Robert N. Barger, University of Notre Dame
             Warren N. Kubitschek, University of Notre Dame
                Josephine C. Barger, Indiana University


A paper presented at the Annual Meeting of the American Educational Research Association, April 13-17, 1998, San Diego, California, USA
1. Introduction Philosophical tendencies and personality types have potential importance for questions in the area of moral development. But before raising questions about either philosophical tendencies or personality types, we want to explore whether there is any statistical correlation between the two. That is the task of this paper. 2. Philosophical Tendencies Before we state our hypotheses, we must define terms. We speak, first of all, of philosophical tendencies. By these we mean the world-views of Idealism, Realism, Pragmatism, and Existentialism. It is not our claim that these are the only world-views that exist, but that these four views are paradigmatic of the basic ways in which people interpret the world. The Idealist believes that reality is basically spirit or idea, that knowledge is gained through the mind, and that value is measured by conformity to ideals. Thus, the Idealist seeks always to treat people as means, never as ends; to treat others as one would wish to be treated oneself; and to judge solely on the basis of the intrinsic nature of actions and not on the basis of the results of actions. The Realist believes that reality is basically matter or the physical universe, that knowledge is gained through the senses, and that value is measured by conformity to nature. Thus, the Realist attempts to act in accord with mental and physical nature; to live in the mean between excess and deficiency. The Pragmatist believes that reality is basically a process or an experience, that knowledge is gained through a trial-and-error approach, and that values are determined by norms established by society. Thus, the Pragmatist tries to promote the most net happiness for the greatest number of people (even though in some cases this net happiness may be more eventual than immediate). The Existentialist believes that reality is self-defined, that knowledge is gained through making personal decisions about what is true, and that value is measured by what an individual responsibly chooses as good. Thus, an Existentialist seeks to be true to him/herself, to exist autonomously, to determine by one's choices whom one becomes, and to take responsibility for one's choices. In summary, the Idealist is invariant and abstract, the Realist is invariant and concrete, the Pragmatist is variant and social, and the Existentialist is variant and individual. Note that each of the four previously-described world-views is stated as a "belief," since none of them can be proven. They are simply fundamental assumptions about the meaning of the world and how it should be negotiated. They are each person's "best guess" at what is at the bottom of reality. 3. Personality Types Carl Jung is generally credited with inventing personality types. He believed that people displayed different psychological or behavioral inclinations. These inclinations, in his view, allowed people to be characterized as belonging to different functional groups, or "types." It was Isabel Myers, however, who devised a way to identify Jung's types in the population using the Myers-Briggs Type Indicator (MBTI). This led the way for subsequent research on personality types through instruments similar to the MBTI. One subsequent researcher was David Keirsey, who, although he retained the typological designations used by Jung and Myers, interpreted the import of some of their types differently. For example, he thought the distinction between Extraversion (E) and Introversion (I) was the least important of the dimensions of personality whereas Jung and Myers thought they were the most important. Keirsey refers to the E type as "expressive" and to the I type as "reserved." Besides the E and I pairs, there are three other pairs of types common to Jung's, Myers', and Keirsey's typology: S and N, T and F, J and P. The Sensing (S) type has a preference for perceiving through the five senses, placing more importance on facts, details, and realities. Keirsey calls this type "observing." The Intuitive (N) type has a preference for perceiving with the mind in the form of ideas, concepts, and imagination. Keirsey calls this type "introspective." The Thinking (T) type has a preference for judging on the basis of impersonal objectivity, seeking consistency and logic. Keirsey calls this type "tough-minded." The Feeling (F) type has a preference for judging on a subjective, personal basis. Keirsey calls this type "tender-minded." The Judging (J) type has a preference for having things planned, scheduled, orderly, and leading to closure. Keirsey calls this type "schedulers." Finally, the Perceptive (P) type has a preference for having things open-ended and flexible. Keirsey calls this type "probers." After excluding the E and I types, which Keirsey does not consider to be fundamental to personality, he uses the remaining six types to identify what he considers to be the four basic types of temperament or character: NF (Idealist), NT (Rationalist), SJ (Guardian), and SP (Artisan). The Idealists (NF) are abstract in communicating, utilitarian in implementing goals, use strategic analysis, and