September 20, 2004
ANNALS OF PSYCHOLOGY
Employers love personality tests.
But what do they really reveal?
When Alexander (Sandy) Nininger was twenty-three, and newly commissioned as a lieutenant in the United States Army, he was sent to the South Pacific to serve with the 57th Infantry of the Philippine Scouts. It was January, 1942. The Japanese had just seized Philippine ports at Vigan, Legazpi, Lamon Bay, and Lingayen, and forced the American and Philippine forces to retreat into Bataan, a rugged peninsula on the South China Sea. There, besieged and outnumbered, the Americans set to work building a defensive line, digging foxholes and constructing dikes and clearing underbrush to provide unobstructed sight lines for rifles and machine guns. Nininger's men were on the line's right flank. They labored day and night. The heat and the mosquitoes were nearly unbearable.
Quiet by nature, Nininger was tall and slender, with wavy blond hair. As Franklin M. Reck recounts in "Beyond the Call of Duty," Nininger had graduated near the top of his class at West Point, where he chaired the lecture-and-entertainment committee. He had spent many hours with a friend, discussing everything from history to the theory of relativity. He loved the theatre. In the evenings, he could often be found sitting by the fireplace in the living room of his commanding officer, sipping tea and listening to Tchaikovsky. As a boy, he once saw his father kill a hawk and had been repulsed. When he went into active service, he wrote a friend to say that he had no feelings of hate, and did not think he could ever kill anyone out of hatred. He had none of the swagger of the natural warrior. He worked hard and had a strong sense of duty.
In the second week of January, the Japanese attacked, slipping hundreds of snipers through the American lines, climbing into trees, turning the battlefield into what Reck calls a "gigantic possum hunt." On the morning of January 12th, Nininger went to his commanding officer. He wanted, he said, to be assigned to another company, one that was in the thick of the action, so he could go hunting for Japanese snipers.
He took several grenades and ammunition belts, slung a Garand rifle over his shoulder, and grabbed a sub machine gun. Starting at the point where the fighting was heaviest—near the position of the battalion's K Company—he crawled through the jungle and shot a Japanese soldier out of a tree. He shot and killed snipers. He threw grenades into enemy positions. He was wounded in the leg, but he kept going, clearing out Japanese positions for the other members of K Company, behind him. He soon ran out of grenades and switched to his rifle, and then, when he ran out of ammunition, used only his bayonet. He was wounded a second time, but when a medic crawled toward him to help bring him back behind the lines Nininger waved him off. He saw a Japanese bunker up ahead. As he leaped out of a shell hole, he was spun around by a bullet to the shoulder, but he kept charging at the bunker, where a Japanese officer and two enlisted men were dug in. He dispatched one soldier with a double thrust of his bayonet, clubbed down the other, and bayonetted the officer. Then, with outstretched arms, he collapsed face down. For his heroism, Nininger was posthumously awarded the Medal of Honor, the first American soldier so decorated in the Second World War.
Suppose that you were a senior Army officer in the early days of the Second World War and were trying to put together a crack team of fearless and ferocious fighters. Sandy Nininger, it now appears, had exactly the right kind of personality for that assignment, but is there any way you could have known this beforehand? It clearly wouldn't have helped to ask Nininger if he was fearless and ferocious, because he didn't know that he was fearless and ferocious. Nor would it have worked to talk to people who spent time with him. His friend would have told you only that Nininger was quiet and thoughtful and loved the theatre, and his commanding officer would have talked about the evenings of tea and Tchaikovsky. With the exception, perhaps, of the Scarlet Pimpernel, a love of music, theatre, and long afternoons in front of a teapot is not a known predictor of great valor. What you need is some kind of sophisticated psychological instrument, capable of getting to the heart of his personality.
