The Myers-Briggs Type Indicator and Applications to Accounting Education and Research.
To address the many challenges facing the accounting profession in the new millennium, the AICPA has issued The CPA Vision Project--2011 and Beyond (ATCPA 1997). Many of the recommendations in The Vision require important changes in the nature of being an accountant; e.g., greater emphasis on professional demeanor, leadership, and interpersonal communications. The stereotypic accountant will play a decreasing role in this new environment, yet two decades of research indicate little change in the personality makeup of those choosing to study or practice accounting (Schloemer and Schloemer 1997; Wolk and Nikolai 1997).
Restructuring accounting education is needed to facilitate these required changes in the natures of accountancy and accountants. The academy of accounting educators is increasingly called upon to teach not only technical and functional abilities, but also personal and interpersonal skills (AAA 1986; AECC 1990; Williams 1999). The AICPA Core Competency Framework for Entry into the Accounting Profession (AICPA 1999), the most recent professional statement regarding the education of accountants, lists "personal competencies" as one of the three major categories of competencies accounting students need to acquire prior to entering the work force.
At core of many of these issues are personality characteristics of accountants, accounting teachers, and accounting students. To address these issues, a research paradigm that enables the academy to ask new questions--questions about the nature of personality in an accounting context--would be beneficial. However, there must be an accompanying psychometric instrument that allows researchers to gather data and test hypotheses developed from the paradigm. This paper discusses a research stream--Jungian psychology--that meets these requirements. Jung's (1921) theory of personality types (hereafter Jung's theory) places the individual as a whole within a broad context, and has given rise to an instrument for studying personality, the Myers-Briggs Type Indicator (MBTI). 
The remainder of this paper examines the MBTI from four viewpoints. The first section provides a brief overview of Jung's theory of personality types. While the emphasis in this paper is on the MBTI, some understanding of theory is necessary for proper use of the instrument. Next, the psychometric properties of the MBTI are examined, especially its reliability and validity. The third section summarizes how the MBIT has been used in accounting research. A section follows that suggests opportunities for MBTI-based research in accounting education and practice.
THE JUNGIAN THEORY OF PERSONALITY TYPES
The Role of Personality in Human Behavior
Explaining human behavior can be attempted at several different levels. At one end of the research spectrum is the neural-physiological approach; at the other is the sociological. Between these extremes are theories of individual psychology, group behavior, and organizational and institutional structure. At the individual level, current research primarily has evolved along three tracks: (1) behaviorism, (2) personality theories, and (3) cognitive science (Engler 1999; Flanagan 1991).
Behaviorism employs an external-stimulus/behavioral-response paradigm that tends to ignore internal psychological processes. Personality theories and cognitive science, on the other hand, place greater emphasis on the internal processes. Behavioral and cognitive paradigms have been used extensively in accounting research; personality theories have not. This paper suggests one way of redressing this imbalance.
In general, personality theories are concerned with listing and defining as exactly and parsimoniously as possible the characteristics that comprise the person as a whole. The following passage from a best-selling undergraduate psychology textbook provides a context in which to place personality theories (Hergenhahn and Olson 1999, 6):
Personality theorists are in the unique position in psychology of studying the entire person. Most other psychologists are concerned with only one aspect of humans such as child development, old age, perception, intelligence, learning, motivation, or memory. It is only the personality theorist who tries to present a complete picture of the human being.
Of the current personality paradigms, Jungian theory is a major representative of the psychoanalytic paradigm (Engler 1999; Hergenhahn and Olson 1999).  Unlike other psychoanalytic personality theories (e.g., Freudian), Jungian psychology de-emphasizes the role of the unconscious and focuses on conscious aspects of personality, decision making, and the effect of personality on understanding. Because of this orientation, Jungian theory continues to be an important influence in psychology and shares many concerns and features with cognitive science (Geyer 1998).
Overview of Jungian Personality-Type Theory
Jung's theory emphasizes the entire personality ("type") of the individual, not the separate characteristics ("traits"). According to theory, a type (e.g., extraverted thinker supported by intuition) consists of various traits (e.g., extraversion, thinking, intuition, and judging) that interact to form the personality. Because of this interaction effect, a trait in one type will have a different effect on the personality than the same trait in another type. 
Jungian theory postulates eight primary personality traits that divide into two dichotomous pairs of mental functions and two dichotomous pairs of attitudes: Extraversion (E) and Introversion (I) denote the El attitude toward the world; Sensing (5) and Intuition (N), denote the SN Perceiving mental function; Thinking (T) and Feeling (F), denote the TF Judging mental function; and Judging (J) and Perceiving (P), denote the JP attitude toward the mental functions.  All eight traits are present in each person to varying degrees. One trait from each of the four dichotomies is "preferred" and predominant--preferred, that is, in the sense that each individual has an innate inclination toward one of the two traits in each dichotomy. The four preferred traits (preferences) interact to "define" the personality type. However, the four nonpreferred traits are still present in the personality, and the individual may be quite competent using them. 
This system of preferences results in 16 possible combinations of the preferred traits. These combinations, referred to as the Jungian personality types, are presented in Table 1.  There are numerous approaches for characterizing the 16 types, e.g., by occupational and organizational traits, educational traits and learning styles, and decision-making traits and cognitive styles (Pascal 1992; Myers and Myers 1995; Myers et al. 1998). Type profiles and descriptions can therefore become very complex and detailed. To illustrate those aspects of personality most relevant to accounting research and education, Table 1 lists only the general personality characteristics and occupational tendencies of each type.
According to Jung's theory, many differences in personality and behavior are explainable as differences in the way individuals perceive the world and form judgments about it. Jungian theory therefore places great emphasis on mental functions. It posits, first, two traits in the Perceiving mental function (the SN dichotomy); viz., Sensing (S) and Intuition (N). These input traits indicate the sources of the data subsequently processed by the Judging function. Sensing refers to direct or objective perceptions made through the physical senses. Intuition refers to subjective or indirect perceptions of the structures and relationships among objects and experiences. For example, comparing the personality types in the Sensing columns in Table 1 to those in the Intuition columns suggests that the former types tend to have factual and observant characteristics, while the latter tend to have insightful and creative ones.
The two traits of the Judging mental function (the TF dichotomy) are Thinking (T) and Feeling (F), which define how inputs from the Perceiving function are transformed into useful output. Thinking refers to processes that link ideas and experiences together by making logical connections, and tends to be impersonal. Feeling refers to processes of decision making that involve weighing the relative merits of personal and group values. Contrasting the personality types in the Thinking columns in Table 1 to those in the Feeling columns indicates that the former types have logical and rational natures, while the latter have idealistic and compassionate natures.
Extraversion (E) and Introversion (I) are attitudes that describe the individual's most fundamental approach toward the world in its internal and external aspects. This pair of traits is referred to as the EI dichotomy. Extraversion refers to an attitude wherein the individual's focus is drawn to the people and objects that comprise the external world. Introversion refers to an attitude wherein the individual's attention is directed primarily to the inner environment of the mind and those things that comprise this subjective world. A comparison of the types in the Introversion rows in Table 1 to those in the Extraversion rows reveals the former types as more detached and contemplative and the latter as more personable and outgoing.
The fourth dichotomy (the JP dichotomy) indicates the individual's overall attitude toward the two mental functions; i.e., the interaction of the SN Perceiving mental function with the TF Judging mental function.  This attitude has two traits: Judging (J) and Perceiving (P). J-preference individuals prefer planning and organizing activities, tending to follow problems to their solution. P-preference individuals tend to gather as much information as possible and thus to leave a problem unsolved as long as possible. Contrasting the types in the Judging rows in Table 1 to those in the Perceiving rows indicates that the former types tend to be decisive and committed, while the latter tend to be open-minded and questioning.
TILE MYERS-BRIGGS TYPE INDICATOR
The MBTI is a questionnaire-style instrument consisting of items arranged in a forced-choice format. For each item, subjects are provided two responses to choose between. The MBTI has undergone numerous modifications since it was first published in 1962. The MBTI versions mentioned in this paper are Forms F, G, J, and M, although there are earlier ones. Form F was published in the early 1970s, Form G in 1977, and the current Form M in 1998. Form J is not for general use; it is an extended version for those doing research on, as opposed to with, the MBTI. Form M is an improvement over Form G in that the wording has been updated, testing indicates increased reliability and validity scores, and the scoring method is more accurate.  Form M takes about 25 minutes to complete, and can be scored and analyzed by the subject, the on-site administrator, or the vendor.
