Powerpoint with speaker notes

The Psychometric Perspective, Part I: Measuring the Capacity to Think Creatively

The discussion in the book so far has centered on the creative thought process. We have examined the cognitive perspective that creative thinking is equivalent to ordinary problem solving (Chapter 3) and the role of knowledge and expertise in problem solving and creative thinking (Chapters 4 and 5). Chapters 6–8 examined three variants on the general notion that creativity depends on extraordinary thought processes—that is, thought processes that are different from ordinary conscious thinking. In Chapter 6, we considered the idea that creative thinking, as exemplified by problem solving, comes about through leaps of insight. In Chapter 7, we examined the hypothesis that creative thinking can be fostered by psychopathology, and in Chapter 8 we explored the hypothesis that the unconscious underlies creative thinking. In each of those chapters, the support was not very strong for the idea that extraordinary thought processes underlie creativity. The upshot of the discussion in those chapters was that it is possible to understand a wide range of phenomena associated with creativity if one assumes that the creative process is based on ordinary thought processes. In this chapter and the next, we will look more broadly at creativity and examine research and theory that go beyond the thought process. We will consider the question of whether one can classify people according to their creative abilities, and whether one can develop tests that enable us to predict who among us will be creative in their lives. Can we predict which third-grader will develop into a world-class scientist and which into the next Picasso? The general view that one can use tests of various sorts to measure people’s creative capacities and potential has been around since the turn of the twentieth century, but the modern interest in such questions began around 1950 and evolved out of the mental-testing or psychometric movement in the United States. In examining research coming out of the psychometric perspective, we will in this chapter consider the issue of whether there is a small universal set of general creative-thinking skills that underlie all creative thinking. Based on the discussion in Chapter 4 of the importance of domain-specific expertise in creative thinking, we would expect that there would not be such a set of general skills, but many researchers believe that general creative skills play a role in creative thinking, so we will examine research addressing that question. In the next chapter we will consider the question of the “creative personality.” We will address the question of whether there is a small set of psychological characteristics—personality characteristics—that distinguish creative from noncreative individuals and play a role in making some people creative. Outline of the Chapter This chapter begins by placing the psychometric perspective on creativity in its historical context, with a brief review of its development. The chapter then examines the work of Guilford (1950), which has been of seminal influence in setting the agenda in this area. Guilford was the first modern researcher to outline possible mental abilities underlying creativity, and he developed a set of tests to measure them. We shall review Guilford’s research and that of those who followed him, paying particular attention to the validity of the tests that have been developed to measure the capacity to think creatively. Guilford and the Modern Psychometric Perspective on Creativity It is not an exaggeration to say that we can trace the beginnings of the modern psychometric perspective on creativity to 1950 and the publication of Guilford’s 1949 presidential address to the American Psychological Association (APA; Guilford, 1950). Guilford was at that time one of the leaders of the psychometric movement in the United States. The president of the APA gives an address at the organization’s annual convention, and in 1949 Guilford used his address to propose that psychology take up the study of creativity. Guilford noted that very few studies of creativity had been carried out in psychology over the 25 years or so before his address, and he presented his ideas in the hope of stimulating other psychologists to take up the challenge of measuring creative-thinking capacity and creative potential, as well as the psychological characteristics of the individuals who succeed in being creative. Guilford (1950) first noted that the term genius had come to be used—incorrectly, in his view—to describe a person of high intelligence. The original referent of that term was someone who made unique creative contributions, and Guilford proposed that psychology should return to that original use. Psychologists should acknowledge that high intelligence quotient (IQ) was something different from the capacity to think creatively, which meant that one could not use existing IQ tests to measure people’s ability to think creatively. IQ tests are composed of problems and questions, each of which has one correct answer (see items in Table 9.1), but situations that demand creative thinking do not necessarily have only one correct answer. Indeed, a situation that demands creative thinking may not be a kind of situation in which there is even a correct answer: When a painter is painting a landscape, is there a correct way to do it? If this analysis is accurate, then creative-thinking capacity must be based on a set of skills different from those measured by IQ tests, and a new set of tests would have to be designed to capture it. Thus, Guilford began his address by trying to convince psychologists—many of whom were, as he was, psychometricians who spent much of their time thinking about and testing IQ—that they needed to develop a whole new set of measures and concepts in order to examine a different and at least equally important set of skills. Table 9.1 Examples of items from IQ tests 1. Two men, starting at the same point, walk in opposite directions for 4 meters, turn left, and walk another 3 meters. What is the distance between them? (a) 2 meters (b) 6 meters (c) 10 meters (d) 12.5 meters (e) 14 meters 2. Which of the following sentences given below means approximately the same as the proverb “Don’t count your chickens until they are hatched”? (a) Some eggs have double yolks, so you can’t really count eggs and chickens. (b) You can’t walk around the henhouse to count the eggs because it will disturb the hens and they won’t lay eggs. (c) It is not really sensible to rely on something that has not yet happened and may not ever happen. (d) Since eggs break so easily, you may not be accurate in your count of future chickens. 3. Following the pattern shown in the number sequence below, what is the missing number? 1 8 27 ? 125 216 (a) 36 (b) 45 (c) 46 (d) 64 (e) 99 4. Which of the designs best completes the following sequence? 5. Continue the following number series with the group of numbers below that best continues the series. 1 10 3 9 5 8 7 7 9 6 ? ? (a) 11 5 (b) 10 5 (c) 10 4 (d) 11 6 Answers: 1. c; 2. c; 3. d; 4. a; 5. a Although he argued that creativity could not be measured with intelligence tests, Guilford nonetheless based his approach to the study of creativity on his extensive experience in psychometrics. Just as psychologists had had success in developing tests that could be used to measure intelligence and to predict performance in school, so Guilford wanted to develop tests to measure creativity and, perhaps more important, creative potential. That is, the tests should not only allow us to designate who is creative now, they should also allow us to predict, for example, which of our children would be creative in the future—as adults. If we were able to do that, we could then provide enrichment opportunities for those individuals so that they could maximize their potential, to the ultimate betterment of all. In addition, we could provide enrichment opportunities for other children so that they might be more creatively productive in their lives than they otherwise would have been. Guilford argued that creative children were our most valuable resource and that it was important that they be identified and nurtured as early in life as possible to maximize the likelihood that they would put their potential to use. At the time of Guilford’s address, there was great concern in the United States that Communism was on the way to overwhelming the democratic West, and Guilford believed that identifying and nurturing creative talent was our best chance of winning the war for people’s hearts and minds. Accordingly, in his address, Guilford (1950) laid out a group of methods that could be used to test people for their capacity to think creatively, just as we test people for their capacity to think intelligently. Most significantly, in a further connection to the psychometric tradition, Guilford (1950) proposed that the entire person was involved in creativity, which meant that creativity was part of the personality of the person. Guilford’s view is thus an early example of what can be called a confluence view of creativity (Sternberg & Lubart, 1996; see Chapters 2 and 11), which assumes that the confluence or coming together of several factors is necessary to produce creativity. According to Guilford, simply possessing creative-thinking ability was not enough to guarantee creative productivity in one’s life: It was also necessary that the individual possess the relevant personality characteristics and motivation before that ability would bear fruit. Guilford (1950) was successful in his desire to stimulate psychologists to think about creativity: The last 50+ years have seen innumerable studies that have owed their existence to Guilford’s ideas, as psychologists attempted to measure the cognitive and personality characteristics of creative individuals. As will be discussed in Chapter 11, numerous confluence theories of creativity have been developed over the years since Guilford’s address, as psychologists have grappled with the complexities of creativity. In addition, many modern theories of creativity have built on the confluence perspective that he advocated. Methods of Measuring Creativity Psychologists and educators have developed many different types of measuring instruments designed to determine whether a given individual is creative (or has been creative), as well as to determine whether a person who as yet may not have produced anything creative has the potential to become creative. Instruments have also been developed that are designed to assess the characteristics of creative individuals once those individuals have been identified (Hocevar & Bachelor, 1989). Those instruments range from tests designed to measure the ability to think in a creative manner to biographical indices that ask the respondent to list accomplishments that would be considered creative. Measures of Creative Accomplishment I will now review some of the measures that have been used to assess an individual’s creative accomplishments and potential. I begin with measures of accomplishment. Achieved Eminence A relatively direct measure of an individual’s creative achievement is the eminence that he or she has achieved in a field that is assumed to require creativity for success, such as architecture, literature, painting, poetry, or the sciences. One can ascertain who is eminent by determining, for example, whose career and work are discussed in standard reference works in the field. One can examine anthologies of modern poetry to determine which living poets are represented. Alternatively, one can ask experts in the field to list people whose work has broken new ground. If one is interested in studying the psychological characteristics of individuals of creative achievement, this method leaves little doubt that one has identified such individuals. Researchers have brought eminent individuals into the laboratory in order to interview them in detail about their life experiences and to assess their personalities, working styles, and so forth (e.g., MacKinnon, 1962). In addition, the work of Simonton (e.g., 1999) and Martindale (e.g., 1990), which uses historiometric methods (and has already been discussed in several places in this text), has directly examined the products of eminent individuals. Similarly, the case studies presented elsewhere in this book have directly studied eminent individuals. In the latter cases, the individuals of eminence usually have been deceased, but the same principles apply: One studies individuals whose achievements within a creative profession have been acknowledged in some way by others in the field. One important issue raised by the study of eminent creative individuals is whether conclusions drawn from such studies are relevant to the general population. For example, if one examined the personality characteristics of eminent poets, say, would any characteristics found in such individuals also be found in “ordinary” people who write poetry only as an avocation? Similarly, are the thought processes used by eminent creative individuals—for example, the thought processes underlying creation of poetry that winds up in an anthology—the same as those used by ordinary individuals when they are producing poetry? As noted in Chapter 1, one cannot assume that the same processes are involved in production of world-class creative products and “ordinary” creative works. Ratings of a Person’s Creativity One way to acquire information about a person’s creativity is to ask others about it. If children are the subjects of study, for example, one can ask teachers to rate their students according to their creativity. The teachers could then be provided with examples of the kinds of behaviors that would be considered indicative of creativity: The teacher might be told that a creative child is one who produces unusual but useful ideas on class assignments, and who produces more ideas than other children. Similarly, to assess the creativity of research scientists at a company, one could ask their supervisors to rate them. One might also assess creativity by asking for peer evaluations—by asking a group of physicists, for example, to rate the most creative people in their field. We have already mentioned this as a method for determining eminence, and rating eminence may simply be one example of rating creativity, assuming that in some fields eminence depends on one’s creative production. Some studies of children’s creativity have also used children’s ratings of their peers as the index of creativity. Judgments of Products: Consensual Assessment Technique A relatively straightforward way to asses the creativity of an individual is to rate the creativity of an actual product that he or she has produced. This method was used in the study by Getzels and Csikszentmihalyi (1976) in which they studied “problem finding” in art students’ production of still-life drawings (see Chapters 3 and 12), and it has been used extensively by Amabile and her colleagues (see Chapters 2 and 11; for reviews, see Amabile, 1983, 1996). Researchers sometimes ask undergraduates to produce a poem, say, or a collage, which is then rated for creativity by poets or artists, respectively. In other studies, the judges have been nonexperts in the area. The technique has also been used with children, both as participants—that is, producers of the creative products—and as the judges of those products’ creativity. Tests of Creative-Thinking Capacity: Testing Divergent Thinking Beginning with the work of Guilford (e.g., 1950), researchers have developed tests designed to measure a person’s capacity to think creatively. Guilford grouped together several different subskills of thinking into the category of divergent thinking, which refers to the capacity to produce ideas that diverge from the ordinary, on the assumption that divergent thinking produces creative ideas. (See items in Table 9.2.) Tests have been designed for assessing divergent-thinking skills in children of various ages, as well as in adults, such as college students. Several critical assumptions underlie the development of such tests. First, it should be noted that materials such as those in Table 9.2 have nothing to do with specific creative domains of the sort in which researchers might usually be ultimately interested, such as painting, literature, science, or entrepreneurship in business. It is assumed that asking people about bricks and white edible things (see Table 9.2) can provide us with information that will be useful in predicting which individuals possess the thinking capacity to accomplish much in those creative domains. This assumes further that the specific processes underlying creative thinking must be relatively general in their applicability. So, for example, the ability to think of many members of a particular class (see Table 9.2F) might be relevant to thinking of a word or concept when one is writing a story. Second, it is assumed that creative thinking is separate from intelligence as a mode of thought. As noted earlier, a critical assumption behind Guilford’s (1950) proposal was that IQ as measured by intelligence tests was a different mode of thinking from the thinking underlying creative accomplishment. Those two sets of assumptions—the generality of creative-thinking skills as measured by divergent-thinking tests and the difference between IQ and creativity—have been the subject of much debate among researchers (see, e.g., Brown, 1989; Hocevar & Bachelor, 1989; see also Kaufman & Baer, 2002). I will examine those issues later in this chapter. Table 9.2 Divergent-thinking exercises (A) Suppose that humans suddenly no longer had to eat. List all the consequences that you can think of that would arise. (Give yourself 5 minutes.) (B) List all the problems or difficulties you can think of with the present-day toaster (5 min.). (C) List all the uses you can think of for a brick (3 min.). (D) List all the uses you can think of for a newspaper (3 min.). (E) List all the uses you can think of for a paper ctdp (3 min.). (F) List all the white edible things that you can (3 min.). (G) List all the words you can think of in response to the word mother (3 min.). (H) Figurative test items In Guilford’s (1950) original suggestions concerning tests to measure creative-thinking capacity, he presented divergent thinking as only one component of that capacity. Other skills also played a role, including “convergent thinking,” as well as the ability to evaluate ideas after they were produced (see Baer, 1993, and Runco, 1991). Convergent thinking is carried out after one has available several possible methods to solve a problem. One selects among them, converging on the final solution. Sometimes researchers have ignored those other aspects of Guilford’s analysis and have taken divergent thinking to be equivalent to creative thinking, and have labeled divergent-thinking tests as “creativity tests.” It must be kept in mind that those tests do not measure creativity: They only measure one component—although, in Guilford’s view, an important one—of the creative capacity. That distinction should be kept in mind while working through the important issues in this chapter. Similarly, some investigations of the personality characteristics of creative individuals have used performance on divergent-thinking tests as the sole criterion for classifying people as creative (for review, see Brown, 1989). Again, those tests do not provide that sort of information. Furthermore, sometimes creativity-training programs have used changes in divergent-thinking performance as evidence that the program has been successful in increasing creativity. Once again, however, increasing performance on divergent-thinking tests is not the same as increasing creativity. In Guilford’s (1950) view, divergent thinking was only one component of creative thinking, and the tests were to be used only to determine one component of the potential to think creatively. That distinction should be kept in mind throughout the discussion. Table 9.3 Sample items from attitude and interest inventories I have a good sense of humor. I like to try new activities and projects. I like to invent things. I like to write stories. I often daydream about unsolved problems. I have to learn things in my own way, rather than accepting ideas or relationships suggested in textbooks. If I had the necessary talent, I would enjoy being a sculptor. I would like to be an inventor. Attitude and Interest Inventories These questionnaires ask the person to agree or disagree with items such as those in Table 9.3, and they are based on the not unreasonable assumption that creative individuals will show interest in creative activities. Personality Inventories As noted earlier, Guilford (1950) proposed that an individual’s being creative was a function of his or her entire personality, and he suggested that personality be examined in order to provide a more complete understanding of the creative person. A number of researchers have developed personality inventories—questionnaires of various sorts—that are designed to measure the personal characteristics of creative people. An example of one sort of such an inventory is shown in Table 9.4. Researchers have also developed scales with which teachers rate the personality characteristics of their students (for review see Hocevar and Bachelor, 1989, pp. 55–56) in an attempt to determine whether a given child possesses personality characteristics assumed to be related to creativity. Many studies have investigated whether there are personality characteristics that distinguish people of creative accomplishment from other individuals (see Feist, 1999). In addition, some researchers have proposed that creativity is critically dependent on personality characteristics rather than being a particular mode of thinking. In this view (e.g., Dellas & Gaier, 1970), the creative individual is one whose personality is oriented toward production of new things and breaking away from the ordinary, and who uses his or her cognitive capacities to accomplish that task. Biographical Inventories and Self-Reports of Creative Activities These questionnaires ask about the person’s life experiences, such as whether he or she attempted to make things as a child, and the hobbies and other interests of the person as an adult. They also ask about the person’s family, such as the education levels of the parents, whether the parents died at an early age, presence of other siblings, and so forth. Items from one such inventory, the Lifetime Creativity Scale, are shown in Table 9.5 (Richards, Kinney, Benet, & Marzel, 1988). The Lifetime Creativity Scale is based on an individual’s responses during an interview and is completed by the interviewer rather than being filled out by the individual. Table 9.4 Gough Adjective Checklist, A Creative Personality scaleSource: Gough (1979). Table 9.5 The Lifetime Creativity scale: Examples and interpretation A. Items from biographical inventory Were you raised in a two-parent household? Do you have any siblings? What are their ages relative to yours? What were the occupations of your parents? What is the highest degree achieved by your parents? What occupations have you pursued in your adult life? Describe your major responsibilities and accomplishments within each occupation listed for the previous question. Describe your main avocational interests (hobbies). Describe your major accomplishments in the avocational interests listed for the previous question. B. Examples of results obtained from analyzing responses to questions such as those in panel A (or from an interview) Level Example 1 Example 2 Vocational history No significant peak creativity Mixed and carried mortar for local bricklayer for 20 years, then inherited a large income-paying trust fund and retired to a passive life on a country estate. Washed store windows for 3 years under foreman’s supervision, spent 5 years on assembly lines in two factories, and for the past 11 years has done routine quality-control tasks in a brewery. Some peak creativity Longtime owner and manager of a small dairy farm who, after 10 years of producing cheese and other dairy products, expanded and began marketing through a local distributor. Optician who spent 4 years selling optical items, then acquired a small optical shop, and now grinds lenses to prescription while managing the retailing of standard optical products. High peak creativity Former avant garde dancer and choreographer who developed and directed a variety of unusual productions for several dance companies but postwar has worked solely as a hotel clerk. Entrepreneur who advanced from apprentice to independent researcher of new products before starting a major paint manufacturing company, and whose operation surreptitiously manufactured and smuggled explosives for the Danish Resistance during World War II. Avocational history No significant peak creativity Once read movie magazines intensively but now spends most evenings with the new family television. Also does much needlepoint following specified patterns and, on weekends, watches spouse play handball. Often reads in spare time, has a standing subscription to the local theatre, belongs to a health club, has been going daily for directed group calisthenics, attends a social club to watch weekly televised sporting events with friends, and occasionally attends local soccer games. Some peak creativity Active church member who has ushered at church services for over 20 years, has sung for the last 10 years in the alto section of the church choir, and recently has been volunteering on a committee designed to expand parish membership. Avid reader and sports enthusiast who previously completed a night school journalism course and now writes brief accounts of sporting events for a monthly community newspaper. On weekends, also volunteers as an assistant coach for a children’s swim team. High peak creativity Amateur archaeologist who for years has spent summers and other free time seeking new sites, initiating archaeological digs with professionals from a nearby university, researching artifacts, reconstructing aspects of primitive societies, and collaborating in writing articles on this work. Invests much spare time in working with own handicapped child and once, over several years, designed and constructed a complex apparatus to help this child with locomotion, gross changes of posture, and fine motor manipulation of objects. Now works as a volunteer teacher twice a week to help other handicapped children use this invention. A relatively direct method of determining the creative capacity of people is to ask them to list their accomplishments in creative domains. One can ask if a person has had poems published, for example, or has had a gallery or museum show of paintings. Interviewers can also ask people about creativity