Over the course of the past century, psychology has been consumed with the search for this kind of magical instrument. Hermann Rorschach proposed that great meaning lay in the way that people described inkblots. The creators of the Minnesota Multiphasic Personality Inventory believed in the revelatory power of true-false items such as "I have never had any black, tarry-looking bowel movements" or "If the money were right, I would like to work for a circus or a carnival." Today, Annie Murphy Paul tells us in her fascinating new book, "Cult of Personality," that there are twenty-five hundred kinds of personality tests. Testing is a four-hundred-million-dollar-a-year industry. A hefty percentage of American corporations use personality tests as part of the hiring and promotion process. The tests figure in custody battles and in sentencing and parole decisions. "Yet despite their prevalence—and the importance of the matters they are called upon to decide—personality tests have received surprisingly little scrutiny," Paul writes. We can call in the psychologists. We can give Sandy Nininger a battery of tests. But will any of it help?
One of the most popular personality tests in the world is the Myers-Briggs Type Indicator (M.B.T.I.), a psychological-assessment system based on Carl Jung's notion that people make sense of the world through a series of psychological frames. Some people are extroverts, some are introverts. Some process information through logical thought. Some are directed by their feelings. Some make sense of the world through intuitive leaps. Others collect data through their senses. To these three categories— (I)ntroversion/(E)xtroversion, i(N)tuition/(S)ensing, (T)hinking/(F)eeling—the Myers-Briggs test adds a fourth: (J)udging/(P)erceiving. Judgers "like to live in a planned, orderly way, seeking to regulate and manage their lives," according to an M.B.T.I. guide, whereas Perceivers "like to live in a flexible, spontaneous way, seeking to experience and understand life, rather than control it." The M.B.T.I. asks the test-taker to answer a series of "forced-choice" questions, where one choice identifies you as belonging to one of these paired traits. The basic test takes twenty minutes, and at the end you are presented with a precise, multidimensional summary of your personality-your type might be INTJ or ESFP, or some other combination. Two and a half million Americans a year take the Myers-Briggs. Eighty-nine companies out of the Fortune 100 make use of it, for things like hiring or training sessions to help employees "understand" themselves or their colleagues. Annie Murphy Paul says that at the eminent consulting firm McKinsey, " 'associates' often know their colleagues' four-letter M.B.T.I. types by heart," the way they might know their own weight or (this being McKinsey) their S.A.T. scores.
It is tempting to think, then, that we could figure out the Myers-Briggs type that corresponds best to commando work, and then test to see whether Sandy Nininger fits the profile. Unfortunately, the notion of personality type is not nearly as straightforward as it appears. For example, the Myers-Briggs poses a series of items grouped around the issue of whether you—the test-taker—are someone who likes to plan your day or evening beforehand or someone who prefers to be spontaneous. The idea is obviously to determine whether you belong to the Judger or Perceiver camp, but the basic question here is surprisingly hard to answer. I think I'm someone who likes to be spontaneous. On the other hand, I have embarked on too many spontaneous evenings that ended up with my friends and me standing on the sidewalk, looking at each other and wondering what to do next. So I guess I'm a spontaneous person who recognizes that life usually goes more smoothly if I plan first, or, rather, I'm a person who prefers to be spontaneous only if there's someone around me who isn't. Does that make me spontaneous or not? I'm not sure. I suppose it means that I'm somewhere in the middle.
This is the first problem with the Myers-Briggs. It assumes that we are either one thing or another—Intuitive or Sensing, Introverted or Extroverted. But personality doesn't fit into neat binary categories: we fall somewhere along a continuum.
Here's another question: Would you rather work under a boss (or a teacher) who is good-natured but often inconsistent, or sharp-tongued but always logical?
On the Myers-Briggs, this is one of a series of questions intended to establish whether you are a Thinker or a Feeler. But I'm not sure I know how to answer this one, either. I once had a good-natured boss whose inconsistency bothered me, because he exerted a great deal of day-to-day control over my work. Then I had a boss who was quite consistent and very sharp-tongued—but at that point I was in a job where day-to-day dealings with my boss were minimal, so his sharp tongue didn't matter that much. So what do I want in a boss? As far as I can tell, the only plausible answer is: It depends. The Myers-Briggs assumes that who we are is consistent from one situation to another. But surely what we want in a boss, and how we behave toward our boss, is affected by what kind of job we have.
