Multiple Intelligences and Personality Type in the College Classroom

by Dario Nardi


An extensive web-based and booklet-based study of 35 students was conducted to determine which of nine theoretical models might be useful in assessing cognitive development and cognitive skill set in liberal arts and sciences college students. Nine models were explored, from “sensory modalities” and “temperament theory” to “multiple intelligences” and “cognitive processes.” Seven of the models proved useful, one did not turn out as expected, and one did not perform well. Three models which, in theory, interrelate, stood-up under scrutiny. The distribution of students within the models also appeared quite close to the general population, and several new theoretical interrelationships between some of the models are offered for future research. In particular, the “story-based and thematic” metacognitive strategy — often not addressed in many educational studies — proved particularly relevant and promising leads for future research into the nature of “second person point of view” were identified.




This study was designed to explore which models of intelligence (cognitive ability, skills, talents, learning styles, and a number of other strategies) are useful, simple, and consistent for use in the college classroom. The emphasis here is cognitive modes and cognitive skills, as opposed to values, attitudes or experiences that traditional college-age students might reference vis-a-vis learning. One question: which models might be useful for further study, such as developing an instrument or more sophisticated web-based assessment for self-discovery and self-evaluation process. And a second question: how do the models links together and what might they really measure? Since this study was an exploration, and not statistically intensive (only 35 students), the emphasis was on

  1. Consistency between cognitive models
  2. Ease of use of models as part of a self-assessment process
  3. Minimization of bias within models

The hope was that with a simple web based self-assessment process, students can learn more about themselves while at the same time providing the instructor with a profile of the class as a whole, so that instruction can be better tailored to students’ strengths and weaknesses.

All of the models, and the self-assessment exercises used with those models, are explained in various referenced sources, primarily “Multiple Intelligences and Personality Type” by Nardi [1], and “Dynamics of Personality Type” by Berens [2].



Experimental Setup


The sample consisted of 35 students, 17 males and 18 females, from ages 18 to 22 (most were traditional college-age juniors and seniors.) The class was an upper-division course in computer programming for the general liberal arts and sciences student (that is, not for engineering or computer science majors, but drawing from the arts and sciences body of the university.) The most represented major was economics. Cognitive science, mathematics, and biology were also highly represented. However, many majors were present in smaller numbers.

Nine models were brought to students attention and a homework assigned using the “Dynamics of Personality Type” booklet and handouts of material now contained in “Multiple Intelligences and Personality Type”.

  1. Points of view
  2. Timeline orientation (a theory of time use developed by Tad James)
  3. Sensory modalities (expanded version)
  4. Metacognitive strategies
  5. Gardner’s theory of multiple intelligences
  6. Temperament theory
  7. Berens’ learning style portraits
  8. Jung’s model of cognitive processes
  9. MBTI snapshot portraits (brief text about each of the 16 types was used over the MBTI instrument itself, since use of the instrument within the class was not feasible.)

Students were instructed to work through the Berens’ booklet and the material now in the Nardi booklet, and then record their responses to the exercises at an instructor-provided web site. Once responses were recorded, the student clicked “submit” and data was transmitted to a server folder for later analysis using Microsoft Excel. Thus, the entire process was self-selection and self-evaluation guided by booklets designed for such a task and by a web site designed to gather the data.



Review of the Literature


At a theory level, the first five of these models stand alone, while the later four models inter-relate. Models one through six can be found in Nardi’s “Multiple Intelligences and Personality Type,” while the remainder can be found in the Berens’ booklet. Temperament is also more thoroughly explained in “An Introduction to Temperament” by Berens, and other references detail the other eight models in various levels of detail.

How are these models currently addressed in the educational literature?

There is study of sensory modalities and MBTI in the classroom. To a far less extent, Gardner’s intelligences and temperament have also been explored. There is one previous study involving Berens’ learning styles portraits. No literature on the “points of view”, timeline orientation, or Jung’s cognitive processes could be found.

In theory, sensory modalities, internal strategies and Gardner’s intelligenes are all separate unique models. In contrast, temperament, the learning style portraits written by Berens, Jung’s cognitive process and the MBTI are all inter-related at a theoretical level in several complex ways, having all sprung originally from Jung’s original theory of mental “functions.”

