Walt Disney World was a curious choice for the 1998 meeting of Rent-A-Center managers. The giddy whirl of theme-park rides and the melting heat of a Florida September made an odd setting for talk about the rent-to-own couches and coffee tables that are the company’s stock in trade.
Things got even stranger early one morning when the managers were herded into a chilly conference room. They had been scheduled to participate in a budgeting workshop, but company leaders instead announced an abrupt change in plans. Employees would immediately begin taking the Management Test, a five-hour battery of nine separate examinations.
“We thought, ‘Uh-oh, this must be that test we’ve been hearing about,’” recalls Art Staples, then a manager of several San Francisco–area stores. Rent-A-Center had been bought by another firm just a month before, and word had gotten out that the new company required all employees to take a long and demanding test. “There was nothing we could really do except refuse to take it, walk out of the room—and find our own transportation back to California,” Staples says.
The managers knew, too, that their scores on the test could determine the course of their careers at Rent-A-Center—or even whether they had jobs at all. “We had been told that if you did not pass the test, then you would not be allowed to work in management,” says Scott Hadley, another store supervisor from the Bay Area. “We were afraid that if we failed, we would be let go.”
An anxious hush fell over the room as the exams were passed out. Within minutes, however, the silence was breached by a stir of astonishment. “People were looking around at each other with this expression of ‘You’ve got to be kidding me,’” Staples recalls. The questions in front of them had nothing to do with renting furniture, or managing employees, or keeping the books.
“My sex life is satisfactory.” “I have diarrhea once a month or more.” “I would like to be a florist.” “Everything tastes the same.” “My mother was a good woman.” “I am a special agent of God.” Arriving at the question “I liked Alice in Wonderland, by Lewis Carroll,” the managers might well have felt that they had slipped down the rabbit hole. The test was the MMPI—the Minnesota Multiphasic Personality Inventory, a personality test created decades earlier at a Midwestern mental hospital.
Hadley and Staples were offended and angered by the personal nature of the items. “These questions had no relation to what we did for a living, to whether we were good managers or not,” Hadley says. “Our employer just didn’t need to have all that information.” He was particularly struck by an item that read, “I like tall women.” “How was I supposed to answer that?” he asks. “My wife is five foot three.”
Things only got worse when the test’s results were returned. Rent-A-Center hired an outfit in Kansas, Associated Personnel Technicians, to produce an extensive personality description of each manager: how self-sufficient, self-confident, and dominant he or she was, how fearful, depressive, and impetuous. Most needed work in one or more of these areas, APT advised, and so each profile included a specific plan for improvement. Employees who were “restless and impatient” should lay off nicotine and caffeine, they counseled; workers with the “tendency to retreat from reality” should drink lots of water.
APT prescribed reading material, too: managers who were found to be “independent and hard-headed” should take in How to Read a Person Like a Book, while those “resistant to rules and codes” should review Your Erroneous Zones and Codependent No More. These self-help books, the anonymous sages at APT directed, should be read “one chapter at a time, with a significant person in his life. After reading and discussing each chapter, he needs to write down what he has learned that is new to him and what he is actually doing differently, then submit these two sentences per chapter to his training supervisor for further discussion.”
Of Scott Hadley, APT concluded: “He makes assumptions without checking them out. We suggest he learn to ask more of the ‘how’ type questions to validate his assumptions and expectations and avoid the ‘why’ type questions.” Hadley was also instructed to read I Ain’t Much, Baby—But I’m All I’ve Got. “Stupid pop psychology,” says Hadley, who declined to do his required reading. “If I actually needed help I would go to a professional, not read some book I could buy at Wal-Mart.”
For his part, Art Staples was admonished to write out daily “self-affirmations” to build his self-confidence. He rather confidently refused. “I thought I was a pretty good person and had been pretty successful,” he says, “and if they felt that they could somehow improve me—you know what, thank you, but I like the way I am.”
The Personal Data Sheet—created in 1917 by Columbia psychology professor Robert Woodworth—was perhaps the world’s first “personality inventory”: a written test in which individuals were asked to respond to questions about their psychological state. Woodworth designed the test in response to urgent messages cabled home by American military commanders in Europe during World War I. The horrors of a new kind of warfare—powerful explosives, poison gas, trench fighting—were breaking down their men’s minds and bodies. Soldiers were suddenly blinded, or lost the ability to smell or taste. Whole chunks of their memories went missing. They forgot how to walk. Some stared mutely, unable to speak. Others seemed to become entirely different people: silent men turned garrulous, prudent ones turned careless; polite men became coarse, aggressive ones became meek. Doctors diagnosed these as cases of “lost personality.”
