found that in the Western world high scores intelligence tests correlated with later career success, but the people he tested live in a culture that places enormous emphasis on achievement on such tests.

Imagine a student with great SAT scores who later goes on to excel in her career. One could say that the student was very smart, and her intelligence led her both to succeed on the test and in her career. But one could also say the student was particularly skilled at test-taking, and since she lived in a society that valued high test scores, her test-taking ability opened up the door to a great college education. That in turn gifted her with the skills and connections she needed to succeed in her chosen field.

Both of these scenarios are over-simplifications. Intelligence--however murky it may be and however many forms it may come in--is undoubtedly a real trait. And intelligence tests have persisted because they do provide a general way to compare people's aptitude, especially in academic settings. But from their invention, intelligence tests have been plagued by misinterpretation. They have been haunted by the false notion that the number they produce represents the pure and absolute capacity of someone's mind, when in reality studies have shown that many other factors are at play.

Donna Ford, a professor of Education and Human Development at Vanderbilt, writes, "Selecting, interpreting and using tests are complicated endeavors. When one adds student differences, including cultural diversity, to the situation, the complexity increases... Tests in and of themselves are harmless; they become harmful when misunderstood and misused. Historically, diverse students have been harmed educationally by test misuse."

If current intelligence tests are subject to factors outside of intelligence, can a new assessment be developed that produces a "pure" measure of innate intelligence? Scientists are starting to examine biology, rather than behavior, to gain a new perspective on the mind's ability. A team of researchers recently set out to understand the genetic roots of intelligence. Their study revealed that hundreds or thousands of genes may be involved. "It is the first to show biologically and unequivocally that human intelligence is highly polygenic and that purely genetic (SNP) information can be used to predict intelligence," they wrote.

But even if researchers discover which genes are related to intelligence, a future in which IQ tests are as simple as cheek swabs seems unlikely. Most contemporary intelligence researchers estimate that genetics only account for around 50 percent of one's cognitive ability; the environment in which one learns and grows determines the other half. One study found that biological siblings, when raised together, had IQ scores that were 47 percent correlated. When raised apart, that percentage dipped to 24.

And, though world knowledge and problem-solving skills are commonly tested, the scientific world has yet to come to a consensus for a precise definition of intelligence. Perhaps such a definition is impossible.

So perhaps Galileo would have stolen Mozart's prestigious preschool seat. But that preschool admissions officer would likely whack herself in the head a year or two later as she heard of the prodigious star wowing audiences with his musical compositions.

Discarding intelligence tests altogether might be too harsh a reaction to their flaws. After all, they are useful when interpreted correctly in the right settings. Instead, we need to understand the tests' limitations and their potential for misuse to avoid determining a person's worth--or future trajectory--with a single number.

Who was smarter--Galileo or Mozart?

Answering that question seems impossible. After all, the former was an astronomer, the latter, a composer. Asking which of the two was smarter seems akin to forcing someone to objectively determine if apples are better than oranges. But while humans have not yet invented a scale to measure the value of fruit, we do have one that measures brain-power: the IQ test. And according to a book by Stanford psychologist Catherine Cox Miles, Galileo's IQ was around 20 points higher than Mozart's. The number may seem trivial, but if both of them were three-years-old today, competing for a slot at a private New York City preschool, Galileo would likely edge out Mozart. But should he? Is there such a thing as a true measure of something as intangible as intelligence?

The notion of assigning a numerical value to intelligence dates back to the early 20th century, when French psychologist Alfred Binet created a series of tests to help Parisian public schools identify "mentally defective" children. Between 1904 and 1911, Binet and his colleague Theodore Simon observed the skills of "average" French schoolchildren, then created a series of tests for students between the ages of three and 12 designed to assess whether their abilities were above or below the norm.

To calculate a student's "intelligence quotient," Binet and Simon simply took his mental age, divided it by his actual age, and then multiplied by 100. For example, if a seven-year-old could perform the tasks required of a nine-year-old, his IQ would be (9 / 7) *100, or around 128.