Around 1920, Jung posited the existence of eight basic cognitive (mental) processes, or “functions.” Since then, his work has come into more use and been explored and tested. However, the cognitive processes model is not used widely in the United States and not used in educational studies. However, with advancements in the discipline of cognitive science and the dawning suggestion that there are in fact cognitive processes, it was considered relevant to include this model with the others. Jung’s model also underlies the MBTI, which is the most widely used psychological instrument in the United States.

Past work with hundreds of thousands of college students since the 1950s using the MBTI very strongly and significantly suggest that the 16 personality types are not evenly distributed throughout the population. Similarly, the 4 Keirseyian temperaments are not evenly distributed, and thus the development and preference of Jung’s cognitive processes and the Berens learning styles would not be evenly distributed either. The expected distribution per the “MBTI Manual” Third Edition [3] was relevant in doing statistical analysis.


Results & Analysis


The results and analysis for each of the nine models are presented one at a time, each in its own section. Statistical results and correlations with other models are indicated when appropriate.





An individual may have mental access (generally, in the mind’s eye) to one or more of three points of view, for the past, the present, and the future. Different points of view may be used or accessible to an individual for different time periods (past, now, future) and in sensory modalities other than visual. The literature [4] states that 1st and 3rd persons are the most common. Students were allowed to select as many points of view as applied to how they remembered the past, experienced the present and foresaw the future. The literature states that ideally, all individuals should have access to all points of view. In fact, however, most people have access to two.

Access to first person Access to second person Access to third person Access to all three points of view
34/35 18*/35 28/35 14/35


Table 1: Number of students with access to each of three points of view.


The results are summarized in table 1. On average, students had access to 2.3 of 3 points of view. Most students (34, and 28) had access to first and third person. In contrast, only 18 had access to second person. The limited access to second person is statistically significant (p<.05) if we expect the sample students to have the same access to 2nd person as they might to 1st and 3rd persons.

Students also recorded their sensory experience (visual, auditory, kinesthetic) for each point of view. For example, did the student see from second person, or hear, or experience a feeling or any combination thereof? Whether looked at as a whole or by points of view, the visual modality was statistically more likely (p<.05) to be the available to students than auditory or kinesthetic. Visual mode was available 35 percent to 88 percent more often. This result matches the literature, which describes human beings as being more visual in general.





In theory, an individual has a single default time orientation. It is possible to shift time orientation. Students were instructed to report their “default.” Table 2 summarizes the results.

Through Time In Time Bent Circular Other
13/35 14/35 4/35 3/35 1/35


Table 2: Distribution of students by timeline orientation.


Tad James, in “Time Line Therapy and the Basis of Personality” [5] states that Through-Time and In-Time are the most common time orientations. The sample clearly matches this statement; 27 of 35 students reported one of the two major orientations as their default — this is statistically significant (p<.05).

One of the applications of becoming aware of one’s timeline is making changes to one’s timeline. In this spirit, the instruction to the students included an exercise to modify their timeline [1]. This modification ability is an example of a metacognitive skill. Traditional models of cognitive development suggest college students should be in the process of developing and refining metacognitive skills. Students were instructed to give one of four responses to the exercise, as summarized in Table 3.

Could alter timeline and wanted to Could alter timeline but didn’t want to Couldn’t alter timeline but wanted to Couldn’t alter timeline and didn’t want to
4/35 19*/35 5/35 7/35


Table 3: Distribution of students by ability and desire to alter timeline.


Two thirds of students were able to alter their timeline, but three-quarters of all students did not want to alter their timeline whether they could or not. This last statistic is significant (p < .05) if we expect a uniform distribution of responses.

Of the 8 students with circular, bent, or “other” timeline orientation, all 8 could alter their timeline, compared to only 56 percent of the other 27 students. If we expected no difference between these groups, then this difference is statistically significant (p < .05). These unusual timelines are considered variations on the basic In-Time and Through-Time configurations, so it might not be surprising that these individuals have more flexibility. Unfortunately, this sample is small.





The term “sensory modalities” generally refers to a preference for auditory, visual or kinesthetic learning modes. In recent years, this model has been questioned as overly simplistic. To help get more out of these three simple modes, twelve items (four for each modality) were used instead. This author developed these twelve but they are not original; rather, they bring together in one place various submodalities mentioned elsewhere [4][5]. Thus, these 4 submodalities were generated based on a review of the literature, and there is an internal logic to the items. Each sensory modality may manifest as:

  • Remembering in the mind (e.g. seeing what was seen earlier in the mind’s eye)
  • With others externally in the here and now (e.g. A desire for thinking outloud — talking – to people to make decisions)
  • Being motivated by stimulus of that modality (e.g. responding when someone touches you or you receive other tactile stimulation)
  • Constructing in the mind using the imagination (e.g. conceptualizing a visual image of the future.)

Submodalities I and IV are internal or mental activities, while submodalities II and III are external to one’s self. Furthermore, submodalities II and IV involve more active constructing, while submodalities I and III are more passive or involve habitual or recalled ways of doing things.

Students were asked to respond to the following question: “Do you often find yourself making decisions by?” The brief snippets of text below are NOT those used with the students. The actual phrasing can be found in [1]. The short phrases here are meant merely as a reference. Students were instructed to indicate “yes” or “no” to each of the 12 items, individually. The results are summarized in Table 4.

Modality Submodality Out of 35 Students
Visual I




In the mind’s eye

Observing others

How seen by others around you

Visualizing the future






Auditory I




Mental tape-recordings

Talking in the here and now

Sounds around you

Self talk





Kinesthetic I




An internal feeling

Taking an action

Tactile stimuli around you

What you “read” some others






Table 4: Number of students who identified with each of twelve submodalities.


The literature in sensory modalities suggests that preferences in the population descend from visual (most preferred) to auditory to kinesthetic (least preferred.) The results here are mostly uniform within each sensory modality, with two exceptions. The fact that no student selected visual submodality III suggests that there may have been a problem with the item. It is also possible that the leading question is misleading — people might not want to believe or admit that they make decisions based on how they appear to others around them. In any case, had this item appealed to students at a rate typical of the average of the other items, then the visual modality would have outranked the others, followed by auditory, and then kinesthetic last. The low appeal of kinesthetic submodality II also begs further study.


If the sensory modality items are typical “internal strategies”, then there are “meta” strategies (strategies about strategies) as well. A review of the literature suggests four very common metacognitive strategies. The wording found on page 11 of [2] was used. Students were instructed to select as many metacognitive strategies as they believed they “excelled” at. The literature suggests that one strategy will be a default or best-developed strategy, and a second and possibly a third will be developed to some extent as well. Fortunately, most students appeared to have a realistic self-appraisal (Table 5), and while the liberal set of instructions to the students made later statistical analysis quite difficult, it did provide this illuminating distribution.

Selected one Selected two Selected three Selected all four
15/35 12/35 6/35 2/35


Table 5: Distribution of students by how many metacognitive strategies they identified with.


Since students selected an average of 1.9 metacognitive strategies, the distribution of strategies in Table 6 adds up to more than 35. Because of rounding, the total appears to add to slightly more than 100%.

Hands on and experiential Step-by-step and memorization Conceptual and theoretical Story based and thematic
21/35 20/35 10/35 14/35


Table 6: Distribution of metacognitive strategies selected by students.


Little could be found in the literature about expected distributions within the population, except the following:

  • The American pre-college school system teaches primarily to a preference for step-by-step learning and memorization.
  • The American college system tends to be more conceptual and theoretical than the pre-college school system.
  • Hands on learners tend to go on to college in fewer numbers than other learners.
  • Story based and thematic learning is given nearly the same importance as the other three metacognitive strategies in education literature; however, it is covered in depth in the psychotherapy literature as an important way to teach clients new behavioral skills.

The course in which this study took place emphasized primarily hands on learning in a laboratory environment, conceptual learning in lecture (the teaching of principles over facts), and also memorization from the textbook. This may be atypical of many college classrooms — without a lab component to a course, there is little opportunity for practical application or hands on learning. And in contrast to this course, some liberal arts classes will offer a more significant set of opportunities for students who prefer a story-based learning strategy.

Data was not statistically significant in any way; nor was there sufficient prior research to suggest there should or should not be. In fact, the lack of a clear frontrunner demonstrates that all four metacognitive strategies are likely important and can benefit students. A larger sample size and more rigorous items would be required to confirm or deny this conjecture.





This exercises consisted of two parts. For part one, students were asked to select one of the eight intelligences and write a paragraph about their experience of that intelligence from an experiential perspective (that is, in terms of sensory modalities and metacognitive strategies.) For part two, students were asked to rank each intelligence along three dimensions (skill level, energy required, and interest.) Unfortunately, an error in the web based data gathering software caused the responses for part 2 to go uncollected. The material gathered for part one was used to flesh out portraits of each intelligence for the book “Multiple Intelligences and Personality Type” and can be found therein. Future research will explore this area further. Additional information can be found in reference [8].





The theory of temperament is detailed in reference [6]. Temperament posits four mutually distinct sets of patterns (categories) that reflect “core needs and values” and the talents that best meet those needs and values. Because this study’s emphasis was on intelligence and cognitive skills, only the “talents” aspect of temperament was explored. Students were presented with 4 snapshots, one of each temperament, and asked to rank them from highest (1st choice) or lowest (4th choice.) The distribution of first choice rankings is summarized in Table 7.

Artisan (_S_P) Guardian (_S_J) Rational (_NT_) Idealist (_NF_)
10/35 15/35 4*/35 6/35
29% 43% 11% 17%


Table 7: Distribution of students by highest ranked temperament pattern.


Research with the MBTI [3], which connects to temperament indirectly through a four-letter code, and data from the “Keirsey Temperament Sorter” [7] suggest that Guardians comprise the largest segment of the general population, from 40 to 45 percent. Artisans are second, appearing at approximately 30 to 35 percent, and then Idealists and Rationals appear in much smaller numbers. The small number of Rationals within this population was statistically significant here (p < .05) if we expected an even distribution. Overall this sample of students is extremely close to the expected distribution, suggesting the snap-shots of temperament talents used with the students may in fact work quite well. The distribution of the “least like me” temperament is also interesting, although there is no definitive research here. The evenness of the “least like me” distribution does possibly suggest, however, that the portraits were all equally likeable (or equally unlikable.) The results are summarized in Table 8.

Artisan (_S_P) Guardian (_S_J) Rational (_NT_) Idealist (_NF_)
8/35 9/35 8/35 9/35
23% 26% 23% 25%


Table 8: Distribution of lowest-ranked temperament patterns.


We can relate the temperament model to other models. Specifically, what was the match between first-ranked temperament and preferred metacognitive strategies, assuming the following correspondence?

Hands-on and experiential = Artisan

Step-by-step and memorization = Guardian

Conceptual and theoretical = Rational

Story-based and thematic = Idealist

No. of students with exact match Number of matches expressed as percentage Percentage of exact matches by chance alone
29*/35 83% 47.5%


Table 9: Match between temperament and metacognitive strategies.


The results of this possible correspondence are summarized in Table 9. Because the average student selected 1.9 metacognitive strategies, the percent for a match by chance alone was expected to be 47.5% (not 25%.) The high degree of correlation (83%) suggests that the metacognitive strategies and temperament talents are quite similar, if not ultimately measuring the same quality. Compared against the expected value (16 students matching) the result of 29 students was statistically significant (p < .05).




The learning style portraits in Berens’ “Dynamics of Personality” booklet are based on interviews with adults (ages 25 to 60 plus) who clearly identify with one of eight personality types corresponding to the last three letters of the MBTI four letter code. For example, ISTJ and ESTJ are grouped together in an STJ learning portrait.

Students were instructed to read all eight portraits, and select the one style that fit them most and the one style that fit them least. The portraits have a general descriptive label, not an MBTI code, although the labels are not used here. The resulting distribution for those learning styles rated “most like me” is shown in Table 10.

4 3 9 3 3 2 6 5


Table 10: Distribution of students vis-a-vis highest rated learning style portrait.


In contrast, Table 11 summarizes the resulting distribution of students vis-à-vis learning style portraits rated as “least like me.”

7/35 1/35 4/35 4/35 4/35 8/35 4/35 3/35


Table 11: Distribution of students to lowest rated learning style portrait.


A quick look at the data shows that STJ and NFJ and NFP were more often selected as best matching portraits, while STP and NTP were more often selected as least matching portraits. While STJ is considered the most common type of the 8 within the general population, NFJ and NFP are definitely not, and STP and SFJ are posited to be quite common but appear unpopular here.

Additionally, Tables 12 and 13 summarize comparisons with temperament and metacognitive strategies, which both match each other and the general population. The results suggest that the learning style portraits do not get at what is hoped here. The correlation to both is poor.

No. of students with exact match Number of matches expressed as percentage Percentage of exact matches by chance alone
21/35 60% 47.5%


Table 12: Learning styles vs. metacognitive strategies


No. of students with exact match Number of matches expressed as percentage Percentage of exact matches by chance alone
12/35 34% 25%


Table 13: Learning styles vs. temperament


Thus, it appears that the NFJ and NFP portraits may be overly appealing, and the STP and NTP under appealing. However, it is important to note that source of the portraits were adults, many of whom well beyond the 18 to 22 years old range of this sample. Thus, these portraits may be aged-biased and inappropriate for young people. In any case, the distribution does not contain statistical significance, so these conclusions are only tentative.





Students were asked to read two-page descriptions of each of the eight cognitive processes [2] and then rate each of the eight as “not like me”, “somewhat like me” or “a lot like me” or “definitely like me.” Since students were free to rate as many or few processes as they liked along this four-point scale, the distributions shown in Tables 14 and 15 are not equal to the number of subjects. Table 14 summarizes the distribution of cognitive processes which students rated as “not like me:”

Se Si Ne Ni Te Ti Fe Fi
2 2 9 10 6 8 4 10


Table 14: Distribution of cognitive processes rated as “not like me.”


The expected distribution was determined from MBTI data, type theory in general, and temperament theory, which state that Se and Se (and thus also Fe and Te) should be more common in the general population. And in fact, these four processes appear least here, and thus are more identified with than the others. Similarly, the literature also suggests that Ni and Ne would be least preferred in the general population, and thus most likely rated last, along with Fi and Ti.

Expected processes Not expected processes
37* 14*


Table 15: Distribution of lowest rated cognitive processes group (four expected processes vs. four not expected processes.)


There is statistical significance (p<.05) if we compare the four cognitive processes we expect against the four we don’t (37 vs. 14.) See Table 15.

The number of cognitive processes rated as “definitely like me” does not show a statistically interesting distribution (see Tables 16 and 17), which is not surprising, since, as other studies suggest, people tend to select where they are developing as well as where their existing preferences lie. However, if we compare the four processes we expect to be preferred (Si, Se, Te and Fe) against the four processes that are expected to not be preferred, then the difference (58 to 39) is more visibly suggestive, although still not statistically significant.

Se Si Ne Ni Te Ti Fe Fi
14 15 8 10 13 11 16 10


Table 16: Distribution of highest rated cognitive processes.


Expected processes Not expected processes
58 39


Table 17: Distribution of highest rated cognitive processes (four expected processes vs. four not expected.)






Finally, students were instructed to four three pages containing brief snapshots of each of the 16 MBTI types. One set of snapshots addressed learning style, while another set addressed problem solving and communication. Given these snapshots, a student was instructed to select a “best fit” type — the one type of the 16 that best captured their learning, problem solving and communication style. The results, and approximate expected values determined from studies of the general population, are summarized in Table 18.