This is the gist of the now famous critique that the psychologist Walter Mischel has made of personality testing. One of Mischel's studies involved watching children interact with one another at a summer camp. Aggressiveness was among the traits that he was interested in, so he watched the children in five different situations: how they behaved when approached by a peer, when teased by a peer, when praised by an adult, when punished by an adult, and when warned by an adult. He found that how aggressively a child responded in one of those situations wasn't a good predictor of how that same child responded in another situation. Just because a boy was aggressive in the face of being teased by another boy didn't mean that he would be aggressive in the face of being warned by an adult. On the other hand, if a child responded aggressively to being teased by a peer one day, it was a pretty good indicator that he'd respond aggressively to being teased by a peer the next day. We have a personality in the sense that we have a consistent pattern of behavior. But that pattern is complex and that personality is contingent: it represents an interaction between our internal disposition and tendencies and the situations that we find ourselves in.
It's not surprising, then, that the Myers-Briggs has a large problem with consistency: according to some studies, more than half of those who take the test a second time end up with a different score than when they took it the first time. Since personality is continuous, not dichotomous, clearly some people who are borderline Introverts or Feelers one week slide over to Extroversion or Thinking the next week. And since personality is contingent, not stable, how we answer is affected by which circumstances are foremost in our minds when we take the test. If I happen to remember my first boss, then I come out as a Thinker. If my mind is on my second boss, I come out as a Feeler. When I took the Myers-Briggs, I scored as an INTJ. But, if odds are that I'm going to be something else if I take the test again, what good is it?
Once, for fun, a friend and I devised our own personality test. Like the M.B.T.I., it has four dimensions. The first is Canine/Feline. In romantic relationships, are you the pursuer, who runs happily to the door, tail wagging? Or are you the pursued? The second is More/Different. Is it your intellectual style to gather and master as much information as you can or to make imaginative use of a discrete amount of information? The third is Insider/Outsider. Do you get along with your parents or do you define yourself outside your relationship with your mother and father? And, finally, there is Nibbler/Gobbler. Do you work steadily, in small increments, or do everything at once, in a big gulp? I'm quite pleased with the personality inventory we devised. It directly touches on four aspects of life and temperament-romance, cognition, family, and work style—that are only hinted at by Myers-Briggs. And it can be completed in under a minute, nineteen minutes faster than Myers-Briggs, an advantage not to be dismissed in today's fast-paced business environment. Of course, the four traits it measures are utterly arbitrary, based on what my friend and I came up with over the course of a phone call. But then again surely all universal dichotomous typing systems are arbitrary.
Where did the Myers-Briggs come from, after all? As Paul tells us, it began with a housewife from Washington, D.C., named Katharine Briggs, at the turn of the last century. Briggs had a daughter, Isabel, an only child for whom (as one relative put it) she did "everything but breathe." When Isabel was still in her teens, Katharine wrote a book-length manuscript about her daughter's remarkable childhood, calling her a "genius" and "a little Shakespeare." When Isabel went off to Swarthmore College, in 1915, the two exchanged letters nearly every day. Then, one day, Isabel brought home her college boyfriend and announced that they were to be married. His name was Clarence (Chief) Myers. He was tall and handsome and studying to be a lawyer, and he could not have been more different from the Briggs women. Katharine and Isabel were bold and imaginative and intuitive. Myers was practical and logical and detail-oriented. Katharine could not understand her future son-in-law. "When the blissful young couple returned to Swarthmore," Paul writes, "Katharine retreated to her study, intent on 'figuring out Chief.' "She began to read widely in psychology and philosophy. Then, in 1923, she came across the first English translation of Carl Jung's "Psychological Types." "This is it!" Katharine told her daughter. Paul recounts, "In a dramatic display of conviction she burned all her own research and adopted Jung's book as her 'Bible,' as she gushed in a letter to the man himself. His system explained it all: Lyman [Katharine's husband], Katharine, Isabel, and Chief were introverts; the two men were thinkers, while the women were feelers; and of course the Briggses were intuitives, while Chief was a senser." Encouraged by her mother, Isabel—who was living in Swarthmore and writing mystery novels—devised a paper-and-pencil test to help people identify which of the Jungian categories they belonged to, and then spent the rest of her life tirelessly and brilliantly promoting her creation.