Intelligence testing reached the United States in 1916, when psychologist Lewis Terman created a new, refined intelligence scale based on the abilities of thousands of students--significantly more than the fifty or so Binet studied. Today, psychologists use a revised version of Terman's scale to evaluate children in five categories: fluid reasoning, knowledge, quantitative reasoning, visual-spatial processing and working memory. Big differences in a student's scores across these categories can help psychologists diagnose learning disabilities.

Try to answer some questions from a real IQ test yourself. Below are some examples, one from each of the five categories.

Fluid Reasoning:
1. "I knew my bag was going to be in the last place I looked, so I looked there first." What is silly or impossible about that?

Knowledge:
2. What does cryptic mean?

Quantitative Reasoning:
3. Given the numbers 3, 6, 9, 12, what number would come next?

Visual-spatial Processing:
4. Suppose that you are going east, then turn right, then turn right again, then turn left. In what direction are you facing now?

Working Memory:
5. Repeat a series of digits (forward or backward) after hearing them once.

Source: Introduction to Psychology by Dennis Coon and John O. Mitterer

Just a few years after Terman brought the IQ test to the United States, it left the classroom--and entered the military. During World War I, the number of army recruits exploded from around 200,000 in March of 1917 to over 3.5 million in November of 1918. As the military grew, so too did the need for trained officers; the most intelligent recruits needed to be identified early so they could enter officer training programs.

Thus Harvard psychologist Robert Yerkes developed the Army Alpha and Beta tests. Modeled after the Stanford-Binet scale, the tests were designed to give commanders a sense of the intelligence of the men they were leading and to screen soldiers for officer potential. Unlike Terman's IQ test, the army exams could be administered to recruits en masse and the results could be summed and interpreted without the expertise of a psychologist. During WWI, over 1.7 million men took the intelligence tests.

Think you have what it takes to be an officer? Try the questions below--they appeared on real Army alpha tests.

1. If you saw a train approaching a broken track you should:
A. telephone for an ambulance
B. signal the engineer to stop the train
C. look for a piece of rail to fit in

2. Why is beef better food than cabbage? Because:
A. it tastes better
B. it is more nourishing
C. it is harder to obtain

3. Why do some men who could afford to own a house live in a rented one? Because:
A. they don't have to pay taxes
B. they don't have to buy a rented house
C. they can make more by investing the money the house would cost

4. A dealer bought some mules for $1,200. He sold them for $1,500, making $50 on each mule. How many mules were there?

5. Unscramble the words to form a sentence. Then indicate if the sentence is true or false.
a. happy is man sick always a
b. day it snow does every not

Answers: 1.) B 2.) B 3.) C 4.) 6 5a.) False - A sick man is always happy. 5b.) True - It does not snow every day.

Sources: historymatters.com and official-asvab.com

While the tests helped educators and administrators in the early 20th century understand more about their students and recruits, they had already begun to stray from Binet's original intention. People began to use them as indicators of general aptitude, removing them from the classroom context for which they were intended. Suddenly, an absolute measure existed for a trait that had never been absolute--adding fuel to the fire of the growing eugenics movement. Rather than simply suggesting the one would have success in grade school, higher scores on Binet's test started to mean that one was more fit for breeding. The Advanced Learning Institute reported that between 1907 and 1965, thousands of people were sterilized on the basis of low scores on intelligence tests that characterized them as "feeble-minded."

In 1924, 18-year-old Carrie Buck became the first person subjected to Virginia's Eugenical Sterilization Act. She was classified as "feeble-minded" after a version of the Stanford-Binet test revealed that she had a mental age of nine. Carrie resided in the State Colony for Epileptics and Feeble-Minded in Virginia, the superintendent of which decided that she would be the first person subjected to the new law.

According to Paul Lombardo, a professor at the University of Virginia and an expert on Buck's history, others arranged a trial for Carrie to challenge the new law. Carrie was unable to convince the court of her mental capacity, but her lawyer appealed the court's decision, arguing the new law was discriminatory. The case, Buck v. Bell, went all the way to the Supreme Court, but ultimately, in 1927, the court deemed that there was nothing unconstitutional about Virginia's new law. Carrie, along with around 8,300 other Virginians, was sterilized.