in their ordinary activities, such as whether they produce new recipes when they are cooking (Carson, Peterson, & Higgins, 2003; see Table 9.6). One can also ask raters to judge the creativity of people’s creative works, based on those people’s descriptions of those works. King and colleagues (King, Walker, & Broyles, 1996), for example, asked a group of college undergraduates to list and describe any creative achievements from the past 2 years. Those lists of items were given to other individuals, some of whom simply counted the items, while others rated them for creative achievement. This method combines aspects of the consensual-assessment technique and self-reports of creative activities. Measuring Creativity: Summary We have now briefly surveyed the wide range of techniques that researchers have used to assess creative potential, creative accomplishments, and the characteristics of creative people. We will encounter many of those techniques in this chapter and the next one, as we examine research that has attempted to measure creative capacity and the creative personality. We turn first to studies of the creative-thinking capacity, which have used tests such as those in Table 9.2 in attempts to identify individuals with the capacity to produce novel ideas. Such individuals would seem to have the potential to be creative. Table 9.6 Creative achievement questionnaire (incomplete) by Shelley Carson, Harvard University I. Place a check mark beside the areas in which you feel you have more talent, ability, or training than the average person. ___ visual arts (painting, sculpture) ___ music ___ dance ___ individual sports (tennis, golf) ___ team sports ___ architectural design ___ entrepreneurial ventures ___ creative writing ___ humor ___ inventions ___ scientific inquiry ___ theater and film ___ culinary arts II. Place a check mark beside sentences that apply to you. Next to sentences with an asterisk (*), write the number of times this sentence applies to you. A. Visual Arts ___ 0. I have no training or recognized talent in this area. (Skip to Music). ___ 1. I have taken lessons in this area. ___ 2. People have commented on my talent in this area. ___ 3. I have won a prize or prizes at a juried art show. ___ 4. I have had a showing of my work in a gallery. ___ 5. I have sold a piece of my work. ___ 6. My work has been critiqued in local publications. *___ 7. My work has been critiqued in national publications. B. Music ___ 0. I have no training or recognized talent in this area (Skip to Dance). ___ 1. I play one or more musical instruments proficiently. ___ 2. I have played with a recognized orchestra or band. ___ 3. I have composed an original piece of music. ___ 4. My musical talent has been critiqued in a local publication. ___ 5. My composition has been recorded. ___ 6. Recordings of my composition have been sold publicly. ___ 7. My compositions have been critiqued in a national publication. C. Dance ___ 0. I have no training or recognized talent in this area (Skip to Architecture). ___ 1. I have danced with a recognized dance company. ___ 2. I have choreographed an original dance number. ___ 3. My choreography has been performed publicly. ___ 4. My dance abilities have been critiqued in a local publication. ___ 5. I have choreographed dance professionally. ___ 6. My choreography has been recognized by a local publication. *___ 7. My choreography has been recognized by a national publication. D. Architectural Design ___ 0. I do not have training or recognized talent in this area (Skip to Writing). ___ 1. I have designed an original structure. ___ 2. A structure designed by me has been constructed. ___ 3. I have sold an original architectural design. ___ 4. A structure that I have designed and sold has been built professionally. ___ 5. My architectural design has won an award or awards. ___ 6. My architectural design has been recognized in a local publication. *___ 7. My architectural design has been recognized in a national publication. E. Creative Writing ___ 0. I do not have training or recognized talent in this area (Skip to Humor). ___ 1. I have written an original short work (poem or short story). ___ 2. My work has won an award or prize. ___ 3. I have written an original long work (epic, novel, or play). ___ 4. I have sold my work to a publisher. ___ 5. My work has been printed and sold publicly. ___ 6. My work has been reviewed in local publications. ___ 7. My work has been reviewed in national publications. F. Humor ___ 0. I do not have recognized talent in this area (Skip to Inventions). ___ 1. People have often commented on my original sense of humor. ___ 2. I have created jokes that are now regularly repeated by others. ___ 3. I have written jokes for other people. ___ 4. I have written a joke or cartoon that has been published. ___ 5. I have worked as a professional comedian. ___ 6. I have worked as a professional comedy writer. ___ 7. My humor has been recognized in a national publication. G. Inventions ___ 0. I do not have recognized talent in this area. ___ 1. I regularly find novel uses for household objects. ___ 2. I have sketched out an invention and worked on its design flaws. ___ 3. I have created original software for a computer. ___ 4. I have built a prototype of one of my designed inventions. ___ 5. I have sold one of my inventions to people I know. *___ 6. I have received a patent for use of my inventions. *___ 7. I have sold one of my inventions to a manufacturing firm. H. Scientific Discovery ___ 0. I do not have training or recognized ability in this field (Skip to Theater). ___ 1. I often think about ways that scientific problems could be solved. ___ 2. I have won a prize at a science fair or other local competition. ___ 3. I have received a scholarship based on my work in science or medicine. ___ 4. I have been author or coauthor of a study published in a scientific journal. *___ 5. I have won a national prize in the field of science or medicine. *___ 6. I have received a grant to pursue my work in science or medicine. ___ 7. My work has been cited by other scientists in national publications. I. Theater and Film ___ 0. I do not have training or recognized ability in this field. ___ 1. I have performed in theater or film. ___ 2. My acting abilities have been recognized in a local publication. ___ 3. I have directed or produced a theater or film production. ___ 4. I have won an award or prize for acting in theater or film. ___ 5. I have been paid to act in theater or film. ___ 6. I have been paid to direct a theater or film production. *___ 7. My theatrical work has been recognized in a national publication. Scoring of the Creative Achievement Questionnaire Each checkmarked item receives the number of points represented by the question number adjacent to the checkmark. If an item is marked by an asterisk, multiply the number of times the item has been achieved by the number of the question to determine points for that item. Sum the total number of points within each domain to determine the domain score. Sum all ten domain scores to determine the total CAQ score. Cognitive Components of the Creative Process: Testing for Creative-Thinking Ability The modern psychometric perspective on creativity began with Guilford’s (1950) discussion of how one might go about measuring ordinary people’s capacity to think creatively. That is, let us say one has a group of college students or a group of schoolchildren, none of whom has yet achieved eminence. How might one determine which of those people is capable of thinking creatively and may develop into a person of accomplishment and perhaps even eminence in some creative domain? Since the people of interest are not yet specialists in any domain, one must develop tests that can be presented to nonspecialists and will tap into what one could call the essence of creative thinking. Thus, Guilford began by analyzing the ability to think creatively into its critical components, and he built his tests to measure those components. It is notable that Guilford’s (1950) analysis of the creative thought processes was not based on the study of creative people thinking creatively: He carried out no investigations of the creative thought processes in action. Rather, his postulation of the necessary components of the creative-thinking processes was based on his intuitions about, and logical analysis of, how the creative mind had to work, given the task of producing novel outcomes. This raises the question of whether it was best to begin by relying on one person’s intuition about the creative process (even if that person was a psychologist who had done creative work). It might have been more useful to first collect information about how the creative process worked and to use that information as the basis for developing tests designed to tap creative thinking (Brown, 1989). We will return to this issue when we evaluate research using Guilford’s tests and others like them. Sensitivity to Problems In Guilford’s view, in order to get the creative process started, an individual must see deficiencies in some aspect of the world (Guilford, 1950). Only then will he or she spend time contemplating what might be done to correct those problems, which is the first step toward producing a creative outcome. This view, of course, is a variation on the old idea that necessity is the mother of invention, and Guilford can be looked upon as making a broader claim: Necessity is the mother of creativity. As an example, a potential inventor might discover a problem with her car: Let us say that on one very hot summer day, when she gets into her car with the take-out lunch she just bought, she notices that none of the cup holders in her car is big enough to hold the extra-large drink she bought to stay hydrated. This discovery could stimulate the search for a way to overcome that problem. Another person might simply decide to stay away from extra-large drinks and not even consider the absence of a large cup holder to be a problem, and so would not think further about it. Only the first person would have the chance to produce a creative idea in that situation (e.g., making some sort of large-sized holder to fit into the existing small one, to accommodate large-sized drinks). Similarly, two scientists might read a research paper in which one result did not come out as expected. The first scientist might simply conclude that that result was due to random error in the experiment, and ignore it. The second scientist, in contrast, might find that aberrant result curious and consider it something to be explained, which might lead him to design a new experiment in order explore that finding further. That new experiment might lead to a creative advance in theory. Thus, as a first step toward creative thinking, a person must be sensitive to problems: Even the most creative thinker will produce nothing new if he or she does not think that there is any difficulty with the present situation. Exercises A and B in Table 9.2 are designed to measure a person’s sensitivity to problems; in each exercise, the person must be able to analyze the situation to determine what follows from it, including possible problems, which could then set the creative process into motion. One can test a large number of people, college students, say, or corporate executives, or advertising executives, on exercises such as those, and one can then determine the average number of problems found for each group. One can then determine if a given person finds more or fewer problems than average by comparing students to students and executives to executives. This would allow one to rank people on their sensitivity to problems. Is Necessity the Mother of Creativity? A question can be raised about the basic notion that sensitivity to problems is critical in setting the creative process in motion. It is no doubt true that many people are motivated to carry out creative activities because of problems they sense in their personal or professional environments. However, there is historical evidence that the creative process can be set in motion without necessity, even in the domain of invention. As one example, consider the invention of the airplane. At the end of the nineteenth century a number of research projects were underway whose purpose was the invention of a flying machine. We discussed several of those projects in Chapter 5. At that time, there was no need for such a machine; only gradually, after the Wright brothers were successful in inventing the airplane, did the broader implications of that invention become apparent. So the driving force behind the invention of the airplane seems not to have been necessity: There was no need to fly; people simply wanted to. Individuals sometimes think creatively because they want to accomplish something that might be possible but that has never been done before, or because they are curious as to what might happen when certain actions are carried out. Thus, necessity is not a mandatory condition for creativity. Components of Creative Thinking Let us now assume that the creative process has been set in motion, either because a problem has been identified, or because the person is challenged to produce something heretofore deemed impossible, or because he or she is curious as to what might happen if certain activities are carried out. The person must then begin to think about how to bring about whatever goal he or she is now working toward. We have now arrived at the