It Will Never Get Well If You Pick It
It Will Never Get Well If You Pick It: Confessions of a Modern-Day Rum-Runner Lewis R. Goldberg Society for Personality and Social Psychology Address at the Presentation of the Jack Block Award Memphis: January 27, 2007 Jack Block Glancing at my title, it dawned on me that it looks like a description of a hypochondriacal psychopath. A better title for my talk might be: My Big Five Habits, around which I’ll try to weave some threads of my professional life. Habit Number 1 Send everything that you write to everyone whom you know For all of my professional career, I’ve sent preview drafts of each of my reports to dozens of friends and colleagues, always asking for their help in making them better. Over the years, this particular bit of chutzpah has gained me an enormous amount of extraordinarily useful feedback; and my published reports have all profited from the wisdom of my peers. Indeed, one might argue that the reason that I’m standing before you now is because many of you taught me to think better and to write better about those thoughts. For that, I thank you all. I’m also tremendously grateful to those who selected me to receive the Jack Block award for lifetime contributions to personality research. There was only one huge problem with this honor: With it came a request to talk to you today, and soon after that came a request for the title of my invited address. Oh??? How would I know what I was going to talk about, all those many months in advance? The obvious answer was to select a title that was on the one hand so ambiguous that I could later talk about anything under the sun and so intriguing that it would serve to attract throngs of listeners, thus sparing me the humiliation of speaking only to the person roped into introducing me. But, how to select so mysterious a title? I did what I always do: I held a contest and asked my GoldNet gang both what I should talk about and how best to title an address on an unknown topic. I received over three dozen title suggestions, a few of which I’ll share with you now. (You might memorize a few of them for use later in your life when you are faced with a similar titling dilemma.) Larry Erlbaum: Multivariate Hierarchical Linear Models and the Jewish Problem Auke Tellegen: Tellegen’s Seminal Contributions and Lasting Legacy Auke Tellegen & Jack Block Jan Macafee: Crested Grebes Still Mate in the West: Learn How Howard Friedman: The King of Proc and Troll: My Accrued Swayed Views Sheldon Cohen: The Truth About Personality: 1. It is important 2. Everyone has one 3. It is hard to measure 4. No one agrees exactly what it is 5. Undergraduates study it 6. Many textbooks are sold about it 7. I take it back, not everyone has one Bill Chaplin: Hanging Around the Edges of the Empirical Dust-bowl, and Sometimes Falling in: Confessions of a Data-Driven Scientist Eric Knowles: Random Thoughts and Systematic Errors: The Sound of One Scientist Clapping Sam Gosling: My Five Big Mistakes Myron Rothbart: Individual Differences: Chimera, Apparition, or Mirage? Tim Goldberg Arctic Weather Forecasts, Shark Wrestling Updates, and The Road to World Peace Tim and Sister Holly Habit Number 2 Don’t be afraid to pick up what’s been dropped by others (I am an inveterate nit-picker.) An Example My very first class as a college freshman at Harvard in 1949 was taught by Gordon Allport, who used for a textbook his classic 1937 volume, “Personality: A Psychological Interpretation.” Chapter 11 in that text, titled “The theory of traits,” concluded with a section entitled “The problem of trait names,” describing the Allport and Odbert (1936) compendium “Trait-names: A psycho-lexical study.” Sounding vaguely familiar? Allport hoped to work with Odbert or others of his students to develop a taxonomy of the nearly 18,000 trait names that they had culled from the 1925 edition of the unabridged Webster’s New International Dictionary of the English Language, rather than leaving that project with merely those four long alphabetical columns of trait terms. But it was not to be: Allport himself was not much of an empirically oriented scientist, and none of his students took on this taxonomic challenge. That remained for Raymond B. Cattell, who saw the wisdom of combining the personality dictionary provided by Allport and Odbert with the techniques of cluster analysis and factor analysis of which he was so enamored. Although Cattell made a heroic beginning, he arrived on the scene too early in technological history, living still in an age when hand-calculations reigned and computers were in their infancy. Raymond Cattell That’s two dropped balls, and one to go. Warren Norman had just obtained his Ph.D. from the University of Minnesota when he took his first and only faculty position, starting as an instructor in 1957 at the University of Michigan. He wasn’t there more than a few weeks when his department head, E. Lowell Kelly, who had just been elected President of APA, asked him to serve on the dissertation committee of a very balding youngster, one of whose committee members had gone away on leave. Thus began a friendship and intellectual collaboration that lasted until Warren Norman’s early death in 1998. It was Warren Norman who took up the Allportian challenge, intending to finally provide a scientifically compelling taxonomy of personality characteristics, based on what I later dubbed the “lexical hypothesis.” And, it was Norman who was the first to collect empirical data on a subset of roughly 2800 of the 18,000 terms in the Allport-Odbert compendium, and it was Norman who persuaded Goldberg that