The objective of the MBTI is to classify an individual into one of the 16 personality types--the four-way interaction of the preferences. The central importance of types does not imply, however, that research cannot be conducted below the type-level; i.e., on preferences separately or two- and three-way interactions of preferences.
The Reliability and Validity of Scores Produced by the MBTI Reliability Issues Related to the MBTI
The MBTI has been subjected to several reliability tests, evidence from which (summarized in Table 2) provides strong support for the reliability of scores produced by the instrument.
The internal consistency reliability studies have typically resulted in correlations and coefficient alpha scores in the 0.80s and 0.90s, with some low scores in the high 0.50s and 0.60s. These results indicate that subjects tend to answer interrelated questions in the MBTI in a consistent manner. The temporal stability correlations range from a low of 0.59 to a high of 0.94. In these studies, from 71.7 percent to 93.0 percent of subjects indicated three or more of the same preferences. These results suggest that subjects select the same answers over time. The scores from all these tests, including the low scores, are within acceptable ranges (Nunnally and Bernstein 1994) and compare favorably to other psychometric instruments (Harvey 1996).
Validity Issues Related to the MBTI
Validity analysis of the MBTI has centered on the four dichotomies as separate, independent factors (Table 3, Panel A) and on interactions among the dichotomies (Table 3, Panel B). The results from the validity studies are not as uniformly supportive of the MBTI as those from the reliability studies, clearly suggesting that the MBTI is not capturing the personality in its entirety. However, the results strongly indicate that the MBTI is measuring most aspects of personality in a way consistent with Jungian psychology. Table 3 summarizes the results of validity tests of the MBTI.
Validity studies of the MBTI dichotomies as independent factors (Table 3, Panel A). Most discriminant validity tests of the MBTI dichotomies as separate, noninteracting variables have produced factor structures consistent with the four-scale MBTI model, although two studies resulted in structures inconsistent with the Jungian model (Sipps and Alexander 1987; Sipps et al. 1985). A limit to these studies, however, is that several models may fit the same data points. The two studies have also been criticized on methodological grounds (Myers et al. 1998).
Convergent validation studies of the dichotomies as separate variables, based on comparison of MBTI traits to non-MBTI traits, indicate positive correlations between the two sets of characteristics. Participants describe themselves and experts describe participants using non-MBTI descriptors in a way consistent with the preference-profiles indicated by the MBTI scores. The positive correlations suggest that both instruments provide coverage of the same phenomena. The non-MBTI descriptions used in these studies were from non-Jungian sources (i.e., the ACL, CPI, 16PF, and NEO-PI)  and from Jungian sources besides the MBTI (i.e., the JTS and Jungian Q-Sorts).
A common result of the convergent validity studies is a positive correlation of the non-MBTI trait with more than one of the MBTI preferences (e.g., the CPI trait flexibility correlates with the SN and JP dichotomies). This result is nonsupportive of convergent validity because it indicates difficulty directly mapping the non-MBTI traits onto the MBTI dichotomies. Further analyses of these nonsupportive results have noted that (1) the correlated non-MBTI and MBTI descriptors provide equal coverage of the same phenomena, and (2) the mapping problems are best understood as due to differences in the underlying theories rather than validity problems (Harvey 1996).
Validity studies of the interactions of the four dichotomies (Table 3, Panel B). Because Jungian theory postulates that the preferences interact to form personality types, complete validation of the MBTI requires evidence for this interaction. Studies testing this aspect of the MBTI are presented in Table 3, Panel B.
In construct validation studies of the interactions of the dichotomies, referred to as "best-fit" studies, participants rank various self-descriptions for accuracy.  The rankings are then compared to the MBTI type results. Using the older Forms F and G, subjects ranked the MBTI type profile as most like themselves from 35 to 62.2 percent of the time. With the new Form M, the MBTI type was ranked most accurate by 58 to 78 percent of the subjects. In assessing these studies, note that the chance of randomly picking the correct type is one in 16 (6.25 percent). Taken as a whole these studies suggest that people identify with descriptions of their own type more than they do alternative descriptions.
Convergent validation studies of the interactions of the MBTI dichotomies compare the interaction aspects of the MBTI to other non-MBTI personality scales (e.g., KTS, ACL, CPI, and LSI). These studies have consistently found positive correlations between the MBTI preferences on an interaction level and non-MBTI scales. The results thus indicate that participants and expert observers identify correspondences between holistic (interactive) personality characteristics in the MBTI and the non-MBTI instruments. Additionally, the ability to find correspondences implies that subjects and experts are able to distinguish between the MBTI types.
APPLICATIONS OF THE MBTI IN ACCOUNTING RESEARCH
The amount of research being done using the MBTI is massive. Myers et al. (1998) estimate that over 4,000 research studies, journal articles, and dissertations have been written using the MBTI since the publication of the first instrument in 1962. According to Hammer (1996), the number of academic articles being published in this area has been averaging over 200 annually since 1985. In this context, the amount of accounting research done with the MBTI is surprisingly small. My own database and reference searches uncovered 16 published articles using the MBTI for accounting research starting with Shackleton (1980)--less than one-half of one percent of the MBTI articles published during the same period.
The accounting areas researched with the MBTI have been fairly narrow. They are: (1) indicated personality types of professional accountants; (2) correlations between performance and indicated personality types of professional accountants; (3) indicated personality types of accounting students; and (4) correlations between performance and indicated personality types of accounting students. Table 4, the basis for the following discussion, summarizes these articles according to this classification scheme.
Personality Types of Professional Accountants
The seven studies in Panel A of Table 4 are primarily concerned with investigating the personality types of professional accountants. The consistency of findings concerning the indicated personality types of accountants is remarkable considering the fact that these studies cover wide ranges of time (1980 to 1997), firm size (from small, local firms to Big 8/Big 6), and location (the U.S., the U.K., and The Netherlands).
Shackleton (1980), the first published MBTI accounting study, investigated 91 accountants and financial managers and found that the STJ-preference was dominant (38.5 percent). He also reported that 58 percent were I-preferenced. All these percentages are above those based on random samples of the population (Myers and McCaulley 1985). Jacoby (1981) examined 333 accountants from Big 8 firms in Washington, D.C. STJ-preference individuals were again dominant (33.6 percent), with ISTJs at 19.8 percent and ESTJs at 13.8 percent.
Otte (1984) studied CPAs in small, local firms. His results indicate that 45.9 percent were STJ-preferenced (ISTJs at 26.7 percent and ESTJs at 19.2 percent). Comparing his results to Jacoby (1981), Otte (1984) found that CPAs in local firms have significantly (p [less than] 0.05) more ST-preferences than CPAs in national firms. Comparisons within the profession suggest that the S-preference is higher (p [less than] 0.05) in accounting and auditing than in tax and consulting. ET-preferences were higher (p [less than] 0.05) in consulting than in accounting, auditing, or tax. Descouzis (1989) found that the 36 tax preparers examined were S-preferenced. ISTJs (25.0 percent) and ISFJs (19.4 percent) were overrepresented. Of the sample, 44.4 percent consisted of the STJ-preference. In addition, there were significant differences between male and female subjects: 83 percent of the females were F and 78 percent of the males were T. These results hold qualitatively for Kreiser et al. (1990), Satava (1996), and Schlo emer and Schloemer (1997).
Satava's (1996) findings, based on 439 CPAs from different-sized firms, suggest that the E-preference is stronger for individuals working in national firms, whereas the I-preference is dominant for personnel in local firms. Schloemer and Schloemer (1997), examining 125 CPAs from large local firms and Big 6 firms, found that the E-preference was greater for more recent partners. Eighty-three percent of partners after 1989 were E-preference, while only 20 percent of pre-1989 partners were E-dominant. Schloemer and Schloemer's (1997) results also indicate a gender effect in the TF-preferences. Thus, despite changes in the accounting environment and calls for greater diversity in the profession, based on research conducted to date there is little evidence of change in the indicated personality types of accountants.
Kreiser et al. (1990) investigated the perception of public accountants by other business professionals. A survey consisting of MBTI preference descriptions was given to 92 bank officers and 113 members of the Financial Executives Institute (FEI). These participants were asked to indicate whether the traits were descriptive of accountants. Accountants were perceived to have STJ-preferences 84 percent of the time (59.1 percent ISTJs and 24.9 percent ESTJs). This statistic is remarkable, considering that the STJ-preference percentage from samples of CPAs is usually about half that amount (38.5 percent in Shackleton 1980; 45.9 percent in Otte 1984; 36 percent in Schloemer and Schloemer 1997). This difference between the actual and perceived images of the accountant has interesting implications for the profession. For example, as accounting moves into areas where "soft" skills are required, it may be difficult to convince potential clients that accountants actually possess these skills because of this apparent b ias.