Working quickly, Woodworth composed a list of 116 questions: “Does the sight of blood make you sick or dizzy?” “Do you nearly always feel that you have strength or energy enough for your work?” “Are you often frightened in the middle of the night?” A quick count of nos and yeses provided an index of “neuroticism,” thought to be a crucial predictor of shell shock. Early tryouts on new conscripts and returning veterans were promising, but peace intervened before it could be put to wider use. Other tests soon followed: tests for maladjustment, for introversion, for dominance. They formed a starting point for the efforts of a University of Minnesota psychologist named Starke Hathaway.
Hathaway was a born tinkerer who always assumed he would become an engineer. When he arrived at Ohio University in 1923, however, he found to his surprise that he was more interested in the delicate machinery of the human mind. Taking up the study of psychology, he saw that it was a disjointed discipline, in need of some tweaking and tightening. After graduation he proceeded to the University of Minnesota, where he acquired a PhD in psychology and then a position at the school’s mental hospital.
As Hathaway began working with patients, he soon discovered that the available personality tests were hopelessly inadequate. To begin with, most of them focused on just one dimension, while he needed an instrument that would survey the whole spectrum of disorders. For another thing, their questions were laughably transparent. The Personal Data Sheet, for example, made inquiries like “Do you usually feel well and strong?” and “Are you happy most of the time?” The test had worked well enough in wartime, when, as Hathaway noted, “men who were afraid and who considered themselves unfit for war were offered symptoms that permitted them to admit the fact.” But when peace returned, so did the vanity, self-consciousness, and shame that made test-takers reluctant to confess such weaknesses. Without imminent combat to coerce honesty, something more subtle was called for.
But perhaps the most puzzling deficiency of these tests was how often they were simply wrong. Most had been developed in the same straightforward fashion: a psychologist had decided which questions to ask and which answers indicated abnormality. Such decisions were really no more than educated guesses, and not infrequently, these guesses were mistaken. Robert Bernreuter was a Pennsylvania State University psychologist who designed one of the era’s most popular tests, the Bernreuter Personality Inventory. In it he posed the question “Are you critical of others?”, surmising that a positive response suggested psychological difficulties. As it turned out, more than two-thirds of normal people taking the test answered “yes,” a reply given much less frequently by neurotics and psychotics. Likewise, the query “Do you daydream frequently?”—assumed by Bernreuter to be a sign of psychological instability—was endorsed more often by healthy individuals than by the mentally ill.
People, it seemed, had the inconvenient habit of contradicting psychologists’ expectations. Hathaway wanted an instrument that didn’t let ideas about human nature get in the way of its reality, and since he couldn’t find the tool he needed, he set out to make it himself. His test would be a machine fueled by solid facts, not by the thin vapors of theory and conjecture.
In this project, Hathaway had a partner: University of Minnesota neuropsychiatrist J. Charnley McKinley. The pair began by collecting as many symptoms of mental illness as they could find. They trawled psychiatric textbooks.
They pored over case studies. They quizzed their colleagues. And they borrowed items from other personality tests, lifting several straight from the Personal Data Sheet. In a move that would cost them much trouble later on, McKinley insisted that they also include questions of a strictly medical nature. Before long the two men had amassed over a thousand potential questions for their test; by eliminating redundant items, they pared the list to about half that.
Then Hathaway and McKinley did something revolutionary. Instead of assuming that they already knew how normal and abnormal individuals would answer their queries, they let the people decide. In this republic of the insane, citizens were allowed to vote—with their fears, their neuroses, and their obsessions. As Hathaway put it: “We permitted the patients to design their own test.”
Using large groups of subjects in psychological research was itself a relatively new phenomenon. In the discipline’s early days, experiments frequently involved just one participant: the investigator himself. His method was that of introspection, of observing and recording his own mental processes. Even after individual differences among people became a point of interest, research was usually conducted with only a handful of subjects, often individually identified by name or initials. Introspection remained the method of choice, and many scientists believed that participants had to be highly trained (generally, psychologists themselves). In 1893 Edward Bradford Titchener, a leading experimental psychologist, insisted it was folly to believe that “since we are all using our minds, in some way or another, everyone is qualified to take part in psychological experimentation. As well maintain, that because we all eat bread, we are all qualified to bake it.”
This approach began to change early in the twentieth century, when advances in statistical procedures—and the need to sort and manage large numbers of people—led to the notion of a comparative norm. Now an individual’s performance was of interest only in relation to the averaged scores of others, a generalized standard one could match or fail to meet. Information was gathered not through introspection but objective measurement, and the subjects of experiments need not be sophisticated self-observers, only naïvely natural participants. In 1904, the American psychologist James McKeen Cattell delivered a cheeky retort to Tichener: “It is usually no more necessary for the subject to be a psychologist,” he said, “than it is for the vivisected frog to be a physiologist.”
For Starke Hathaway, of course, abnormal subjects were easy to come by. Relying on his own diagnostic judgment, he selected groups of patients with “pure,” uncomplicated cases of particular disorders: first hypochondriacs, later depressives, psychasthenics, hysterics, psychopathic deviates, hypomanics, paranoiacs, and schizophrenics. (Some of these classifications are no longer in use or go by other names.) But where to find a sufficient number of normal people?
Hathaway hit upon the idea of using the friends and family members of patients at the university’s regular hospital as a control group, and soon he and McKinley were “testing visitors and anybody else we could get hold of who came into the hospital and said he was not under a doctor’s care.” By trolling hallways and lobbies, Hathaway found 724 presumably sane subjects for his research.
The members of his scavenged comparison group had a few things in common with each other: They were all Minnesotans. They were all white. Most were Protestant, and many were of Scandinavian descent. They ranged in age from sixteen to fifty-five, with an average age of thirty-five. Almost all were married, and many were parents. The majority were rural people, employed as farmers, blue-collar workers, and housewives. They had an average educational level of eighth grade. This rather small and provincial group soon came to be called the “Minnesota Normals,” and though no one could have guessed it at the time, they were to form psychology’s major benchmark of normality for the next fifty years.
Now Hathaway had his subjects, normal and abnormal, and he had his long list of symptoms. All that remained was to put the latter to the former and find out which items “discriminated”: which questions effectively distinguished the Normals from the psychiatric patients. The answer (“true” or “false”) given by the majority of Normals to a particular question became, ipso facto, the normal answer; the answer offered by the majority of, say, depressives became, ipso facto, a sign of depression. The items that discriminated best—that produced the cleanest divisions between normal and abnormal—were the ones included in the questionnaire.
This was a radically new way of creating a test, and its results were often confounding. Early personality tests like the Bernreuter Personality Inventory and the Personal Data Sheet were “face-valid,” meaning the intent of the questions was clear on their face. Test-givers, and more importantly test-takers, understood what was being asked and why. Not so with this test. Even its creators were often at a loss to understand why the normal answer to one question should be “false” instead of “true,” or why a second question seemed to tap one form of mental illness rather than another. “Frequently the authors can see no possible rationale to an item in a given scale,” Hathaway and McKinley noted in an early paper, adding stoically, “it is nevertheless accepted if it appears to differentiate.”
This aspect of the test developed at Minnesota was to inspire much controversy in coming years, not only among the public but among Hathaway and McKinley’s professional colleagues. One critic assailed their “who-gives-a-damn-why-it-works attitude”; another, the prominent psychologist Gordon Allport, denounced what he called “galloping empiricism,” which “dashes forth like a headless horseman.” Such an approach, he protested, “has no rational objective; uses no rational method other than mathematical; reaches no rational conclusion. It lets the discordant data sing for themselves.”
To Hathaway, however, the “discordant data” sounded sweeter than any mellifluous strains of theory. His test was the epitome of “dustbowl empiricism,” the just-the-facts style of science practiced in the nation’s sturdy Midwest. Let the New York psychoanalysts have their tortuous theories about Oedipal complexes and death drives; Hathaway just wanted something that worked.
But for all the plain pragmatism that went into its development, the test Hathaway and McKinley ultimately produced was without a doubt one of the weirdest creations in the history of man’s attempts to understand himself. One observer has puckishly, though accurately, described it as “a Joycean soliloquy in Whitmanic rhythms, the interior monologue of a neurotic modern Everyman.”
There was the content of the items (written as bald statements, to be countered “true,” “false,” or “cannot say”). McKinley’s insistence on the inclusion of medical topics meant there were lines such as “I have never had any black, tarry-looking bowel movements” and “I have had no difficulty starting or holding my urine.” There were questions about sex: “I have never indulged in any unusual sexual practices,” “There is something wrong with my sex organs.” There were questions about religion: “I believe there is a Devil and a Hell in the afterlife,” “Everything is turning out just like the prophets of the Bible said it would.” And there were questions that were simply odd: “I think I would like to belong to a motorcycle club,” “Often I feel as if there were a tight band around my head.”
Even stranger than the content were the items’ flat, affectless tone and their careless alternation of the weightiest subjects with the most banal. “I loved my father.” “I like to flirt.” “I believe my sins are unpardonable.” “I have a good appetite.” The items seemed stripped of context, as though one were overhearing a conversation on Mars: “I think Lincoln was greater than Washington.” “Women should not be allowed to drink in cocktail bars.” Some statements sounded pious and prim: “A large number of people are guilty of bad sexual conduct.” Others sounded almost frisky. “If the money were right, I would like to work for a circus or carnival.” Altogether, the test—504 items long—presented its takers with an often distasteful, often tedious, always bewildering experience.