It took the rise of Nazi Germany for people in the United States to recognize the horrific consequences of eugenics. But, chillingly, though the sterilization of individuals in mental institutions came to a halt in the 1970s, the Buck v. Bell decision has never officially been overruled.

Editor's note: NOVA scienceNOW explores the science of intelligence on "How Smart Can We Get," premiering Wednesday, October 24 at 10 pm ET on most PBS stations.

19 worst passwords
The 19 worst passwords, as identified by SplashData.

If your go-to password is "password" or "123456," you should probably stop reading and go change it--both topped SplashData's list of the most common passwords of 2011. The software company was able to create this list by analyzing millions of passwords that hackers had stolen and posted online. According to SplashData CEO Morgan Slain, "password" and "123456" were among the ten most common passwords in over 90% of the individual files hackers posted.

Microsoft researcher Cormac Herley, who specializes in computer security, said based on the lists of stolen passwords he has seen, he suspects that between 0.2 and 0.5 percent of all passwords are the word "password." If that's true, a hacker trying to break into 1,000 different accounts using "password" will likely gain access to between two and five of them.

If you didn't find your password in the above image, don't applaud yourself quite yet. Having a strong, hard-to-guess password can only do so much. "It's definitely better, but it's better against one particular type of attack," Herley said, referring to the "guessing attack" described above in which a hacker tries different username and password combinations to gain access to an account.

Unfortunately, hackers usually take more efficient routes. Sometimes they exploit weaknesses in applications within internet browsers to secretly install keystroke-logging software onto users' computers; other times they create seemingly-innocent websites that lure users to directly reveal their login information. "If you [inadvertently] install something that's a password stealer, it doesn't matter how strong your password is," Herley said.

To protect yourself from these types of attacks, Herley recommends that you run up-to-date software. Hackers often exploit known vulnerabilities in applications like Flash or Java, and the longer a version of a program has been available, the more likely it is that hackers have learned how to break into them. He also recommends installing antivirus software.

Hackers also target servers directly. In 2009, someone broke into the server of the social-gaming site RockYou and released over 32 million passwords, including user login information for partner sites like MySpace. Websites are supposed to encrypt passwords, meaning that rather than storing a user's actual password, they store a "hash"--the result of a specific function calculated on it. But RockYou had stored the passwords in plain text, meaning when hackers gained access to the database they found an easily exploitable list of usernames and corresponding passwords. Herley called this "really, really bad practice."

Although there is no foolproof way to protect yourself from an attack on a server, Herley said you should take measures to ensure that a breach on a low value site does not compromise their security on a more important one. Although creating new passwords for each site one uses is sound advice, in practice, people may not be able to choose unique passwords for every single account they have. Herley recommends that users rank them in order of importance--creating a unique password for one's email account is likely more important than making five different passwords for five different online magazine subscriptions.

Given the various ways in which hackers can exploit traditional password login systems, one may wonder why we still use them. Researchers have proposed and developed new ways of verifying an individual's online identity. Sites could require users to scan their fingerprints or insert personalized smart cards into specialized readers attached to their computers. But Herley said he doesn't see the new technologies replacing traditional login systems any time soon for one main reason: cost. No new technology, especially one that requires a physical device, will ever be cheaper than passwords, which essentially cost websites and users nothing.

He used Facebook's "explosive growth" as an example. The site grew to nearly one million users before it received any big investments, a feat that would have been impossible had authenticating the identity of each user cost even as little as ten cents.

Sites' continued reliance on traditional verification methods does not mean that internet security has stopped improving. Companies are constantly working behind the scenes to augment password systems, implementing new techniques of verifying user identity such as tracking the locations of the computers in which a user logs in. If a website detects something unusual--like a login on a new computer in a different country--it can boost its security by requiring the user to enter additional information, like a code that is text messaged to the number linked to the user's account.

As the methods sites use to protect users' information advance, so too will hackers' determination to crack them. But before you spiral into panic, rest assured that although a spammer might find his way into your Facebook or email accounts, most Internet users will never suffer from a devastating theft of cyber identity. Still, "Anyone who has been online for any amount of time should have their guard up," Herley said. He means you, "password" users.