Jacoby (1981) also investigated the preference structure of accountants within various areas of the profession. Audit and tax partners and managers were more likely to have TJ- preferences than those at lower professional levels in these areas. Also, audit partners and managers were more likely to have IS-preferences than partners and managers in tax and consulting. Jacoby's (1981) study overlaps the articles in the category of MBTI accounting research that examine the relationship between performance and the personality structure of accountants (Table 4, Panel B). However, Jacoby (1981) did not examine the dynamic aspects of the relationship between performance and preferences, but instead confined his research to classifying where the preferences occurred in the accounting organization.
Performance and Personality Types of Professional Accountants
The three articles in Panel B of Table 4 use the MBTI to investigate functional relationships between performance and indicated aspects of personality. Chenhall and Morris (1991) used the MBTI (Form G) to investigate the effect of "cognitive style" on opportunity-cost decision making in 64 middle- to senior-level managers. Cognitive style may be defined as the way an individual processes and transforms stimuli from the environment to shape a response. The SN scale was used to measure cognitive style. The results indicate that those managers with an N-preference tended to incorporate opportunity costs in their resource-allocation decisions. Those with the S-preference tended not to identify the opportunity-cost implications. The introduction of project sponsorship (whether managers supervise the project) moderated the influence of the SN scale on decision making. Project sponsorship is therefore hypothesized to proxy for accountability. Chenhall and Morris (1991), in short, illustrate how the MBTI can be used to investigate the cognitive styles of accountants, the importance of a personality trait on professional behavior, and a better understanding of decision making within a broader context than that provided by a strictly information-processing approach.
Vassen et al. (1993) examined 25 auditors from the three largest accounting firms in The Netherlands. Like Chenhall and Morris (1991), Vassen et al. (1993) did not use all four MBTI preference scales. For cognitive style, they used the SN scale to represent "information acquisition" and the TF scale to represent "information processing." Vassen et al. (1993) found that S-preference auditors accessed more information than N-preference auditors, and that T-preference auditors took longer to process the information accessed and had a lower tolerance for ambiguity than those with a F-preference. The auditors in their sample were predominantly 5-preferenced (68 percent, p [less than] 0.05). Consistent with U.S. studies (e.g., Schloemer and Schloemer 1997), most of the auditors in the Vassen et al. (1993) Dutch study were T-preferenced (14 of 25; 56 percent), although this statistic was not significant (p = 0.13).
Scarbrough (1993) investigated the relationship between job satisfaction, measured by the Minnesota Satisfaction Questionnaire, and personality type, measured by the MBTI, Form G. The sample consisted of 255 mid-level accountants from a Big 8 firm in the U.S. Once again, the STJ-preference was overrepresented (34.9 percent). Prior to dividing the sample by gender, the only difference in job satisfaction was on the JP scale, where J-preference individuals were marginally more satisfied than P-preference individuals (p [less than] 0.07). Additional results appeared by comparing male to female mean scores. Females were more satisfied than males on the I-preference (p [less than] 0.005), the T-preference (p [less than] 0.08), and the J-preference (p [less than] 0.03) scales. On the personality-type level, the only significance difference was between ISTJ females and males, with women indicating more job satisfaction (p [less than] 0.05).
Personality Types of Accounting Students
Panel C of Table 4, concerning the personality types of accounting students, lists two studies. Laribee (1994) administered Form G to 320 undergraduate accounting students (105 sophomores, 165 juniors, and 50 seniors). These students were compared to traditional-age college students, to accounting professionals, and to one another across the three classes. Compared to traditional-age students, the STJ-preference was overrepresented among accounting students (37.3 percent vs. 20.5 percent). Despite the accountant stereotype (Kreiser et al. 1990), the I-preference was not dominant (56 percent). Among males, accounting students had higher levels of E-, S-, T-, and J-preferences than traditional-age male students, although E decreased with class level. Among females, accounting students had higher levels of S and T than traditional-age female students. A comparison of accounting students and professionals revealed an interesting filtering-out effect. Sophomores, compared to accountants, were different (p [less t han] 0.05) in the E (60 percent vs. 44 percent), S (61 percent vs. 72 percent), T (60 percent vs. 76 percent), and J (60 percent vs. 76 percent) preferences. However, senior accounting students and accountants were not different statistically. Thus, a change occurred between the sophomore and senior levels in which the E-, N-, F-, and P-preferences decreased. This filtering-out effect is a major focus of suggested research in the next section.
Wolk and Nikolai (1997) used the MBTI to investigate differences between 152 accounting undergraduates, 94 accounting graduates, and 98 accounting faculty members. There were no statistically significant differences between accounting undergraduates and accounting graduates on the EI and JP scales. There were, however, differences on the SN and TF scales. Undergraduates were higher in the 5and F-preferences than graduates. The results were qualitatively the same in a comparison of undergraduates and accounting faculty, except that undergraduates were higher on the E-preference than faculty. There were no statistically significant differences between accounting graduate students and accounting faculty on the four dichotomous scales. Wolk and Nikolai (1997) found similar results when looking at the distribution of personality types. Undergraduates were different from both graduates and faculty in that the undergraduates were more likely to be ESTJ. Graduates and faculty showed no statistically significant diff erences. It is interesting that, in terms of the MBTI results, undergraduates were closer to professional accountants than were either graduates or faculty. Wolk and Nikolai (1997) argue that these results have important implications for devising strategies for attracting the right types of students to accounting and for teaching accounting more effectively.
Performance and Personality Types of Accounting Students
The final category of MBTI accounting research consists of four articles that studied various relationships between the performance of accounting students and their indicated MBTI preferences (Table 4, Panel D). Ott et al. (1990) investigated the relationship between personality traits and the effectiveness of different methods of classroom instruction. Instructional effectiveness was based on student performance as measured by exam grades. Ott et al. (1990) administered the MBTI to 89 undergraduate students enrolled in the first introductory accounting course. The students were assigned randomly to one of two teaching methods: lecture groups or computer-assisted instruction (CAI) groups. Of the four preference scales, only the SN (p [less than] 0.001) and TF (p [less than] 0.05) had significant interactions with teaching method. S-preference students performed better with the lecture method than with the CAI method, while N-preference students performed better with the CAI method. T-preference students perf ormed better with the lecture method; F-preference students performed better with the CAI method.
Nourayi and Cherry (1993) tracked the performance of 103 accounting students after completing a first intermediate financial accounting course. The authors examined the relationship between the MBTI preferences and performance in seven accounting courses. As with Ott et al. (1990), performance was measured in terms of course grades. The only statistically significant results (p [less than] 0.05) were found on the SN scale, with S-preference students outperforming N-preference students in three of the seven courses, and in an average of the seven accounting classes. The three courses were second intermediate, tax, and auditing. S-preference student performance in tax and auditing was especially strong (p = 0.002 and p = 0.006, respectively), consistent with Jacoby's (1981) results.
Landry et al. (1996) compared computer usage to the MBTI personality structure using two instruments. One instrument measured student attitude toward learning about and using computers. The second instrument simulated computing situations to classify students as either a user or nonuser. The sample of participants consisted of 88 junior and senior-level accounting students. Landry et al. (1996) found the STJ-preference overrepresented (42 percent of the accounting students were STJ; ISTJ at 17 percent; ESTJ at 25 percent). As with Ott et al. (1990), only the SN (p = 0.001) and TF (p = 0.064, marginal) scales were correlated with computer usage. Also like Ott et al. (1990), Landry et al. (1996) show that those with the N-preference are less "computerphobic" than those with the S-preference. However, not in agreement with Ott et al. (1990), the results of Landry et al. (1996) suggest that those with the T-preference are more likely to be computer users than those with the F-preference. Landry et al. (1996) exa mined the two-way interactions of the SN and TF scales and found that NT-preference students were less computerphobic than were the ST-, SF-, and NF-preference students.
