How to Defeat Intelligence Tests

by David Ricardo

There is a plethora of psychological tests that purport to measure every conceivable aspect of how and why you think the way you do.

In this article, we will look at some of the most important, best known, and most controversial of these: intelligence tests.

Let's see how this article shakes out: I have no desire to overstay my welcome and the easiest way a writer can do that is by writing things in which the beginning and the ending are just too far apart!

Before we get too deeply into this matter, you should know that all of these tests are of at least some importance in your life, even if you are not aware of them.  Intelligence tests do not necessarily measure what they purport to measure, because there really isn't a rigorous universally accepted definition of just what intelligence is, so it seems highly unlikely that intelligence tests can measure that, whatever it is.

Instead, we will just say that intelligence tests measure whatever it is that they do measure, which is the test taker's skill in certain particular cognitive abilities: his or her reading comprehension, the ability to discern analogies, and vocabulary size, to name just three of them.  Any numbers that are derived from taking this test are based on the test taker's mastery of those skills relative to the population as a whole.

In psychometrics, that is, the field of psychological measurements, validity is the extent to which theory and, thankfully to an ever-increasing degree, evidence support the interpretation of test scores.

There is another related concept, reliability, which is the extent to which the measure produces similar results under similar conditions and these conditions must be standardized, which is why those intelligence and personality tests you see in magazines are really worthless, except as entertaining diversions.  If you take the test again in six months or a year, will you receive a score that is in the same ballpark as the first score?  If the test gives a nearly equivalent score when it is retaken, and it does this consistently with a large enough and random enough sample of people, then the test is reliable, and that is wonderful.

This is certainly easy enough if rather time consuming to demonstrate, but that does not answer the question of whether it is valid: is this intelligence test measuring intelligence, that which it claims to measure?  Even if it does measure what it measures, that could be something completely different.  This remains an unanswered question in the field of intelligence measurement.

So, how do you define "intelligence?"

My definition of intelligence is that it is a certain degree of agility in people's thinking that allows them to solve problems and to better adapt to the environment.

Your own experience has surely revealed to you people who are more adept at doing this than others, but remember that is my definition.  What about the young person who is very skilled at making plastic model automobiles, which is certainly a form of hacking?  That requires a degree of manual dexterity that I no longer have (if, in fact, I ever had it) combined with something most people might recognize as at least some form of intelligence.

I will tell you that not long ago I met a child who was, oh, maybe eight years old, and I say that because I don't believe he thought in terms of how old he was, or was even aware of his age.  What is amazing is that this child knew everything there is to know about deep fat fryers, the machines used to make french fries, to the point that if you named a manufacturer and a serial number, he could tell you when that machine was made!  Now, he couldn't be trusted to operate or repair a deep fryer because this child is absolutely wrapped up in his own field of interest to the exclusion of everything else.

So, I pose the question to you: is this intelligence in action?  This is hardly my area of interest and, after the initial novelty wore off, which did not take long, listening to him was rather boring and yet, if he stumbled upon a convention of deep fat fryer enthusiasts, I am sure he would have enraptured them as rapidly as he bored me.

So, what we consider to be intelligence is closely related to the environment and the situation into which it comes into play.  I tend to think that living on the streets of a large city in this cruel nation of ours requires more sheer intelligence than running a large corporation because running the corporation, while exacting, is fairly well defined, but a homeless person is constantly presented with new and different situations, all requiring a response, and many times survival hinges upon that response.

The important part of hacking is understanding why things are as they are, and that means realizing that any creation of the human mind is subject to the foibles of the human mind.

I don't believe it is original to me, but I am fond of saying that anything created by the human mind can be defeated by the human mind.  All manufactured objects are products of human thinking and, to hack that thing, you must understand the thinking behind the creation of what you are hacking.

Intelligence tests, like all the other tests, are the products of patterns of thinking by educated, upper-middle class people who look upon the masses as objects to be poked, prodded, studied, and measured.  I know this because I once was one of those upper-middle class types and, while I would like to think that I have outgrown the attitude, I am aware that collectively they think that they are better and smarter than most people, and this is the mentality that creates psychological tests.

Psychological tests can be beaten because the people creating them think that the people taking them are incapable of understanding how the tests work.

In the case of intelligence, it is all due to Charles Spearman, the noted statistician of factor analysis fame and the person who first thought of intelligence in terms of something called g or a general factor.

Let's take a moment to understand Spearman's thinking: as early as 1904, he noticed that there was positive correlation between children's performances across a wide variety of tasks, including what were seemingly unrelated school subjects.

After all, if you do well in history, chances are good that you will also do well in science, and this extends to intelligence tests, too - if you score well on one, you will almost certainly score well on others.  It is claimed (and may very well be true) that this positive correlation is the most replicated finding in all of psychometrics.  Well, if this is true, and for the moment we will accept that it is, then there must be that thing called g and the object of the intelligence test is to measure it in a replicable way, even if there is no test that directly measures it.

Because g has so many components, the only way to arrive at its value is to take a battery of tests, or a test composed of numerous components.

Then it is necessary to determine how well any chosen test correlates with this thing we call g.