At the time, no one worried much about how peculiar and potentially offensive the test might be. This was a minor venture, designed specifically for mental patients, intended for use at a thirty-eight-bed hospital in snow-swept Minnesota. But that was all about to change.
The test’s formal introduction to the world was not auspicious. Hathaway and McKinley gave it a long, awkward name: “Minnesota,” because that’s where it was made, “Multiphasic,” because it measured a number of different dimensions. (“I remember being in McKinley’s office one day,” trying to come up with a term to describe this feature of the test, said Hathaway. “We scrounged around and finally came up with this name, ‘multiphasic,’ which is a Greek-Latin bastard, and that satisfied us for the moment.”) Lastly, “Personality Inventory,” because it was meant to catalogue all relevant aspects of abnormal personality. The resulting mouthful—the Minnesota Multiphasic Personality Inventory—was almost immediately shortened to MMPI.
The MMPI had to overcome obstacles other than its name. Hathaway had chosen an unconventional format for the test: instead of listing the items on a sheet of paper, he had typed each one on a three-by-five-inch slip, creating something like a deck of cards. Test-takers were instructed to sort the deck into three piles according to their answers: “true,” “false,” or “cannot say.” Hathaway had speculated that people would be more willing to admit to embarrassing thoughts or practices if they didn’t have to put it in writing, but many users found the card format confusing and cumbersome.
Perhaps most discouraging for the test’s prospects, the country was still emerging from the Great Depression, and few companies had the resources or the inclination to reproduce Hathaway’s strange invention. He and McKinley approached three major test-publishing firms about distributing the MMPI. All three declined (“unduly awkward and impractical,” demurred one). Finally, the University of Minnesota Press agreed to take it on, if its authors would put up half the money. Hathaway and McKinley found a patron to contribute the requisite sum, and in 1942 the presses started to roll—at the university printing shop. “We were beggars, not choosers,” Hathaway wryly observed.
But their test—what Hathaway and McKinley called “an easily applicable measuring device”—soon piqued the interest of psychologists and psychiatrists around the country. There were a number of reasons for its unexpected appeal, beginning with the fact that there were few other options available to test-givers who wished to survey the entire personality. The MMPI had distinct advantages over the Rorschach inkblot test and over earlier personality questionnaires: its relentless emphasis on quantification avoided the taint of unscientific artistry that still clung to the Rorschach, and its scrupulously empirical construction made it more valid and reliable than tests based on clinicians’ best guesses.
The MMPI’s major innovation, however, were the clever devices Hathaway had built into the test to detect false or careless responses. A common charge made against self-report questionnaires (especially by Rorschach proponents) was that these tests were easily faked. After all, the critics demanded, would we trust an intelligence test that asked people how smart they were? In response, Hathaway designed special scales—sets of questions with separate scoring keys—that would alert the test-giver to a possibly duplicitous subject.
Actually, Hathaway had long had an interest in uncovering deception. As a senior in college, he devised a rudimentary lie detector and “got a little notoriety,” he recalled, when he used his machine to help find a thief who’d stolen a diamond ring. He became a consultant to the police in Athens, Ohio, assisting them on a murder case, and later worked with law enforcement in Minneapolis and St. Paul to develop psychological profiles of at-large criminals.
The first of the traps Hathaway laid in the MMPI was the Feeblemindedness Scale, meant to identify test-takers who could not read the items or who were responding randomly. Individuals who offered answers different from an overwhelming majority—at least 90 percent—of the Minnesota Normals could be identified by an elevated score on this measure. Answering “true” to “Evil spirits possess me at times,” for example, would raise a red flag (but so would responding “false” to “I believe in law enforcement”).
Even more ingenious was the Lie Scale, intended to foil those who would deliberately place themselves in a flattering light. Hathaway included in the test a number of items like “I do not always tell the truth,” and “I gossip a little at times.” A truly conscientious person might honestly answer “false” to a higher-than-average number of these items, noted Hathaway and an associate, “but for a person to have six or eight of them seemed almost impossibly good.” (Interestingly, high scores on the Lie Scale are not infrequently obtained by members of the clergy—who might really be saintly, or who might feel the need to present themselves that way.)
Most sophisticated of all was the Correction Scale, which came about after Hathaway noticed that obviously disturbed patients occasionally generated perfectly normal results. With a student, Paul Meehl, he developed a scale to detect “defensive” responding by mentally ill people wily enough to answer as if they were sane.