Oswick and Barber (1998) compared indicated personality preferences (based on the MBTI, Form G) to performance in an introductory accounting course. The sample consisted of 344 U.K. traditional-age nonaccounting majors. Performance was measured by course grade. Oswick and Barber (1998) found that there were no statistically significant correlations (p [less than] 0.05) between indicated personality preferences and performance. There was a marginal difference (p [less than] 0.10) for the TF-preference, with T doing better. These results appear to conflict with those of Nourayi and Cherry (1993), which found that S-preference students earned higher grades in accounting courses than N-preference students. This inconsistency may be explained by differences in the samples of the two studies. Oswick and Barber (1998) examined nonaccounting majors, while Nourayi and Cherry (1993) studied accounting majors. Cultural differences between the two samples may also account for some of the inconsistency. Nourayi and Cherr y (1993) conclude by making the interesting observation that although STJ may be the most prevalent type among accountants and accounting majors, STJs are not necessarily the best accounting students. "It is important to distinguish between aptitude and inclination" (Oswick and Barber 1998, 253).
Summary of Results from MBTI-Based Accounting Research
Despite two decades of change in the accounting profession, research indicates that the distribution of personality types among accountants is remarkably stable (STJ) across time, location, and firm size. However, there is evidence of differences in the indicated personality types of accountants within the levels of a firm's organization and among the various specialties in the profession. The E-preference has increased over time and is more likely in a national firm than in a local firm.
Undergraduate accounting students have a personality-type distribution similar to that of accountants. There are apparently differences between the personality types of accounting undergraduates, accounting graduates, and accounting faculty. Undergraduates differ from the latter two groups, while there is little difference between accounting graduates and accounting faculty. Of the three groups, undergraduates appear closest to the indicated personality types of accounting professionals. In accounting education, a filtering-out process may be occurring that decreases the E-, N-, F-, and P-preferences in accounting students.
Two studies indicate significant relationships between the MBTI preferences and decision making and cognitive style in practicing accountants (Chenhall and Morris 1991; Vassen et al. 1993). Results of the relationship between MBTI preferences and academic performance in accounting are mixed. A study of accounting majors shows a correlation between performance and the SN scale only (Nourayi and Cherry 1993). A study of nonaccounting majors found no significant correlations with any of the scales (Oswick and Barber 1998). Results on the relationship between indicated personality types and teaching styles suggest that N-preference students responds more favorably to computer-assisted teaching than to lecture-style classes. S-preference students performed better with the lecture method. The results from two studies (Ott et al. 1990; Landry et al. 1996) are inconsistent in regard to the relationship between computer usage and the TF scale.
The MBTI can be used to address four general areas of accounting research. First, research on the psychometric properties of the new MBTI version (Form M) can be examined in the context of college-aged accounting students. Second, MBTI-based research can be conducted on accounting students, especially on the prevalence of indicated personality types, academic performance, and choice of major. Third, pedagogical research can be carried out on fine-tuning teaching styles; e.g., to match teaching styles with learning styles based on the respective personality types of teachers and students. Fourth, MBTI research in the accounting profession can be conducted on the prevalence of indicated personality types, career success, job placement, and uses of the MBTI within companies.
Although psychometric data has been collected on the MBTI from the population at large (Myers et al. 1998), little, if any, has been gathered from the populations of those choosing to study or practice accounting. This is especially important in light of the newness of Form M (1998). As with any instrument, the reliability and validity of Form M need to first be established for the population of interest prior to conducting basic research on that population. For example, temporal stability studies can investigate changes in accounting students' preferences and types as they transition through school or into entry-level positions. Theory predicts that preferences are highly stable but that the strength of the preferences (or preference clarity) can change, especially in young adulthood. Accounting psychometric studies can also be conducted on the convergent validity properties of the MBTI and a non-MBTI instrument. The characteristics of the scales have specific meanings that vary with the subjects being inve stigated. For example, feeling among accounting students may carry different connotations than among nonaccounting subjects. Comparison of the MBTI F-preference to related non-MBTI traits may reveal these differences.
The above analysis of prior MBTI accounting papers is suggestive of numerous specific opportunities for applying the MBTI to the latter three research areas. For example, many institutions have initiated changes in their accounting curricula and teaching methods. Studies can investigate whether these changes result in greater diversity in accounting students and attracting the types of students needed in the profession's new environment (Wolk and Nikolai 1997). Similar studies can explore whether these institutional changes have affected the filtering-out process that decreases the percentages of E-, N-, F-, and P-preferences among accounting students. However, accounting researchers need to first understand why this filtering is occurring. Consideration of ways to attenuate the filtering-out process so that diversity can increase among accounting students can then be addressed in a systematic manner. It may, for example, be related to teaching style, although accounting teachers tend to be N-preferenced.
The crucial question is to ask what is lost to the profession by this filtering process. Halting this process may help provide the profession with the wide range of skills needed (AICPA 1999). Research can seek to understand how to relate the indicated personality types to professional accounting skills. Perhaps MBTI-based research can help the academy understand how to attract the "right" students to accounting (Wolk and Nikolai 1997). Ultimately, stopping the filtering-out of certain traits could decrease the declining selection of accounting as a major. In any event, research is needed to address these questions.
The mixed results from the two studies of the academic performance in undergraduate accounting courses are especially interesting (Nourayi and Cherry 1993; Oswick and Barber 1998). Nourayi and Cherry (1993) found a correlation between indicated personality traits and performance; Oswick and Barber (1998) did not. Research should consider whether these inconsistencies are due to differences in the samples (for example, cultural differences) or differences in student major (accounting vs. nonaccounting majors). Furthermore, if there are no correlations between personality traits and performance in introductory accounting courses, then research should explore why there are correlations between personality and performance in later accounting courses. Finally, if it is not performance, then it would be interesting to know what personality traits are determining the self-selection of those proceeding to the later accounting courses (and becoming accounting majors and accountants).
The learning styles of individuals affect their performance and are, in turn, affected by their personalities. The MBTI has the potential to further improve our understanding of learning styles. The only accounting learning style issues investigated using the MBTI have been computerphobia (Landry et al. 1996) and student receptiveness to lecture-style classes vs. computer-assisted instruction (Ott et al. 1990). There are many other learner characteristics that can be investigated using the MBTI, e.g., written vs. oral, team vs. individual, concrete vs. theoretical, and structured vs. open-ended problems (Myers et al. 1998). Knowing such characteristics of students can assist instructors in deciding on the best teaching method to use. However, accounting researchers need to first investigate the relationships between student preferences and teaching styles before they can be used effectively in the classroom. Many of these relationships are not understandable at the present. For example, the conflicting resul ts on the use of computers and TF scales need to be explained.
In the classroom, students are only half the equation. The one MBTI accounting study (Wolk and Nikolai 1997) that examined the indicated personality types of accounting faculty did not investigate the relationship between faculty personality and teaching method. Perhaps there is a relationship between accounting faculty personality types and their effectiveness in using certain teaching methods. Also, there may be a relationship between accounting faculty personality types and learning styles. The relationship between accounting faculty personality types and the types of accounting courses they teach, and how effectively they teach them, can also be studied. In short, MBTI studies can potentially improve our understanding of classroom dynamics, i.e., the way accounting professors teach and the way their students learn. Research is needed to provide understanding of these admittedly complex relationships.
In the area of how accountants perform their tasks, much more can be done than the research conducted by the two studies to date (Chenhall and Morris 1991; Vassen et al. 1993). Both studies examined only auditors, cognitive processes, and the SN and TF scales as noninteractive variables. Other professional areas in accounting are open to research on the relationship between performance and personality types. These research opportunities are particularly numerous at present because of the many changes in the tasks and responsibilities of accountants accompanying changes in the work environment. The new nomenclature in the profession--assurance services, information-technology consulting, professional service firm--is indicative of requirements for new performance capabilities on the part of accountants.
Additionally, research into the relationships between the performance of accountants and two-way, three-way, and four-way preference interactions needs to be done; i.e., beyond the level of separate preferences. A strength of MBTI research is that it can examine aspects of the mind besides cognition and information processing. According to Jungian theory, the interactive aspects of the attitude preferences (EI; JP) embed the mental functions (SN; TF) within a larger context of noncognitive, personality characteristics.
Group dynamics is one of the more promising areas of MBTI research open to the accounting researcher (Hammer and Huszczo 1996). The AICPA's (1999) Vision lists teamwork and, in particular, the ability to work "within diverse, cross-functional teams" as required skills for accountants in the new environment. The Bedford Committee report (AAA 1986) promotes increased group work and group-based learning in the accounting curriculum. The MBTI can address both cognitive and noncognitive aspects of groups. From a cognitive perspective, groups may be studied as problem-solving units by focusing on the distribution of MBTI-indicated mental functions among group members. Perhaps certain personality mixes lead to greater efficiency or effectiveness in solving problems. Noncognitive aspects of groups also may be addressed by the MBTI--an advantage the MBTI offers the accounting researcher compared to cognitive-science instruments. For example, because communication in the group affects overall group performance, certai n combinations of types may be better at communicating than others.