These correlations are always positive and they can be as low as about 0.10, while other tests such as the complete Wechsler test have a correlation as high as 0.95 (and I will be talking more about the Wechsler test in a little while), with most people thinking that tests of general knowledge and vocabulary correlate well with g.

Now, recall my young acquaintance (and I will not call him a "friend" since I do not think he has the capacity to form friendships) with all that knowledge of deep fat fryers.  He was not an astute speaker or an organized thinker, except when he was seized by the enthusiasm of talking about deep fat fryers - he couldn't talk about dogs or the weather or cars or anything really, just deep fat fryers.  It is easy to say that he is autistic or a reasonably high-functioning person afflicted with Asperger syndrome and, just as it is easy to attach similar cut and dry labels to everyone we meet, even if our diagnoses are correct, it is also wrong.

I am convinced that there is some form of intelligence at work there, but not the variety of intelligence that intelligence tests claim to measure.

Next, intelligence tests are commercial items and they must sell in sufficient volumes and at a sufficient markup to justify their development, production, and marketing costs.

This means that there must be enough demand for them and, for there to be enough demand for them, they must be very general in how they measure intelligence and arrive at that all-important number.  Is the conclusion that this number measures something useful justified?  It is, to an extent.  The skills that are tested on modern Intelligence Quotient (IQ) tests are highly valued in American society and, if we think of the IQ test in terms of measuring skills rather than ability, then they are of genuine value and they are reasonably good at predicting academic performance, income, and even health.

This is a limitation of IQ tests: the scores are good indications of one's potential to succeed in modern America, but underneath this the scores are some imponderable mishmash of ability, opportunities, and motivations.  On this basis, it is not possible to compare IQ scores across ethnicities, cultures, nations, or periods of time simply because these skills are not as highly valued across cultures and ethnicities, just as they have not been as highly valued through history: remember that universal literacy is a relatively recent phenomena.

The development and need for this degree of abstract thinking skills are a cultural adaptation to the complexities of our science and technology-driven society, but the complexities of this modern life are far from evenly distributed throughout our society.  These skills are more necessary in the developed world and, even then, they are used and prized more by the upper socioeconomic strata than any others, hence their appearance in intelligence tests.

This is also why there is no IQ test based on knowledge of deep fat fryers.  If there was enough demand for such a test, then that test would exist and my young acquaintance would be one of humanity's intellectual superstars: Albert Einstein, Enrico Fermi, and John von Neumann all rolled into one!  Alas, despite the existence of a few very specialized tests, most intelligence tests must be more generalized than this, and so compromises must be made just like that television I have which is made down to a price, rather than up to a standard.  This is not to say that psychological tests are shoddy: quite the contrary in fact, and Rorschach cards alone are proof of this, but it is the thinking behind them that leaves a great deal to be desired.

Intelligence testing in a big way all started with the United States Army in World War I.

There were intelligence tests before then, but this was when they took center stage in American life to an unprecedented extent and now, a hundred years later, they are still here.  The Army needed lots of conscripts and they had to be able to rapidly find officer material as, at that time, there were not enough college graduates to fill the officer ranks.  So the military turned to native ability, or at least this is what they thought they were turning to, so nearly two million conscripts might be properly assigned to appropriate jobs.

In time, intelligence tests found their way into schools, so students could be segregated into groups based on their mental ability, though originally the tests were intended to identify those in need of special education.  For the most part, intelligence tests offered in schools are machine scored, fill in the bubble with a #2 pencil type of test because these are quick and cheap to administer and they are good enough for their purpose.  But what you will soon discover is that there is a very big gulf between "good" and merely "good enough."

Let's dig a little deeper into this.

I think we can all agree that at least some of what anyone will regard as intelligence is the ability to use language.  There are many ways to determine how well someone uses language.  For example, you can have the test taker define certain words, like "edifice," "tirade," or "ominous."  That's just a matter of knowing them or not knowing them and you then define them in your own words.  Those three words at one time or another were found on the Wechsler Adult Intelligence Scale, or the WAIS.

What I can say is that if you are familiar with a list of words used on the SAT to test your verbal reasoning, then you can easily handle anything the WAIS will throw at you in that section of the test.  On the WAIS there are other tests of verbal reasoning.  You are asked, for example, to explain the meaning of the proverb "still waters run deep" or the similarity between praise and punishment.

To my way of thinking, these are better tests of verbal intelligence than whether you know or don't know the meaning of a group of specific facts or words.  I will leave the proverbs to you to figure out, but the similarity between praise and punishment is that they are both used as behavior modifiers and, when you use that line on a psychologist while taking the WAIS, just watch that person sit up and take notice.  The WAIS is an excellent test, but it requires the services of a trained professional for about an hour, though as with everything else involving the human mind, this time varies and this is why it is not appropriate for administering to a school full of children.

Even more extreme is the Luria-Nebraska Neuropsychological Battery (LNNB), which is intended to determine whether the person taking the test has suffered any brain damage, what this brain damage is and its extent.