With such fine tuning, Hathaway succeeded in building a better test than any other available. Though he largely declined to promote it, his graduate students were not so modest. As they departed the university to take teaching jobs all over the country, they carried with them their faith in the “Minnesota Way”: quantified, empirical, objective. If the “Multers,” as they called themselves, lacked the cultish fervor of the Rorschachers, they did have an unshakable confidence that science was on their side.
Early on, the test also received midcentury America’s ultimate salute: a story in Time magazine. “By using the inventory, a physician may know a new patient, in an hour or two, as well as his family doctor does,” exclaimed the article, which also featured a photograph of Hathaway and his trademark fluff of Strangelovian hair. (“This picture aroused a certain amount of amusement among my colleagues around here,” Hathaway remarked dryly. “Those who knew me best said, ‘Why did he slick his hair down so?’ and those who didn’t know me at all said, ‘Why didn’t he comb his hair?’”) The newsmagazine noted that the MMPI “is already the most widely popular test for its purposes… in use in universities, penitentiaries, corporations, [and] clinics.”
Three years later, in 1946, a medical journal reported that the MMPI “is used routinely by hundreds of private clinics and individual doctors; it is part of the personnel procedure in some of our largest corporations; it was used by individual medical and psychological personnel in all theaters of war… [and] it is used today in all veterans’ administration medical clinics.” Demand for the test soon grew so intense that Minnesota’s press couldn’t keep up. “The university printing office was doing nothing but Multiphasics. The whole doggone place was covered with them,” Hathaway marveled. “A new producer had to be found.” In 1947, just five years after Hathaway and McKinley had gone begging, the distribution of the MMPI was assumed by The Psychological Corporation, one of the country’s largest test-publishing companies. By then, Hathaway noted with some satisfaction, “the Corporation was glad to get what they had turned down earlier.” Within fifteen years, the MMPI would become the world’s most widely used objective personality test.
In a sense, however, the MMPI that achieved such popularity was not the same test that Hathaway and McKinley designed. In the years following its initial publication, a subtle but important transformation took place, a shrewd shift that obscured a simple fact: judged on its own terms, the MMPI was a failure.
The original purpose of the MMPI, of course, was to sort a group of mental patients into diagnostic categories. A high score on the Depression scale would indicate the presence of depression; a high score on the Schizophrenia scale would point directly to schizophrenia. If a test-taker’s responses were plotted on a graph, Hathaway expected to see a large spike on a single dimension. But that’s not what happened. As the MMPI became more widely used, psychologists began reporting that their patients showed such elevations on several scales at once, and that normal people were often elevated as well. Hathaway’s sleek machine was looking more like a lemon.
By now, however, the MMPI had a core of committed users—many of them Hathaway’s former students—and they were determined to get their test up and running again. Perhaps there was meaning, they suggested, in the patterns of scores that test-takers produced. Such patterns would describe not simple categories but complex syndromes, and would therefore be more accurate than Hathaway’s original groupings. A test taker who rated high on both the Psychopathic Deviate and the Hypomania scales, for example, was a very specific sort of person: aggressive and impulsive, with a careless conscience and an inclination to bend the rules. Likewise, the individual who achieved elevations on the Depression and the Hysteria scales was something more than the sum of these two conditions: she was likely to be moody, insecure, and dependent, with a sensitivity to criticism and a tendency to play the martyr. This new use of the old MMPI provided not an elementary diagnosis but a comprehensive profile.
As the approach gained ground, an even grander possibility came into view: what if the MMPI could be used to describe normal personality types? The scores of normal test-takers, though less elevated, also fell into distinctive patterns. By this measure, every one of us was an assemblage of ailments, mild or severe—some composite of tendencies toward mania, hysteria, paranoia, depression. A test that had been designed to spot pathological extremes of behavior was now expected to sort out the much subtler differences among normal people.
Soon MMPI users abandoned the original names of the scales, replacing them with numbers: the Hypochondriasis scale became Scale 1, the Depression scale became Scale 2, and so on. A neat shorthand was thereby created: a person could now be referred to as a “4-9” or a “2-3,” the numbers summing up the personality with brisk economy. Minnesota graduate students and others trained to think in terms of the MMPI found its abbreviations becoming a reflex, a language to describe not only patients but acquaintances, friends, and family. Guides to these numerical sequences and what they meant, called “codebooks” or “cookbooks,” began to be compiled. Paul Meehl, once Hathaway’s student and now his colleague, applauded the fact that at last personality description could become “an automatic, mechanical, clerical kind of task.” Each person was a lock, and the MMPI held the precise combination.