MBTI-based group research can address other pedagogical questions. The accounting instructor may be able to improve the effectiveness of group learning and team projects by selecting membership based on personality traits or types. An examination of whether groups self-select in an equally advantageous manner would be of interest. Personality traits and types of effective groups may vary with types of accounting course. For example, the characteristics of effective financial accounting teams could be different from those of auditing teams.
A parallel set of questions can be asked about group dynamics in the accounting profession. Research could reveal how much forethought should go into deciding about team composition. Professional accountants would benefit by knowing which traits and trait interactions are most closely related to effective team performance in making personnel assignments. Contextual effects are potentially important here; some membership arrangements may be better for certain types of tasks than others. Perhaps the traits and types of effective groups vary with the type of engagement, i.e., tax, auditing, financial, or consulting. Finally, client personality traits may have an impact on the accounting team's effectiveness and could influence personnel assignments. The profession begs for research to address these issues. Ultimately, MBTI-based research is a fertile ground for exploring many unanswered questions related to teaching, learning, and professional performance. 
Jungian personality-type theory is well developed and established. It does not focus on any one aspect of the personality but instead examines the personality as a whole, both in its structure and its relationship with the world. This comprehensiveness is an advantage that Jungian theory and its associated instrument, the MBTI, offer researchers addressing the professional and pedagogical issues facing accounting. The MBTI research stream is also a rich source from which accounting researchers can draw. In contrast to the large amount of MBTI research that has been conducted in nonaccounting areas, the amount of published MBTI accounting research is surprisingly small. The personalities of accounting students, teachers, and professionals have been investigated in only a few areas. Accounting education researchers and teachers thus have numerous opportunities available to them using the MBTI.
Patrick Wheeler is an Assistant Professor at the University of South Florida.
The author gratefully acknowledges the many helpful comments on earlier versions of this paper by Barbara Apostolou (associate editor), David E. Stout (editor), Joseph DaCrema, Donald Jones, and an anonymous reviewer.
(1.) There are several other personality instruments developed from Jungian theory; e.g., the Jungian Type Survey and the Keirsey Temperament Sorter. The Jungian Type Survey (JTS) predates the MBTI and is no longer widely used or easily available. The Keirsey Temperament Sorter (KTS) is easily available, but has not been tested for reliability and validity. For more on the Jungian Type Survey refer to Wheelwright et al. (1964), and for more on the Keirsey Temperament Sorter refer to Keirsey (1998).
(2.) According to Hergenhahn and Olson (1999), there are six paradigms for explaining personality: (1) psychoanalytic, (2) socio-cultural, (3) trait, (4) learning, (5) socio-biological, and (6) existential-humanistic.
(3.) This interaction determines which traits will be dominant, inferior, auxiliary, and tertiary. An understanding of these aspects of the traits is important for an overall understanding of the personality type. For a complete discussion of the roles of the dominant, inferior, auxiliary, and tertiary traits in Jungian personality theory refer to Pascal (1992) and Myers et al. (1998).
(4.) To avoid confusion between the everyday and the specialized use of terms for the eight personality traits, I have employed the convention employed by Myers et al. (1998) of capitalizing the specialized usage. For example, Thinking (the Jungian trait) vs. thinking (the everyday concept).
(5.) Left-handedness and right-handedness are often used to illustrate the concepts of preference and nonpreference. Most people have a preference for one hand over the other. This preference is not consciously chosen but is primarily genetically determined. Furthermore, proficiencies with the preferred and nonpreferred hands vary widely. Some individuals have better handwriting with their nonpreferred hand than others with their preferred hand. These characteristics of handedness apply by analogy to trait preferences.
(6.) Jung's original theory contained eight personality types. Myers and Briggs expanded Jung's theory to include a new dichotomy, the JP attitude (Myers et al. 1998). This addition doubles the number of personality types to 16. Most Jungians accept this interpretation as consistent with Jung's implicit views (Hergenhahn and Olson 1999; Pascal 1992). It is the one adopted for this paper because this theory underlies the MBTI.
(7.) As noted previously, the JP dichotomy was not explicitly developed by Jung (1921), but was the work of I. Meyers and K. Briggs. It was introduced in response to mixed results from attempting to develop a psychometric instrument from Jungian theory (Meier and Wozny 1978; Rosenak and Shontz 1988).
(8.) Form G scoring is based on Classical Test Theory (OTT). This approach contends that a single measure can describe the relationship between the observed responses and the hypothetical (but unobservable) true response. According to CTT, a test measures the individual with equal precision across the entire range of possible score values on the test's scales. Form M is based on Item Response Theory (IRT). IRT contends that a test is not necessarily equally reliable at different ranges of scores on the test's scales. Thus, IRT models the relationship between observed responses and latent true responses in a more complex manner than that of the single-number OTT approach. Refer to Harvey (1996) and Myers et al. (1998).
(9.) The abbreviations stand for the Adjective Check List (ACL), the California Personality Inventory (CPI), the Sixteen Personality Factor (16PF), and the Neuroticism Extraversion Openness Personality Inventory (NEO-PI). For more information on these instruments refer to Table 3, footnotes b-f.
(10.) For example, Carskadon's (1982) seminal best-fit study had students first take the MBTI to obtain their personality types. The subjects were then given five MBTI personality descriptions: (1) the individual's MBTI type, (2) the opposite type (all four preferences reversed), (3) the type with the lowest scored preference reversed, (4) the type with the two mental functions reversed (SN; TF), and (5) the type with the two attitudes reversed (EI; JP). Participants then ranked the provided profiles as to how closely they matched their self-perceptions. These rankings were then correlated to the MBTI scores.
(11.) For many of these research suggestions, there are examples of similar MBTI research from nonaccounting areas. Refer to the literature reviews by Gardner and Martinko (1996), Hammer (1996), and Myers et al. (1998).
Accounting Education Change Commission (AECC). 1990. Objectives of education for accountants: Position Statement Number One. Issues in Accounting Education (Fall): 307-312.
American Accounting Association (AAA), Committee on the Future Structure, Content, and Scope of Accounting Education (The Bedford Committee). 1986. Future accounting education: Preparing for the expanding profession. Issues in Accounting Education (Spring): 168-195.
American Institute of Certified Public Accountants (AICPA). 1997. AICPA Vision Project--2011 and Beyond. Available at: http://www.cpavision.org.
-----. 1999. The AICPA Core Competency Framework for Entry into the Accounting Profession. Available at: http://www.aicpa.org/edu/corecomp.htm.
Brooks, F. R., and R. W. Johnson. 1979. Self-descriptive adjectives associated with a Jungian personality inventory. Psychological Reports 44: 747-750.
Carskadon, T. G. 1982. The Myers-Briggs Type Indicator characterizations: A Jungian horoscope? Research in Psychological Type 8: 87-88.
-----. and D. D. Cook. 1982. Validity of MBTI type descriptions as perceived by recipients unfamiliar with type. Research in Psychological Type 8: 89-94.
Cattell, Raymond B. 1980. Handbook for the Sixteen Personality Factor Questionnaire. Revised edition. Champaign, IL: Institute of Personality and Ability Testing.
Chenhall, R., and D. Morris. 1991. The effect of cognitive style and sponsorship bias on the treatment of opportunity costs in resource allocation decisions. Accounting, Organizations and Society 16: 27-46.
Descouzis, D. 1989. Psychological types of tax preparers. Journal of Psychological Type 17: 36-38.
Engler, B. 1999. Personality Theories: An Introduction. 5th edition. Boston, MA: Houghton Muffin College Press.
Flanagan, O. 1991. The Science of the Mind. 2nd edition. Cambridge, MA: MIT Press.
Gardner, W. L., and M. J. Martinko. 1996. Using the Myers-Briggs Type Indicator to study managers: A literature review and research agenda. Journal of Management 22: 45-83.
Geyer, P. 1998. Science and culture in action: An historical examination of the acceptance of the work of C. G. Jung and Isabel Myers. Presented at the Third Multicultural Research Conference: Psychological Type and Culture East and West, Honolulu, HI. Available at http://www.ozemail.com.au/-alchymia/library/science.html.