have an impersonal objectivity. Keirsey calls them "Identity seeking" personalities. The Rationalists (NT) are abstract in communicating, cooperative in implementing goals, use diplomatic integration, and have a personal subjectivity. Keirsey calls them "Knowledge seeking" personalities. The Guardians are concrete in communicating, cooperative in implementing goals, use logistical maintenance, and are scheduled and orderly. Keirsey calls them "Security seeking" personalities. Finally, the Artisans (SP) are concrete in communicating, utilitarian in implementing goals, use tactical variation, and are open-ended and flexible. Keirsey calls them "Sensation seeking" personalities. In summary, the NF type is abstract and cooperative, the NT is abstract and utilitarian, the SJ is concrete and cooperative, and the SP is concrete and utilitarian. For further information on Keirsey's work on personality types, see 4. Hypotheses Regarding Tendencies and Types Both philosophical tendencies and personality types are ways of trying to understand people and their behavior. The MBTI and its variants have become popular ways of investigating people's personality types. But there has been less interest in investigating people's philosophical tendencies in such an empirical way. Comparing the philosophical tendencies and personality types explained above, one might hypothesize that they would have certain correlations. If the four philosophical tendencies are mutually exclusive, one might expect them all to correlate negatively with one another. Likewise, if the personality types are mutually exclusive one might expect that they would also all correlate negatively with one another. Since personality types NF (Idealism) and NT (Rationalism) are linked to perceiving with the mind, one might expect that one or both of them may correlate with the philosophical tendency of Idealism which is concerned with the abstract or immaterial dimension of things and gaining knowledge through the mind. Note: it is unfortunate that the label "Idealism" is employed, using a somewhat different definition, by both the personality researchers and traditional philosophers. Since personality types SJ (Guardian) and SP (Artisan) are linked to perceiving with the senses, one might expect that one or both of them may correlate with the philosophical tendency of Realism which is concerned with the concrete aspects of reality and gaining knowledge through the senses. Since personality type SP (Artisan) is linked with being sense-oriented and flexible, one might expect a negative correlation with the philosophical tendency of Idealism which is invariant and abstract. Since personality type SJ (Guardian) is linked with being cooperative, one might expect a correlation with the philosophical tendency of Pragmatism which is social. Since personality type SJ (Guardian) is linked with being orderly, one might expect a correlation with the philosophical tendency of Realism which attends to order in nature. 5. Differences Between Tendencies and Types Keirsey and Myers and Jungian psychology in general say there are distinct types of personalities that are essentially fixed. They say that if you are a type X, no amount of change can make you anything else. There are different shades and variations within a type (each of the four main types is broken down into four more), but if you are one type you cannot be another type. Many social psychological scales measure tendencies, e.g., you are more of this or less of that. But they don't preclude one person from being more of this and more of that, or another person from being less of both. Other scales are situational, e.g., one can be more extroverted in one situation and more introverted in another. Those who believe in types don't believe in tendencies, or perhaps even in valid situational tendencies (cf., passim, Those who believe in tendencies see types as an oversimplification, perhaps useful as an indicator of typical behavior/personality, but as incomplete and not very useful in situations that vary. Difference is crucial in terms of interpretation. Types and tendencies are two different things, based on two different underlying assumptions. Difference is also crucial in terms of measurement. Keirsey and Myers-Briggs give forced choice questions. One has to rank four alternatives or has to choose only one of two options. As a result, those who end up high on NT must end up low on NF, and vice versa. Those who end up high on SJ must end up low on SP, and vice versa. On the other hand, we measured philosophical tendencies by a standard Likert "strongly agree"-to-"strongly disagree" scale. Those who end up high on one philosophical tendency may also end up high on any or all other philosophical tendencies. Those who end up low on one may end up low on any or all others. Any combination of high and low philosophical tendency scores is possible. One could pick a person's highest philosophy score and claim that it is the person's "philosophical type." Or one could create additional types like "absolutist" and "relativist" by adding together a person's Idealist and Realist scores and the person's Pragmatist and Existentialist scores. Unfortunately, people don't always answer forced choice questions in the same way that they answer combinations of unforced questions. Thus, combining these tendencies may not reveal the respondent's "philosophical type." We will examine a person's philosophical type later, as a way of interpreting some results, but this will be an oversimplification of the way it was measured. 