As you might well imagine, this test takes hours of a trained professional's time to administer.  These tests are used in those cases in which someone really wants to know a lot about you, such as in the criminal justice system.  And if you are unlucky enough to find yourself in the criminal justice system, then a high score on the WAIS will get you different treatment from a person receiving a lower score: you will be working in the prison library or given administrative tasks, rather than being out in the hot sun performing manual labor.

So, how do you beat the WAIS?

It certainly helps to see one.  Theoretically, these tests are sold only to licensed professionals, but I have found them for sale on eBay, at second-hand bookstores, and even at yard sales!  There is no guarantee that you will find one this way and any that you do find in these places will probably be obsolete versions, but they are still informative.  You can use them to work on your vocabulary, practice building puzzles with blocks, and to study proverbs.

Alternatively, you can see a recent edition of the Wechsler just by asking.

Simply go to the psychology department of any university near you and place a notice on a bulletin board that you are willing to volunteer to take psychological tests.  Such volunteers are in very short supply, especially people who are willing to take the tests without compensation from impecunious graduate students.  If they ask you why you are doing this, and they probably will, just say that you are interested in psychology and you are considering it as a career, and they will think they are talking to one of them.  You will not get to pick and choose the tests you take, but you will soon be exposed to the Wechsler.  Take it.  Look at it.  Study it.  You will be surprised at how much you remember because it is so logical in its content and structure.

Believe me, if you can do many of the feats discussed in this magazine, before long you will be able to manipulate the WAIS to the point that you can get the score that you want.

Nothing on the test is difficult, but when you take the test, you are playing beat the clock.  On your own, you can take your time and reflect on it.  It does not take much practice to become a "natural" at it, even when you do have to beat the clock.  Keep in mind that doing this does not make you more intelligent or smarter.  It just gives you a higher numeric score on a test.  For most situations, this is all you need.  This alone can open up many opportunities in education and employment, and I cannot stress that enough.

And now, a word about those scores.

Generally, IQ tests are normed to give an average or mean score of 100 when applied to the population as a whole, with the distribution of scores following that infamous bell curve.  We also speak of standard deviations, and, in the case of the Wechsler, the standard deviation is 15.

The total area under that bell curve is 1, corresponding to a probability of 1, meaning that it is an absolute certainty that everyone is somewhere under that curve.

Starting from the mean of 100, one standard deviation below the mean is 85 and one standard deviation above the mean is 115.

According to the empirical rule, 68.26 percent of everyone everywhere will score within 85 and 115 on the Wechsler.  Two standard deviations below the mean is 70 and two standard deviations above the mean is 130 and 95.44 percent of all people will score between 70 and 130.  Three standard deviations below the mean is 55 while three standard deviations above the mean is 145 with 99.72 percent of people obtaining scores from 55 to 145.  Four standard deviations below the mean is 40 and four standard deviations above the mean is 160 and fully 99.98 percent of all people will receive scores between these two values.  We often think of where 95 percent of people are and that is within 1.98 standard deviations above or below the mean, while 99 percent of all people are within plus or minus 2.58 standard deviations of the mean.

There is a person who is said to have an IQ of 228, and this is based on an old formula of dividing mental age by chronological age and multiplying by 100, but this person was ten years old at the time, which will give an inflated score and this is an outmoded, dubious method in any event.

With a mean of 100 and a standard deviation of 15, an IQ of 228 is 8.53 standard deviations above the mean and this corresponds to "one in I do not know how many" (but several) hundred trillion people, or more people than have ever lived on this Earth, so it is highly unlikely that this score is accurate.

The current Wechsler test really doesn't work above a score of 155, though older versions worked to 160, so don't count on obtaining a score in the 180s, though there are novelty IQ tests that purport to measure IQ scores at least this extreme.  I don't believe it.

I do, however, believe that there is this thing called intelligence, and I believe that having it comes in handy when addressing the vexations of modern life.  That said, I think that intelligence is one of those things where you know it when you encounter it, but it is so fluid and changing that it cannot be rigorously defined.

The law in this country dictates that everyone must have access to a free and appropriate public education, hence the widespread availability of special education based on IQ.  But there are limits.

The Army has found that it cannot train people with IQs under about 80 to perform the tasks a soldier will have to do and so, such people are rejected from military service and that is a hard and fast rule.  What is frightening is that this is about one in 11 people (actually 9.12 percent of the population), and if we accept the reasonable premise that the military is similar to society in general, then this means that about one in 11 people is unable to handle our modern technological society on the basis of an IQ score.

The problem is that in our society which is so interested in measuring things, we want to measure even that which is unmeasurable because it cannot be adequately defined.  Too many people tend to place too much faith in those numbers and feel that being a 145 is somehow better than being a 128 when, in truth, being a 128 will get you through life quite well, while being a 145 can actually be an impediment.  If that was as far as it went, it would be fine, but like your credit score, that number can determine how far you go in life by opening up educational and employment opportunities that would otherwise be unavailable to you, and this is why you should be aware of how intelligence tests work and how they can be defeated.

I assure you that if you are conscientious about the test when you are initially taking it, test/retest reliability does not apply; you will score ten points higher on a retake of the test and, while there will be diminishing returns on subsequent retakes, you will soon be in control of the test, rather than the other way around.