The problem, as Paul points out, is that Myers and her mother did not actually understand Jung at all. Jung didn't believe that types were easily identifiable, and he didn't believe that people could be permanently slotted into one category or another. "Every individual is an exception to the rule," he wrote; to "stick labels on people at first sight," in his view, was "nothing but a childish parlor game." Why is a parlor game based on my desire to entertain my friends any less valid than a parlor game based on Katharine Briggs's obsession with her son-in-law?
The problems with the Myers-Briggs suggest that we need a test that is responsive to the complexity and variability of the human personality. And that is why, not long ago, I found myself in the office of a psychologist from New Jersey named Lon Gieser. He is among the country's leading experts on what is called the Thematic Apperception Test (T.A.T.), an assessment tool developed in the nineteen-thirties by Henry Murray, one of the most influential psychologists of the twentieth century.
I sat in a chair facing Gieser, as if I were his patient. He had in his hand two dozen or so pictures—mostly black-and-white drawings—on legal-sized cards, all of which had been chosen by Murray years before. "These pictures present a series of scenes," Gieser said to me. "What I want you to do with each scene is tell a story with a beginning, a middle, and an end." He handed me the first card. It was of a young boy looking at a violin. I had imagined, as Gieser was describing the test to me, that it would be hard to come up with stories to match the pictures. As I quickly discovered, though, the exercise was relatively effortless: the stories just tumbled out.
"This is a young boy," I began. "His parents want him to take up the violin, and they've been encouraging him. I think he is uncertain whether he wants to be a violin player, and maybe even resents the imposition of having to play this instrument, which doesn't seem to have any appeal for him. He's not excited or thrilled about this. He'd rather be somewhere else. He's just sitting there looking at it, and dreading having to fulfill this parental obligation."
I continued in that vein for a few more minutes. Gieser gave me another card, this one of a muscular man clinging to a rope and looking off into the distance. "He's climbing up, not climbing down," I said, and went on:
It's out in public. It's some kind of big square, in Europe, and there is some kind of spectacle going on. It's the seventeenth or eighteenth century. The King is coming by in a carriage, and this man is shimmying up, so he can see over everyone else and get a better view of the King. I don't get the sense that he's any kind of highborn person. I think he aspires to be more than he is. And he's kind of getting a glimpse of the King as a way of giving himself a sense of what he could be, or what his own future could be like.
We went on like this for the better part of an hour, as I responded to twelve cards—each of people in various kinds of ambiguous situations. One picture showed a woman slumped on the ground, with some small object next to her; another showed an attractive couple in a kind of angry embrace, apparently having an argument. (I said that the fight they were having was staged, that each was simply playing a role.) As I talked, Gieser took notes. Later, he called me and gave me his impressions. "What came out was the way you deal with emotion," he said. "Even when you recognized the emotion, you distanced yourself from it. The underlying motive is this desire to avoid conflict. The other thing is that when there are opportunities to go to someone else and work stuff out, your character is always going off alone. There is a real avoidance of emotion and dealing with other people, and everyone goes to their own corners and works things out on their own."
How could Gieser make such a confident reading of my personality after listening to me for such a short time? I was baffled by this, at first, because I felt that I had told a series of random and idiosyncratic stories. When I listened to the tape I had made of the session, though, I saw what Gieser had picked up on: my stories were exceedingly repetitive in just the way that he had identified. The final card that Gieser gave me was blank, and he asked me to imagine my own picture and tell a story about it. For some reason, what came to mind was Andrew Wyeth's famous painting "Christina's World," of a woman alone in a field, her hair being blown by the wind. She was from the city, I said, and had come home to see her family in the country: "I think she is taking a walk. She is pondering some piece of important news. She has gone off from the rest of the people to think about it." Only later did I realize that in the actual painting the woman is not strolling through the field. She is crawling, desperately, on her hands and knees. How obvious could my aversion to strong emotion be?
The T.A.T. has a number of cards that are used to assess achievement—that is, how interested someone is in getting ahead and succeeding in life. One is the card of the man on the rope; another is the boy looking at his violin. Gieser, in listening to my stories, concluded that I was very low in achievement:
Some people say this kid is dreaming about being a great violinist, and he's going to make it. With you, it wasn't what he wanted to do at all. His parents were making him do it. With the rope climbing, some people do this Tarzan thing. They climb the pole and get to the top and feel this great achievement. You have him going up the rope—and why is he feeling the pleasure? Because he's seeing the King. He's still a nobody in the public square, looking at the King.