Gough, H. G. 1981. Studies of the Myers-Briggs Type Indicator in a personality assessment research institute. Paper presented at the Fourth National Conference on the Myers-Briggs Type Indicator, Palo Alto, CA.
-----, and A. B. Heilburn. 1983. The Adjective Check List Manual. Palo Alto, CA: Consulting Psychologists Press.
-----, and P. Bradley. 1996. CPI Manual. 3rd edition. Palo Alto, CA: Consulting Psychologists Press.
Grant, W. H. 1965. Comparability of the Gray-Wheelwright Psychological Type Questionnaire and the Myers-Briggs Type Indicator. Research report, Student Counseling Services, Auburn University.
Hammer, A. L., ed. 1996. MBTI Applications: A Decade of Research on the Myers-Briggs Type Indicator. Palo Alto, CA: Consulting Psychologists Press.
-----, and G. E. Huszczo. 1996. Teams. In MBTI Applications: A Decade of Research on the Myers-Briggs Type Indicator, edited by A. L. Hammer. Palo Alto, CA: Consulting Psychologists Press.
Harvey, R. J., W. D. Murry, and D. Stamoulis. 1995. Unresolved issues in the dimensionality of the Myers-Briggs Type Indicator. Educational and Psychological Measurement 55: 535-544.
-----. 1996. Reliability and validity. In MBTI Applications: A Decade of Research on the Myers-Briggs Type Indicator, edited by A. L. Hammer. Palo Alto, CA: Consulting Psychologists Press.
Hergenhahn, B. R., and M. H. Olson. 1999. An Introduction to Theories of Personality. 5th edition. Upper Saddle River, NJ: Prentice Hall, Inc.
Hoy, F., and D. Hellriegel. 1982. The Kilman and Herden model of organizational effectiveness criteria for small business managers. Academy of Management Journal 25: 308-322.
Jacoby, P. F. 1981. Psychological types and career success in the accounting profession. Research in Psychological Type 4: 24-37.
Johnson, D. A. 1992. Test-retest reliabilities of the Myers-Briggs Type Indicator and the Type Differentiation Indicator over a 30-month period. Journal of Psychological Type 24: 54-58.
Jung, C. G. 1921. Psychological Types. Princeton, NJ: Princeton University Press.
Keirsey, D. 1998. Please Understand Me II: Temperament, Character, Intelligence. Del Mar, CA: Prometheus Nemesis Book Company.
Kreiser, L., J. M. McKeon, and A. Post. 1990. A personality profile of CPAs in public practice. The Ohio CPA Journal (Winter): 29-34.
Kummerow, J. M. 1988. A methodology for verifying type: Research results. Journal of Psychological Type 15: 20-25.
Landry, R. M., Jr., R. L. Rogers, and H. W. Harrell. 1996. Computer usage and psychological type characteristics in accounting students. Journal of Accounting and Computers 12: Article 4. Available at http://www.swcollege.com/acct/jac/jac12.html.
Laribee, S. 1994. The psychological types of college accounting students. Journal of Psychological Type 28: 37-42.
McCrae, R. R., and P. T. Costa. 1989. Reinterpreting the Myers-Briggs Type Indicator from the perspective of the five-factor model of personality. Journal of Personality 57: 17-37.
-----, and O. P. John. 1992. An introduction to the five-factor model and its applications. Journal of Personality 60: 175-215.
Meier, C. A., and N. A. Wozny. 1978. An empirical study of Jungian typology. Journal of Analytical Psychology 23: 3-15.
Myers, I. B., and M. H. McCaulley. 1985. Manual: A Guide to the Development and Use of the Myers-Briggs Type Indicator. Palo Alto, CA: Consulting Psychologists Press.
-----, and P. B. Myers. 1995. Gifts Differing: Understanding Personality Type. Palo Alto, CA: Davies-Black Publishing.
-----, M. H. McCaulley, N. L. Quenk, and A. L. Hammer. 1998. MBTI Manual: A Guide to the Development and Use of the Myers-Briggs Type Indicator. 3rd edition. Palo Alto, CA: Consulting Psychologists Press.
Nourayi, M. M., and A. C. Cherry. 1993. Accounting students' performance and personality types. Journal of Education for Business (November/December): 111-115.
Nunnally, J. M., and I. H. Bernstein. 1994. Psychometric Theory. 3rd edition. New York, NY: McGraw-Hill Book Company.
Oswick, C., and P. Barber. 1998. Personality type and performance in an introductory level accounting course: A research note. Accounting Education 7: 249-254.
Ott, R. L., M. H. Mann, and C. T. Moores. 1990. An empirical investigation into the interactive effects of student personality traits and method of instruction (lecture or CAI) on student performance in elementary accounting. Journal of Accounting Education 8: 17-35.
Otte, P. 1984. Do CPAs have a unique personality: Are certain personality types found more frequently in our profession? The Michigan CPA (Spring): 29-36.
Pascal, E. 1992. Jung to Live By. New York, NY: Warner Books, Inc.
Pearman, R. R., and J. Fleenor. 1996. Differences in observed and self-reported qualities of psychological types. Journal of Psychological Type 39: 3-17.
Rich, B. 1972. A correlational study of the Myers-Briggs Type Indicator and the Jungian Type Survey. Research paper, University of Missouri-Columbia.
Rosenak, C. M., and F. C. Shontz. 1988. Jungian Q-Sorts: Demonstrating construct validity for psychological type and the MBTI. Journal of Psychological Type 15: 33-45.
Ruble, T. L., and R. A. Cosier. 1990. Effects of cognitive style and decision setting on performance. Organizational Behavior and Human Decision Processes 46: 283-295.
Russell, M. T., and D. L. Karol. 1994. l6PF Fifth Edition Administrator's Manual. Champaign, IL: Institute for Ability and Personality Testing.
Satava, D. 1996. Personality types of CPAs: National vs. local firms. Journal of Psychological Type 36: 36-41.
Scarbrough, D. P. 1993. Personality types and job satisfaction of accountants. Journal of Psychological Type 25: 3-10.
Schloemer, P. G., and M. S. Schloemer. 1997. The personality types and preferences of CPA firm professionals: An analysis of changes in the profession. Accounting Horizons (December): 24-39.
Schweiger, D. M. 1985. Measuring managerial cognitive styles: On the logical validity of the Myers-Briggs Type Indicator. Journal of Business Report 13: 315-328.
Shackleton, V. 1980. The accountant stereotype: Myth or reality? Accountancy (November): 122-123.
Sipps, G. J., R. A. Alexander, and L. Friedt. 1985. Item analysis of the Myers-Briggs Type Indicator. Educational and Psychological Measurement 45: 789-796.
-----, and -----. 1987. The multifactorial nature of extraversion-introversion in the Myers-Briggs Type Indicator and Eysenck personality inventory. Educational and Psychological Measurement 47: 543-552.
Thorne, A., and H. Gough. 1991. Portraits of Type: An MBTI Research Compendium. Palo Alto, CA: Consulting Psychologists Press.
Tischler, L. 1994. The MBTI factor structure. Journal of Psychological Type 31: 24-31.
Tzeng, O., D. Landis, and J-M. Chen. 1989. Measurement and utility of continuous unipolar ratings for the Myers-Briggs Type Indicator. Journal of Personality Assessment 53: 727-738.
Vassen, E., C. Baker, and R. Hayes. 1993. Cognitive styles of experienced auditors in the Netherlands. British Accounting Review 25: 367-382.
Ware, R., and C. Yokomoto. 1985. Perceived accuracy of Myers-Briggs Type Indicator descriptions using Keirsey profiles. Journal of Psychological Type 10: 27-31.
Wheelwright, J. B., J. H. Wheelwright, and H. A. Buehler. 1964. Jungian Type Survey: The Gray-Wheelwright Test. 16th edition. San Francisco, CA: Society of Jungian Analysts of Northern California.
Williams, J. R. 1999. President's message. Accounting Education News (Fall): 1-2.
Wolk, C., and L. A. Nikolai. 1997. Personality types of accounting students and faculty: Comparisons and implications. Journal of Accounting Education 15: 1-17.