Profile coding, as the method came to be called, was a brilliant stroke, for it didn’t just preserve the MMPI as a test for the psychologically troubled. It expanded its use exponentially, to spheres far beyond the mental ward: business suites, army barracks, courtrooms, high schools, doctors’ offices. By the early 1960s, the MMPI was given at least as often to normal people as to psychiatric patients, used to screen job applicants, offer vocational advice, settle custody disputes, and determine legal status. The MMPI had expanded in a philosophical sense as well: from a utilitarian tool designed to solve a specific problem, the test had become a key to personality, a code breaker that would decipher at last the enigma of human nature.
So sweeping were the test’s claims, and so heavy-handed were its methods, that the MMPI was bound to inspire a backlash. In the middle of the 1960s, it arrived.
“I have never seen anything to equal the outrage and indignation from government employees, their families and their friends,” thundered Senator Sam J. Ervin Jr. The date was September 24, 1966, and the setting was a hearing on Capitol Hill. Ervin, a Democrat from North Carolina and chairman of the Senate Constitutional Rights Subcommittee, had called the hearing to address complaints that the government’s use of personality inventories was invading its workers’ privacy. Ervin had been an early critic of the communist-hunting tactics of Senator Joseph McCarthy, and of wiretapping and eavesdropping by government and industry. He saw similar dangers in personality testing. “Clearly, the government should not send out an investigator to peer through an employee’s or an applicant’s bedroom window,” he had declared before the hearing. “Neither should the government ask, through subtle psychological questioning, what a person does and thinks after he draws the curtains.”
Unease about the increasingly prevalent use of personality tests like the MMPI had been growing for more than a decade. In 1956’s The Organization Man, William Whyte sounded a warning about the stifling conformity such tests imposed on American executives. Whyte even included a half-serious guide to giving the boss the answers he wants: “I loved my father and my mother, but my father a little bit more. I like things pretty much the way they are. I never worry much about anything. I don’t care for books or music much. I love my wife and children. I don’t let them get in the way of company work.” More explicit critiques, with titles like The Brain Watchers and The Tyranny of Testing, appeared in the early 1960s. Newspapers and magazines ran investigative exposés under such headlines as “The Snoops” and “The ‘I Love My Mother’ Tests.” There were even reports of public burning of test data by parents angry that their children had been given the questionnaires at school.
In 1965 and 1966, a series of congressional hearings compelled the testers to respond to their critics. For decades, psychologists were the ones asking the questions; now they would be interrogated about their own motives and methods. They knew the stakes were high: Ervin had introduced a bill, cosponsored by thirty-five senators, that would severely limit or even eliminate personality testing in government. This “barrage,” wrote one psychologist, “is the most serious attack that has ever been launched by citizen groups or by government against any part of psychological research or services.”
The testers mounted an impassioned self-defense, both at the hearings and in the pages of professional journals and popular magazines. They began by insisting that their intentions were pure. Hathaway himself published a long “Letter to Mr. R,” replying to a correspondent’s questions about the MMPI: “I hope you will see,” he implored, “that its origins were motivated toward virtue.” They were no “peeping toms,” the psychologists protested, but serious scientists doing worthy work. Many people are made nervous by doctor’s visits, noted Hathaway’s former student W. Grant Dahlstrom, “but there is no corresponding legislative pressure now to outlaw the use of needles and scalpels in the practice of medicine.” The apparent intrusiveness of personality tests was all for your own good, the testers told the public, and with a little education you’d see that, too.
“To change even a comma in an item may change its meaning,” Hathaway explained, adding that though he would alter some MMPI items “if I could,” the value of decades’ worth of research depended on the test remaining exactly the same. Besides, shrugged another psychologist, “as one cannot make an omelet without breaking eggs, so one cannot find out anything important about personality without some pointed questions.” Growing confident, the testers turned the tables on their critics, insinuating that their opposition to the MMPI was evidence of prudishness or irrationality. They even hinted that it grew out of the accusers’ own (presumably warped) personalities.
Their tactics worked. Senator Ervin’s law failed to pass, the media frenzy died down, and the picketers outside the American Psychological Association (“Help stop psychological sex tests. Write your Congressman”) packed up their signs and went home. Lost in the hoopla over the supposed immorality of using the MMPI to test workers was “the real issue,” noted psychologist Malcolm Gynther a few years later, “which is that the test wasn’t devised for this purpose and isn’t effective for predicting who will be a good employee.”
Like Scott Hadley and Art Staples, an increasing number of American workers today are finding that an intrusive personality test is the price of getting or keeping a job. Currently the MMPI (in an updated version) is employed by 86 percent of clinical psychologists and administered, by one estimate, to fifteen million Americans each year. Some of these are mental patients like those for whom Hathaway originally designed the test. But many others are perfectly sane individuals: people who want to become doctors, psychologists, firefighters, airplane pilots, paramedics, nuclear power plant operators, law enforcement agents. The MMPI is used by 60 percent of police departments evaluating prospective officers, for example, and by 91 percent of psychologists screening applicants to Roman Catholic seminaries and religious orders. Personality tests featuring MMPI-like items are given to even more people, mostly job applicants in industries from retail to banking to security services.