the next great challenge of a scientific personality psychology was to solve the riddle of personality structure. Then, he too dropped that ball. What makes an extraordinarily brilliant young scientist burn out at an early age? If we can answer that question, we may come to understand how Warren Norman correctly identified a crucial scientific problem, contributed enormously to its solution, but then stopped in mid-tracks. Warren Norman In his explosive burst of brilliance and intellectual exuberance, he infected me, and I came to believe that it was my calling to carry on his vision. I always believed that the problem was going to be too difficult for me to solve, but that it was important enough to devote a lifetime to working on it. A Second Example Here is another example of picking up what others have, in this case inadvertently, dropped. From 1961 to 1965, I worked as a consultant to the U. S. Peace Corps, serving as a Field Selection Officer, selecting Peace Corps trainees to serve in various countries in Southeast Asia, including the Philippines, Thailand, and Malaysia. All of these projects trained on the Big Island in Hawaii, and in my role as a selection officer, I met with each training group at the beginning, middle, and end of their three-month training period— virtually always in beautiful Hilo, Hawaii. A rough assignment? You bet. How did I get so cushy an arrangement? By luck, the first full-time director of selection for the Peace Corps was my dissertation advisor at Michigan, E. Lowell Kelly. Being in Hawaii about every six weeks, I met some colleagues teaching at the University of Hawaii, one of whom, Jack Digman, became one of my dearest friends. Unknown to me at the time, during the years from 1959 to 1967 Digman had persuaded 88 elementary-school teachers on the islands of Oahu and Kauai to describe each of their students at the very end of the school years. All of the teachers rank-ordered their students on each of about 50 personality traits, originally selected by Cattell but later augmented by Digman so as to be as comprehensive as possible. In total, Digman obtained personality descriptions on about 2,400 kids in the first, second, fifth, and sixth grades. By analyzing some subsets of this rich data pool, Digman discovered five robust factors that generalized across subsamples, thus making an important contribution to what I later dubbed the “BigFive” factor representation. Jack Digman (1923-1998) Scholar Colleague Friend The years went by. Digman’s wife moved to Lorane, Oregon, where she ran a horse ranch, and Jack followed her. He needed something of his own to do, and so elected to work with us at the Oregon Research Institute in Eugene. Our colleague, Sarah Hampson, and others at ORI, kept urging Jack to apply for funds to study those elementary school children 40 years later when they were now adults. It took awhile, but in 1998 he received a research grant from NIMH, entitled “Personality and Health: A Longitudinal Study,” to locate as many as possible from that original child cohort. Within a few months of receiving this grant, Jack died— suddenly and unexpectedly. We were all devastated. The NIMH research grant needed a Principal Investigator, and although I had not been part of this project, I was asked to take over. Now, eight years later the team which includes Sarah Hampson, Tom Vogt, and Joan Dubanoski has located over 80% of the original cohort; has administered a series of health-related questionnaires to most of them; and has recruited many of them to participate in a comprehensive physicalmedical-psychological examination at the KaiserPermanente research center in Honolulu. Some Recent Findings: 1 Already, at this relatively early stage of a project that we hope will follow this cohort to the end of their lives, we have made some important discoveries: We have found strong links between childhood personality traits (such as Conscientiousness) and adult health-related behaviors such as smoking, and health outcomes such as obesity. Some Recent Findings: 2 And, in a first study of personality-trait stability over the 40-year span between elementary school and middle adulthood, we found substantial differences in long-term stability between different types of traits. Traits related to Extraversion and Conscientiousness seem to have considerable stability, whereas traits related to Emotional Stability (versus Neuroticism) and Agreeableness seem to have virtually no stability at all. Picking Up Dropped Balls Example Number 3 As an undergraduate at Harvard, I took no courses in the Psychology Department, which at that time included Skinner, Boring, Stevens, and a host of other such luminaries. Yet as a college junior faced with a decision about what to do with my life, I decided to obtain graduate training in psychology—for the obvious reason that psychology would teach me all about people, and knowledge of people would be useful in any profession I might later elect to enter. Oh, sure. Once in the graduate program in Clinical Psychology at the University of Michigan, which was highly psychodynamic in orientation, I found myself skeptical about the scientific status of all that Rorschach inkblot prognosticating to which I was being exposed. So, I set out to test whether clinical psychologists making predictions from projective protocols were experts or quacks, and these early excursions into the accuracy of human judgments remained a theme of my research for the next two decades. After a first study examining whether brain-damaged patients could be distinguished from psychotic ones by their after-images