The 16 Personality Types with General Characteristics and Occupational Tendencies Sensing (S) Thinking (T) Feeling (F) Judging (J) ISTJ ISFJ Practical, sensible, Practical, concrete, decisive, logical, cooperative, sensitive. detached. Education, health Management and care, and religion. administration. Perceiving (P) ISTP ISFP Detached, logical Trusting, kind, problem solvers, sensitive, observant pragmatic, factual. practical, concrete. Skilled trades and Health care and technical fields. business. Perceiving (P) ESTP ESFP Observant, active, Observant, specific rational problem active, sympathetic, solvers, assertive, idealistic, warm. Marketing, business, Health care and skilled trades. and teaching. Judging (J) ESTJ ESFJ Logical, decisive, Factual, personable, objectively critical, cooperative, practical practical, systematic. decisive. Management and Education, health administration. care, and religion. Intuition (N) Feeling (F) Thinking (T) Judging (J) INFJ INTJ Insightful, symbolic, Insightful, long-range idealistic, committed, thinkers, clear, compassionate. rational, detached. Religion, counseling, Science, computers, and teaching. and technical fields. Perceiving (P) INFP INTP Sensitive, caring, Logical, curious, idealistic, curious, detached, insightful, creative, visionary. contemplative. Counseling, writing, Scientific and and arts. technical fields. Perceiving (P) ENFP ENTP Curious, creative, Creative, imaginative, energetic, friendly, theoretical, analytical, cooperative, warm. rational, questioning. Counseling, religion, Science, management, and teaching. and technology. Judging (J) ENFJ ENTJ Compassionate, loyal, Analytical, assertive, imaginative, likes conceptual thinkers, variety, supportive. innovation planners. Arts, religion, Management and and teaching. leadership. Judging (J) Introversion (I) Perceiving (P) Perceiving (P) Extraversion (E) Judging (J) General personal characteristics are shown in regular font style; occupational tendencies are shown in italics. Source: Myers et al. (1998), with modifications. Reliability Issues Related to the MBTI Type of Reliability Type of Test Reference Myers et al. (1998) Myers and Split-half McCaulley (1985) procedure Hoy and Internal Hellriegel (1982) Consistency Myers et al. (1998) Harvey (1996) Alpha Ruble and coefficient Cosier (1990) Tzeng et al. (1989) Schweiger (1985) Myers et al. (1998) Johnson (1992) Temporal Test-retest Stability procedure Myers and McCaulley (1985) Type of Reliability Key Findings Logical and consecutive split-half correlations exceed 0.90 for Form M for all four scales. Split-half correlations exceed 0.80 for Form G for all four scales. Split-half correlations of 0.67 for Form G Internal for SN anf TF scales only. Consistency Test of a national sample (n = 2,859) with coefficient aplha scores of 0.91 (EI), 0.92 (SN), 0.91 (TF), anf 0.92 (JP). Meta-analysis of three Form G studies (n = 3,991) showed coeficient alpha scores of 0.86 (EI), 0.87 (SN), 0.85 (TF), and 0.87 (JP). Coefficient alphas for Form G: 0:85 for SN scale and 0.79 for TF scale. Coefficient alphas 0.89 or higher for Form G for all four scales. Coeffiicient alphas for Form G: 0.74 for SN scale and 0.69 for TF scale. Correlations for Form M (n = 424) over a four-week period: 0.88 (TF), 0.94 (EI), 0.93 (JP), and 0.92 (SN); 65% with same personality type. Correlations for Form J (n = 74), over a 30- Temporal month period: 0.62 (TF), 0.79 (EI), 0.82 Stability (JP), and 0.83 (SN). Correlations for Form G (n = 1, 139), interval [greater than] nine months: 0.59 (TF), 0.70 (EI), 0.63 (JP), and 0.68 (SN); 35.8% with same personality type. For n = 559, inerval [less than] nine months: 0.77 (TF), 0.84 (EI), 0.82 (JP), and 0.81 (SN); 50.5% with same personality type. Validity Issues Related to the MBTI Panel A: Tests of Independent Dichotomies Type of Validity [a] Type of Test Reference Harvey et al. (1995) Tischler (1994) Discriminant Factor analysis Tzeng et al. (1989) Myers et al. (1998) Sipps and Alexander (1987) Sipps et al. (1985) Gough (1981) Self-assessment Convergent and rater- assessment Brooks and Johnson (1979) Rosenak and Comparison to Shontz (1988) other Jungian personality Rich (1972) scales Grant (1965) Type of Validity [a] Key Findings Confirmatory factor analysis of Form G provides support for the theoretical four-scale model vs. two competing five-scale models. Strong supportive analysis of Form G found four distinct scales highly correlated to theoretical constructs. Discriminant Exploratory factor analysis of Form G found four distinct factors correlated to the four MBTI constructs accounting for 56.4% of the variation. Confirmatory factor analysis (Form M) using a national sample (n = 3,036) produced a four-factor model with 0.949 adjusted goodness-of-fit. Partial supportive analysis found two factorially distinct scales in Form F similar to the EI and JP scales, but with some significant differences from the theory constructs. Nonsupportive factor analysis of Form F yielded six distinct factors. Four of the six resembled the four Jungian scales. Staff assessments of 244 males and 186 females at the Institute for Personality Assessment and Research using the ACL [b] Convergent and Q-Sorts correlated positively with MBTI preferences (p [less than] 0.05). Self-assessments of 209 undergraduates and graduate students using the ACL correlated positively with MBTI preferences (p [less than] 0.05). Positive correlations were found between Jungian Q-sorts [c] and MBTI types (p [less than] 0.001) where n = 120. Statistically significant correlations (p [less than] 0.01) between the JTS and the MBTI along the six preferences shared between the two instruments. Statistically significant correlations (p [less than] 0.01) between the six common preferences of the JTS and the MBTI. However, only 21% of subjects showed the same overall personality type on both instruments. Comparison to non- Jungian personality scales Panel B: Tests of Inter- actions among Dichotomies Construct Self-assessment and rater- assessment Convergent Comparison Myer et al. (1998) Gough and Bradley (1996) Russell and Karol (1994) McCrae and Costa (1989) Panel B: Tests of Inter- actions among Dichotomies Construct Carskadon (1982) Carskadon and Cook (1982) Kummerow (1988) Myers et al. (1998) Convergent Ware and Yokomoto (1985) Thorne and Gough (1991) Pearman and Fleenor (1996) Where n = 247, 32 of the 33 personality traits of the ACL correlate with correspon- ding MBTI dichotomies, but 28 of the 32 correlations occurred with more than one MBTI dimension. Where n = 693, 23 of the 27 CPI [d] person- ality traits correlate with the MBTI dichoto- mies; 14 of the 27 traits correlate with more than one of the MBTI dichotomies. Where n = 119, 31 statistically significant correlations of 16PF [e] with MBTI Form G (p [less than] 0.05). Fourteen of the 16 per- sonality factors correlate with the four MBTI dichotomies with strong indications of a positive/negative dichotomy. Ten of the 16 correlate with more than one MBTI dimension. Compares NEO-PI [f] to MBTI Form G. All five NEO-PI traits correlate with the four MBTI dichotomies (p [less than] 0.05). Three of the five correlate with more than one dimension. Panel B: Tests of Inter- actions among Dichotomies Construct Using Forms F/G on 129 undergraduates, the MBTI type was selected as the best fit 35% of the time (p[less than]0.001). The type with the four preferences reversed was selected as the best fit 4% of the time. Using Forms F/G on 118 MBTI-naive undergra- duates the MBTI type was selected as the best fit by participants 50% of the time. The type with the four preferences reversed was selected as the best fit 13% of the time. Using Form G and 387 working professionals found 62.2% best-fit agreement with all four preferences and 93.4% agreement with three or more preferences (p [less than] 0.001). Two studies using Form M and nonstudent adults. First study (n = 158) had 78% best- fit agreement between the MBTI type and the selected description. Second study (n = 92) had 58% best-fit agreement (compared to 53% best-fit agreement using Form G). Convergent Significant correlations between Keirsey temperament theory and MBTI types used Form G on 343 undergraduate students performing self-assessments (p [less than] 0.001). 189 raters classified 614 subjects using 20 adjectives from the ACL. Unique adjectives for MBTI (Form F) types ranged from 12 to 20, with an average of 15.35. Interjudge reliability was 0.76. Using 2,398 managers, comparisons of the 16 MBTI types were made using CPI and LSI. [g] ANOVAs by MBTI type were statistically significant for both instruments (p [less than] 0.05).
(a.)Validity studies are classified into three categories. The first categorization is whether a test examines the four preference scales as independent dichotomies or the interaction of preference scales (Panels A and B). The second categorization is the type of validity tested, viz., i, convergent, or construct validity (far left column). The third categorization is the type of test used (second column from left).