Index Actual Expected Index Actual Expected
Extraverting 17 26 Introverting 18* 8.75
Sensing 22 26 Intuiting 13 8.75
Thinking 16 17 Feeling 19 18
Judging 22 18 Perceiving 13 17


Table 18: Student distribution by the four MBTI indexes


Within the general population, we expect 25 percent Introverting preference; we find a greater than 50 percent preference here, which is statistically significant (p < .05). None of the other numbers are significant. Overall, this distribution also paints _S_J as the most common set of types, which is consistent with the distributions of metacognitive strategies, temperament and cognitive processes data. However, a similar distribution is not the same as the percentage of one-to-one matches. For example, a student who selects ESTJ should, ideally, also select the Guardian (_S_J) temperament, the “step-by-step and memorization” metacognitive strategy, and the _STJ learning style portrait. Tables 19, 20 and 21 summarize these comparisons.

No. of students with exact match Number of matches expressed as percentage Percentage of exact matches by chance alone
15*/35 42% 25%


Table 19: Match between whole type and first-ranked temperament


No. of students with exact match Number of matches expressed as percentage Percentage of exact matches by chance alone
15*/35 42% 25%


Table 20: Match between whole type and learning style


No. of students with exact match Number of matches expressed as percentage Percentage of exact matches by chance alone
25*/35 71% 47.5%


Table 21: Match between whole type and meta-strategy


The chance of a whole-type vs. metacognitive strategy match is 47.5% because on average students selected 1.9 metacognitive strategies.

These statistically significant numbers show there is (as expected) a relationship between whole type and temperament, learning style and apparently metacognitive strategy as well. The correlation between whole-type and learning style portrait is also surprising, since the learning style portraits did not correlate with temperament or metacognitive strategy. This suggests that the learning style portraits do link to the other models, but probably not as closely as intended. In any case, all three comparisons are statistically significant (p < .05) in that all exceed what is expected by chance alone.


Exploring Second Person Point of View


Many aspects and inter-relationships between the models might be explored. The nature of second person point of view and timeline orientation is of particular interest because they are, in theory, distinct models.

No statistically significant correlation was found between time-orientation and the other models. The literature states that In-Time and Through-Time might match, in theory, to the J-P index of the MBTI, but this was not the case with this sample.

Correlation between second person point of view and several other models was found.

Eighteen of 35 students indicated access to second person point of view. These 18 students indicated 37 metacognitive strategies. Each of these 18 students indicated approximately 2.1 metacognitive strategies, higher than the 1.6 average for the remaining 17 students. This difference, while not significant, might be an area of future research. Table 22 shows that the “experiential hands-on” metacognitive strategy might also relate to the second person point of view. We would expect step-by-step learning to be co-equal to experiential learning if there was no relationship between the point of view model and this model, as these two strategies are equally common in this student population. However, the emphasis on the experiential strategy is not statistically significant.

Experiential Step by Step Concepts Story based
15 8 6 8


Table 22: The distribution of students with access to second person by metacognitive strategies.


Table 23 shows the statistical correlation (p<.05) between second person and kinesthetic submodality IV (getting a feeling about what someone else feels.) Eighteen students identified with submodality IV. Of those 18, 12 also indicated access to second person point of view. In contrast, only 6 of the remaining 17 students identified with submodality IV. We might expect comfort with submodality IV to be evenly distributed; instead, there is a big difference. This suggests, for example, that access to second person point of view (“empathy”) is a largely kinesthetic experience, or that kinesthetic information adds a bit to the experience.

Percent of those with access to second person and kinesthetic submodality IV Percent of those without access to second person, but with access to kinesthetic submodality IV.
12*/18, or 67% 6*/17, or 35%


Table 23: Relationship between second-person point of view and kinesthetic submodality IV.


As shown in Table 24, there is a potential link between second person and temperament identification. For example, 5 of 6 Idealists recorded access to second person, and 4 of 4 who identified as Rational did, while only 3 of 15 Guardians identified with second person point of view. Unfortunately, the sample size appears too small and muddied to really capture a statistical significance. A larger sample might yield very interesting results.

Artisan (_S_P) Guardian (_S_J) Rational (_NT_) Idealist (_NF_)
6 3 4 5


Table 24: The relationship between second person and temperament items



Table 25 summarizes the relationship between second person and the learning style portraits. It appears that STP, SFP, NFJ and NFP (the Artisan and Idealist learning styles) might be related to second person. According to temperament theory, the Artisan and Idealist temperaments share a “motive” dynamic, in contrast to the Guardian and Rational temperaments, which share a “structure” dynamic. However, the distribution here is only suggestive and not statistically significant. Perhaps a larger sample would fare better.