core of the creative thought processes. In this context, Guilford (1950) concentrated on the generation of ideas. It seemed obvious to him that the more ideas a person can produce, the greater the chances that he or she will produce a useful one. That is, generating more ideas results in greater chances of a creative outcome. Therefore, he proposed that fluency of thought—that is, the capacity to produce a large number of ideas in a given period of time that are relevant to some situation—would be one characteristic of creative thinkers. In order to test for fluent thought, one can give people a task that requires that they produce multiple ideas, such as exercises C–H in Table 9.2. As with sensitivity to problems, one can determine if a given person has produced more or fewer ideas than average for his or her peer group. The exercises in Table 9.2 are of two kinds, verbal (C–G) and figurative (H). The inclusion of figurative exercises was designed to allow the assessment of creative-thinking capacity in individuals whose language might be less than fully developed, such as young children. In addition to being fluent in idea production, the creative thinker must also be a flexible thinker. According to Guilford (1950), creative thinking may require that one break away from one’s habitual ways of thinking and strike out in new directions. (This is a familiar idea by now.) One can score exercises C–F in Table 9.2 for flexibility of thought by determining whether a given person tends to stay within a category when generating ideas or to switch from one category to another. For example, in producing names of white edible things, a hypothetical person who says milk, cottage cheese, sour cream, vanilla ice cream, vanilla frozen yogurt would be relatively inflexible, since at most only two categories are used. A flexible person might say milk, white-chocolate-covered raisins, white corn, white wine, tofu, lobster, and in so doing would use six categories. The second person, in Guilford’s view, is more flexible in thinking, and so would be more likely to break away from habitual modes of thought and come up with novel ideas to deal with difficulties. Presumably, the person’s propensity to change categories when listing white edible things would be related to his or her ability to take a new perspective when, say, solving a problem or when creating a work of art. The creative thinker will also produce original ideas: ideas not produced by many other people. In addition to scoring the exercises in Table 9.2 for fluency, one can also score all of them for originality of ideas. For example, if a student lists a particular use for a brick that most students list, then there is little originality in that idea; however, if he or she lists a use for a brick that was thought of by no one else in a large group of people, that is an original idea. A person who produces many original ideas would in Guilford’s view be likely to produce creative solutions to problems, because difficult problems are presumably difficult because the solutions that are obvious to everyone do not work on them. Again, the underlying assumption is that producing an unusual use for a brick is related to producing unusual ideas in response to a problem or developing an unusual perspective in artistic endeavors. Guilford took measures of fluency, flexibility, and originality and combined them into divergent thinking, a mode of thinking that in his view plays a critical role in the creative process by enabling the person to produce ideas that diverge, or move away, from the usual. He assumed that divergent thinking is a general characteristic or trait of people, and that it is relevant across a broad range of activities that might be approached creatively. Thus, one can take a person’s performance on a battery of tests, such as those in Table 9.2C–H, and use them to predict that person’s performance in situations far removed from the testing context. As noted earlier, divergent thinking is contrasted with convergent thinking, which occurs when one uses available information to converge on the single answer that solves a problem. In a situation demanding a creative response, one would not have a method available, so one would first use divergent thinking to produce many possible ideas. Convergent thinking can then be used to narrow down those ideas to something potentially useful. As we have seen, over the years since Guilford’s original proposal, the notion of divergent thinking has gradually become the core of the psychometric view of creative thinking. (See discussion in Brown, 1989.) In a strict interpretation of Guilford’s views, this is not correct, since divergent thinking is only one component of a multicomponent process. However, in keeping with the thrust of the literature, much of this chapter’s discussion of the creative thought process will focus on divergent thinking. We noted earlier that Guilford used his expertise with IQ tests in developing tests to measure creative-thinking capacity. In a further analogy to the psychometrics of IQ, Guilford (1950) believed that each person is capable of creative thinking, at least to some degree. That is, divergent-thinking ability is normally distributed among the population. People who produce great creative advances (e.g., Picasso, Edison, Mozart) may possess the divergent-thinking capacities to a great degree, but we all have some of that ability. Indeed, if one did not make such an assumption, one could not talk about testing “creativity” and studying “the creative personality,” since usually one is not able to test people like Picasso and Mozart. Therefore, one must assume that there is continuity between the processes that the great creative thinkers use and those used by the rest of us when we produce our small examples of creativity, such as when we think up a new use for a paper clip. It should be kept in mind, however, that this assumption of continuity of process is just that—an assumption. It does not necessarily follow that the processes involved in creative thinking are the same for all of us, as has already been discussed in several earlier chapters. It is entirely possible, for example, that Picasso and Mozart thought differently from the rest of us. In order to conclude that there is continuity of process involved across all humans, it is necessary to provide evidence that the same sorts of processes are actually used. That is one reason why case studies of seminal creative advances are presented in many places in this book, since such case studies can provide evidence for continuity of thought across people. It is interesting to note further that, while I believe that the case studies provide evidence of continuity of thinking, I do not believe that they provide evidence for divergent thinking as the basis for creativity. This point will be elaborated shortly. Other Tests of Creative-Thinking Capacity Guilford’s proposed tests of creative-thinking capacity were elaborated and extended by him and his research group over the years (e.g., Guilford, 1967), but other researchers also developed batteries of tests to measure creative-thinking abilities (e.g., Kogan & Wallach, 1965; Torrance, 1974). Those batteries owed much to Guilford; when one examines the tests, one finds the same basic types of items, with variations in content rather than basic differences in the logic underlying the design of the tests. In addition to tests measuring divergent thinking in adults, researchers also developed comparable tests designed for children (Kogan & Wallach, 1965) to allow educators to determine as early as possible who among their students was capable of producing creative ideas if given the chance. This would allow educational institutions to provide additional support and enrichment for such children, thereby ensuring that society made the best use of this valuable resource. As noted earlier, Guilford’s early work on testing and measuring creativity was carried out during the cold war era, when there was much concern that the United States and its allies might be overrun by the Communists, and there is an urgency in some of Guilford’s writing as he warns about the possible dire consequences if the Western democracies do not make the best use of their creative thinkers (e.g., Guilford, 1950). One aspect of the administration of divergent-thinking batteries became a source of further research. Guilford’s tests were given in an atmosphere much like that in which IQ tests are given: a formal classroom atmosphere. However, Wallach and Kogan (1965) proposed that measurement of creative capacity, especially in children, should be carried out in a different sort of environment. They developed a battery of tests that were similar to Guilford’s in content but designed to be given in a playlike atmosphere. Numerous studies have examined the effects of different testing environments on performance on divergent-thinking tests (for review, see Barron & Harrington, 1981). Mednick’s Remote Associates Test A different slant on assessing the capacity to think creatively was developed by Mednick (1962) from the idea that creative people are able to generate unusual ideas because of the structure of the associations in their memories. Mednick’s view has already been briefly discussed in Chapter 8, in the context of Poincaré’s (1913) ideas concerning remote associations and creativity and Simonton’s (1988) theory of unconscious processing in creative thinking. Mednick approached the question of creative thinking from the then-dominant S-R perspective in American psychology. He proposed that one could analyze a situation that requires creative thinking as a stimulus, which produces associated responses in the thinker. Creative thinking means that one produces unusual or infrequent responses to a stimulus. The critical question is what enables some people to produce those responses. According to Mednick’s analysis, the crucial difference among thinkers is the organization of their associative hierarchies, or sets of associative responses to stimulus situations. Noncreative people, shown in Figure 9.1A (presented earlier as Fig. 8.2), have steep hierarchies, with a strong or dominant response to a given situation, that will tend to be given all the time, and producing it will make it harder for the thinker to produce a less frequent response. As an example, if someone says up, all I can think of is down; it is my dominant response to that stimulus. The creative person, on the other hand, has a flat hierarchy (see Fig. 9.1), with several responses available to the situation, none of which is particularly strong. Therefore, this person will have a greater tendency to produce unusual responses and thus may think of an original response to the situation, which could result in a creative solution to a problem. Mednick (1962) was proposing a theoretical mechanism to explain differences among people in Guilford’s divergent-thinking ability: People with flat associative hierarchies would be divergent thinkers. Mednick’s (1962) distinction between steep and flat associative hierarchies has been influential in psychology. Simonton (e.g., 1988, 1995) incorporated those ideas into his theory, discussed in detail in Chapter 11 (see also Chapter 8), and other psychologists also refer to Mednick’s ideas when discussing factors that influence the ability to produce novel ideas (e.g., Ansburg & Hill, 2003). Figure 9.1 Mednick’s theorySource: Adapted from Mednick (1962). In order to measure people’s associative hierarchies, Mednick developed the Remote Associates Test (the RAT). Some examples of the items on this test are shown in Table 9.7. The three words in each RAT item are each related to a single target word, but only weakly. Therefore, in order to think of the answer, one has to be able to move beyond the dominant responses to each of the words to the less common ones. We are already familiar with RAT-like problems from the discussion in Chapter 6 of Bowden and Beeman’s (1998) work on insight in problem solving. The verbal problems that they used were based on RAT items (see Figure 6.5D, p. 299). Mednick assumed that the person who is able to answer RAT items correctly possesses flat associative hierarchies, which should enable him or her to produce original and potentially creative responses in other situations. One might think that the RAT is curious as a test of creative-thinking capacity, since the RAT itself does not seem to require creative thinking in that there is a single correct response to each item (e.g., Ochse, 1990). That is, correctly answering an item on the RAT does not seem to require divergent thinking. This would seem to go against the whole purpose of testing creative thinking. However, in Mednick’s defense, it should be emphasized that the RAT itself is not directly measuring creative thinking; it is measuring the capacity to think creatively. This point has already been discussed in the context of Guilford’s tests. If one can answer RAT items correctly, it is presumably because one possesses flat hierarchies, which can then be put to use in other situations, such as those demanding open-ended creative thinking. As an analogy, if one wants to test the potential of someone to do well in marathons, one might measure lung capacity rather than running speed. Lung capacity is presumably related to the ability to successfully compete in a marathon, and so measuring lung capacity might be a good test