The same principle applies to scientific research. Any individual study might be affected by inadvertent biases or errors that distort its results, leading to an imprecise estimate of the true effect (here, how much your IQ increases after listening to Mozart). By averaging across a number of studies, though, any random errors that led to over-or underestimates of the size of an effect will tend to average out, leaving a better estimate of the truth. Because they are based on all of the relevant studies, the results of a meta-analysis are not unduly influenced by a single memorable or well-publicized finding, such as the original Rauscher-Shaw article.
After scouring scientific journals for experiments like the original one, Chris noticed that—aside from Steele’s article in Psychological Science—all of the followup studies were published in journals that most researchers never read, and many have never even heard of. He wrote to the authors of many of the articles to request additional data or information he needed to evaluate their results. In total, he found sixteen experiments (including the original) that tested the Mozart effect and were published in peer-reviewed scientific journals. All of them used the same sonata and compared it with silence, relaxation, or both. For each experiment, Chris calculated the size of the difference in performance between those subjects who had listened to Mozart and those who had not. When compared with silence, Mozart improved performance by the equivalent of 1.4 IQ points, only one-sixth as much as the Rauscher-Shaw team had found. For experiments comparing the sonata with relaxation, the advantage for Mozart turned out to be three IQ points, about a third as much as the original article reported, but still twice as large as in the comparison between Mozart and silence. There may be good reason for this small benefit: Relaxation reduces anxiety and arousal, but being in a “laid-back” state is not ideal for solving difficult problems on IQ tests. Nor is being excessively anxious, of course—a happy medium is best. Compared with relaxation, sitting in silence likely has a similar, but weaker effect—without external stimulation, your mind may wander, making you less prepared for hard work.
Chris concluded that the entire “Mozart effect” might have nothing to do with a positive effect of listening to music. Rather than Mozart making you smarter, sitting in silence or getting relaxed might make you dumber! Viewed this way, Mozart’s music is a control condition that resembles the general level of mental stimulation we encounter during everyday life, and silence and relaxation are “treatments” that reduce cognitive performance. In either case, though, there is little or no Mozart effect to explain.
Several additional studies could not be included in Chris’s meta-analysis because they did not include the relaxation or silence control conditions. However, they did reveal another possible explanation for the apparent benefit of Mozart. In one, British researcher Susan Hallam arranged for the BBC to conduct a massive experiment on eight thousand children in two hundred schools around the United Kingdom. The children listened to either a Mozart string quintet, a discussion about scientific experiments, or three popular songs (“Country House” by Blur, “Return of the Mack” by Mark Morrison, and “Stepping Stone” by PJ and Duncan), and then performed cognitive tests like those originally used by Rauscher. The children who listened to popular music did the best, and there was no difference in performance between those who listened to Mozart and those who heard the science discussion. An article on this finding cheekily dubbed it the “Blur Effect.”22
A second study by Kristin Nantais and Glenn Schellenberg of the University of Toronto found no overall difference in cognitive task performance after listening to the Mozart sonata or the short story “The Last Rung on the Ladder” by Stephen King. But subjects did do better after listening to what they liked best.23 The most sensible explanation for this finding, as well as for the “Blur Effect,” is that your mood improves when you hear what you like, and you do modestly better on IQ tests when you are in a better mood. The effect has nothing to do with increasing your intelligence per se.
Chris submitted his meta-analysis to Nature, the journal that published the initial 1993 article. He did not expect the editors to accept it, because its conclusion—that any small benefits that do exist result from arousal and positive mood rather than any special property of Mozart’s music—could be interpreted as questioning the journal’s decision to publish the first paper. To his surprise and delight, they accepted the paper and published it in August 1999 alongside another report of a failure to replicate by Kenneth Steele and his colleagues. Rauscher was given space to reply, and Nature highlighted the exchange in its weekly press bulletin. The media, loving a good fight, even among staid academics, sprang into action: Chris was interviewed for CNN, CBS, and NBC news programs. Rauscher and Steele debated on the Today show, with Matt Lauer as referee. Chris’s article even earned him a short appearance on an episode of Penn and Teller: Bullshit! entitled, charmingly, “Baby Bullshit.”
Recall the media analysis done by Adrian Bangerter and Chip Heath. They found a spike in coverage of the Mozart effect in 1999, coincident with these articles in Nature, and then things died down again. Did Chris’s meta-analysis, and the studies by Steele and Schellenberg, finally debunk the Mozart effect? Yes and no. Bangerter and Heath found that news articles mentioning the positive effect of listening to Mozart for adults became less and less frequent, but that articles falsely claiming that Mozart made babies smarter became more common! Indeed, this trend started just one year after the original Rauscher-Shaw report. To be clear, we repeat that no published studies had ever examined the effect with babies!24 Our national survey of fifteen hundred adults was conducted in 2009, ten years after Chris’s meta-analysis was published. It found that 40 percent of people agreed that “listening to music by Mozart will increase your intelligence.” A majority disagreed, but keep in mind that the scientific evidence does not support this claim at all. It would be better if almost everybody disagreed, as they would with a statement like “on average, women are taller than men.”
