Personality and Consumer Behavior: an Operational Approach

ABSTRACT - This is a report on the promising preliminary results of the Q project at Kent State University to operationalize the personality organization from psychoanalytic object relations theory for empirical research on consumer behavior using the California Q Sort. The validation process is well under way. As a future extension of this research, the template matching technique is presented for relating the personality to a general pattern of behavior by relating these patterns of behavior to variations in the level of personality development, thereby paving the way for an application to the critical social issues of addictive and compulsive behavior.

Citation:

Paul J. Albanese (1993) ,"Personality and Consumer Behavior: an Operational Approach", in E - European Advances in Consumer Research Volume 1, eds. W. Fred Van Raaij and Gary J. Bamossy, Provo, UT : Association for Consumer Research, Pages: 513-517.

European Advances in Consumer Research Volume 1, 1993      Pages 513-517

PERSONALITY AND CONSUMER BEHAVIOR: AN OPERATIONAL APPROACH

Paul J. Albanese, Kent State University, USA

ABSTRACT -

This is a report on the promising preliminary results of the Q project at Kent State University to operationalize the personality organization from psychoanalytic object relations theory for empirical research on consumer behavior using the California Q Sort. The validation process is well under way. As a future extension of this research, the template matching technique is presented for relating the personality to a general pattern of behavior by relating these patterns of behavior to variations in the level of personality development, thereby paving the way for an application to the critical social issues of addictive and compulsive behavior.

INTRODUCTION: PERSONALITY AND CONSUMER BEHAVIOR

According to the paradox of personality in marketing, we all have a personality, but we do not know how it is systematically related to our consumer behavior (Albanese 1989). Efforts to operationalize the personality have had a nefarious history in marketing. The most notorious effort was that of Franklin B. Evans (1959) to use the Edwards Personal Preference Schedule to differentiate the personalities of Ford and Chevrolet owners. Some effort was made to reinterpret Evans's results in a positive way (Jacoby 1969); however, it would be hard to demonstrate on any grounds that there was a wits worth of difference between Ford and Chevrolet owners in the late 1950's, or for that matter, in the early 1990's.

The literature review on personality and consumer behavior can be broken down into the Dark Ages and the Renaissance. The Dark Ages refer to the negative attitude within the field of marketing toward the area personality and consumer behavior and is most notably represented by Kassarjian and Sheffet (1991). First published in 1971, republished virtually unchanged in 1981, and then again 1991, there are no consequential differences in any of the versions of this survey, thereby dismissing two decades of research which has steadily moved toward revitalizing this area.

The Renaissance in marketing began most notably with Holbrook (1988), and continues with Foxall and Goldsmith (1989), and Albanese (1990). Holbrook (1988) uses the study of individual personality to go into greater depth in the analysis of consumer behavior, while Foxall and Goldsmith (1989) and Albanese (1990) go beyond the personality of the individual to explore wider applications to more general patterns of behavior.

At the 1991 conference of the Association for Consumer Research in Chicago, William Wells made the following observation with respect to the work of Freud, Abraham, Jung, and Adler:

"It seems to me that it is time to go back and take a look at what these seminal thinkers had to say and take them seriously, which we have never really done." [Wells made this comment as a discussant at a Special Topics Session on Personality and Psychoanalytic Theory: Current Theoretical and Empirical Research.]

Rather than trying to squeeze out any empirical direction from the marketing literature on personality and consumer behavior, it is far more fruitful to "strike another match, go start anew" (Dylan 1985, p. 179).

The Q Project

A project has been established to operationalize the personality organization from psychoanalytic object relations theory using the California Q Sort. The primary purpose of this report is to describe setting up the Q Project and to discuss some theoretical implications of the preliminary results. It is not meant to be a comprehensive survey or summary of efforts to operationalize psychoanalytic theory, or to review the extensive literature on the Q methodology, or even the California Q Sort.

At this stage of my research, I am using the technique to classify an individual's personality organization along the PERSONALITY CONTINUUM for the purpose of doing further research on behavior related to that personality organization (Albanese 1990). The PERSONALITY CONTINUUM is the integrative framework behind this research. If this technique proves to be useful to clinical psychologists for diagnostic purposes, I would be delighted; however, that is not my primary purpose.