even though no running is involved. Similar logic lies behind the design of the RAT. Table 9.7 RAT-related items Each of the three words in each stimulus set is related to the same fourth word, with which it can be combined to produce a familiar English phrase. Stimulus set Solution piggy / green / lash mark / shelf / telephone surprise / political / favor stick / maker / tennis cream / cottage / cloth back book party match cheese Tests of Creative-Thinking Capacity: Summary The tests developed by Guilford and those who were influenced by his work and Mednick’s (1962) RAT can be looked upon as measuring complementary aspects of the creative process. Guilford (1950) developed tests to measure, among other things, divergent thinking. In his view, divergent thinking is made up of the basic components of creative thinking, which is built on fluent, flexible, and original thinking. He developed a number of different measures that allow one to determine relatively easily how a given person compares with the average in this ability. Those tests, and others like them developed by other researchers, have been used in many different settings: from schools, where they are used in screening children for gifted programs; to industry, where they are used in hiring decisions; to research settings, where they are used to select “creative” participants for research studies. Mednick’s RAT is designed to illuminate the mechanism assumed to underlie the capacity for divergent thinking: the associative hierarchies of the would-be creative thinker. The person with steep hierarchies will not be able to produce the remote associates necessary in confronting situations that demand divergent thinking. It should be clear from this discussion that the basic perspective outlined by Guilford and adopted by those who use tests to measure creative-thinking capacity is that creative thinking involves a set of traits that are different from those underlying ordinary thinking. As noted in Chapter 2, this perspective has been adopted by many psychologists who do not directly use tests to measure creative thinking. Guilford’s notion of divergent thinking as the cognitive capacity forming the foundation of creative thinking, and the flat hierarchies proposed by Mednick as the basis of that skill, are assumed to occur in various degrees across people. The creative thinker of the first rank is assumed to possess those capacities to an extreme degree, so much so that such an individual can be assumed to be basically different from the average person (e.g., Csikszentmihalyi, 1996; Simonton, 1999). Testing the Tests: The Reliability and Validity of Tests of Creative-Thinking Capacity I have now outlined the structure and the logic behind tests designed to measure creative-thinking capacity. As noted earlier, since Guilford’s (1950) address and his pioneering work in developing tests to measure the capacity to think creatively, there has developed a wide and deep stream of research that has followed his lead. That research, which has concentrated on using tests to measure the creative capacity as well as relating test performance to other aspects of personality, has become a dominant stream in the psychological study of creativity. However, along with this stream of research carrying Guilford’s ideas forward, there have developed questions as to the usefulness of tests designed to measure creative capacity. Because of the important role of those tests in the modern study of creativity, it is critical that we examine in detail both sides of the question of the usefulness of the purported creative-thinking tests. Are Tests of Creative-Thinking Capacity Reliable? The first question that one must ask about any measurement instrument, whether it is a bathroom scale or a test of creative-thinking capacity, is whether it is reliable. That is, does the test produce consistent outcomes? For example, if you weighed yourself three times within a minute and your scale gave readings of 150 pounds, 100 pounds, and then 200 pounds, it would be obvious that the time had come to replace the scale: It is not a reliable instrument. Similarly, if you test a group of eighth graders using a battery of divergent-thinking tests and come back in a few weeks and do it again, the scores should be about the same (assuming for the sake of discussion that there is not much carryover from the first test and also that the students did not do any special preparation in the interval). This stability of a test over several administrations is called test-retest reliability. Demonstrating test-retest reliability is of critical importance for any test, because it means that we can have confidence in the scores obtained by people when they take it. Another form of reliability becomes important when a psychological measuring instrument contains multiple items, as do the divergent-thinking tests presented in Table 9.2. One would like to combine all those items together in scoring each person, because more items would mean a more stable score. That means that one would hope that the various items would give similar scores, since they have been designed to measure the same capacity (i.e., the capacity to think creatively). In order to determine the consistency of the various items on the test, one can divide the test into halves, by taking alternating items, say. One can then determine each person’s score on each half of the test. If the two sets of items are reliable in measuring the same capacity, a given person’s score on the two halves of the test should be about the same. This is called split-half reliability. Research studies have found that divergent-thinking tests are reliable; that is, they produce outcomes that are reasonably consistent (Barron & Harrington, 1981). That means, as noted, that we can be confident that a person’s score is representative of his or her performance. There is, however, one caution that must be given here: It is sometimes found that performance on divergent-thinking tests is affected by the conditions under which the tests are given. For example, if you instruct people to be creative in their responses, they may score higher than if you say nothing about being creative on the test. This seems to indicate that divergent-thinking ability is a strategy that can be applied to a testing situation, rather than some automatically engaged and unchanging trait or characteristic of a person. It is interesting to consider further what we might conclude from the finding that one can change people’s performance on a divergent-thinking test by telling them to be creative. Consider performance on an intelligence test as an analogous situation. Presumably, performance on IQ tests would not be affected by instructions to “be smart.” One does not have to tell schoolchildren to try to do their best to answer the questions when they take an IQ test, or that they are not to produce multiple answers for each question, or that they are to mark the answer they think is best, or that they should try to be accurate. There is only one strategy that seems reasonable on such a test, and schoolchildren know what it is. On the one hand, that result might lead one to the conclusion that an IQ test is measuring some constant characteristic of a person, and creative-thinking tests are therefore deficient in some way. However, the reason that children’s performance on IQ tests would not be affected by instructions to be smart might be that they have been exposed to such tests from an early age, and so they have been socialized to deal with them in the ways that the administrators of the tests expect them to. Tests designed to measure creative-thinking capacity are a different matter, for several reasons. First, since there is no single correct answer for the items on such tests, the test taker has at least two strategies available: to produce common responses or to produce unique responses. In addition, divergent-thinking tests are not usually given in most schools. Given the typical child’s lack of familiarity with the tests, it may not be surprising that additional instruction may sometimes be needed. Thus, it is not necessarily a flaw in the tests that they might be affected by instructions. On the basis of the available evidence, it seems that creative-thinking tests are reliable instruments. Are Creative-Thinking Tests a Valid Measure of Creative Capacity? The conclusion that tests designed to measure creative-thinking capacity are reliable leads to a second question: Just what is it that those instruments measure? Obviously, use of the tests is based on the assumption that they measure the capacity to think creatively, which is what they were designed to measure. However, we do not as yet have any evidence that that is what the tests do. As was mentioned earlier, Guilford (1950) designed the tests on the basis of his analysis of what must be involved in creativity. However, because he did not study creative thinkers to determine how they actually think (Brown, 1989), we have no idea whether Guilford’s intuitive analysis of the creative process was accurate; accordingly, we have no idea whether the creative-thinking tests actually measure creative-thinking capacity. The question of whether a test measures what it was designed to measure is the question of whether the test is valid: A valid test measures what it is supposed to. If a test is not valid, then it may be reliable, but it will be useless. Your bathroom scale may be extremely reliable, but it is useless if you want to measure your IQ or the amount of money in your savings account. It is possible to discern three periods in the development of tests of creative-thinking capacity. The first period began with Guilford’s (1950) APA address, which stimulated a great deal of research on creativity using his tests and similar ones developed by others (e.g., Kogan & Wallach, 1965; Torrance, 1974). However, some researchers began to raise questions about the usefulness of such tests, including some of the researchers who developed them (e.g., Crockenburg, 1972; Kogan & Pankove, 1974). By the 1980s, strong reservations were voiced concerning the value of that direction of research, and some researchers in the field began to move away from these instruments (e.g., Brown, 1989; Weisberg, 1986). In response to those criticisms, supporters of the psychometric perspective have presented new findings and have reanalyzed old results in an attempt to demonstrate that rejection of divergent-thinking tests was an error (e.g., Plucker, 1999a). I will trace these developments in the debate on the validity of tests designed to measure the capacity to think creatively. As we just saw in the case of reliability, there are several different sorts of validity that we can examine. Discriminant Validity Tests of creative-thinking capacity should measure something unique, especially something different from what intelligence tests measure. That is, these tests should be discriminating between what they measure. If a divergent-thinking test measures a capacity that IQ tests do not measure, then the former test is said to have discriminant validity. Barron and Harrington (1981) reviewed many studies that examined the relationship between the two types of tests, and the results varied. Some studies reported relatively high correlations between divergent-thinking scores and IQ scores, which might mean that they do not involve two different abilities. However, other studies reported nonsignificant relations between the two, indicating that they might be independent. Barron and Harrington noted that the inconsistent results might be due to the specific sample of individuals being tested, to the specific tests used (especially the divergent-thinking tests), and to the conditions under which the tests were administered. As we saw earlier, for example, scores on divergent-thinking tests can be significantly affected by the testing environment. It is also interesting that it has been found that ratings of students’ creativity by their teachers are sometimes very highly and positively correlated with the teachers’ ratings of the students’ intelligence (for review, see Hocevar & Bachelor, 1989). Although that result has nothing to do with the validity of creative-thinking tests, per se, it does raise an interesting question as to whether creativity and intelligence are independent dimensions of a person’s performance. Some have proposed that creative thinking is carried out by the same processes as those involved in performance on IQ tests—that is, that there is no special “creative-thinking process,” per se. According to this view, the only difference between the creative and the noncreative individual is the difference in attitude that they bring to tasks: The individual who produces creative outcomes wants to use his or her intelligence to produce new things (e.g., Dellas & Gaier, 1970). The cognitive position that underlies this book is related to that view (see Chapters 3 and 4), and this issue will be discussed further in the next chapter. Face Validity Sometimes test items obviously measure what they were designed to measure. Under such circumstances, one can say that the test is valid on its face; that is, the test can be said to have face validity. For example, if one had designed a test to measure a person’s knowledge of professional basketball, and some of the test items required recalling the names of the professional teams or of the recent