Indeed, the Mozart effect still resonates with many. Eric Mangini must have been a believer in 2007 when he made classical music the new workout soundtrack for the New York Jets. Until we each had our first child, we didn’t realize the extent to which the Mozart-for-babies myth has permeated the child-care industry. Intelligent, highly educated friends sent us toys that included—as a matter of routine, not a special feature—a “Mozart” setting that played classical music. The Baby Einstein company was founded in a basement with $5,000 in capital in 1997 (hot on the heels of the initial burst of Mozart effect publicity) and grew to sales of $25 million in 2001 before it was acquired by Disney.25 The names of its DVDs—Baby Mozart, Baby Einstein, Baby Van Gogh, and so on—imply that by watching them, your child will become more like a genius and less like an ordinary baby. Videos designed to be watched by babies are now a $100-million-a-year business,26 even though the American Academy of Pediatrics currently recommends that children younger than two years old watch no television or videos whatsoever.
A research group led by Frederick J. Zimmerman, a pediatrician at the University of Washington, attempted to test the effect of the products inspired by the Mozart effect on children’s cognitive abilities. The researchers commissioned a telephone survey of parents of children less than two years old in the states of Washington and Montana. Each parent answered a series of questions about how much time his or her child spent watching educational television, movies, and other media, with a separate category for “baby DVDs/videos.” Later in the survey, the parents were asked whether their children understood and/or used each of ninety words typically found in the vocabularies of young children. There were separate vocabulary lists for infants (age 8–16 months) and toddlers (age 17–24 months), so the researchers looked at these age groups separately. For the infants, each additional hour per day spent watching baby DVDs was associated with an 8 percent reduction in vocabulary. For the toddlers, there was no significant relationship between DVD viewing and vocabulary size.27
If you have become sensitive to the illusion of cause that we discussed in Chapter 5, you will notice that this is just a correlational study. The researchers couldn’t randomly assign some babies to watch videos and others to not watch videos, so a headline of “Watching Baby DVDs Will Make Your Child Dumber” is not justified. The family environments of infants who watched more videos might be less conducive to vocabulary building in other ways. In their statistical analysis, Zimmerman and his colleagues accounted for some of the most likely factors that could make the DVD-watching children different, such as how much education their parents had, how much their parents read to them, how much other media they watched, whether they watched alone or with their parents, and so on. Even after all of those factors were accounted for, DVD watching was still associated with smaller vocabularies. Although we cannot make a strong causal inference from this study, it certainly provides no support for the belief that watching videos or listening to Mozart improves cognition.
Disney, which was getting $200 million in annual revenue from the Baby Einstein brand when the Zimmerman group published its article, reacted sharply. Its CEO, Robert Iger, publicly criticized the study as “flawed” for not differentiating between different baby DVD products, implying that other DVDs might lead to smaller vocabularies, but not those made by his company.28 A Disney spokesman pounced on a statement by one of Zimmerman’s coauthors, who told a newspaper that the study had found “harm” to children’s vocabularies from baby DVDs. The company had a point here: As we have noted, the study was correlational, not causal, so strictly speaking, harm was not found.
Unfortunately, Disney’s spokesman undermined his defense of scientific rigor by making an even more fallacious argument himself: “‘Baby Einstein’ has been so well-received, and if properly used, they do have an impact on infants’ health and happiness.”29 In other words, the product must be good for kids because it has been “well-received” (presumably by parents, many of whom might be understandably grateful for something that absorbs the attention of a crying baby for a few minutes, and who would like to believe that a product they spent money on with good intentions really did benefit their child). The spokesman offered no evidence, either correlational or causal, to support his claim that using the DVDs “properly” is beneficial.
In the end, Eric Mangini’s own Mozart experiment did not succeed. In 2006, he had guided the Jets to a 10–6 record and a playoff appearance. He added classical music to the practices the next season, and his team went 4–12. Mangini lasted just one more year as the Jets’ head coach before being fired.30
What Lies Beneath
Why does the Mozart effect find such a ready audience? Why do so many people buy classical CDs for their infants and DVDs for their toddlers? Why are people so willing to believe that music and videos can effortlessly raise their children’s IQs? The Mozart effect masterfully exploits the illusion of potential. We all would like to be more intelligent, and the Mozart effect tells us that we can become more intelligent just by listening to classical music. The subtitle to Don Campbell’s book The Mozart Effect directly appeals to the illusion: Tapping the Power of Music to Heal the Body, Strengthen the Mind, and Unlock the Creative Spirit.
We already mentioned that 40 percent of people still believe in the Mozart effect, despite the scientific evidence against it. Lest you think that this is just a silly belief that has no real importance, consider some of the implications. Parents holding this belief might think that they are doing just as much, if not more, for their children by sitting them in front of a baby DVD or playing classical music than by interacting with them. Daycare centers, schools, and other institutions might follow suit. The fad of playing Mozart to babies could substitute for much better practices, ones that might actually help the social and intellectual development of children. In other words, a belief in the Mozart effect might make children worse off than they would have been otherwise, as suggested by the Zimmerman group’s study of baby DVDs.