from watching a rotating Archimedes spiral, I studied clinical expertise in diagnosing brain damage using a popular projective procedure of that era, the Bender VisualMotor Gestalt test. In my first contribution to the scientific literature on clinical judgment, I showed that experienced clinicians, clinical psychology graduate students, and hospital secretaries were all equally valid diagnosticians, but that the secretaries were more confident in their judgments than the other two groups. Thus, I argued, for this task one should use one’s secretaries because with no loss in accuracy they would feel a lot better about what they were doing. At that time, it was not common for graduate students to publish reports of their research, but before I finished my degree I had published two articles, the one on clinical accuracy being widely reprinted. Among those who read that article was Paul J. Hoffman, who was starting to achieve some fame from a classic Psychological Bulletin article entitled “The paramorphic representation of clinical judgment” in which he argued that human judgments could be “captured” by multiple regression equations linking the cues that they used to their predictions from those cues. A few years later in my life I was to join Hoffman in Eugene, Oregon, where he had just started a new organization, which he called the Oregon Research Institute. At ORI we were joined by Leonard Rorer (“the great response-style myth”), later by Paul Slovic and Sarah Lichtenstein, and in one of the headiest years of our young lives by Amos Tversky and Dan Kahneman who conducted all of their early studies of judgmental heuristics and biases with us at ORI. But that is another story. Paul Hoffman: ORI Founder The First ORI Building: 1960 Now back to my penchant for picking up stuff dropped by others. As a graduate student at Michigan, the only clinical psychologist who made any sense to me was the one who became my advisor and friend, Lowell Kelly, about whom you’ve heard before. Kelly, with his student Donald Fiske, had conducted a large-scale study of all incoming graduate students in clinical psychology throughout the U.S. immediately after World War II, and their 1951 report of that study entitled “The prediction of performance in clinical psychology” became a milestone in the history of psychological assessment. However, despite the title of that classic volume, Kelly and Fiske had obtained no information about the actual work “performance” of their graduate-student cohort, intending to collect such criterion information in the years ahead. Another ball was about to get dropped. For my doctoral dissertation, I picked up this particular ball, and began a massive detective hunt to find the members of this graduate student cohort 10 years later in their lives. In this my first longitudinal investigation, I managed to locate about 95% of the original cohort, and 100% of those who had gone on to obtain their Ph.D. degrees in psychology. I showed that the kind of work that these individuals ended up doing (clinical practice, academic teaching and/or research, or administration) was predictable from their personality and interests when they first entered graduate school. This early research stimulated an interest in occupational choices, an interest which lay dormant for many decades until recently when I began developing new public-domain measures of vocational and avocational preferences. Habit Number 3 Don’t Be Afraid to Go Away and Then to Stay a Long Time My first academic job was at Stanford University where the faculty superstars included Leon Festinger, Richard Sears, Jack Hilgard, Quinn McNemar, and Alex Bavelas. In the office across from me was a youngster named Albert Bandura and a few doors away was an even younger Jerry Wiggins, with whom I co-taught a graduate assessment course and with whom I outlined a textbook in assessment that would eventually become that great Wiggins classic, “Personality and Prediction.” I had been hired as a temporary one-year replacement which then got renewed for a second year. When the time came to take a permanent position, I joined the faculty at the University of Oregon in Eugene, which has remained my residence for almost 50 years. But, unlike most folks I know, whenever I could take a leave, I stayed away for a year or more at a time, and these years away from home have been the most stimulating ones in my life. Jerry S. Wiggins Jack Digman and Jerry Wiggins First was Nijmegen in The Netherlands, where as a Fulbright Professor I got to visit the five other psychology departments in Holland at that time and to meet most if not all of their faculty members. The most famous psychologist of that era was Adriaan de Groot (Thought and Choice in Chess, Perception and Memory in Chess) whom I invited to spend a year with us at ORI, during which time I helped edit his classic volume on Methodology which he had translated from Dutch to English. But it was a young whippersnapper at the University of Groningen, Willem Hofstee, who ended up as one of my closest and most stimulating of colleagues. Wim Hofstee visited us at ORI on a number of occasions, along with his students Arend Tomas and Frank Brokken, the later whose doctoral dissertation was an early milestone in lexical research has now visited the United States so frequently that he owns a truck parked in Eugene. It is Wim Hofstee who deserves primary credit for the development of the Abridged Big-Five dimensional Circumplex (AB5C) model integrating dimensional and circular representations of the Big-Five domains. Willem K. B. Hofstee By the end of my first