championship teams, then the test would seem to have face validity. As for creative-thinking tests, I feel that one can question their face validity. Most researchers accept on its face Guilford’s (1950) analysis of the creative process—that is, most researchers agree that divergent thinking, defined as the production of many and varied ideas, is the foundation for creative thinking. Similarly, there is near-universal acceptance that “remote” associations, à la Poincaré (1913) and Mednick (1962), are critical in bringing about production of creative ideas. Those two assumptions are, as noted earlier, two facets of the same general perspective: The remote-associates concept can be looked upon as the mechanism underlying divergent-thinking ability. However, there is evidence that this remote-associates view of creative thinking may not be correct. Based on the information that we have available concerning the way the creative process works, from the case studies presented earlier as well as from studies of problem solving discussed in Chapters 3 and 6 (e.g., Fleck & Weisberg, 2004), one can challenge the theory of remote associates as the basis for creative thinking, and, ipso facto, the face validity of the creative-thinking tests that we have examined. In considering in this context the case studies from the arts (e.g., Picasso’s creation of Guernica; Calder’s invention of mobiles), the sciences (the discovery of the double helix), and invention (Edison’s invention of the light bulb; the Wright brothers’ invention of the airplane), we have seen that the creative process does not seem to work in the way that Guilford (1950) and others assumed when they developed divergent-thinking tests. The basic assumption behind the divergent-thinking tests in Table 9.2 is that producing many ideas is critical for producing creative ones. This is based on the further assumption that the creative process works in two stages: first by producing many ideas and then by keeping the good ones. The creative products that we have examined, however, were not produced by such a two-phase process. Rather, from the beginning the creative process was focused in a top-down manner, so that promising directions were taken right from the beginning. There was in none of those cases an initial period during which ideas were simply randomly produced in the hope that something relevant might be found at a later stage. If the conclusions from the case studies are valid, they indicate that most tests of creative-thinking capacity possess only apparent face validity. That is, although Guilford’s description of the creative process seems plausible to many people, it still may not be accurate. Laboratory studies of problem solving, including studies of solution of insight problems, have produced similar conclusions (see Chapter 6). The study of insight problem solving by Fleck and Weisberg (2004, 2006; see also Weisberg, 1980; Weisberg & Suls, 1973) has shown that those problems too are not approached in two stages, the first of which involves something like divergent thinking. Top-down processes are operating from the very beginning in problem solving also, meaning that tentative solutions that are produced are based on the analysis of the problem and the match between that analysis and the person’s knowledge, rather than on some divergent-thinking process that is independent of the problem the person is facing. Mednick’s RAT (1962) was designed to assess an individual’s potential to think creatively, as hypothesized by Mednick in his analysis of creative thinking in S-R terms. In Mednick’s view, the ability to produce creative responses in some situation is the result of a flat associative hierarchy. One can raise the same criticisms about the face validity of Mednick’s analysis of the creative process as have been raised concerning that of Guilford (1950). That is, Mednick did not obtain detailed records concerning how the creative process actually worked when people were producing new things; he assumed that the process worked through remote associations. He made no attempt to set up laboratory situations or to examine case studies to determine if there was evidence that the creative process could be described as relying on remote associations. In contrast to Mednick’s assumptions, however, results from case studies of seminal creative advances, as well as from laboratory studies of problem solving (all of which were reviewed earlier in this book), do not support the notion that creative thinking depends on remote associations. For example, we saw that Watson and Crick did not think remotely when they were deciding on a general strategy for pursuing the structure of DNA: They used Pauling’s modeling approach, which came from the content of their expertise. In addition, in choosing the specific path that they would take, they assumed that DNA was helical, again an idea that was not remote from their expertise. Similarly, Picasso’s creation of Guernica was deeply rooted in his expertise and required no radical break into some area remote from his ordinary ways of thinking. So, although many researchers seem to accept the face validity of Mednick’s (1962) approach, and his notion that creative thinking depends on remote associations (which, as noted earlier, Mednick traces back to Poincaré), the face validity of those ideas may be illusory. In sum, one can on several grounds raise questions about the face validity of divergent-thinking tests of creative capacity. It should be noted, however, that a lack of face validity is not necessarily a critical flaw in a test. A test can still be useful even if it does not possess face validity: It still might serve as the basis for classifying and selecting individuals for some purpose. As a hypothetical example, let us say that there is a significant positive correlation between the number of letters in a horse’s name and his or her performance in the Kentucky Derby. (Don’t concern yourself with why that might be the case; for the time being let us simply assume that it is true. I would guess that it is not true, although I have not checked.) If one were interested in wagering on the Derby, one could then use name length as the basis for choosing the horse on which to bet. In such a case, the “test” does not possess face validity, but it is nonetheless useful. It has criterion validity: It correlates with some criterion performance in which one is interested (i.e., running fast in a horse race). One could also say that the name-length test possesses predictive validity: It enables us to predict performance in some other situation. So we can also review studies that have examined the predictive validity of divergent-thinking tests. Criterion and predictive validity are usually discussed separately in the literature, but they are closely related, since in both cases one is using performance in one situation to predict performance in another. The only difference between them is whether the second performance (the predicted performance) is assessed in the same situation or at a later time. So we now turn to the question of whether divergent-thinking tests possess criterion and predictive validity. Criterion and Predictive Validity of Divergent-Thinking Tests If tests of creative capacity possess criterion validity, performance on those tests should be correlated with some external criterion of creative performance. As an analogy, when IQ tests were developed around the turn of the twentieth century by Alfred Binet, their initial purpose was to differentiate between bright and dull schoolchildren (Brown, 1989). The success of the tests was relatively easy to determine: One had only to look at how well the tests differentiated children who performed well in school from those who performed poorly. The reason Binet’s test was adopted was that it was possible to determine that it was successful. That is, the test demonstrated criterion validity. When Guilford (1950) presented his suggestions for creativity tests, in contrast, his bases for the various measures, such as those outlined in Table 9.2, were his intuitions on how the creative process must work: It seemed reasonable to Guilford that the creative person would first produce many ideas and then select the good ones for use. Thus, there was no external criterion of creative performance that guided Guilford’s work and that could be pointed to as an indicator of success of the divergent-thinking tests as predictors of creative-thinking capacity. He and his colleagues did not attempt to correlate their tests with criterion measures of creative performance. Indeed, Guilford noted that before bothering creative individuals by asking them to submit to testing, it was first necessary that the tests be developed to a considerable degree. Over the years there have been studies on the relationship between performance on creative-thinking tests and various criterion measures of creativity, and the results have been mixed (for reviews, see Brown 1989; Hocevar & Bachelor, 1989). In one example, Baer (1993) carried out a number of studies in which people of various ages, ranging from schoolchildren to adults, were asked to produce creative products of various sorts (e.g., write a poem or a story in response to a picture; construct a collage). The participants were also given a divergent-thinking test, and it was consistently found that divergent-thinking performance was not significantly correlated with performance on the creative tasks. Such results raise doubt about the criterion validity of divergent-thinking tests. Guilford (1950) originally proposed that tests of creative-thinking capacity would be useful in allowing us to predict who among us would develop into creative adults. That is, he believed that such tests would have predictive validity. In addition to criterion validity, the question of predictive validity is critical for any creative-thinking test. As noted, criterion and predictive validity are closely related: One could say that predictive validity is criterion validity with a time delay. As with other aspects of the validity of divergent-thinking tests, the support for their predictive validity is mixed. It is somewhat difficult to review this literature, because consistent findings are not easy to discern. Baer (1993) reviews the literature on the predictive validity of divergent-thinking tests and concludes that support is poor. In contrast, Plucker (1999) has concluded that the tests are of value. I will review several studies that have used divergent-thinking tests to try to predict the future creative accomplishments of individuals and will demonstrate the mixed outcomes of such studies (for further review, see Brown, 1989; Hocevar & Bachelor, 1989). Cropley (1972) gave divergent-thinking tests to junior high school students. Five years later, the creative achievements of the students were assessed in four areas: art, drama, literature, and music. Cropley found no significant correlations between divergent-thinking scores and creative achievements, so the predictive validity of the tests was not supported. Similar results were reported by Kogan and Pankove (1974), who used the Wallach and Kogan tests (1965) to measure creative-thinking capacity in children. The children were tested in the 5th and 10th grades, and they filled out a questionnaire about their overall accomplishments when they were high school seniors. The question of interest was whether their divergent-thinking scores had predicted the creative accomplishments they reported as seniors. The 10th-grade divergent-thinking scores made what was called a “marginally significant contribution” in predicting activities and accomplishments reported 2 years later. That is, there was a weak effect. However, there were no significant predictions from the 5th-grade tests to the senior year. Thus, while there might have been a small amount of predictive usefulness of the test scores, it was only helpful over a relatively short period of time. In addition, Wallach and Pankove found that IQ scores obtained at both grades made stronger predictive contributions than did the divergent-thinking scores for both grade levels at 12th grade. This takes us back to the question of the discriminant validity of divergent-thinking tests: In this study, it looks like obtaining divergent-thinking scores did not provide any information above what would have been provided by IQ scores. In another investigation of the criterion validity of divergent-thinking tests, Hocevar (1980) tested 94 university students and obtained measures of intelligence and ideational fluency, which, as noted above, is one component of divergent thinking. A self-report index of creative activities (similar to that in Table 9.6) was used to assess creative achievement. Results indicated that neither kind of test—IQ or divergent thinking—predicted creativity with more than modest accuracy. In addition, Hocevar reported that there were large differences in the predictive ability of the two types of tests across domains, and in some cases IQ was the superior predictor. So this study also fails to provide strong support for the predictive validity (or for the discriminant validity) of divergent-thinking tests. In a study mentioned in Chapter 3, Getzels and Csikszentmihalyi (1976) had fine-arts students draw a still life that they had first created out of objects that the experimenters made available to them. In addition to carrying