Now, this is a little strange. I consider myself quite ambitious. On a questionnaire, if you asked me to rank how important getting ahead and being successful was to me, I'd check the "very important" box. But Gieser is suggesting that the T.A.T. allowed him to glimpse another dimension of my personality.
This idea—that our personality can hold contradictory elements—is at the heart of "Strangers to Ourselves," by the social psychologist Timothy D. Wilson. He is one of the discipline's most prominent researchers, and his book is what popular psychology ought to be (and rarely is): thoughtful, beautifully written, and full of unexpected insights. Wilson's interest is in what he calls the "adaptive unconscious" (not to be confused with the Freudian unconscious). The adaptive unconscious, in Wilson's description, is a big computer in our brain which sits below the surface and evaluates, filters, and looks for patterns in the mountain of data that come in through our senses. That system, Wilson argues, has a personality: it has a set of patterns and responses and tendencies that are laid down by our genes and our early-childhood experiences. These patterns are stable and hard to change, and we are only dimly aware of them. On top of that, in his schema we have another personality: it's the conscious identity that we create for ourselves with the choices we make, the stories we tell about ourselves, and the formal reasons we come up with to explain our motives and feelings. Yet this "constructed self" has no particular connection with the personality of our adaptive unconscious. In fact, they could easily be at odds. Wilson writes:
The adaptive unconscious is more likely to influence people's uncontrolled, implicit responses, whereas the constructed self is more likely to influence people's deliberative, explicit responses. For example, the quick, spontaneous decision of whether to argue with a co-worker is likely to be under the control of one's nonconscious needs for power and affiliation. A more thoughtful decision about whether to invite a co-worker over for dinner is more likely to be under the control of one's conscious, self-attributed motives.
When Gieser said that he thought I was low in achievement, then, he presumably saw in my stories an unconscious ambivalence toward success. The T.A.T., he believes, allowed him to go beyond the way I viewed myself and arrive at a reading with greater depth and nuance.
Even if he's right, though, does this help us pick commandos? I'm not so sure. Clearly, underneath Sandy Nininger's peaceful façade there was another Nininger capable of great bravery and ferocity, and a T.A.T. of Nininger might have given us a glimpse of that part of who he was. But let's not forget that he volunteered for the front lines: he made a conscious decision to put himself in the heat of the action. What we really need is an understanding of how those two sides of his personality interact in critical situations. When is Sandy Nininger's commitment to peacefulness more, or less, important than some unconscious ferocity? The other problem with the T.A.T., of course, is that it's a subjective instrument. You could say that my story about the man climbing the rope is evidence that I'm low in achievement or you could say that it shows a strong desire for social mobility. The climber wants to look down—not up—at the King in order to get a sense "of what he could be." You could say that my interpretation that the couple's fighting was staged was evidence of my aversion to strong emotion. Or you could say that it was evidence of my delight in deception and role-playing. This isn't to question Gieser's skill or experience as a diagnostician. The T.A.T. is supposed to do no more than identify themes and problem areas, and I'm sure Gieser would be happy to put me on the couch for a year to explore those themes and see which of his initial hypotheses had any validity. But the reason employers want a magical instrument for measuring personality is that they don't have a year to work through the ambiguities. They need an answer now.
A larger limitation of both Myers-Briggs and the T.A.T. is that they are indirect. Tests of this kind require us first to identify a personality trait that corresponds to the behavior we're interested in, and then to figure out how to measure that trait—but by then we're two steps removed from what we're after. And each of those steps represents an opportunity for error and distortion. Shouldn't we try, instead, to test directly for the behavior we're interested in? This is the idea that lies behind what's known as the Assessment Center, and the leading practitioner of this approach is a company called Development Dimensions International, or D.D.I.
Companies trying to evaluate job applicants send them to D.D.I.'s headquarters, outside Pittsburgh, where they spend the day role-playing as business executives. When I contacted D.D.I., I was told that I was going to be Terry Turner, the head of the robotics division of a company called Global Solutions.