But there are no congressional hearings, no newspaper headlines, no protesters with picket signs; whatever redress workers have received in recent years has come through the courts. That was the case for Scott Hadley and Art Staples, who filed a class-action suit against Rent-A-Center in 1999. The company agreed to a $2-million settlement, distributed among the employees forced to take the test. More important than the money, say the two men, was that Rent-A-Center promised to destroy all test records in California and to cease administering the test nationwide.
A small number of other cases have yielded similar results. The one that paved the way for the rest was Soroka v. Dayton Hudson Corporation, settled in 1993. A security guard named Sibi Soroka had applied for a job at a Target store in California, and was asked to take an exam called the PsychScreen—a combination of the MMPI and another test. “These are questions you wouldn’t even answer for your own mother if she asked you, let alone some personnel director at some company you’re not even sure you want to work for,” said Soroka. “It doesn’t take Einstein to figure out that these questions really don’t have any bearing on our world and life today, or certainly on a job walking around looking for shoplifters.”
Soroka’s case was taken on by Berkeley lawyer Brad Seligman, who likens Target’s use of the MMPI in a workplace setting to “killing a gnat with an atomic bomb.” He presented to the court evidence indicating that the PsychScreen had a 61 percent false-positive rate, rejecting as unfit six out of ten psychologically healthy applicants.
Seligman wrested a $1.3-million settlement from the company, the first of several he would win from organizations accused of misusing personality tests. In 1995, he filed suit against Contra Costa County, a local government in California that had administered an MMPI-like test to welfare applicants. The questionnaire, which included items like “I believe everything is turning out just the way the Bible said it would” and “Pornography and obscenity have become serious problems and must be curbed,” was intended to pick out substance abusers. Anyone who failed it was labeled “chemically dependent,” obligated to sign statements admitting to addiction, and required to participate in a six-month rehab program (if they refused, they were denied welfare). The problem was that—according to a study jointly commissioned by plaintiff and defendant—the test incorrectly classified 44 percent of all applicants as addicts. The county agreed to pay $1.2 million to the mislabeled test-takers, and it stopped using the test.
Most recently, Seligman and three other lawyers initiated a lawsuit against Burns International Security Services, part of the nation’s largest security-guard company, on behalf of a job applicant named Mel Thompson. Thompson was given a personality test that included pointed questions about his political beliefs, including items like “Workers usually come last as far as most companies are concerned,” “Most companies make too much profit,” “The drinking age should be lowered,” and “Marijuana should be legalized.” The test amounted to “a true-or-false pledge of allegiance to corporate America,” says Seligman. “Anyone who believes that employers are not perfect, or who believes drug and alcohol laws should be modified in some way, flunks this test.” Burns agreed to pay up to $2.1 million to the eight thousand people who’d answered the questionnaire.
Seligman has had some success fighting intrusive personality tests with a handful of applicable laws: a California constitutional provision that limits the invasion of privacy, a section of the California labor code that bars discrimination against employees based on their political beliefs, and the Americans with Disabilities Act (ADA), which restricts employers’ ability to make a medical examination a condition of employment. But this legal patchwork, different in every state, is hardly an impermeable defense against the spread of personality testing in the American workplace, and a few out-of-court settlements are no match for what has become a vast and growing industry.
As successful as the MMPI has been, its origins in the mental ward have imposed certain limits on its use: it must be administered by a psychologist, for example, and it is bound by the ADA’s restrictions on medical examinations. (These restrictions may be waived, however, in the case of workers charged with public safety—such as airplane pilots, police officers, and firefighters—and indeed many of these workers are required to take the MMPI.) But no such caveats apply to the hundreds of MMPI imitators now on the market: tests that borrow its format, its language, even the unusual way it was constructed.
Though these tests are cheaper, shorter, and more readily available than the MMPI, all of the quirks of the original are preserved intact. There is the MMPI’s smug piety: “This country needs higher moral standards.” There are its opportunities for guilty confession: “I guess I know some pretty undesirable types.” There are its muddled propositions: “Illegal use of marijuana is worse than drinking liquor.” There are its suspiciously scrupulous assertions: “I have never wanted to buy things I couldn’t afford.” And there are its sweeping, context-free statements: “Most people would lie to get what they want.” Anyone familiar with the MMPI will recognize the provenance of these items: Starke Hathaway’s peculiar test, as reflected in the funhouse mirror of American capitalism.