out the drawing task, the artists were also given personality tests and divergent-thinking tests. The quality of the still life drawings was judged by a panel of artists. It was found, as noted earlier, that the student artists who spent the most time preparing the still life produced paintings that were judged most favorably by the panel of artists. This indicated that preparation was critical in determining the quality of the final product. Getzels and Csikszentmihalyi (1976) called this preliminary time period “problem finding,” although, as noted in Chapter 3, in my view the student artists were simply solving the problem of arranging a still life that was to be drawn. (For further discussion of this issue, see Chapter 12.) In addition, the artists who produced the highest-quality paintings also had a greater tendency to change their arrangements as they were working; they were more open to changes in the structure as they were working. The other student artists tended to close off their options earlier. As just noted, Getzels and Csikszentmihalyi (1976) also had the student artists take divergent-thinking tests, and none of those scores was related to the ratings of the drawings, so they were not effective predictors of artistic performance. In addition, a follow-up study was carried out 7 years later, when the students had had a chance to establish their careers. Although 7 years is not a lot of time, there were already some clear differences in career success among the artists. Some had dropped out of the art world; others were working as commercial artists, which meant that they had for the time being changed their career path within art in a direction away from fine art. One artist had had a solo show, and one of his paintings had been purchased by a museum, which indicated that he had achieved early success. Getzels and Csikszentmihalyi reexamined the measures obtained in the initial study to see which of them, if any, might predict career success. The preparation-time (“problem-finding”) measure, which was related to the quality judgments of the original drawings, was also related to career success several years out of art school. This indicates that the student artists’ early habits and attitudes toward their works were also important in their long-run success. Of direct relevance to the present discussion, performance on the divergent-thinking tests was not related to career success, so in this study the tests were not valid predictors of later creative performance. As noted earlier, Csikszentmihalyi (1999) reported an 18-year follow-up study, and again only the problem-finding measures significantly predicted career success. We have now reviewed several studies that have provided at best weak support for the predictive validity of divergent-thinking scores. As noted earlier, however, this area is not one in which simple conclusions can be drawn. Several other studies have produced results that have been taken as support for the validity of divergent-thinking tests. Reviewing those studies will provide evidence of the complexities involved in assessing people’s creative-thinking capacities. Soon after Guilford’s (1950) discussion of creativity testing, Torrance (e.g., 1974) developed a battery of tests of a similar nature. Torrance was concerned about the predictive validity of his tests and carried out several studies examining whether the tests were useful in predicting creative accomplishment, both short term and long term. Torrance reported 12 studies examining the short-term predictive validity of the Torrance tests. The participants ranged from kindergarten children to adults, with sample sizes from 12 to 133. Torrance examined a wide range of behaviors, some of which have been called into question by other researchers as criteria for creative accomplishment. The predicted behaviors included psychiatrists’ assessments of fourth graders’ sense of humor and strength of self-image. Those measures were obtained by having the children respond to projective techniques, such as the Thematic Apperception Test (TAT). The originality of stories written by sixth-grade students was assessed. For younger children—kindergartners and first graders—one criterion behavior was development of Piagetian conservation (e.g., conservation of liquid: the child’s developing belief, first examined by Piaget, 1965, that the amount of liquid stays constant when it is poured from a tall thin container, say, into a wide low one). For adults, one criterion was success in teaching in inner-city schools. Torrance reported significant correlations between divergent-thinking scores and various criterion variables, but one can question the validity of those criteria as measures of creativity (Baer, 1993). What do the strength of a child’s self-image and the development of conservation of liquid have to do with creativity? Similarly, one might successfully teach in an inner-city school without being creative. Given the questions that can be raised about the criteria, one might conclude that those short-term studies do not provide evidence either for or against the predictive validity of the Torrance tests. Torrance (1974) also reported six long-term correlational studies, which are more germane to the present discussion. Two of those studies are major studies, in that they involved more participants, the participants were followed over longer time periods, and more rigorous procedures were used to measure creative accomplishment. I will examine in more detail those two studies, because of their importance and because one of them has become the focus of a recent reanalysis. The first study began in 1959, when all 329 students in grades 7–12 in the University of Minnesota High School were given Torrance Verbal Tests. They were very able students: Their mean IQ was 118. Twelve years later, 234 of those students completed questionnaires about their activities in a variety of areas. Examples of activities that were included in the assessment of creative achievement were subscribing to a professional journal; learning a new language; writing songs, plays, and so on; changing one’s religion; performing on TV or radio; and publishing a professional paper. The researchers obtained three key indices from those responses. First, the quantity of creative achievements was determined. Second, the quality of creative achievements was measured. Independent judges scored each respondent’s three most creative achievements. Finally, respondents were scored for the creativity of their aspirations: what they would like to do in the future. The results of the study were consistently positive: The three criterion indices correlated +.27 to +.45 with scores from tests taken 12 years earlier. So this study indicates that one can use divergent-thinking scores (at least, verbal divergent-thinking scores) to predict creative achievements in the future. However, as noted for the short-term studies, one can raise questions about the results here also (e.g., Baer, 1993). First of all, one can again question the relevance of the criterion variables to creativity, and this seems to be especially true with the quantity measures. Are those achievements creative achievements? If not, then the more or less accurate predictions are irrelevant. A different problem concerns the relationship of those achievements to general intelligence. The criterion variables (e.g., subscribing to a professional journal, learning a foreign language) are things intelligent people do. In this regard, it is important to note that in this study the correlations between intelligence and the criterion creativity variables were as high as those between the Torrance tests and the criteria. In addition, the Torrance tests also correlate with intelligence, so perhaps the predictability of the creative achievements (ignoring for the moment the question of whether those achievements on the whole involve true creativity) is based on intelligence and not on divergent-thinking ability. One problem with the design that Torrance used is that it does not allow one to determine which variable (Torrance scores or IQ) is the stronger predictor of the criterion variables. Torrance did not use a multiple-regression design, which would have allowed that more sophisticated question to be answered. Baer (1993) makes several other criticisms of this study, and the interested reader should see his discussion. It thus seems that one of the major studies of the long-term predictive validity of divergent-thinking tests has possible design flaws that render it of little use in drawing firm conclusions concerning the usefulness of such tests. However, Plucker (1999a) recently reanalyzed the data from a second long-term study conducted by Torrance (1974) and found more positive results concerning predictive validity of the divergent-thinking tests. In Torrance’s second study, all the children in two elementary schools in Minnesota were given the Torrance test battery multiple times over the years 1958–1965. Other measures, including IQ scores, were also obtained. The study focused on 400 children who had been given the divergent-thinking tests for three consecutive years over grades 5–8, because multiple measures provide the most reliable estimate of creative-thinking potential. In 1980, 22 years after the study was begun, the students, now approximately 28 years of age, were contacted, and 220 participated in a follow-up study, in which they were asked about creative achievements, such as inventions, published scientific articles, awards for creative work, and so forth. Each participant also listed his or her most creative achievements, which were rated on overall creativity by three judges. Plucker (1999a) used multiple-regression analysis to examine the predictive power of the Torrance tests over 22 years, and the results provided support for the tests’ usefulness. The verbal test score predicted creative achievement, and it did so more strongly than intelligence did. The figurative component of the divergent-thinking score did not successfully predict later creative achievement, however. From these results Plucker concluded that the Torrance tests do possess predictive validity and that such tests provide useful information about people’s creative potential. Validity of Divergent-Thinking Tests: Conclusions and a Remaining Question We have now reviewed several studies on the validity of divergent-thinking tests as measures of creative-thinking capacity, and it should now be clear why it is difficult to draw firm conclusions in this area. Studies can be brought forth to support a whole range of seemingly contradictory conclusions, ranging from (1) divergent-thinking scores have little or no validity as even short-term predictors of creative achievement, to (2) divergent-thinking scores are valid predictors of people’s future creative accomplishments, even over many years. At present, the most reasonable conclusion may be that the relationship between divergent-thinking scores and creative achievement is complicated and that simple conclusions may not be possible. Divergent-thinking scores may predict creative achievements in only some people, and then in only some circumstances. The validity of the divergent-thinking scores may be related, for example, to the general intelligence level of the individuals being tested (Runco, 1991) as well as to the specific types of achievements that are being measured. Plucker’s (2003) reanalysis of Torrance’s (1974) longitudinal predictive-validity data using multiple-regression methods indicates that the divergent-thinking tests have predictive validity in that sample. However, in several other studies that have reported negative results, Plucker’s methods are not available to rescue the validity of the divergent-thinking tests, because the researchers reporting negative findings did use multiple regression when they carried out their original analyses. In their often-cited review of the validity of divergent-thinking tests, Barron and Harrington (1981) drew the following conclusion: “Some divergent thinking tests, administered under some conditions and scored by some sets of criteria, do measure abilities related to creative achievement and behavior in some domains” (p. 447). That very conservative statement—which, as Brown (1989) noted, does not instill great confidence in researchers who might be thinking about using those tests—still seems to be valid 25 years later. The criterion / predictive validity of the RAT has been subject to less investigation than that of divergent-thinking tests, but overall the results do not support the predictive validity of the test (see Brown, 1989, and Mansfield & Busse, 1981, for review). A Remaining Question: What Do Divergent-Thinking Scores Mean? If we assume for the sake of discussion that some divergent-thinking tests in some circumstances are valid predictors of creative achievement, that still leaves us with the question of why that should be so. Divergent-thinking scores might be correlated with later creative achievements for any of a number of reasons. It might be that divergent-thinking scores predict later creative achievement because divergent thinking forms part of the creative thought processes, so possession of divergent-thinking skill might make it more likely that one would produce creative responses in various situations. However, as noted earlier in the discussion of the face validity of divergent-thinking