European sabbatical, it had become obvious to Jerry Wiggins and to me that my heart lay in empirical analyses, not in textbook writing, and thus that Jerry should complete our assessment volume on his own. When that book finally came out in 1973, it was dedicated “To ORI: The people and the concept.” During the 1971-1972 academic year, I got to hang out with Ravenna Helson and Harrison Gough (the 4th and 5th Jack Block awardees) at the Institute for Personality Assessment and Research (IPAR) at the University of California in Berkeley, and with Jack Block himself, usually at his home. Jack and I had met years earlier at a convention in Honolulu, where we spent hours on a beach discussing the crazy response-set wars of that period, ongoing discussions that culminated in Block’s remarkable 1965 critique, “The Challenge of Response Sets.” While spending a year with Block in Berkeley, I accompanied him on his weekly excursions to San Francisco to pick up, each from a different small specialty shop, his weekly assortments of wines, pastas, cheeses, sausages, bagels, rolls, and bread-sticks. Once through those demanding errands, it was time for our weekly dim-sum Chinese lunch. IPAR at UC-Berkeley Harrison Gough The Netherlands was cool, Berkeley was hot (especially during these years of frenzied counter-cultural protest), but Turkey was the cat’s meow—the very best. During a Fulbrightsupported year in Istanbul in the mid-1970’s, I got to explore a city diametrically opposite my home in Eugene, Oregon, in so many ways that it would take my breath away. I taught courses at Istanbul University, the oldest university in the middle east, and certainly the most entertaining. During the days, I walked the streets of this ancient city, and when the sun went down I savored the sights and sounds along the mighty Bosporus. Roughly once a month, I wrote a raki-enfused account of our adventures, and the collection of these unpublished “Letters from Turkey” soon became an underground collector’s item in the rum-running world. In my second letter, I described a period of “Unmitigated gawking, while walking every sidestreet, sniffing every smell, visiting every mosque, riding every ferry, bus, and dolmus I could find.” It was one hell of a ride, and one hell of a year. Had I not spent that year in Istanbul, would I have still devoted so much time to studying the Turkish language of personality? Strangely, the two periods of my life were not connected. Two decades after my Istanbul odyssey, a young Turkish psychologist, Oya Somer, discovered the lexical hypothesis, the Big-Five factor structure, and me. She obtained the data; I planned the analyses; and together we published two articles describing our many findings from this ancient Altaic language. But, another full year abroad was connected to my earlier Dutch sabbatical. Probably because of my friendship with Wim Hofstee, I was invited to spend a year as a fellow-in-residence at the Netherlands Institute for Advanced Study (NIAS), in Wassenaar, near the Hague in the Netherlands. Frank Brokken set up a computer connection between my NIAS office and his Groningen computer center, which had the most advanced academic computer system in Holland at that time. Where I had once I had once learned to program in Fortran II, I could now use SPSS to do it for me, and so once again I could function as my own data analyst, which I did all day most of the days of that wonderful year. When not computing, I was driving—from Bielefeld in Germany to Wassenaar in Holland with stops in Groningen along the way—with my new graduate student, Oliver John. Oliver had been an undergraduate at Bielefeld with Alois Angleitner, who sent Oliver on a scouting expedition to interview Block, Wiggins, Fiske, Jackson, Mischel, and Goldberg in search of an ideal graduate program. Somehow I won out and Oliver came to Oregon, and then returned to Bielefeld during my year at NIAS. Through Angleitner, Oliver and I met Sarah Hampson, and together John, Hampson, Chaplin, and Goldberg spent eight heady years studying personality traits as semantic categories, eventually discovering a basic level in personality trait hierarchies. Oliver P. John John and Chaplin (& Friend) William Chaplin & Sarah Hampson Sarah Hampson Habit Number 4 Remember Good Ideas— They Can Be Useful in the Future Back when I was an undergraduate at Harvard, I read Henry Murray’s (1938) classic volume, “Explorations in Personality.” Perhaps I didn’t get much past the Preface of that long tome, at least at that early stage of my life, but it was in that short preface that Murray raved about the advantages of using the same sample of research participants to study many different topics. He argued that if different investigators collected data from the same sample, the resulting data pool would be rich enough so that one could check up on possible explanations for one’s findings using data obtained by others. Murray implemented this plan with Harvard undergraduates, although that kind of sample has a shelf life of at most four years. But, the general idea is a winner. More than 40 years after I first came across that idea, Gerard Saucier and I obtained an NIMH grant, “Mapping Personality Trait Structure,” which had elicited a high enough priority score to win a Merit Award, thus doubling the length of our grant. Now with eight years of guaranteed funding, the time seemed right to solicit our own pet sample. We needed to find individuals who would be living stably in the community for at least the next ten years, and therefore university students and other transients were out. Instead, we