A personality organization is the central analytic construct and it is defined by the level of intrapsychic structural formation and the predominant defense used against anxiety in interpersonal relationships. Personality organizations at different levels of development can be arrayed into discrete ranges along the PERSONALITY CONTINUUM. Each range of the PERSONALITY CONTINUUM is demarcated by a specific interpersonal achievement in personality development, therefore, each range represents a different level of personality development. [The personality organizations along the PERSONALITY CONTINUUM represent the chronologically adult person, and should not be construed as a presentation of an object relations theory of personality development; i.e., the early infant does not start out life as psychotic. This is a critical caveat in using the PERSONALITY CONTINUUM. It does not mean, however, that a psychotic adult does not have something in common with an early infant; in fact, they both face the world without having established boundaries between themselves and other people. For the infant this represents a normal stage of personality development--the primary undifferentiated phase--but for the psychotic person, this represents the absence of even rudimentary reality testing.] The PERSONALITY CONTINUUM divided in four discrete rangesCNormal, Neurotic, Primitive, and Psychotic. [The PERSONALITY CONTINUUM has been largely adapted from the work of Otto F. Kernberg (1975, 1976), particularly from his three contributions to intimacy in interpersonal relationships (1974A, 1974B, 1977).]

Personality organizations with a similar intrapsychic structure and the same predominant defensive techniques used against anxiety in interpersonal relationships can be characterized by a common, or general pattern of behavior. This facilitates the comparison of variations in patterns of behavior with the personality organization of the individual, and results in the recognition of characteristic variations in individual patterns of behavior, which can be classified, related to particular personality organizations, and arrayed along the personality continuum.

The PERSONALITY CONTINUUM is not a one dimensional concept of the personality organization: each panel arranged across the top represents a different aspect of the intrapsychic structure, a pattern of behavior, or a specific capacity achieved by a human being. An important methodological point is being pressed here: for the purposes of hypothesis testing, it does not matter where in the PERSONALITY CONTINUUM one begins, or in what direction one's subsequent research moves, everything is systematically related to everything else.

As an application of my research, I am interested in compulsive, and in the more extreme case, addictive patterns of consumption behavior. If I locate compulsive and addictive behavior and read across the PERSONALITY CONTINUUM, these patterns of behavior are related to the Primitive Range, and characterize the Borderline, Narcissistic, Schizoid, as well as the Infantile Personality Organizations. For the purposes of hypothesis testing, it does not matter whether I identify people who are compulsive and addictive and infer that they should fall into the Primitive Range, or whether I find people who fall into the Primitive Range and infer that they will have compulsive and addictive patterns of behavior.

Individual Difference Measures: Treating People as Parts

Individual difference measures treat people as parts. An objective of this research is to construct a whole and more realistic conception of the person upon which to base the study of consumer behavior. What would be interesting and useful is to look at how individuals within one range of the PERSONALITY CONTINUUM score on an individual difference measure. If personality organizations within one range have a characteristic pattern of scoring which differentiates them from other ranges of the personality continuum, two consequences result: 1. The scores within a range can be used to elucidate and elaborate the behavior of the personality organizations within that range; 2. The individual difference measure can then be used to classify, or at least contribute to the classification of, an individual's personality into the appropriate range of the PERSONALITY CONTINUUM. It is quite possible for different persons with different personality organizations to score the same way on an individual difference measure, but for entirely different reasons; the interpretation would depend on the level of personality development achieved, represented by the range of the continuum occupied by the personality organization.

At the same time, the behavior being measured by various individual difference indicators can be organized into a larger, more comprehensive pattern of behavior characteristic of that range of the personality continuum. In "Personality as Traits," Arnold H. Buss (1989) argues that personality psychologists should seek ways of grouping individual responses in classes of responses. After arguing that a range of behavior should be studied, Buss concludes that, "each time one adds a trait, the prediction of behavior becomes more precise, and one moves closer to a major goal of personality trait research: to understand persons as combinations of traits (Buss 1989, p. 1385)."

This amplifies the importance of every individual difference measure used in marketing, and overcomes the chief liability of these measuresCindividual difference measures treat the person as parts, not as a whole. Thus, as the researcher progresses from classes of responses, to ranges of behavior, to combinations of traits, ultimately one arrives at a comprehensive picture of the personality (Buss 1989, p. 1385). In contrast, I start with a comprehensive picture of the personality, the personality organization, and work in the opposite direction, from a range of behavior down to the interpretation of single individual difference measures. This underscores the importance of the personality organization as an organizing framework and the personality continuum as an integrative framework: no useful approach is excluded from this framework.

The present endeavor to operationalize the personality organization using the California Q Sort provides a promising opportunity to make the entire PERSONALITY CONTINUUM useful for empirical research. If the use of the California Q Sort as a personality assessment instrument can be validated, and, therefore, a person's personality organization can be classified into the appropriate range of the PERSONALITY CONTINUUM, then every related dimension of that person's personality organization arranged in the panels across the continuum becomes a testable hypothesis.

THE TECHNIQUE

Why the California Q Sort, and not the MMPI-2, or the CPI, or even the Edwards Personal Preference Schedule? The California Q Sort consists of 100 personality characteristics which are highly descriptive of behavior, and therefore, can be used by researchers interested in a wide variety of behavioral issues. The 100 descriptive personality characteristics are printed on cards which are sorted into nine categories, with each item receiving a score from 1 to 9, from least to the most characteristic of the person respectively, forming a forced symmetric distribution with a mean of 5 and a standard deviation of 2.08. Daryl J. Bem modified the California Q Sort by adding clarifying statements to the bottom of some cards to make it possible for nonclinicians to use the technique, and for an individual to do a self sort. In 1990, Jack Block, who developed the California Q Sort in the early 1960's, brought this process of modification to fruition in the CAQ-90 version which is being utilized for this project.

I am adapting the California Q Sort to psychoanalytic object relations theory by constructing templates based on the clinical descriptions of personality organizations; i.e., creating Q Sorts for the personality organizations that have been extensively studied in the object relations literature. A sort for a particular person would then be correlated with each of the templates to see which of the personality organizations it approximates most closely. The intense scrutiny of human behavior in the clinical relationship provides the clearest portrait and the deepest understanding of human behavior; the Q technique offers a promising opportunity to open this wealth of information to other researchers. This adaptation of the Q Sort to object relations theory offers the promise of operationalizing the concept of the personality organization.

THE ANALYSIS OF TWO ACTUAL SUBJECTS

The Correlation Matrix

The first two subjects were contributed by a clinical psychologist who sorted clients which had been in his practice for some time. The analysis first involved a simple correlation analysis, and then a number of factor analyses were performed.

The results of the correlation analysis are presented in the following table. My initial reaction to the correlation matrix is that the technique did not work. I was expecting the correlation coefficients to be .99 on a single template or prototype, precisely positioning the personality organization of the subject on the PERSONALITY CONTINUUM. (See Table 1).

My initial expectation was, of course, highly unrealistic; a far more realistic expectation is that the technique would crudely classify an individual's personality organization into the right range of the continuum. This expectation was borne out by the correlation analysis. Subject GS showed a significant correlation at the 0.001 level of 0.4329 with the Schizoid template, and a 0.2685 correlation with the Paranoid template, significant at the .01 level. In the diagnosis of the clinical psychologist, Subject GS, a female patient in his private practice, has a predominantly schizoid personality organization. I would argue that the technique correctly classified her personality organization into the Primitive Range of the PERSONALITY CONTINUUM.

TABLE 1

CORRELATION MATRIX

Subject PY showed 0.2708 correlation with the Obsessive template and a 0.2755 correlation with the Paranoid template, both significant at the .01 level. I would argue that Subject PY should be classified into the Neurotic Range of the PERSONALITY CONTINUUM. In the clinical psychologist's diagnosis, Subject PY, a male patient in his private practice, had barely achieved a neurosis. I would argue that the Q technique has correctly classified this person's personality organization in to the right range of the continuum under ordinary functioning. In an extraordinary situation fraught with intolerable anxiety, this person is likely to regress to a lower level of functioning, and therefore, to earlier patterns of behavior. This is the meaning of the 0.25 correlation with the Schizoid template, significant at the .05 level.

Subject PY is a real test of the technique, since he had barely achieved a neurosis, under difficult circumstances he is likely to break down to a lower level of functioning. While the classification takes place at one point in time, this does not mean that the personality is immutable. The essence of a mental "break down" is a regression to a lower level of functioning and a return to earlier patterns of behavior under trying and difficult circumstances. A clinical psychologist working with a client over any length of time and having made a diagnosis of that person's personality is capable of making a distinction between ordinary functioning and functioning under extraordinary circumstances.

It is important to note that each subject was negatively correlated with the Optimally Adjusted Personality template. The factor analyses also supported a clear distinction between normality and pathology. This will be discussed further in the section on the factor analysis.

The Interpretation of the Correlation Matrix

Upon further inspection of the correlation matrix, the technique yielded even more refined results than anticipated. The substance of psychoanalytic object relations theory is the intrapsychic structure of the personality. I would argue that it is the intrapsychic structure of the personality organization that is being manifested by the Q technique, representing the internalization of interpersonal relationships into patterned and persisting structures, and therefore, the enduring aspects of the personality (Rapaport and Gill 1959). The innovator of Q methodology, William Stephenson, believed that Q methodology offered the best opportunity to operationalize psychoanalytic theory because it transformed the unconscious into an operant factor structure (Stephenson 1987). This is precisely what appears to be happening in this application of the California Q Sort: the manifestation of the intrapsychic structure of the personality.

The personality organization is defined by the level of intrapsychic structural formation and the predominant defense used against anxiety in interpersonal relations, and the predominant defense also appears to be manifesting on the templates as well. W.R.D. Fairbairn argued that what appear to be neurotic disorders for personality organizations in the Primitive Range of the PERSONALITY CONTINUUM are actually techniques for dealing with severe anxiety in interpersonal relationships (Fairbairn 1952, p. 30). This view is borne out for Subject GS: the fundamental nature of the schizoid position is apparent from the .4329 correlation coefficient, and she is using a predominant paranoid defense against anxiety, a correlation of.2685.

For Subject PY, having barely achieved a neurosis, the basic position is clearly schizoid, but, under favorable circumstances, this man is able to function at a neurotic level of personality development using obsessional and paranoid techniques. If these techniques fail when this person is experiencing severe anxiety in an interpersonal situation, he would regress to a lower, more primitive level of functioning. I have no doubt that a Q sort done on this person under these circumstances would yield a higher correlation coefficient on the Schizoid template.

Two important observations appear to be warranted by the preliminary correlation results for these two subjects: First, pieces of the personality organization are manifesting on several templates, not just one, and I believe that these pieces represent the intrapsychic structure and the predominant defense. Second, at the same time, the technique is crudely, but clearly, classifying the personality organization of the subjects into the correct range of the PERSONALITY CONTINUUM. The latter result lends support to the demarcation of the PERSONALITY CONTINUUM into discrete ranges, and the hierarchical arrangement of the ranges based the level of personality development that has been achieved.

The Factor Analysis

Factor analyses were performed and three factors emerged, Normal, Neurotic, and Primitive. The results of the factor analyses initially looked promising, but the factors were not orthogonal; i.e., the factors were not independent. This is in line with the previous correlation analysis: pieces of the personality organization load on different factors. This is an appealing result theoretically, because the basis for interpretation from object relations theory varies by range with the level of intrapsychic structural formation and the predominant defense. I find this result theoretically very satisfying.

Factor arrays were produced from the factor analyses, which represent rankings of California Q Sort items for the Normal, Neurotic, and Primitive factors. Subsequently, Subjects PY and GS were correlated with the factor arrays representing the ranges. The correlation coefficient of Subject PY with the Neurotic Factor was 0.28, significant at the 0.01 level. The correlation of Subject GS was not statistically significant. Again, I believe a similar interpretation applies: pieces of the personality organization for an individual are loading on more than one factor. This means that it is probably not appropriate to construct a single template for each range; as far as prototype personality organizations go, the more the merrier. [In a personal correspondence, Jack Block revealed that he works with over 50 prototypes in his research.]

THE LIMITATIONS OF THE STUDY

The limitations of the study are many at this point in time. First and foremost, of course, is the limited number of subjectsCtwo. I have no doubt that the results will improve with a larger subject pool. I will continue to work with clinical psychologists who will contribute a few sorts on patients in their practice with known personality organizations. The second strategy for acquiring new subjects is to establish a working relationship with the Ph. D. program in clinical psychology at Kent State University in order to work with doctoral students doing their clinical practicum under supervision. I am actively analyzing sorts generated from this alliance now. One advantage of this approach is that it would yield a stream of subjects who have undergone standardized diagnostic tests like the MMPI-2 and have been given a DSM III diagnostic label. Another advantage of using DSM III terminology is that it facilitates communication, especially across disciplines, and it stems the temptation to be neologistic in labeling personality organizations.

Missing templates or prototypes

More prototype personality organizations are required. A glaring omission at this point in time is a template for the Depressive Personality Organization. The demarcation of the ranges and the hierarchical arrangement of the ranges along the PERSONALITY CONTINUUM is supported by the preliminary results; however, the ranking of personality organizations within the ranges requires refinement.

Confirmation of existing prototypes

In Q methodology, the "specimen" is a subject who fits the pure type, and therefore, can be used to refine the template. I prefer this technique to that used by Block (1978) in which sorts from nine psychologists were averaged together into one template or personality prototype.

A BRIEF SUMMARY

I believe strongly that the preliminary results are promising and bode well for the future of the Q Project. The strongest result of the analysis is that the Q technique is capable of correctly classifying an individual's personality organization into the appropriate range of the PERSONALITY CONTINUUM. For the establishment of a relationship between the level of personality development and the general pattern of behavior, and the subsequent explorations of the behavioral implications of personality, this is the most important result. Yet, at the same time, the way in which the pieces of the personality organization manifested on a variety of templates speaks well for a more refined use of the technique, and the interpretation of the meaning of these manifestations based upon the level of personality development as represented by the range of the PERSONALITY CONTINUUM. The immediate potential of the technique for diagnostic purposes is there as well: If a clinician could discern the level of personality development of a client, it could move the therapy in the right direction more quickly, and avoid some potential problems. All in all, I am confident that the Q technique will progress toward a highly useful approach for operationalizing the personality organization for empirical research.

Finally, a template for a pattern of behavior can be constructed from a subset of California Q Sort items, which can then be matched to sorts for the individual subjects. This is known as the template matching technique, and was developed by Bem and Funder (1978) for the study of patterns of behavior. This is the way in which I intend to study compulsive and addictive behavior, and to relate these patterns of behavior to particular personality organizations and to the appropriate ranges of the PERSONALITY CONTINUUM.

Personality and consumer behavior has emerged from the Dark Ages as a viable area of research in marketing, and we must now forsake the naysayers and go forth into the Renaissance.

REFERENCES

Albanese, Paul J. (1988), "The Intimate Relations of the Consistent Consumer: Psychoanalytic Object Relations Theory Applied to Economics," in Psychological Foundations of Economic Behavior, Paul J. Albanese, ed., New York: Praeger, 59-79.

Albanese, Paul J. (1989), "The Paradox of Personality in Marketing: A New Approach to the Problem" in Enhancing Knowledge Development in Marketing, Paul Bloom, et al. eds., Chicago, Illinois: American Marketing Association, 245-249.

Albanese, Paul J. (1990), "Personality, Consumer Behavior, and Marketing Research: A New Theoretical and Empirical Approach," in Research in Consumer Behavior, Vol. IV, Elizabeth C. Hirschman, ed., Greenwich, CT.: JAI Press, 1-49.

Bem, Daryl J. and David C. Funder (1978), "Predicting More of the People More of the Time: Assessing the Personality of Situations," Psychological Review, 85, 6(November), 485-501.

Block, Jack (1978), The Q-Sort Method in Personality Assessment and Psychiatric Research, Palo Alto, California: Consulting Psychologists Press.

Buss, Arnold (1989), American Psychologist, 44, 11 (November), 1378-1388.

Dylan, Bob (1985), Lyrics, 1962-1985, New York: Alfred A. Knopf.

Evans, Franklin B. (1959), "Psychological and Objective Factors in the prediction of Brand Choice," Journal of Business, 32 (October), 340-369.

Fairbairn, W. R. D. (1952), Psychoanalytic Studies of the Personality, London: Routledge & Kegan Paul.

Foxall, Gordon R. and Ronald E. Goldsmith (1989), "Personality and Consumer Research: Another Look," Journal of the Market Research Society, 30, 111-125.

Jacoby, Jacob (1969), "Personality and Consumer Behavior: How to Find Relationships," Purdue Papers in Consumer Psychology, No. 102. Lafayette, IN: Purdue University.

Kernberg, Otto F. (1977), "Boundaries and Structure in Love Relations," Journal of the American Psychoanalytic Association, 25, 81-114.

Kernberg, Otto F. (1976), Object Relations Theory and Clinical Psychoanalysis, New York: Jason Aronson.

Kernberg, Otto F. (1975), Borderline Conditions and Pathological Narcissism, New York: Jason Aronson.

Kernberg, Otto F. (1974A), "Mature Love: Prerequisites and Characteristics," Journal of the American Psychoanalytic Association, 22, 743-68.

Kernberg, Otto F. (1974B), "Barriers to Falling and Remaining in Love," Journal of the American Psychoanalytic Association, 22, 486-514.

Rapaport, David and Merton M. Gill (1959), "The Points of View and Assumptions of Metapsychology," International Journal of Psycho-Analysis, 40, 3-4(May-August), 153-161.

Stephenson, William, (1987), "Q Methodology: Interbehavioral and Quantum Theoretical Connections in Clinical Psychology," in New Ideas in Therapy: Introduction to an Interdisciplinary Approach, Douglas H. Ruben and Dennis J. Delprato, (eds.), New York: Greenwood Press, 97-106.

----------------------------------------