tests, results from the case studies presented earlier in this book, as well as from laboratory studies of problem solving (e.g., Fleck & Weisberg, 2004, 2006), provide support for the conclusion that the creative-thinking process is not as Guilford (1950) assumed it was. Most important, if the conclusions from the case studies and laboratory research are accurate, the first stage of the creative process does not entail production of large numbers of possible responses, as Guilford proposed. In addition, as discussed earlier, creative responses do not seem to be based on remote associations (consider again Picasso’s Guernica and Minotauromachy; Watson and Crick’s adoption of Pauling’s helical idea; Edison’s carbon- and platinum-filament electric lights; and the Wright brothers’ conception of the airplane as a bicycle with wings). If those analyses are correct, divergent thinking is not a critical component of the creative process. This conclusion leaves open the question of why divergent-thinking tests might in some cases predict later creative achievement. This question will be dealt with in the next chapter. The Generality versus Domain Specificity of Creative-Thinking Skills As noted earlier, one critical assumption underlying the development of tests to measure creative potential is that creativity is a general trait of people. That is why researchers assume that they can use responses on an unusual-uses-for-a-brick test to predict creative production in areas such as literature, poetry, painting, and so on. However, here too there is controversy in the literature, because research results are contradictory concerning whether creativity is a general trait. First of all, it is notable that one can find very few people who make significant contributions in multiple creative domains, especially domains that are very different in subject matter. For example, it is very rare to find a creative scientist who also makes significant contributions to any of the arts. Such negative findings would seem to indicate that creativity is relatively domain-specific, but there is one problem in interpreting the finding that people are usually creative in only one domain. As we saw in Chapter 4, it takes many years of dedication to reach world-class levels in any domain (the 10-Year Rule). We also saw in Chapter 5 that the 10-Year Rule is relevant to creativity, which means that we might not be able to find people who are creative in more than one domain due to time constraints. That is, there might not be enough time available for a person to develop the expertise needed to be creative in multiple domains. Creativity might be a general trait, but exhibiting that generality might be limited by constraints of time and effort rather than anything intrinsic to the creative process. If so, then we must turn elsewhere for evidence concerning the generality versus specificity of creativity. However, before we move on, one further point should be noted. The finding that creative achievement depends on domain-specific expertise that takes years to acquire, if valid, by itself casts doubt on any theory that assumes that creative achievement depends on general creative-thinking skills. The finding of the necessity of domain-specific expertise for creative achievement, on its face, raises problems for the notion of generality of creativity skills, since the 10-Year Rule indicates that specific skills are crucial in creative achievement. Baer (see Baer, 1998, for review) has taken a different approach to assessing the possible generality of creativity, by asking ordinary people, ranging from schoolchildren to adults, to produce creative products in a variety of domains, such as poems, short stories, collages, and mathematical puzzles. The products in each domain were then rated for creativity by judges. Results from several studies indicated that performance across domains was not related; that is, the person who wrote the best poem did not necessarily construct the best collage. Baer concluded that creative ability is specific in nature. Indeed, creative production in Baer’s studies was extremely domain-specific: Even the domains of poetry and fiction, which one might think would be related, were not. The best poets did not produce the best stories. Baer noted that this result raises problems not only for theories that assume that creativity is a general trait, but also for Gardner’s (e.g., 1993) theory of multiple intelligences. Gardner assumed that linguistic intelligence, which would encompass poetry and fiction, was one category of intelligence. Baer’s finding that performances in those domains were independent does not support Gardner’s assumption. From these results, which are based on ratings of products that people produce, one would conclude that creativity is a domain-specific trait. Baer (1998) also used training in divergent-thinking skills in an attempt to examine the general versus specific nature of creative-thinking skills. In one study, seventh-grade students were taught divergent-thinking skills that were related to poetry (e.g., the students practiced producing words that rhymed with a given stimulus word). The trained students and a nontrained control group then wrote poems and short stories, which were evaluated by judges. The training had a very specific effect: The poetry-writing skills increased for the trained group compared with the control group, but the story-writing skills did not. This is evidence that training in skills potentially relevant to creative thinking has effects only within the domain to which the training is directed. One does not seem to get general increases in creative thinking from such training. An opposite conclusion concerning the generality versus specificity of creative thinking comes from another set of studies, which collect data using questionnaires—such as that in Table 9.6—that ask people to list their past creative accomplishments (for reviews, see Plucker, 1999b, 2004). When one analyzes data from such questionnaires, one finds that there is a general tendency toward creative accomplishment across domains: People who list accomplishments in one field tend to list accomplishments in other fields as well. Plucker (1999b) carried out research in which he reanalyzed data from earlier studies (e.g., Runco, 1987) as well as collecting new data, and he has found evidence for generality in people’s responses to questionnaires. In addition, Plucker carried out a study in which people provided both questionnaire responses and descriptions of their most creative achievements in several different domains for judges to evaluate. In this dual-method study, Plucker found evidence for generality from the questionnaires and for domain specificity from the judges’ ratings of the concrete products. A similar conclusion arose from Plucker’s reanalysis of data from the study by Runco (1987), in which questionnaire responses and creative products were also judged. As a result of those various analyses, Plucker (2004) has concluded that there is a “method effect” that affects the conclusion of generality versus specificity in creative achievement. Traditional psychometric methods, such as questionnaires, produce evidence for generality; newer measures based on specific creative performances, such as the consensual assessment technique, produce evidence for domain specificity. Assuming that Plucker’s (2004) “method effect” conclusion is valid, it leaves us with the question of why such an effect might be found. Plucker does not provide any specific reasons for why it might occur, although he notes that the consensual assessment technique is subjective in nature, since it is based on judges’ ratings of creative products. This might indicate that Plucker believes that there is an inherent weakness in such studies. However, in the study that he reports, in which both questionnaire responses and judges’ ratings were obtained for the same group of people (Plucker, 2004), the reliabilities of the judges’ ratings were higher than that for the responses on the questionnaires. That finding seems to indicate that there is not an inherent weakness in the consensual-assessment technique. In addition, one might raise the question of whether judges’ ratings of the creativity of products are any more subjective than people’s answers to questions concerning how many times they have written poems, for example. While it is obvious that frequency of poems could be measured objectively—that is, one could videotape a person’s life and then go over the tape and count every time the person writes a poem—that is not how those data are collected. If we ask a person to report frequencies of various activities, then a subjective element is inevitably added. In a related vein, Baer (e.g., 1998), as might be expected from the earlier discussion, has argued that evidence for generality of creative achievement arising from questionnaire studies might be due to factors not related to creativity at all. He takes the position opposite to Plucker and raises questions about subjective factors not related to creativity that might affect people’s responses to questionnaires about their creative accomplishments. There might be some aspects of the strategies that a person could use in responding to a questionnaire that might be critical here. For example, as seen in Table 9.6, questionnaires sometimes ask about awards that people have earned for their creative works. However, the term award is not usually specifically defined, so it may be that some people are more likely than others to credit themselves with creative achievement simply because of the vagueness of some of the terms involved. Other people might be more reluctant to do what they might see as blowing their own horn. Similarly, if one is asked whether one has taken lessons in painting, say, does spending a couple of evenings in an adult-education drawing class count as lessons? Thus, aspects of people’s response styles, such as how likely they are to give themselves credit for some accomplishment, might affect conclusions concerning the generality or domain specificity of their creative achievements. Some people might report creative accomplishments across a wide range of areas because they tend to feel positively about their own accomplishments. More reticent people might report fewer activities. That pattern of responses would tend to support the conclusion that creativity is a general trait, because people might report creative achievements across the whole set of domains or in none of them. However, a conclusion of generality might not be warranted by the actual achievements of the respondents. In the next chapter, I will examine another noncreativity factor that might contribute to findings of generality of creative achievement. As this discussion has shown, there is some controversy as to whether creativity is based on general or specific skills. Evidence indicates that the method through which creative achievement is assessed plays a large role in one’s conclusion concerning this issue. We have examined several possible reasons for the controversy. In addition, evidence from the case studies of creative thinking presented elsewhere in this book, including evidence concerning the 10-Year Rule and expertise in creativity, supports the premise that domain-specific skills underlie creativity. Additional evidence concerning this question will be discussed in Chapter 11, in the context of the discussion of confluence models of creativity. Testing Creativity: Conclusions The review in this chapter has indicated that serious questions can be raised about the divergent-thinking tests that have been used in a wide range of studies. Those tests have demonstrated mixed levels of discriminant and predictive validity. Questions can also be raised about the face validity of divergent-thinking tests, based on reservations about Guilford’s divergent-thinking analysis of the creative process. Similarly, questions can be raised about Mednick’s (1962) postulation of flat versus steep associative hierarchies in creative versus noncreative thinking. In sum, empirical and theoretical questions can be raised about the premise that divergent thinking is the basis for the capacity to think creatively. This is indirect support for the “ordinary-thinking” perspective. In the next chapter, we examine the hypothesis that a unique set of personality characteristics (the “creative personality”) play a role in determining whether a person will be creative. This chapter has focused on development of the divergent-thinking tests by Guilford and others. However, Guilford’s analysis of the components of the creative thought process was more complex than the distinction between divergent and convergent thinking. Indeed, Guilford proposed a complex model of cognition—the structure of intellect model—based on many factors. In recent years, that model has received much less interest than has the role of divergent thinking in creativity, which is why that aspect of Guilford’s research has been discussed here. Mumford (2001) has recently presented a review of the wider range of Guilford’s thinking vis-à-vis creative thinking in an attempt to make more researchers aware of the potential value in the broad range of Guilford’s ideas. It is too early to determine if Mumford’s effort will result in a renewal of interest in aspects of Guilford’s research that have more or less been put aside.