4 2 1 2 1 1 3 3


Table 25: The relationship between second person and learning style portraits


As Table 26 shows, there appears to be little correlation between second person and cognitive processes, as the distribution for those who indicated second person is quite similar to the overall cognitive processes distribution for all 35 students regardless of access to second person point of view.

Se Si Ne Ni Te Ti Fe Fi
1 1 4 4 1 4 1 4


Table 26: The relationship between second-person and “not like me” cognitive processes


Finally, as Table 28 summarizes, with respect to MBTI type indexes, there is no significant correlation between an MBTI preference and access to the second-person point of view. On the other hand, four particular types (ISFP, ISFJ, INFJ and INFP) have the best access to second person. A quick look at the whole-type data (not shown here) shows 7 I_F_ types, 5 E_F_ types, and 5 __TP types. The I_F_ types normally comprise only about 12 percent of the general population. We might expect that since second person is described as accessing information about others and experiencing “empathy”, the Feeling function would play a role. Similarly, for example, Introverting, while thought of colloquially as not being people-oriented, does in fact point to qualities such as reflection, sensitivity, and introspection — qualities related to or enhancing empathic understanding. It may be that all aspects of one’s type can potentially contribute to second person, and thus no correlation will appear. However, a large sample and more thorough study will be required to confirm any conjectures.

Index Actual Expected Index Actual Expected
Extraverting 7 8.25 Introverting 10 8.75
Sensing 10 10 Intuiting 7 7
Thinking 5 7.75 Feeling 12 9.25
Judging 9 10 Perceiving 8 7



Table 28:

The relationship between second person and MBTI type




Seven of the nine models proved appropriate to this group of college students. Students’ responses to these seven mirrored distributions in the general population, and were consist with each other as expected. For example, the relationship between SJ and Se with Guardian, and SP with Se with Artisan were supported, lending credence to Jung’s original model. One of the 9 models, the learning style portraits by Berens, did not match with the other three models it was designed with (temperament, cognitive processes and whole-type.) It is likely inappropriate for traditional college-age students, perhaps because it was designed from interviews with adults ages 25 to 60 plus. A strong correlation emerged between metacognitive strategies and temperament, suggesting these might be capturing aspects of the same underlying model. Correlations also emerged between second person and the kinesthetic sensory modality. The uniqueness of the timeline model suggests that James’ conjecture of a relationship to the MBTI J/P scale may be in error, and is appears that an atypical timeline orientation may be indicative of cognitive flexibility. Whether all these correlations hold under future and more extensive research remains to be seen. All in all, the temperament portraits, metacognitive strategies, and three sets of whole-type descriptions (communication, problem solving and learning) are probably good ways to get at a “best fit” type, if the instructor is disinclined or unable to use an instrument like the MBTI. Finally, the strong presence of the “story-based and thematic” metacognitive strategy suggests that educational research should pay more attention to this, in addition to step-by-step and other metacognitive strategies.





1] Nardi, D. Multiple Intelligences and Personality Type. Telos Publications, 2001.

2] Berens, L. Dynamics of Personality Type. Telos Publications, 2000.

3] Myers, I.B., McCaulley, M.H. MBTI Manual, Third Edition. Consulting Psychologists Press, 1998.

4] O’Connor, J and Seymour, J. Introducing NLP. The Acquarian Press, 1993.

5] James, T., and Woodsmall, W. Timeline Therapy and the Basis of Personality. Meta Publications, 1988.

6] Berens, L. Understand Yourself and Others, An Introduction to Temperament. Telos Publications, 1998

7] Keirsey, D. Please Understand Me II. Prometheus Nemesis Books, 1998.

8] Gardner, H. Frames of Mind: The Theory of Multiple Intelligences. Basic Books, 1993.


This Paper Originally Presented as Proceedings of The Power of Self-Leadership Conference, Huntington Beach, CA, May, 2001.