(b.)The Adjective Check List (ACL) is an adjective format instrument, not a questionnaire like the MBTI or NEO-PI (Neuroticism Extraversion Openness Personality Inventory). The ACL consists of 300 adjectives subjects are to select as descriptive or nondescriptive. These selections collapse into 33 personality characteristics. As suggested by the large number of characteristics, the ACL is more descriptive and less theory-based than the MBTI. Psychological inventories may be divided into questionnaires, which ask subjects to respond to questions and situations, and checklists (or lexical instruments), which ask subjects to evaluate themselves in relation to single words or short phrases (Gough and Heilburn 1983; McCrae and John 1992). Some instruments include both types of items; e.g., the NEO-PI. The MBTI, Form M, includes some checklist items but is predominantly a questionnaire.
(c.)The Jungian Q-Sort instrument is an extension of the generic Q-Sort testing method. Subjects are given a stack of cards of varying content and asked to sort them along certain criteria. For more on Jungian Q-Sorts, refer to Rosenak and Shontz (1988).
(d.)The California Personality Inventory (CPI) is a checklist-type instrument with 434 items. It results in a detailed personal portrait of an individual by describing personality characteristics across 36 different scales (3 structural scales, 20 folk scales, and 13 special-purpose scales). Note the large number of characteristics compared to the MBTI. For more on the CPI, refer to Gough and Bradley (1996).
(e.)The Sixteen Personality Factor (16PF) questionnaire consists of 185 questions and requires approximately one hour to complete (Cattell 1980). It measures 16 factors that are considered to be universal traits. The 16PF is widely used and has been extensively tested for reliability and validity. Like the ACL, the relatively large number of characteristics (16) makes the 16PF more descriptive and less theory-dependent than the MBTI.
(f.)The NEO-PI (Neuroticism Extraversion Openness Personality Inventory) is a questionnaire format instrument (as opposed to a checklist-type instrument, like the ACL) developed in accordance with the Five Factors Model, also known as "the Big Five" (McCrae and John 1992). The five factors are neuroticism, extraversion, openness, agreeableness, and conscientiousness. There are significant differences in the theories underlying the NEO-PI and the MBTI. For example, unlike the MBTI, the NEO-PI scales are viewed as continuous, and not bipolar or dichotomous. Also, unlike the MBTI, these scales imply a value judgment on the subject's personality (Myers et al. 1998). For example, a low score on the extraversion scale implies a deficit or lack of this characteristic.
(g.)The Learning Style Inventory, LSI, contains 48 leadership adjectives yielding 24 scales. For more on the LSI, refer to Pearman and Fleenar (1996).
Accounting Research Using the MBTI Panel A: Personality Types of Professional Accountants Reference Journal Shackleton (1980) Accountancy Jacoby (1981) Reserach in Psychological Type Otte (1984) The Michigan CPA Descouzis (1989) Journal of Psychological Type Kreiser et al. (1990) The Ohio CPA Journal Satava (1996) Journal of Psychological Type Schloemer and Schloemer (1997) Accounting Horizons Panel B: Performance and Personality Types of Professional Accountants Chenhall and Morris (1991) Accounting, Organizations and Society Vassen et al. (1993) British Accounting Review Panel A: Personality Types of Professional Accountants Reference Sample Shackleton (1980) 91 accountants and financial managers Jacoby (1981) 333 accountants in three Big 8 firms in the Washington, D.C. area Otte (1984) 494 CPAs in small, local Michigan firms Descouzis (1989) 36 H&R Block tax preparers, average seven years' experience Kreiser et al. (1990) 96 CPAs not in public practice; 86 CPAs in public practice; 92 bank officers; 113 FEI members Satava (1996) 439 CPAs in local and national firms throughout the U.S. Schloemer and Schloemer (1997) 125 accounts from Big 6 and large local firms Panel B: Perference and Personality Types of Professional Accountants Chenhall and Morris (1991) 64 middle- to senior-level managers participating in an executive program Vassen et al. (1993) 25 experienced (5-39 years auditing experience) auditors from the three largest public accounting firms in The Netherlands Panel A: Personality Types of Professional Accountants Reference Findings Shackleton (1980) Subjects were primarily STJ- preferenced (38.5%), with ISTJs at 25.3% and ESTJs at 13.2%. I- preferences were at 58%. Jacoby (1981) STJ dominant (33.6%), with ISTJ at 19.8% and ESTJ at 13.8%. Audit and tax partners and managers were more likely to be TJ than those at lower professional levels in these areas. Audit partners and managers were more likely to be IS then partners and managers in tax and consulting. Otte (1984) CPAs in small local firms were primarily STJ (45.9%). Stronger ST- preferences at local level than at national level. Descouzis (1989) All tax preparers were S-perfer- enced; 58% were I-preferenced; 81 percent J-preferenced; 44.4% STJ- preferencd. Kreiser et al. (1990) CPAs perceived by other business professionals as primarily I-pre- ferenced and STJ-preferenced. These perceptions were stronger than the actual occurrences of the preferences among CPAs. Satava (1996) Comparison of local and national firms indicates no differences on SN, TF, and JP scales. Predominance of STJ in both types of firms. E- preferences tend to prefer national firms, whereas I-preferences prefer local firms. Schloemer and Schloemer (1997) CPAs primarily STJ, although some increase in E was observed over the past decade. Panel B: Perference and Personality Types of Professional Accountants Chenhall and Morris (1991) Found that managers with an N- preference included opportunity cost in resource allocation decisions, while those with an S- preference tended to ignore opportunity cost. Vassen et al. (1993) Subjects had significantly higher S-perferences (68%) over N-prefer- ences. T-Preferences were higher (56%) but not significantly. Those with the T-preference accessed more information, took longer to process the information accessed, and had a lower tolerance for ambiguity. Scarbrough (1993) Journal of Psychological Type Panel C: Personality Types of Accounting Students Laribee (1994) Journal of Psychological Type Wolk and Journal of Nikolai (1997) Accounting Education Panel D: performance and Personality Types of Accounting Students Ott et al. (1990) Journal of Accounting Education Nourayi and Journal of Education Cherry (1993) for Business Landry et al. Journal of Accounting (1996) and Computers Oswick and Accounting Education Barber (1998) Scarbrough (1993) 255 mid-level Big 8 accountants Panel C: Personality Types of Accounting Students Laribee (1994) 320 undergraduate accounting students (sophomores to seniors) from a midwestern AACSB-accredited school of business Wolk and 152 undergraduate Nikolai (1997) students; 94 graduate students; and 98 faculty members Panel D: performance and Personality Types of Accounting Students Ott et al. (1990) 89 undergraduate students enrolled in first elementary accounting course at Texas Tech University Nourayi and 103 undergraduate accounting Cherry (1993) majors who completed first intermediate accounting course Landry et al. 88 junior- and senior-level (1996) accounting students Oswick and 344 U.K. undergraduate Barber (1998) students who had completed introductory- level accounting course Scarbrough (1993) The J-preference had marginally more job satisfaction. Gender was an important factor in determining the relationship between preferences and job satisfaction. Females were more satisfied than males on I, T, and J scales, and as ISTJs. Panel C: Personality Types of Accounting Students Laribee (1994) Undergraduate accounting students were primarily STJs and were different from other college students. A filtering- out process occurred in which E-, N-, F-, and P-preferences declined, while I-, S-, T-, and J-preferences increased. Wolk and Undergraduate accounting Nikolai (1997) students are different from graduate accounting students and accounting faculty. Graduate students and faculty members were not significantly different. Panel D: performance and Personality Types of Accounting Students Ott et al. (1990) S-preference and T-preference students performed better with the lecture method. N-preference and F-preference students performed better with the computer-assisted method. Nourayi and The only significant results Cherry (1993) were found on the SN scale. S- preference students did better than N-preference students in three of the seven accounting courses and in an overall accounting average score. Landry et al. 42% of students were STJs. N- (1996) and T-preference students scored significantly lower on the computer anxiety instrument. Oswick and No statistically significant Barber (1998) relationships between the MBTI preferences and performance were found.
|Printer friendly Cite/link Email Feedback|
|Publication:||Issues in Accounting Education|
|Date:||Feb 1, 2001|
|Previous Article:||Pecos Products: A Project Introducing Complexity into the Study of Activity-Based Costing.|
|Next Article:||CCH Federal Taxation: Basic Principles.|