Instead of measuring depression or mania or hypochondria, these personality tests set out to identify qualities of interest to employers: dependability, honesty, friendliness, and the “desire to follow rules, policies, and procedures,” as well as the appropriate “customer service attitude.” More important, such tests promise to screen out those who will be chronically late or absent, who will engage in theft or “computer abuse” (emailing and Internet surfing on company time), who will have “personal and/or transportation problems” and who will indulge in “counterproductive behavior” or “alienated attitudes.” Still more ambitiously, these questionnaires purport to predict whether employees will be injured in a job-related accident, will file a fraudulent workers’ comp claim, will abuse drugs or alcohol, or will engage in workplace violence. (One workplace personality test alerts employers to applicants who fit a “Litigious Profile” or a “Corporate Stalker Profile.”)
The MMPI mimics even have their own versions of mental-patient groupings and Minnesota Normals. As promotional materials for one such test, the Step One Survey, explain, it “compares the applicant’s responses with those of two distinctly different groups—recently released ex-convicts and long-term retail store employees with excellent work histories. The ex-convicts’ crimes were predominantly offenses dealing with theft and drug use. Test items for the Step One Survey consist of questions in which there were statistically significant differences between the two groups.” As another test purveyor puts it, their questionnaire will tell you “whether job applicants think like trusted employees or convicted felons.”
Aggressively marketed to managers and business owners, these tests are often promoted with a dose of alarmism. One test-provider’s website features a photograph of smiling people dressed in business attire, over which flashes the message: “Which one physically hurt an employee? Which one faked an injury? Which one uses illegal drugs? Which one will be late for work? Which one stole $500 from their employer? We can tell you!” The terrorist attacks of September 11, 2001, gave the personality-testing industry an additional boost. In a speech given five months after the tragedy, David Arnold, general counsel of the Association of Test Publishers, told an audience that “invasion of privacy—or rather, what the courts deem to be an invasion of privacy—may be a whole new ball game since September 11.” Increased fears of terrorism, Arnold predicted, would both induce employers to do more testing and put them on safer legal ground when they did so. Indeed, testing companies around the country reported increased interest in personality tests following September 11.
The test publishers’ hard sell seems to be working: a 2003 survey by Management Recruiters International showed that 30 percent of companies now administer personality tests. But the tests’ popularity is no guarantee of their value. To begin with, many of the characteristics they claim to measure are broad, fuzzy categories covering many kinds of behavior. For example, “the construct of honesty or integrity remains vague and ill-defined after more than fifty years of research,” note two experts writing in the journal American Psychologist in 1995. Nevertheless, so-called honesty tests or integrity tests are administered by an estimated five thousand to six thousand U.S. organizations and taken by as many as five million Americans each year. These tests are scattershot in their attempts to target liars, cheats, and thieves. According to a review conducted by the federal government’s Office of Technology Assessment, 95.6 percent of people who fail integrity tests are incorrectly classified as dishonest—an error rate far worse than that of the notoriously unreliable polygraph machine.
These “false positives” may be turned away from jobs and other opportunities before they’ve had a chance to demonstrate their probity. One test-provider recommends that its questionnaire be administered in advance of an interview. “After all,” goes the sales pitch, “if a candidate’s integrity, reliability, work ethic, and attitudes toward substance abuse are below your requirements, why waste your valuable time even talking to them?” Full information about integrity tests (and about most other workplace personality tests) is often unavailable, since these are proprietary instruments of profit-making companies disinclined to submit their products to scientific scrutiny. For some integrity tests, “almost no evidence at all is available beyond assurances that evidence exists,” reported a task force appointed by the American Psychological Association. The task force also found that more than half of integrity-test publishers do not require any training or other qualifications of people who administer their tests.
Less obvious, though perhaps more troubling, are several other aspects of workplace personality screens. First, the reflexive testing of job applicants may reduce employers’ incentive to provide a constructive workplace. If emphasis is placed from the beginning on an employee’s supposedly inherent, unchanging qualities, it’s likely that little energy will go into on-the-job training, mentoring, or development. Second, the intrusiveness of the tests’ items may itself carry a message. As social critic Barbara Ehrenreich, author of Nickel and Dimed, has observed: “Maybe the real function of the tests is not to convey information to the employer but to the potential employee, and the information being conveyed is, You will have no secrets from us. We don’t just want your time and your effort, we want your entire self.” And lastly, recent financial scandals at Enron, WorldCom, Tyco, Adelphia, and other American companies should clue many employers that they’re testing the wrong people. The billions of dollars allegedly looted by these firms’ high-ranking executives make tardiness or “computer abuse” look like small change. A true test of integrity would be holding boss and worker to the same ethical standard.