solicited participants from lists of home-owners, asking folks if they wanted to help science and get paid (at least modestly) for their time. Roughly 500 men and 500 women between the ages of 18 and 85 initially expressed an interest in the project and completed a mini-inventory of 360 traitdescriptive adjectives, and around 850 of them then completed our first real inventory, a set of 858 items that became the kernel of the International Personality Item Pool (IPIP). With Gerard Saucier: The Eugene-Springfield Community Sample (ESCS) The International Personality Item Pool (IPIP) Gerard Saucier Thus was born the Eugene-Springfield Community Sample (ESCS), a loyal stable of research participants who have now completed by mail 29 questionnaires covering an enormous range of topics, including personality traits, values and attitudes, vocational and avocational interests, possessions, current and past activities, aspects of psychopathology and of physical health, talents and skills, and exposure to potentially traumatic events both in childhood and adulthood. Remarkably, over the decade between 1994 and 2004 virtually all participant attrition was due to death or disability. In addition to the extraordinary collection of self-report measures we have obtained, participants wrote projective stories to TAT pictures, and most of them have been described by two or three individuals who knew them well. Over the years, members of this sample have completed a wide array of current personality inventories, including the NEO, California Psychological Inventory, 16PF, Hogan Personality Inventory, MPQ, Jackson Personality Inventory, TCI, and HEXACO. The availability of so many instruments, all developed to provide broad bandwidth assessments of normal personality functioning, suggested the possibility of a comparative-validity horse race. In a study conducted with Rick Grucza of Washington University medical school in St. Louis, we used multiple regression analyses and bootstrap resampling procedures to provide cross-validity coefficients for 11 inventories predicting each of three kinds of criteria: (a) The frequencies of occurrence of diverse activities; (b) personality descriptions by knowledgeable informants; and (c) clinical indicators known to be associated with various kinds of emotional disorders. Perhaps the most remarkable of our many findings is that inter-inventory differences across the entire range of criteria were quite small, suggesting some sort of common core of personality-trait variance measured by these seemingly quite diverse inventories. Data from our community sample have been provided freely to a host of investigators throughout the world. Yet, the many who have used ESCS data are but a tiny fraction of those who have used the public-domain measures available at the International Personality Item Pool (IPIP) web-site. The same research grant that funded our community sample has supported this international collaboratory, which now includes around 2,500 personality items and around 250 personality scales. English IPIP items have now been translated into more than 25 other languages, and the IPIP web-site already lists around 100 publications that have used IPIP scales. It was Willem Hofstee of the University of Groningen in the Netherlands who called attention to the fact that those single trait-descriptive adjectives used in lexical studies of personality structure are not ideal stimuli for use in personality inventories because they are too broad and abstract in their nature, and perhaps because of this it is often difficult to find exact one-to-one translations across languages; in many cases terms that seem descriptively identical differ in their evaluations. Hofstee and his students pioneered the use of short verbal phrases, which are more contextualized than single adjectives but still more compact than many items included in popular inventories, and these verbal phrases are used for all IPIP items. Examples include: “Believe in an eye for an eye.” “Can read people like a book.” “Dislike being the center of attention.” “Enjoy the beauty of nature.” “Forget appointments.” “Get upset easily.” “Have gotten better with age.” By far my favorite IPIP item is “Am able to disregard rules,” and that explains the reference to rum-running in the title of my address. During the period of prohibition in the United States, those who smuggled alcoholic drinks into the country were referred to as rum-runners. Although I am not a proponent of lawbreaking in general, as a booze-lover I can understand why some folks might want to defy the blue-noses who sought to prohibit others from drinking. More generally, however, I like the structure of that IPIP item: Note that it describes an ability (“am able to”), not a failing (“can’t help”) nor a propensity (“tend to”); and it concerns “rules” rather than “laws.” Breaking rules can be good for you, if the rule-breaking hurts no one else and if one knows what one is doing and why one is doing it. If you are the kind of person who is not able to disregard rules, perhaps you might give it a try. Goldberg-Saucier Jam (with fireplace tools) Actually, rule-bound conformity is better than taking oneself too seriously. What I really believe is that in the battle between frivolity and ponderousness, it is better to veer to the light-hearted. When in doubt, just remember that old maxim from our childhood: “It will never get well if you pick it.” A Gang of Four • Goldberg Habit Number 5: How to Win the Block Award Shave Your Head, Stay Healthy, and Outlive the Competition Thank you everso-much for this award.