I arrived early in the morning, and was led to an office. On the desk was a computer, a phone, and a tape recorder. In the corner of the room was a video camera, and on my desk was an agenda for the day. I had a long telephone conversation with a business partner from France. There were labor difficulties at an overseas plant. A new product—a robot for the home-had run into a series of technical glitches. I answered e-mails. I prepared and recorded a talk for a product-launch meeting. I gave a live interview to a local television reporter. In the afternoon, I met with another senior Global Solutions manager, and presented a strategic plan for the future of the robotics division. It was a long, demanding day at the office, and when I left, a team of D.D.I. specialists combed through copies of my e-mails, the audiotapes of my phone calls and my speech, and the videotapes of my interviews, and analyzed me across four dimensions: interpersonal skills, leadership skills, business-management skills, and personal attributes. A few weeks later, I was given my report. Some of it was positive: I was a quick learner. I had good ideas. I expressed myself well, and—I was relieved to hear—wrote clearly. But, as the assessment of my performance made plain, I was something less than top management material:
Although you did a remarkable job addressing matters, you tended to handle issues from a fairly lofty perch, pitching good ideas somewhat unilaterally while lobbing supporting rationale down to the team below. . . . Had you brought your team closer to decisions by vesting them with greater accountability, responsibility and decision-making authority, they would have undoubtedly felt more engaged, satisfied and valued. . . .In a somewhat similar vein, but on a slightly more interpersonal level, while you seemed to recognize the value of collaboration and building positive working relationships with people, you tended to take a purely businesslike approach to forging partnerships. You spoke of win/win solutions from a business perspective and your rationale for partnering and collaboration seemed to be based solely on business logic. Additionally, at times you did not respond to some of the softer, subtler cues that spoke to people's real frustrations, more personal feelings, or true point of view.
Ouch! Of course, when the D.D.I. analysts said that I did not respond to "some of the softer, subtler cues that spoke to people's real frustrations, more personal feelings, or true point of view," they didn't mean that I was an insensitive person. They meant that I was insensitive in the role of manager. The T.A.T. and M.B.T.I. aimed to make global assessments of the different aspects of my personality. My day as Terry Turner was meant to find out only what I'm like when I'm the head of the robotics division of Global Solutions. That's an important difference. It respects the role of situation and contingency in personality. It sidesteps the difficulty of integrating my unconscious self with my constructed self by looking at the way that my various selves interact in the real world. Most important, it offers the hope that with experience and attention I can construct a more appropriate executive "self." The Assessment Center is probably the best method that employers have for evaluating personality.
But could an Assessment Center help us identify the Sandy Niningers of the world? The center makes a behavioral prediction, and, as solid and specific as that prediction is, people are least predictable at those critical moments when prediction would be most valuable. The answer to the question of whether my Terry Turner would be a good executive is, once again: It depends. It depends on what kind of company Global Solutions is, and on what kind of respect my co-workers have for me, and on how quickly I manage to correct my shortcomings, and on all kinds of other things that cannot be anticipated. The quality of being a good manager is, in the end, as irreducible as the quality of being a good friend. We think that a friend has to be loyal and nice and interesting—and that's certainly a good start. But people whom we don't find loyal, nice, or interesting have friends, too, because loyalty, niceness, and interestingness are emergent traits. They arise out of the interaction of two people, and all we really mean when we say that someone is interesting or nice is that they are interesting or nice to us.
All these difficulties do not mean that we should give up on the task of trying to understand and categorize one another. We could certainly send Sandy Nininger to an Assessment Center, and find out whether, in a make-believe battle, he plays the role of commando with verve and discipline. We could talk to his friends and discover his love of music and theatre. We could find out how he responded to the picture of the man on a rope. We could sit him down and have him do the Myers-Briggs and dutifully note that he is an Introverted, Intuitive, Thinking Judger, and, for good measure, take an extra minute to run him through my own favorite personality inventory and type him as a Canine, Different, Insider Gobbler. We will know all kinds of things about him then. His personnel file will be as thick as a phone book, and we can consult our findings whenever we make decisions about his future. We just have to acknowledge that his file will tell us little about the thing we're most interested in. For that, we have to join him in the jungles of Bataan.