If such a sizable number of people continue to believe in the Mozart effect despite its debunking, what about other beliefs in hidden mental powers that have not received as severe a public lashing as the Mozart effect? In our national telephone survey, we asked several questions that touched on other manifestations of the illusion of potential.
Sixty-one percent of our respondents agreed that “hypnosis is useful in helping witnesses accurately recall details of crimes.” The idea that hypnosis can put the brain into a special state, in which the powers of memory are dramatically greater than normal, reflects a belief in a form of easily unlocked potential. But it is false. People under hypnosis do generate more “memories” than they do in a normal state, but these recollections are as likely to be false as true.31 Hypnosis leads them to come up with more information, but not necessarily more accurate information. In fact, it might actually be people’s beliefs in the power of hypnosis that lead them to recall more things: If people believe that they should have better memory under hypnosis, they will try harder to retrieve more memories when hypnotized. Unfortunately, there’s no way to know whether the memories hypnotized people retrieve are true or not—unless of course we know exactly what the person should be able to remember. But if we knew that, then we’d have no need to use hypnosis in the first place!32
Seventy-two percent of people agreed that “most people use only 10 percent of their brain capacity.” This strange belief, a staple of advertisements, self-help books, and comedy routines, has been around so long that some psychologists have conducted historical investigations of its origins.33 In some ways, it is the purest form of the illusion of potential: If we use only 10 percent of our brain, there must be another 90 percent waiting to be put to work, if we can just figure out how. There are so many problems with this belief that it’s hard to know where to begin. Just as some laws cannot be enforced because they are written too imprecisely, this statement ought to be declared “void for vagueness.” First, there is no known way to measure a person’s “brain capacity” or to determine how much of that capacity he or she uses. Second, when brain tissue produces no activity whatsoever for an extended time, that means it is dead. So, if we only used 10 percent of our brain, there would be no possibility of increasing that percentage, short of a miraculous resurrection or a brain transplant. Finally, there is no reason to suspect that evolution—or even an intelligent designer—would give us an organ that is 90 percent inefficient. Having a large brain is positively dangerous to the survival of the human species—the large head needed to contain it can barely exit the birth canal, leading to a risk of death during childbirth. If we used only a fraction of our brain, natural selection would have shrunk it long ago.
This “10 percent myth” surfaced long before brain-imaging technologies like MRI and PET scanning even existed, but misunderstandings of neuroscience research might reinforce it. In the pictures of brain activity (“brain porn”) that appear in media reports about neuroscience research, large areas of the brain are dark, or not “lit up” with blobs of color. However, the blobs don’t indicate the “active” areas of the brain—they indicate areas that are more active in one situation or group of people than in another. For a neurologically normal person, the entire brain, including the dark areas, is always “on,” with at least a baseline level of activity, and any task you can perform will raise activity in many brain areas. So, needless to say, “using more of your brain” will not help you avoid everyday illusions.
Sixty-five percent of people apparently believe that “if someone behind you is staring at the back of your head, you can sense that they are looking at you.” Although it would be nice if we could reach out and touch someone with our eyes, our eyes do not emit any such rays, and there are no receptors in the back of our head that can detect someone’s stare. This false belief rests on the idea that people have hidden, previously unmeasured perceptual abilities that function independently of our standard five senses, and that this sixth sense can prove usef
ul. The idea has been thoroughly debunked, though. A prominent psychologist named Edward Titchener wrote, in the journal Science, “I have tested this … in a series of laboratory experiments conducted with persons who declared themselves peculiarly susceptible to the stare or peculiarly capable of ‘making people turn round’ … the experiments have invariably given a negative result.”34 We can’t make people turn around by looking at them, and we can’t tell when someone else is looking at us, at least not without first looking back at them.35
Why would people come to believe in such extrasensory perception? We tend to remember those cases when we turned around and saw someone, but not those cases when we turned around and nobody was there (nor the times when someone was there and we didn’t notice, and certainly not the “times” when nobody was there and we didn’t notice anyone). Recall from Chapter 5 that we are also prone to infer a causal pattern when the sequence of events is consistent with a narrative. If you start staring at someone and then they happen to turn around, the illusion of cause would lead you to the false inference that you caused them to turn. And when you infer a cause, you are especially likely to remember it.
Since it was utterly obvious to him that people actually couldn’t feel the stares of others, Titchener felt the need to explain why he bothered to conduct studies to debunk the idea in the first place. He noted that the experiments “have their justification in the breaking-down of a superstition which has deep and widespread roots in the popular consciousness.” He was absolutely right about the prevalence of the “sixth sense” belief. Unfortunately, Titchener’s attempts to eradicate this superstition through experimentation were ineffective.36 The prevalence of the false belief about feeling the stares of others has been remarkably stable over time—Titchener’s article in Science was published in 1898.
The Invisible Gorilla: And Other Ways Our Intuitions Deceive Us Page 23