6. Measurement Methods of the Tendencies and Types An inventory of 40 Likert-scale items, the Ross-Barger Philosophy Inventory, (Ross & Barger, 1990) was used to measure philosophical tendencies. Groups of ten items surveyed tendencies toward each of the four philosophies of Idealism, Realism, Pragmatism, and Existentialism. Sample items from this inventory are the following: "Knowledge is found by considering the practical consequences of ideas" and "Reality originates in the material and physical world." An inventory of 36 forced-choice items, the Keirsey Character Sorter, (Keirsey, 1997) was used to measure personality types. Each of the four responses to the first 16 items were to be ranked according to the respondent's preference. A sample of this type of item follows: "I respect myself more for: - being autonomous and independent; - doing good deeds; - having good intentions; - being bold and adventurous." These initial items were followed by a set of 20 binary-choice items. A sample of this type of item follows: "Do you think of yourself as: - a private person; - an outgoing person." There were no noticeable overlaps in content in these inventories. 7. Reliability of Scales It is not known whether reliability measures exist for the Keirsey Character Sorter. Previous research has been done with the Ross-Barger Philosophy Inventory, however. An expanded 80 item version of this scale was used with N=347 students at a midwestern university. The inter-item reliability was moderately high: Idealism: alpha=.77, Realism: alpha=.74, Pragmatism: alpha=.69, Existentialism: alpha=.72. Factor analysis showed some support for a four factor solution, although it was not "clean." The current research was done with N=50 students from two midwestern universities. A shortened Ross-Barger Philosophy Inventory with ten items per scale was used. These items were chosen to maximize corrected item-total correlations for factor analysis in previous research. Inter-item reliability was mixed: Idealism: alpha=.82, Realism: alpha=.71, Pragmatism: alpha=.56, Existentialism: alpha=.61. Evidence shows that these scales are measuring something. Some scales are measuring "whatever they measure" more consistently than do other scales. Differences in reliability between studies and range of reliability across scales show that more work is needed to refine scales and measurements. 8. Descriptive Statistics 8.1. Personality types Keirsey Type N % NF 20 40% NT 10 20% SJ 13 26% SP 7 14% Total 50 100% Two-fifths of these respondents are NF (Idealists). About a quarter are SJ (Guardians). One-fifth are NT (Rationalists). One-seventh are SP (Artisans). 8.2 Philosophic tendencies Philosophic Tendency Mean Std. Dev. Minimum Maximum Idealism 33.0 6.3 18 47 Realism 30.9 5.2 22 49 Pragmatism 35.0 4.2 23 45 Existentialism 36.5 4.7 26 46 These respondents, on average, had high existentialism and pragmatism scores, moderate idealism scores, and lower realism scores. They varied the most in their idealism scores, then their realism scores (std. dev. and range). They varied least in their existentialism and pragmatism scores. Respondents were more likely to agree with any statement than disagree with it. (Means are all greater than 30.) 9. Correlation of Philosophic Scales, Personality Types 9.1. Correlation of personality types in these data Four dichotomous variables were created, indicating Keirsey personality type: Variable NT = 1 if Keirsey Type is NT, otherwise = 0; Variable NF = 1 if Keirsey Type is NF, otherwise = 0; Variable SJ = 1 if Keirsey Type is SJ, otherwise = 0; Variable SP = 1 if Keirsey Type is 1, otherwise = 0. Note that these variables are not mathematically independent: if we know the value of three of these variables, we know the value of the fourth. 9.1.2. Correlations NF NT SJ NF NT -.41** SJ -.48** -.30* SP -.33* -.20 -.24 N=50 ** = p<.01 * = p<.05 All correlations among these variables are negative. This is because of the way the Keirsey scale asks questions and measures personality types. If one is an NF (NF = 1), one cannot be an NT, SJ, or SP (NT=0, SJ=0, SP=0). That is, the correlations are negative by construction. (So these results say more about measurement, than they do about whether there are "personality types" or that if there are personality types, Keirsey is correct about what they are.) The correlations are not perfectly negative (i.e., =-1) because of the way correlations are computed. (Some people with NT=0 have NF=1, but more people have NT=0 and NF=0.) 9.2. Correlation of philosophic tendencies in these data In the philosophy scales, a high score indicates greater tendency to agree with ideas of that philosophy. Since this was not a forced choice, respondents who agreed with Existential statements could also agree with Idealistic statements, and so on. 9.2.1. Correlations Existentialism Idealism Pragmatism Existentialism Idealism -.05 Pragmatism .22 -.01 Realism -.05 -.02 .64** N=50 ** = p<.01 * = p<.05 The only statistically significant correlation is between Pragmatism and Realism. Those who had high Pragmatism scores are likely to have high Realism scores. Those who had low Realism scores are likely to have low Pragmatism scores. All other correlations are not significantly different from 0. No other correlations are greater in absolute value than .22 (p > .10). Knowing how someone responded to Existential statements, for example, tells us nothing about how they responded to Idealistic statements. Put another way, some of those who had high Existentialism scores had high Idealism scores and some of those with high Existentialism scores had low Idealism scores, and vice versa. In the previous study, all philosophic scales were correlated with the minimum r = .24. Part of this difference is a function of removing half the items from the previous scales, in order to maximize item correlations. By maximizing the inter-item correlation within a scale, one expects the between-scale correlations to be reduced. Part of this may be due to differences in samples, particularly if there was differential exposure to consideration of philosophic tendencies (e.g., class lectures and discussions come to mind). This difference in studies again indicates the need for further research on and refinement of the philosophic scales. Philosophical types cannot be clearly distinguished. While some respondents had one high score and three lower scores, most had several high scores. Indeed, three respondents did not have a single "high score" but instead had the same high score on two scales. 9.3. Correlation of personality types and philosophic tendencies 9.3.1. Correlations NF NT SJ SP Idealism .09 -.18 .16 -.13 Realism -.45** .36* .11 .08 Pragmatism -.38** .22 .08 .17 Existentialism -.06 .12 -.01 -.04 N=50 ** = p<.01 * = p<.05 There is a pattern of statistically significant correlations relating the NF and NT Keirsey personality types with the philosophical tendencies of Pragmatism and Realism. Keirsey Idealists (NF) are significantly more likely to have low Pragmatism and Realism tendencies. Keirsey Rationalists (NT) are significantly more likely to have higher Realism tendencies. Note that they also are likely to have higher Pragmatism tendencies, although this correlation is not statistically significant at any of the usual levels (p = .13). We know that Pragmatism and Realism scores are strongly correlated. We know that the Keirsey scale has set up F (Feeling) and T (Thinking) as opposites, so this pattern of correlations is consistent. Non-significant correlations: No other correlation is statistically significant at the usual levels. Indeed, no other correlation is as large in magnitude as the correlation of NT and Pragmatism (.22, p=.13). 9.4. Personality types and their relation to philosophic tendencies We think of personality as being an important characteristic of individuals. We expect their personality to influence how they act and react, what they like and dislike, what they find good and bad. Philosophers think in much the same way about an individual's philosophy. How one views the world and how one relates to the world should influence how people act and react, what they believe, and what they find good and bad. Are an individual's personality and philosophy simply two names for the same thing? Or, if not the same thing, is one dependent on the other? If we know someone's personality, do we know that person's philosophy? If we know that person's philosophy, do we know her/his personality? Keirsey, in describing the different personality types which he claims exist, sometimes uses statements which we would claim are statements of philosophy. Thus, we would expect a correlation between personality types and philosophical beliefs. Our results show that this is not the case. While some personality types are moderately related to certain philosophical beliefs, for the most part personality and philosophy are essentially unrelated. Even where personality and philosophy are related, it is not a strong relationship. A number of different conclusions may explain this finding: (1) Personality and philosophy are truly different. Attempting to understand people solely in terms of their personality is at best an incomplete understanding. (2) Personality type and philosophical tendency are more closely related than indicated here, but our measurement of philosophy is too prone to error to show it. Certainly, more work needs to be done to refine our measures of philosophical tendency. (2a) In particular, it may be that there are philosophic types, rather than philosophical tendencies. If people were forced to choose one set of beliefs over others, rather than just agreeing or disagreeing with certain statements, the results might be quite different. (3) It may be that there are not personality types, as opposed to the holding of Keirsey. That is, people may not fall into four, clearly separated categories. Rather, they may have differing tendencies along certain dimensions, some of which may be consistent with Keirsey's types, but some of which may be inconsistent. It may be that personality tendencies (e.g., aggression, depression, etc.) may be more closely related to philosophical beliefs. (4) Philosophical tendencies have a moral component, whereas personality types do not. If we are interested in moral development, philosophical tendencies appear to have more to contribute than do personality types. References Keirsey, D.M. (1997). The Keirsey Character Sorter. (Online: Ross, C., & Barger, R.N. (1990). The Ross-Barger Philosophy Inventory. (Online: