Title: Nature, nurture, and temperament: Comparisons of temperament styles displayed by U.S. students
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00102714/00001
 Material Information
Title: Nature, nurture, and temperament: Comparisons of temperament styles displayed by U.S. students
Physical Description: Book
Language: English
Creator: Bassett, Kyle D
Copyright Date: 2004
 Record Information
Bibliographic ID: UF00102714
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.

Full Text








I extend sincerest appreciation to Dr. Thomas Oakland, my dissertation chair

and mentor, who gave generously of his time, knowledge, and resources throughout my

graduate education and the completion of this dissertation. His commitment to students

is irreproachable and extends so much further than the classroom. His commitment to

the school psychology profession is extraordinary.

Appreciation also is extended to Dr. Tina Smith-Bonahue for her support

throughout my graduate education and the dissertation process. I wish to thank Dr.

Jennifer Asmus and Dr. Maureen Conroy for their guidance in this project as well as the

Autism Inclusion Project.

I extend special thanks to all of my family members. I am so fortunate to have

each of them in my life. Their unwavering support is always an inspiration.

Laura, words alone...I extend my deepest thanks for your patience, care, and




ACKNOWLEDGMENTS ................................................................. ii

LIST OF TABLES ............................................................ v

LIST OF FIGURES ..................................................................... vi

ABSTRACT .............................................................................. vii


1 INTRODUCTION ...................................................................... ... 1

Points of Convergence ................................................................. 2
Points of Divergence ................................................................ 2
Convergence Revisted ............................................................. 2
Utility of Understanding Children's Temperament ................................. 3

2 REVIEW OF LITERATURE ........................................................ 7

Ancient and Pre-Moder Perspectives ................................................. 7
Current Perspectives ................................................................. 16
Developmental Perspectives .......................................................... 34
Gender Perspectives .................................................................. 43
Purposes of the Study ..................................................................... 47
Research Questions ................................................................... 49

3 METHODS ............................................................................ 50

Participants ............................................................................ 50
Instrumentation ........................................................................ 51
Data Analyses Procedures .................................................. ............ 58

4 RESULTS .............................................................................. 62

5 DISCUSSION ................................................................... ...... 77

Preferences for an Extroverted or Introverted Style ................................ 77


Preferences for a Practical or Imaginative Style ...................................... 83
Preferences for a Thinking or Feeling Style ........................................... 84
Preferences for an Organized or Flexible Style .................................... 90
Implications of the Study ............................. ............................ 95
Limitations of the Study .................................................................... 96



B ADMINISTRATION PROCEDURES ............................................... 105

REFERENCES ....................................................... 108

BIOGRAPHICAL SKETCH .............................................................. 125



Table page

1 Preference frequency percentages by age and gender ............................. 68

2 Summary of chi-square results for extroversion/introversion dimension ......... 69

3 Summary of chi-square results for practical/imaginative dimension .............. 70

4 Summary of chi-square results for thinking/feeling dimension ................... 71

5 Summary of chi-square results for organized/flexible dimension ............... 72


Figure page

1 Percentage of children who display an extroverted style across ten age groups ... 73

2 Percentage of children who display a practical style across ten age groups ........ 74

3 Percentage of children who display a thinking style across ten age groups ........ 75

4 Percentage of children who display an organized style across ten age groups ..... 76

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy



Kyle D. Bassett

December 2004

Chair: Thomas Oakland
Major Department: Educational Psychology

Temperament preferences displayed by U.S. students ages 8 through 17 were

examined. Temperament styles of 7,902 U.S. males and females were compared to

identify possible age and gender differences in temperament preferences. In general,

both male and female students prefer extroversion to introversion, imaginative to

practical, and organized to flexible styles. Males prefer a thinking style while females

prefer a feeling style.

From a developmental perspective, preferences for an extroverted style increase

between ages 8 and 13 and decrease thereafter. A shift from an initially balanced

perspective on the practical-imaginative dimension toward an increase in preferences

for an imaginative style is evident between ages 8 and 10. A return toward a more

balanced perspective on the practical imaginative dimension occurs between ages 10

and 15. Preferences for an imaginative style increase again between ages 15 and 17. The

prevalence of a thinking style among males generally is stable between ages 8 and 17.

In contrast, age level differences in the direction of an increasingly stronger preference

for a feeling style are evident among females. A shift from an initially strong preference

for an organized style toward a more balanced perspective that includes both organized

and flexible styles is evident between ages 8 and 15. A return to a preference for an

organized style is observed between ages 15 and 17. Biological and cultural

explanations were considered in interpreting the findings. Results were compared with

existing data in related adult-based literature.


Interest in temperament probably is as old as mankind. Notions about

differences in temperament are found in the ancient writings of eastern and western

philosophers, narratives of writers and historians, and abstracts of present day

researchers and experimenters. Various terms and concepts have been used to describe

and explain obvious interpersonal differences. For example, temperament has been

described, among other things, as dispositions, character, behaviors, styles, abilities,

motivations, emotions, and personality. "This state of affairs has led to some confusion

for scholars who seek to offer an overview of the field of temperament research"

(Garrison & Earls, 1987, p.13), especially when relating the concept of temperament to

personality. Temperament qualities have been of interest, in part, because of

expectations that they may influence the development of personality (Rutter, 1987).

Given the absence of a consensus construct and definition for temperament,

temperament commonly has been defined operationally, based on the particular

approach of a researcher (Hubert, Wachs, Peters-Martin, & Gandour, 1982). Indeed, a

roundtable discussion of eight contemporary researchers offered eight different

definitions of temperament (Goldsmith et al., 1987). However, several points of

convergence exist among contemporary temperament theories (Goldsmith & Rieser-

Danner, 1986).

Points of Convergence

Temperament is an inclusive term referring to several dimensions of behavior

characterized by individual differences. Most temperament characteristics appear early

in life and form the foundation upon which personality characteristics later emerge.

Temperament dimensions are relatively stable across time and context. Although

temperament traits have a strong biological basis, their expression can be influenced by

contextual factors and personal choice (Goldsmith & Rieser-Danner, 1986).

Points of Divergence

Issues representing points of divergence among contemporary theories also can

be identified: the actual dimensions of temperament, the role and emphasis of biological

factors in the determination of temperaments, the role of motivation in temperament, the

fundamental definitions of temperament-especially at the behavioral and

psychophysiological level, the role of temperament in behavior, the role of contextual

influences on temperament, and relationships between temperament and personality

(Goldsmith & Rieser-Danner, 1986).

Convergence Revisited

While one accepted construct of temperament may not exist, the essential

rationale for studying a concept such as temperament is entrenched firmly. The

fundamental significance of temperament lies in its proposed role as a biologically-

rooted determinant of personality. Tegalsi (1998a) states:

The identification of temperament traits fills a gap in our understanding of
human development and..... the idea that individuals influence their own
development by bringing unique attributes to their experiences contrasts the
view of children as "blank slates" to be inscribed by the environment. By
assuming a bidirectional interplay among intrinsic attributes of a person and
external demands, supports, and circumstances, temperament theory supports


the widely accepted assumption that development is propelled by the person and
environment interaction. (Lewin, 1935, cited in Tegalsi, p. 475)

Utility of Understanding Children's Temperament

The person-environment schema implies that those who are responsible for

children may foster their development with an understanding of temperament theory.

Mental health professionals, medical practitioners, educators, parents, and other

caregivers who use temperament theory may be more sensitive to the needs of

individual children. This belief is reflected in the following statement by McClowry


Many children's behavioral problems, according to temperament theory, could
be averted if the environment provides goodness of fit, which is achieved when
demands, expectations, and opportunities are consonant with the child's
development (Chess & Thomas, 1984). Parents and other caregivers of children
with challenging temperaments are often relieved to learn that some children
require considerably more effort to rear than others do. Once this insight is
achieved self-blame is reduced and caregivers can redirect their energies to
resolving the conflict their child has with his/her social environment. (Bates,
Wachs, & Emde, 1994, cited in McClowry, p. 553)

An illustration of the positive effects of goodness of fit on later personality has

been provided by Wachs (1994). The inhibited child who experiences a social

environment with sensitive parents that lets the child set one's own pace, a physical

environment with "defensible spaces" to which the child can retreat, peer groups with

other inhibited children and common interests, and a school environment that is

"undermanned" so the child is tolerated and feels one makes a contribution, is likely to

develop into a healthy adult with behavioral manifestations close to extraversion and

emotional stability.

However, the inhibited child who encounters controlling parents, chaotic

environments that allow no escape from stimulation, peer groups that reject the child,

and an "overmanned" school where the child is less tolerated and does not feel valued is

likely to develop behavioral manifestations of neuroticism.

While the examples provided above are hypothetical, evidence supports the

notion that consonance between the person and environment, or goodness of fit,

promotes optimal development (Chess & Thomas, 1996).

Taken as a whole, goodness of fit appears to be important for understanding

children's experiences in school (Keogh, 1986). Given the importance of school

psychologists in the development of educational programs for children, the conclusion

that school psychologists can play a key role in promoting goodness of fit in school

settings is reasonable. This implies that a school psychologist's assessment,

intervention, and consultation efforts with teachers and parents are informed by

knowledge about children's temperament styles.


From a psychoeducational assessment perspective, assessment of children and

youth traditionally has been limited to variables of cognitive ability, academic skills,

and behavior problems (Martin, 1994). While important to understanding children's

progress, these domains are insufficient for describing the child's behavioral style. A

school psychologist's primary goal is to accurately describe the child and provide a

realistic picture of the child's developmental trajectory (Martin, 1994). The inclusion of

temperament measures in a psychologist's assessment battery provides avenues for

evaluating styles, predicting future performance and behavior, and developing

interventions that can improve school experiences for students and teachers (Carey,

1998; Keogh, 2003; McClowry, 1998; Tegalsi, 1998b).


Goodness of fit theory infers that school psychologists move beyond exclusive

focus on the assessment of the child and into the child's learning environments to

understand how social and physical contextual variables interact with the child's

temperament (Keogh, 1989), leading to the development of environments that are

consistent with a child's needs (Martin, 1994). Additionally, awareness of temperament

differences may provide teachers important information about students' learning styles

that may be used to enhance environments where learning occurs. For example, the

extroverted child may be provided opportunities to work in cooperative groups whereas

the introverted child may be allowed to work independently.


When consulting with caregivers, school psychologists can provide

temperament information that may have a profound effect on clarifying or enhancing

relationships between the consultee and child, especially with children displaying a

difficult style. Consultee knowledge that children's behavior may be biologically rooted

may reduce self-blame or blaming the child for challenging and seemingly willful

behavior (Tegalsi, 1998b). Furthermore, "acknowledging that some temperament

characteristics are challenging for caregivers normalizes some of the frustrations

experienced in caring for the child" (Tegalsi, 1998b, p. 580). Thus, an understanding of

children's temperament characteristics may help caregivers develop realistic

expectations, leading to reduced stress in situations in which the caregiver's

requirements are inconsistent with the child's behavioral style (Carey, 1998). Moreover,

through a consultative relationship, teachers may be able to identify and develop a

wider range of instructional and management strategies that are more responsive to

children's learning styles.

The purpose of this study is to describe and explain possible developmental

change and stability of temperament preferences displayed by U.S. male and female

students age 8 through 17. An overview of historical perspectives on the construct of

temperament is provided. Contemporary theories are reviewed and how various

developmental and gender considerations flow from the theoretical orientations are

discussed. The research questions that guide the study are set forth. The methods used

to conduct the investigation are described. The dissertation will conclude with a results

section and a discussion of the findings.


Ancient and Pre-Moder Perspectives

In Psychology of Character (1927), Roback wrote, "the ancients have given

evidence of an uncanny insight in many of the scattered observations on both character

and temperament" (p. 4). Roback probably was referring to the contributions of such

ancient Greeks as Hippocrates, Plato, and Aristotle. Hippocrates seemingly first

recorded notions about temperament around 350 B.C. in On the Nature ofMan. He

wrote of four humors or temperaments that were associated with body fluids: black and

yellow bile, phlegm, and blood. These humors were thought to control the whole

existence and behavior of mankind. An individual's character and health were

determined according to the manner in which the humors were combined.

A generation later, Plato and Aristotle each described four types of character

akin to Hippocrates' temperaments (Keirsey, 1998). Galen, around 200 A.D., ultimately

delineated a theory of temperament that would remain influential through the nineteenth

century (Strelau, 1998). Galen described a quartet of temperament styles derived from

Hippocrates' humors: the melancholic, sanguine, choleric, and phlegmatic styles. These

styles resulted from an excess of one of the bodily fluids. The melancholic person,

always serious and downcast, owed his nature to an excess of black bile. Sanguine's

impulsivity and enthusiasm was thought to be due to the strength of his blood. The

choleric's ease of irritability and sensitivity was traced to the influence of yellow bile.

The phlegmatic person's slowness and impassivity were attributed to the predominance

of the phlegm (Roback, 1927).

During the Middle Ages, Galen's humoral doctrine penetrated everyday life.

Moreover, his concept of the four temperaments made a deep impression among some

Renaissance and Enlightenment scholars. For example, Paracelsus, a mid-sixteenth

century physician, described four human natures that were comparable to Galen's

temperament types (Keirsey, 1998), despite the fact he disputed Galen's interpretations

of Hippocrates' writings. Literary greats, including Shakespeare and his contemporary,

Ben Jonson, made numerous references to the humors in their plays. Philosophers

during the Enlightenment period (e.g., Bruno, Hume, Voltaire, Rousseau, and Kant) did

not question the use of humoral theory (Keirsey, 1998).

Nevertheless, the widespread acceptance of Galen's inborn predispositions did not

go entirely unchallenged. As an omen of things to come, the great British empiricist

John Locke (1690/1979) wrote in an oft-quoted passage from his An Essay Concerning

Human Understanding:

Let us suppose the mind to be, as we say, white paper, void of all characters,
without any ideas: How comes it to be furnished? Whence comes it by that
vast store which the busy and boundless fancy of man has painted on it with an
almost endless variety? Whence has it all the materials of reason and
knowledge? To this I answer, in one word, from experience.

Galen's views gradually began to erode around the time of the American and

French revolutions (Kagan, 1994a). Another Greek design, democracy, captured the

attention of the western world. With it came a shift in thinking that stood in direct

contrast to Galen's inborn temperament types: people were fundamentally alike rather

than different. The idea of innate differences in humans, as Galen described, was

thought by some to violate the democratic principle of egalitarianism. People were

conceived to be essentially alike, to be molded by their experiences. These ideas had

particular meaning to the new middle class whose aspirations for power and influence

now could be realized legitimately (Kagan, 1994a). By the end of the nineteenth

century, inquiry into the nature of persons emphasized the influence of the environment

in the development of character and personality. At about this time Sigmund Freud

proposed his highly influential psychoanalytic theory of development that embraced

early experience as a determinant of adult personality.

Environmentalism Undaunted

Freud developed a theory that described personality in terms of three

components: the id, ego, and superego. During the development of the id, ego, and

superego, the child proceeded through a sequence of five psychosexual stages. In each

stage the child was faced with problems caused by the conflict of wanting to satisfy

instincts, especially sexual desires, with the need to comply with the demands of

society. Childhood events played a significant role in the resolution of these conflicts

and, in turn, the acquisition of personality traits. For example, an infant deprived of

satisfactions during the oral stage was likely to be gloomy, doubting, and somewhat

pessimistic. Satisfaction of oral needs resulted in a self-assured personality with an

optimistic outlook on life. Freud's inclusion of the environment as an influence on the

development of personality was clearly distinct from the innately derived temperaments

described by the Greeks a millennium earlier.

Many of Freud's contemporaries took issue with his focus on sexual instincts

but retained his idea of a single motive for behavior. Alfred Adler, a colleague of


Freud's, replaced sexual instincts with a desire for superiority as the prime motivator in

life. Adler proposed that man was born into the world feeling incomplete and

unfulfilled. Feeling inferior, he was motivated to emulate the strength and capabilities

of others. Once having attained those capacities, he again felt inferior. In order for man

to achieve superiority, however temporary, it was necessary to conduct the self in a

singular style of life. Style, or personality, was uniquely determined by forces from the

social environment that aided, interfered, or reshaped one's quest for superiority.

Harry Sullivan described a theory of personality that put forth anxiety reduction

as the basic motive for behavior. Sullivan proposed that man was surrounded by

feelings of anxiety the moment he entered life. Real or imagined anxiety situations were

alleviated best by the formation of safe interpersonal relationships with nonanxious

others. Over time, the number of relationships increased and interpersonal patterns

eventually emerged to form the core of one's personality. Adler and Sullivan's theories

lent further credence to the notion that the social environment was a primary molder of


Interest in the environment as the primary determinant of personality flourished

in analytic circles during the remainder of the first half of the twentieth century. Eric

Fromm concluded that the quest for self-preservation was shaped by parental and

societal influences. Karen Homey reasoned that the human endeavor toward self-

realization was either aided or blocked by experiences, at the crux of which were human

relationships. Erik Erikson's theory of personality, based on the tenets of Freud, placed

greater emphasis on the roles of society and culture on the formation of the ego.

About the time analytic approaches to personality development were enjoying

their greatest popularity, the noted American behavioral psychologist, B.F. Skinner,

sought to redefine personality theory as the study of human behavior. Emphatically

rejecting hypothetical underlying causes of personality (e.g., instincts, feelings of

inferiority or superiority, self-preservation, and self-realization) as described in analytic

theories, Skinner believed psychology could become scientific only if it restricted its

attention to observable behaviors. Based on principles from John Watson, Skinner

concluded that all human behavior was learned through prior conditioning, that is, as a

function of the forces of the external environment that reinforced behavior. Behavioral

patterns are the lasting result of reinforcement schedules. Skinner's behaviorist

approach departed significantly from the popular psychoanalytic theories. However, he

also emphasized the role of environmental events in explaining human behaviors.

The idea of inborn dispositions was not abandoned altogether in the early

twentieth century. Interestingly, Carl Jung, a psychologist well-known for his

psychoanalytic orientation, revived interest in innately derived temperament types.

Whereas many psychoanalysts of his era stressed childhood determinants of personality,

Jung believed the newborn infant was by no means a tabula rasa.

Jungian Typology

Predating temperament theories developed latter in the twentieth century, Jung

attributed individual differences in personality to a pair of inborn, possible genetic or

physiological dimensions mediated by environmental influences. The first dimension

concerned the direction of psychic energy and consisted of two attitudes: extraversion

and introversion. The second dimension involved the apprehension of stimuli and


consisted of four mental functions: thinking, feeling, sensation, and intuition. From this

point of view, Jung differentiated eight basic types of people. The following section

provides a more detailed description of Jung's theory of psychological types.

Attitude types. In his now classic book, Psychological Types, Jung (1921/1971)

distinguishes between two basic typological attitudes of extraversion and introversion.

In general, attitude is conceived to be a state of psychological readiness in the

individual to act or react in a certain direction.

Jung (1953) first took notice of attitudes when trying to understand why Freud's

view of human psychology was so different from that of Adler. From Jung's

perspective, Freud saw his patients as defining themselves in relation to significant

objects in the environment, especially parents. Conversely, Adler saw his patients

defining themselves in relation to a self-determined quest for supremacy. Thus, Freud

stressed a movement of psychic energy toward the outer world; Adler stressed a

movement of psychic energy toward the inner world or self. The contrasting

perspectives of his contemporaries essentially laid the foundation for Jung's extraverted

and introverted attitudes.

In the extraverted attitude, the movement of energy is toward the outer world

(i.e., from the person to the object). The extravert's interest and attention follow events

and activities in the immediate environment. Persons and things regulate the actions of

the extravert. Moreover, the extravert's philosophy of life is governed by the collective

standards of society. According to Jung, the extreme extravert is at risk of losing his or

her own identity as a separate person.

In contrast, the introvert's movement of energy is toward the inner world (i.e.,

from the object to the person). Subjective factors are the chief determinant in the

interest and attention of the introvert. Thus, personal decisions and convictions guide

the introvert's actions and morality. The extreme introvert is at risk of retreating into a

world of emotional and behavioral isolationism.

However, a pure and exclusive attitude never occurs (Jung, 1921/1971). Jung

acknowledged that every individual possesses attitudes of extraversion and introversion.

A healthy individual adapts in both ways, depending on one's suitability to the

environment. For example, the introvert occasionally exhibits an extraverted orientation

and vice versa. Indeed, most people probably can identify both introverted and

extraverted behaviors in their personality (Spoto, 1989). Still, a healthy individual

ultimately develops an attitude type consonant with his or her prevailing and preferred

nature. This reasoning similarly applies under the discussion of function types.

Function types. Jung's early formulations of typology centered almost

exclusively on the polarity of introverted and extraverted attitudes (Spoto, 1989).

However, Jung eventually concluded that attitude alone was insufficient for explaining

psychological types. Jung (1921/1971) wrote:

I had tried to explain too much in too simple a way, as often happens in the first
joy of discovery .... What struck me now was the undeniable fact that while
people may be classed as introverts and extraverts, this does not account for the
tremendous differences between individuals in either class. (p. 535)

Based on his many years of clinical experience, Jung (1953) ultimately

concluded the existence of four mental functions (i.e., thinking, feeling, sensation, and

intuition) in addition to the two attitudes. The functions are conceived to be forms of

psychological orientation, each with its own area of expertise (Jung, 1921/1971).


The thinking function refers to the process of cognitive thought. It organizes and

decides things based on analysis and logic. The feeling function refers to the process of

subjective judgment. It organizes and decides things on the basis of values and

individual worth. According to Jung, the thinking and feeling functions are

characterized as rational (i.e., judging) functions because they are based on linear

reasoning and judgment.

The sensation function refers to apprehension via the five sense organs. It

operates primarily through experience of the physical world. The intuition function

refers to apprehension through unconscious experience or instincts. Jung characterized

both sensation and intuition as irrational (i.e., sensing) functions because they operate

outside the realm of reasoning. In other words, perceptions by these functions simply

happen upon one's self and are made independent of rational thought processes.

Here again, Jung acknowledged that every individual possesses each of the four

mental functions. Further, the more developed each of the four functions, the more

comprehensive the individual's understanding of the world. Indeed, Jung (1921/1971)


These four together produce a kind of totality. Sensation establishes what is
actually present, thinking enables us to recognize its meaning, feeling tells us its
value, and intuition points to its possibilities as to whence it came and whither it
is going in a given situation. In this way we can orient ourselves with respect to
the immediate world. (pp. 540-541)

Still, Jung (1921/1971) theorized that attempts to bring all psychological

functions into simultaneous development are virtually impossible. Instead, the

individual typically develops one superior function. The superior function ordinarily is

that with which the individual is most gifted by nature. It possesses its own inherent

energy and is the most efficient means for one's social success. Specifically, it is the

individual's preferred function.

Individuation and psychological type. The concept of individuation is perhaps

the glue that binds one's psychological type. According to Jung (1921/1971), the goal

of individuation is to develop one's innate dispositions so that the superior function

combines with the dominant attitude. In this manner, the extraverted or introverted

attitude is paired with one of the four mental functions to yield one of eight possible

personality types.

If extraversion is the superior attitude, a person with a dominant thinking

function will likely pursue a lifestyle according to intellectual considerations of

objective facts. In a similar manner, the introverted thinking type also is concerned

mainly with ideas, with the ideas springing from the subjective domain.

The extraverted feeling type is characterized by a desire to create harmonious

conditions in the surrounding environment, particularly in social situations. The

introverted feeling type also seeks harmony but is less concerned with the comfort and

well-being of others. The introverted feeling type often appears to be silent and


The extraverted sensation type displays a strong inclination for objects and

people in the environment that arouse the senses. This type is highly motivated by

physical or concrete enjoyment. In a similar manner, the introverted sensation type also

is concerned with physical sensitivity to objects and people but is subjectively oriented

toward what is perceived.


The extraverted intuitive type is characterized by an extraordinary ability to see

possibilities in external situations, especially the ideas and creations of others. The

introverted intuitive type also possesses the uncanny ability to see the possibilities,

although the visions are likely to originate from ideas and creations formulated within.

Falsification of type. A healthy personality emerges if the individuation process

was not hindered by external forces or misguided attempts to make another attitude or

function superior (Jung, 1921/1971). A child could be forced into assuming a type that

violates the innate disposition under abnormal conditions (e.g., coercive parenting). The

consequences of such a situation are deleterious (Jung, 1921/1946):

As a rule, whenever such a falsification of type takes place as a result of external
influence, the individual becomes neurotic later, and a cure can successfully be
sought only in a development of that attitude which corresponds with the
individual's natural way .... a reversal of type often proves exceedingly harmful
to the physiological well-being of the organism, often provoking an acute state
of exhaustion. (pp. 415-416)

Jung described a theory that included both innate dispositions and environmental

factors in the determination of personality. His views were forerunners to the person-

environment schema described in many contemporary temperament theories.

Current Perspectives

Myers-Briggs Theory

Around the time Jung was developing his theory of psychological types,

Katherine Briggs was attempting to develop her own typology. Her interest in

personality types was sparked when she noted that her daughter, Isabel Myers, had

fallen in love with a man so different from others in the Briggs family. Thus, Briggs

began reading a multitude of varied biographies in an effort to develop a typology based

on the patterns she found. She first delineated four types: the meditative, spontaneous,


executive, and sociable. When Briggs discovered Psychological Types, she abandoned

her own typology in favor of Jung's.

Working in the late 1940s and throughout the 1950s with the "belief that many

problems might be dealt with more successfully if approached in the light of C.G.

Jung's theory of psychological types" (Myers & Myers, 1980, p. xiii), Myers and

Briggs developed their own conceptual framework of Jung's typology. Thus, although

their framework is strikingly similar to Jung's theory, several modifications can be


They accepted the two attitudes to life (i.e., extraversion and introversion) and

the four basic functions (i.e. sensing, intuition, thinking, and feeling) postulated by

Jung. However, Myers split the functions into two dichotomized and separate qualities:

sensing/intuition and thinking/feeling. She also added a fourth dimension: judging vs.

perceiving. Thus, the test based on this model, the Myers-Briggs Type Indicator

(MBTI), operatizes and describes individual differences along four temperament

dimensions, resulting in sixteen types instead of Jung's eight types. Consistent with

Jung's model, these temperament qualities are conceived to arise from inborn

differences (Myers & Myers, 1980).

The attitudes of extraversion (E) and introversion (I) refer to ways people orient

themselves to the world. In extraversion, energy flows outward, in action, to objects

and people. In introversion, energy is directed inward, to reflection on thoughts, ideas,

and concepts.

According to Myers and Myers (1980), the two modes of perception, sensing (S)

and intuition (N), compete for a person's attention. One mode is preferred from infancy


onward. Sensing is the process of becoming "aware of things directly through our five

senses" (p.2). The process of intuition is "indirect perception by way of the

unconscious, incorporating ideas or associations that the unconscious tacks on to

perceptions coming from outside" (p.2). Temperament traits associated with the

preferred perspective emerge as the SN preference becomes more refined (Myers &

Myers, 1980).

Myers also describes two modes ofjudging: thinking (T) and feeling (F). These

two functions process the perceptions or information gathered to form judgments or

decisions. Thinkers use logical analysis to make decisions while feelers use personal

values and priorities. Myers and Myers (1980) noted "the child who prefers thinking

develops along divergent lines from the child who prefers feeling, even when both like

the same perception process and start with the same perceptions" (p.4).

The final and added dimension, judgment (J) and perception (P), Myers' most

radical departure from Jung's typology, concerns the preferred "method of dealing with

the world around us" (Myers & Myers, 1980, p. 8). The judger deals with the outside

world with regiment and purpose whereas the perceiver is more flexible and adaptable.

Importantly, users of type theory need to be clear that the judging/perceiving scale

refers only to the manner in which one conducts the self in the external world,

regardless of position on the extraversion/introversion dimension.

Like Jung, Myers concluded the environment impacts inborn differences along

the dimensions. In this respect, experiences could be considered obstacles that

discourage individuals from developing natural preferences (Myers & Myers, 1980).

Examples of environmental obstacles include a lack of faith in one's own type due to its


infrequency in the general population, its lack of acceptance in one's environment, and

a lack of opportunity or incentive to exercise one's favored process or attitude. On the

other hand, Myers believed type development is encouraged when the environment met

the needs of a child's native capacities.

Given the relative similarity in the conceptual frameworks of Jung and Myers

and Briggs, one might be tempted to discount Myers and Briggs contributions to type

theory. Such a perspective would be extremely short sighted. Importantly, with the

creation of the Myers-Briggs Type Indicator (MBTI), Myers and Briggs brought Jung's

typological theory to life and made available applications of type theory to individuals

everywhere (Keirsey & Bates, 1978).

Criticisms. Critical remarks regarding the MBTI model have focused mainly on

the reliability and factor structure of the instrument. For example, one-half of subjects

were found to be classified as a different type when they took the MBTI five weeks

apart (McCarley & Carskadon, 1983). Accordingly, Myers' types could not represent

stable personality characteristics. Furthermore, item analysis of the MBTI revealed a

six-factor structure, thus inconsistent with that developed by Myers (Sipps, Alexander,

& Friedt, 1985). Recently, however, pointing to omissions and inappropriate techniques

used by the researchers, the above claims have been rebutted (Hammer, 1996). Despite

conflicting reports regarding its psychometric properties, the MBTI has enjoyed

enormous success with counselors, educators, and employers since its publication in

1962. "Today the MBTI is the most widely used personality instrument in the world -

two million administrations are given each year" (Myers, McCaulley, Quenk, &

Hammer, 1998, p. 9).

Keirseian Theory

David Keirsey began a 50 year career in typewatching after a chance encounter

in the mid-1950s with a psychologist familiar with the Myers-Briggs Type Indicator

(Keirsey, 1998). Working with concepts developed by Jung and refined by Myers and

Briggs, Keirsey eventually developed his own theory of temperament.

At least one major modification distinguishes Keirseian theory from his

predecessors (Keirsey, 1998). Troubled by what he deemed subjective typologies of

Jung and Myers that distinguish personality types almost exclusively by people's

mental makeup, Keirsey developed a typology based on observable human behavior.

Specifically, temperament types are derived from two basic human actions: how people

communicate with others and how people use tools to accomplish their goals.

According to Keirsey, people are prone to use either abstract or concrete speech.

Those who use abstract speech communicate in fictional, figurative, symbolic, and

theoretical ways. Those who use concrete speech communicate in detailed, factual,

literal, and specific ways. Thus, Keirsey delineates a dichotomous model captured by

and reflected through word usage.

A dichotomy in the tools people use to accomplish goals also is postulated.

People are predisposed either to cooperative or utilitarian tool usage. The cooperator, in

his desire to get along with others, is inclined to use tools that have been approved by

the social groups to which they belong. The utilitarian, stressing usefulness over social

considerations, is inclined to choose tools that result in success with minimum effort

and cost.

Keirsey postulated that the combination of one's preferences for word and tool

use are present in the individual from birth and result in one of four possible

temperament types: the Artisan (Myers's sensing-perceiving), Guardian (sensing-

judging), Idealist (intuition-feeling), and Rational (intuition-thinking).

The Artisan, a concrete utilitarian, is artistic, audacious, and adaptable. Artisans

are likely to be interested in the acquisition and perfection of technique. They also are

likely to be impulsive, enjoy freedom, and have a strong play ethic.

The Guardian, a concrete cooperator, is dependable, beneficent, and respectable.

Guardians are likely to enjoy accommodating others and belonging to groups.

Moreover, Guardians often have a strong sense of morality.

The Idealist, an abstract cooperator, is characterized by empathy, benevolence,

and authenticity. Idealists are prone to be concerned about the morale of those around

them. They also are likely to champion the underdog and become involved in social


Finally, the Rational, an abstract utilitarian, is characterized by ingenuity,

autonomy, and self-confidence. Rationals are likely to be preoccupied with technology

and the logic of building. They often put work before play and have a keen drive for


Not unlike the views of the other theorists presented in this paper, Keirsey

considers the environment to be a powerful mediator in the expression of temperament.

Accordingly, personality and character emerge through the interaction of one's

temperament and environment (Keirsey, 1998).

Criticisms. Keirsey (1998) has operatized his theory of temperament with the

creation of the Keirsey Temperament Sorter II. However, the author does not view the

instrument as a "test". Thus, there are no statistics regarding its reliability or validity.

Given the lack of its empirical support, the utility of the Keirsey Temperament Sorter II

for research and practice is limited.

Stylistic Theory

Temperament has been discussed previously in light of its expression in adults.

Given a consensus that temperament traits are early appearing substrates of personality,

research with children is expected.

Thomas and Chess were at the forefront of this research. Unable to make sense

of their clinical findings through standard applications of psychoanalysis, behaviorism,

and learning theory (Chess & Thomas, 1996), these researchers turned toward

biological explanations of individual differences. Thomas and Chess infer nine early

appearing, constitutionally rooted temperament categories from their now famous New

York Longitudinal Study (NYLS): activity level, rhythmicity, approach/withdrawal,

adaptability, sensory threshold, quality of mood, intensity of mood expression,

distractibility, and persistence/attention span (Thomas, Chess, Birch, Hertzig, & Korn,

1963). The dimensions are seen as the stylistic components of behavior that result from

innate features within the brain (Goldsmith et al., 1987).

Using factor analysis, Thomas, Chess, and Birch (1968) found these dimensions

are at least partially independent and cluster around three basic characteristics: easy

temperament, slow-to-warm-up temperament, and difficult temperament. The easy

temperament comprised about 40% of the NYLS sample and is characterized by a

positive approach to new stimuli, quick adaptability, and positive mood (Chess &

Thomas, 1996). The slow-to-warm-up temperament comprised about 15% of the NYLS

sample and is characterized by negative responses of mild intensity to new experiences

and slow adaptability. With repeated exposure, these children gradually adapt to new

experiences with "quiet and positive interest and involvement" (Chess & Thomas, 1996,

p. 37). The final constellation, the difficult temperament, comprised about 10% of the

sample and is characterized by withdrawal responses to new situations, irregularity in

biological functions, slow adaptability, and frequent, intense negative reactions. This

group represents a normal and healthy segment of the population despite these negative

connotations (Chess & Thomas, 1996). About one-third of the NYLS sample was

difficult to classify and did not fit into one of the three temperament categories, a result

of varying combinations of traits (Chess & Thomas, 1996).

Central to Chess and Thomas' analysis is the notion that temperaments are not

entirely fixed or immutable. Using the term goodness offit (Thomas et al., 1968), they

describe a view of temperament wherein environmental and maturational variables

interact with the organism to play a pivotal role in the expression of temperament.

According to Chess and Thomas (1996):

Goodness of fit results when the organism's capacities, motivations, and style of
behavior and the environmental demands and expectations are in accord. Such
consonance between organism and environment potentiates optimal positive
development. Should there be dissonance between the capacities and characteristics
of the organism, on one hand, and the environmental opportunities and demands,
on the other hand, there is poorness of fit, which leads to maladaptive functioning
and distorted development. (pp. 52-53)

The goodness of fit model implies that caregivers who recognize and respect

temperament traits can affect change in children's environments that promote the

healthy development of individual personality (Chess & Thomas, 1996).

Criticisms. Major criticism of Thomas and Chess' approach centers on a

weakness in the theory itself due to consistent failure to replicate their nine dimensional

model of temperament (Slabach, Morrow, & Wachs, 1991). In addition to these

empirical and theoretical issues, more than a few researchers level their criticism on the

poorly structured parental interviews used to derive the NYLS dimensions (Bates, 1980;

Buss & Plomin, 1984; Garrison & Earls, 1987; Goldsmith & Campos, 1982; McNeil,

1976; Persson-Blennow & McNeil, 1979; Rothbart & Derryberry, 1981). Like Slabach

et al., Garrison and Earls (1987) pointed out that these methodological limitations

challenge the essential theory underlying Thomas and Chess' approach. Despite these

criticisms, Thomas and Chess' pioneering work, begun in the late 1950s, continues to

evoke thoughtful consideration among current researchers and is widely regarded as the

foundation for the modem study of temperament in young children (Bates, 1987;

Plomin, 1986; Strelau, 1991; Wilson & Matheny, 1986).

Biological Theory

Eysenck, another of the more outspoken critics of psychoanalysis, employs

stringent statistical methods to arrive at his biological explanation of human behavior.

Partially influenced by Jung, Eysenck describes personality, which he equates with

temperament, along three dimensions: extraversion-introversion (E), neuroticism (N),

and psychoticism (P) (Eysenck, 1952). Eysenck's dimension of extraversion-

introversion is comparable to Jung's description of these terms. Neuroticism is defined

as a general tendency to respond with excess emotion and anxious reactions under

stress. Psychoticism refers to a condition characterized by a loss of contact with reality

marked by episodes of delusions and hallucinations. At the opposite end of the neurotic

and psychotic spectrums are normal personalities. Eysenck proposed that these

dimensions have a complete biological basis, either through "physiological factors,

neurological structure, biochemical and hormonal determinants, or other biological

features of the organism" (Eysenck, 1991, p. 88).

A number of findings support Eysenck's claims that organic constituents

primarily influence these dimensions. First, using the Eysenck Personality

Questionnaire (EPQ; Eysenck & Eysenck, 1975), Barrett and Eysenck (1984)

demonstrated the universality of the dimensions in an extensive cross-cultural study

involving more than 25 countries. The emergence of these dimensions across such a

wide array of cultures provides compelling evidence for biological determinants of

personality (Eysenck, 1991). Further evidence for biological influences is found in

research that indicates a person's position along these dimensions remains relatively

stable over long periods of time (Conley, 1984a, 1984b). Results of this nature are

highly unlikely if everyday influences play a primary role in the formation of

personality. Finally, using modem genetic methods, a number of studies indicate the

likelihood that extraversion, neuroticism, and psychoticism are somewhat inherited

(Eaves, Eysenck, & Martin, 1989).

Criticisms. Criticism of Eysenck's work has been leveled against his extraction

of only three factors from the personality questionnaire data. For example, Cattell

maintained that the demands of the data suggest that there are at least 16 dimensions of

personality. The enormous amount of work Eysenck has produced over the past 50

years has provided many opportunities for further scrutiny. The resulting dialog

between Eysenck and his opponents clearly reflects the controversy associated with

modem conceptualizations and has shown that alternative temperament perspectives are

possible (Strelau, 1998).

Genetic Theory

The contributions of Eysenck as well as Thomas and Chess created a favorable

context for Buss and Plomin's behavioral genetic approach to temperament (Strelau,

1998). In their view, temperaments are "inherited personality traits present in early

childhood" (Buss & Plomin, 1984, p. 84). Based on their review of the literature, they

inferred four broad traits that met their criteria: emotionality (E), activity (A),

sociability (S), and impulsivity (I) (Buss & Plomin, 1975). A decade later they dropped

the impulsivity trait on the grounds that its heritability was not firmly established (Buss

& Plomin, 1984). Accordingly, emotionality is defined as the tendency to become upset

easily and intensely. Activity refers to total energy output and is equated to physical

movement. Sociability is defined as the tendency to prefer the presence of others to

being alone. According to these researchers, EAS "traits are likely to be the foundation

on which later personality traits are built" (Buss & Plomin, 1984, p. 84). Using

behavioral genetics methods including twin, family, and adoption studies, Buss and

Plomin (1975, 1984) provided considerable evidence for the heritability of the EAS


Like Thomas and Chess, Buss and Plomin do not consider temperaments to be

fixed identities. Pointing to the fact that heredity accounts for only a portion of the

variance in temperaments, they postulated that environments also must exert an

important influence in temperament development (Buss & Plomin, 1984). EAS

temperaments in the middle range of the dimensions are conceived to be most

modifiable by the environment. Extreme temperaments tend to affect the environment.

Thus, temperament and environment mutually influence one another.

Criticisms. Criticism of Buss and Plomin's theory center mainly on their claims

regarding the heritability of the EAS traits. Independent researchers investigating the

genetic contribution to the variance of these traits produced only moderate support for

these conclusions (Goldsmith & Campos, 1982; Goldsmith & Gottesman, 1981;

Matheny, 1980). Moreover, contending that some temperament dimensions may be

genetically influenced but not heritable, Goldsmith (1986) questioned Buss and

Plomin's view that heritability should be the defining criteria for temperaments.

Criticisms aside, the methodology behind genetic analysis permits researchers a rare

opportunity to systematically test the assumptions of EAS theory. The use of behavioral

genetic techniques represents a promising field for the exploration of temperament and

other behavioral phenomena.

Developmental Theory

Taking advantage of the growth in research on temperament, Rothbart and

Derryberry used many of the findings as well as their own analyses to elaborate their

developmental approach to the concept of temperament. They define temperament as

constitutionally based, individual differences in reactivity and self-regulation (Rothbart

& Derryberry, 1981). The term constitutionally based refers to relatively enduring

biological aspects of the organism influenced over time by heredity, maturation, and

experience (Rothbart & Derryberry, 1981). The term reactivity is defined as

arousability of the physiological and behavioral systems of the organism. The term self-

regulation refers to processes that act to facilitate or inhibit reactivity including, among

others, attention, approach, avoidance/withdrawal, and selection. With growth, the

modulating affect of self-regulation on reactivity increases as these processes became

more under effortful control. This unique perspective implies that genetic and

developmental determinants lead to affective, motivational, and cognitive influences on

temperament as well. Together these influences are integral in the later development of

adult personality and temperament (Rothbart, 1991).

Criticisms. Rothbart and Derryberry's model has been criticized as too wide-

ranging and difficult to operationalize and investigate (Buss & Plomin, 1984; Garrison

& Earls, 1987; Goldsmith & Campos, 1982; Strelau, 1998). For example, Buss and

Plomin (1984) and Goldsmith and Campos (1982) pointed out that most personality

dispositions meet Rothbart and Derryberry's constitutionally based criteria for

temperament. As many of these behaviors also involve reactivity and self-regulation,

they concluded that the developmental approach does not adequately distinguish

temperament traits from other behaviors. Despite these criticisms, this theory has

continued to gain popularity with child-oriented temperament researchers (Strelau,


Emotion Centered Theory

Goldsmith and Campos' definitions of temperament are the most radically

behavioral (Bates, 1987). Quite unlike the position of the contemporary theorists

presented thus far, Goldsmith and Campos do not consider the biological aspects of

temperament as a definitional component of the construct. Instead, they center their

research on behavioral aspects of the organism, based on the theoretical perspective that

the "temperament phenomena of prime importance during infancy are those which have

socially communicative functions" (Goldsmith & Campos 1982, p. 176). Additionally,

noting that underlying temperaments may become masked by socialization influences

and cognitive processes as children mature, Goldsmith and Campos further narrow their

focus to the infancy period. With these criteria in mind, they describe temperament as

"individual differences in the probability of experiencing and expressing the primary

emotions" (Goldsmith et al., 1987). Primary emotions (e.g., fear, anger, joy, and

excitement) are expressed behaviorally in the infant's facial, vocal, postural, and motor

systems. Thus, temperaments are delineated by individual differences in the temporal

and intensive parameters (i.e., latency, rise time, decay time, magnitude, shape) when

expressing the emotions. Not unlike the views of other contemporary researchers

presented in this section, with the exception of Eysenck, Goldsmith and Campos (1982)

consider the environment to be a powerful contextual mediator in the expression of


Criticisms. Goldsmith and Campos' approach, like Rothbart and Derryberry's

model, has resulted in comparatively few critical remarks (Strelau, 1998). Nevertheless,

their theory has not gone entirely unchallenged. Buss and Plomin (1984) have found it

difficult to operationalize the emotion oriented model. Pointing to the seemingly

arbitrary inclusion of interest as a primary emotion, which they saw as more cognitive

in nature, Buss and Plomin questioned what behaviors would not be examples of

emotions. Strelau (in Matheny, 1986, p. 62) criticized their exclusive focus on

emotional behaviors as too narrow and their assessment techniques as incomplete and

inconsistent (Strelau, 1998). Still, Goldsmith and Campos' focus on socially relevant

behaviors (i.e., emotions) has added to the overall understanding of temperament and

children's development (Bates, 1987).

Regulative Theory

Strelau (1983) developed his regulative theory of temperament based heavily on

Pavlov's discussion of nervous system properties. Strelau (1985) defines temperaments

as relatively stable personality traits revealed in energetic and temporal characteristics

of behavior. Temperament traits are a product of inborn biological mechanisms and are

subjected to slow changes caused by maturation and environmental factors. At the

energetic level, behavior is described along the dimensions of reactivity and activity.

Reactivity refers to differences in the intensity or magnitude of reaction to stimuli and is

determined directly by physiological mechanisms of the organism. Activity refers to

individual differences in the amount of stimulation needed to maintain or achieve the

optimal level for activation. These two dimensions are conceived to be inter-related.

That is, an individual's reactivity determines his or her optimal stimulation load. For

example, a highly reactive individual regulates his or her acute sensitivity and intensity

of reaction with a preference for activities of low stimulative value. Strelau (1998)

devoted less attention to the temporal traits that he describes as individual differences in

the speed, tempo, persistence, recurrence, regularity, and mobility of reactions.

Strelau essentially presents a relationship between temperament and personality

where inborn traits determine the activity level that is aimed at regulating stimulation or

organizing the environment. Over time, these physiologically determined preferences

of activity and environment result in patterns that affect the development of personality

(Strelau, 1985).

Criticisms. A variety of models linking biological processes and temperament

have been postulated (e.g., Strelau, Rothbart and Derryberry, Eysenck). Bates, Wachs,

and Emde (1994) cautioned that, while biobehavioral systems may be relevant, such

relationships remain tentative. Goldsmith and Campos (1986) found Strelau's concept

hard to operationalize based on what they saw as a difficulty of separating energetic

from nonenergetic behaviors. Garrison and Earls (1987) questioned the empirical

validity of Strelau's adult oriented physiological approach, especially as it related to

infants and children. Despite these criticisms, the eastern European orientation of

Strelau's research has served to broaden the international scope of temperament


Student Styles Theory

Working under the premise that "goodness of fit" can be promoted in

educational settings, Oakland postulated that children are best able to develop their

talents and abilities when they and those around them have an understanding of type

theory (Oakland et al., 1996). To that end, Oakland has delineated and operatized a

theory of temperament in the Student Styles Questionnaire (SSQ; Oakland et al., 1996).

Based on the tenets of Jung, Myers and Briggs, and Keirsey, the Student Styles

Questionnaire is designed to detect differences along the dimensions of extroversion-

introversion, practical-imaginative (Myers's sensing-intuitive), thinking-feeling, and

organized-flexible (Myers's judging-perceiving).

The extroversion-introversion (EI) dimension identifies where one acquires

energy. Children with an extroverted style receive energy from and are oriented to the

outer world of people, things, and events. They typically seek interaction with others

and are easy to get to know. Extroverted children also are likely to have many friends

and a wide variety of interests. Conversely, children with an introverted style receive

energy from and are oriented to their inner world of thoughts and ideas. Introverted

children are likely to prefer solitude and private time. They also are likely to have a few

close friends and a small number of interests that they pursue in depth.

The practical-imaginative (PM) dimension identifies whether one attends to

facts or theories. The practical child is likely to pay attention to the real world and

information that comes through one's senses. Children with a practical style typically

take a realistic and pragmatic approach to life. They often pay attention to the present,

value tradition, and work to maintain things that already exist. They often enjoy using

and practicing skills already learned and may become discouraged when things get too

complicated. Children with an imaginative style prefer to attend to the possibilities.

They are likely to be creative and have many original ideas. They also are likely to

prefer theory over facts. Children with an imaginative style are likely to be visionaries

with a keen interest in improving the future.

The thinking-feeling (TF) dimension describes how one makes decisions.

Children with a preference for a thinking style are likely to use objective standards to

make decisions. They also are likely to be critical and have a talent for analyzing

problems. Children with a thinking style typically value logic over sentiment.

Conversely, children with a feeling style typically use personal and subjective standards


to make decisions. They are likely to be sympathetic and have a talent for understanding

relationships. Children with a feeling style are comfortable with emotional issues and

typically value sentiment over logic.

The organized-flexible (OL) dimension indicates when one likes to make

decisions. Children with an organized style are likely to settle things and seek closure in

their affairs. They like to have a plan and often schedule their activities. They typically

are persistent and dependable. Children with a flexible style are likely to put off

decision-making and often are comfortable leaving matters undecided. They adapt to

life as it comes. They also are likely to be flexible in their attitudes and commitments.

See Appendix A for a more comprehensive review of characteristics associated with

each of the four dimensions measured by the SSQ.

Once again, a theory has been postulated that emphasizes the role of the

environment in the expression of temperament. Indeed, a basic assumption of student

styles theory asserts that parents and teachers can nurture children's growth by creating

environments that are consonant with their temperament styles.

The applied nature of student styles theory makes it appealing both to students

and teachers. For example, students may enhance personal development by identifying

talents and possible weaknesses. Social development may be enhanced through self-

understanding and understanding of others. Moreover, student styles theory provides

teachers important information regarding students' learning styles and their optimal

learning environments. Teachers also may acquire a better understanding of their

students' communication patterns, and academic and personal problems in light of the

information provided by student styles theory.

Criticisms. To date, the SSQ has not been subject to critical review by outside

investigators; thus critical commentary does not exist in the literature. However, during

the course of the current investigation, this researcher identified an error in the scoring

protocol for item 10 of the SSQ. Still, extensive factor-analytic research indicates the

SSQ's internal structure is consistent with its theoretical framework and supports use of

the instrument for measuring temperament styles displayed by U.S. children and


Developmental Perspectives

As earlier referenced in this paper, most contemporary researchers agree that

temperament dimensions are biologically rooted determinants of adult personality and

therefore show stability and continuity over time. From a seemingly contradictory

perspective, these researchers also agree that such dimensions are influenced by

contextual factors. Such a view implies that temperament traits can be modified by the

cognitive, social, and emotional qualities experienced throughout childhood and later.

In other words, temperament develops and evolves. The purpose of this section is to

provide an overview of temperament development and to describe the functional

significance of individual differences in temperament dimensions.

Rothbart's (1989) extensive review of the literature reported the developmental

sequence of temperament dimensions from the newborn period through childhood.

According to Rothbart, the newborn period is marked by the emergence of individual

differences in negative emotionality (i.e., reactivity), activity level, distractibility,

attention span (i.e., task persistence), and approach-withdrawal. Early infancy is

associated with all components of newborn temperament and the emergence of

individual differences in positive emotionality (i.e., reactivity). The period of late

infancy is associated with all components observed in earlier periods as well as the

emergence of inhibition of approach, sociability (i.e., introversion-extraversion), and

self-regulation (i.e., effortful control). During the preschool to late adolescent years,

temperaments are refined through experience and practice and resulted in the adult


The preceding topical review provides insight into the emergence of

temperament dimensions. However, it is incomplete without a discussion of the

fundamental significance of individual differences along temperament dimensions and

later development.

Activity and development. "Activity level is a dimension of individual

difference that appears as a primary component of most scales of infant and child

temperament... in large measure because it is a highly salient feature of child

behavior" (Eaton & Enns, 1986, p. 19). As a cornerstone dimension of temperament,

several studies in temperament research have focused exclusively on activity level and

developmental implications (Buss, 1981; Buss, Block, & Block, 1980; Eaton, 1994;

Eaton & Enns, 1986; Halverson & Waldrop, 1976). For example, Buss et al. (1980)

found that activity level at ages three and four is associated with motor and behavioral

activity at age seven. More specifically, in the motor domain, highly active preschool

children, in contrast to their same age peers, are more physically active, energetic, and

fidgety at age seven. Highly active preschoolers also are less likely to be inhibited and

physically cautious in adolescence. Interpersonally, at age seven, they are more likely to

stretch limits, take advantage of others, try to be the center of attention, compete with

others, assert themselves, and display physical and verbal aggression. They also are less

likely to be obedient or shy than their same age peers. Finally, at age seven, highly

active preschoolers are less likely to keep thoughts and feelings to themselves, delay

gratification, and think ahead. These results are consistent with findings that show

preschool activity is positively correlated with dominance in interpersonal relationships

and independence at age seven (Halverson & Waldrop, 1976).

A relationship between activity and intellectual ability also has been reported

(Halverson & Waldrop, 1976). Specifically, high activity during the preschool period is

negatively correlated with scores on both verbal and performance scales of the

Wechsler Intelligence Scale for Children at age seven. In a similar study, a relationship

between activity and intelligence was not confirmed (Czeschik, 1993). However,

significant overlap between high intelligence and temperament dimensions of low

distractibility and high persistence were reported. Academic achievement and scholastic

ability are negatively correlated with activity and distractibility and positively correlated

with persistence (Martin, 1989).

Relationships between children's' activity level and parental behaviors also have

been examined (Buss, 1981). Parents with preschoolers who display low activity levels

are more resourceful, responsive, and successful in aiding their children than parents of

preschoolers who display high activity levels. The parents low active preschoolers also

are described as more patient, supportive, and encouraging. Conversely, parents of

highly active preschoolers have a greater tendency to get in power struggles, intrude

physically into tasks, and have more difficulty establishing good working relationships

with their children. Impatience, frustration, and hostility also often mark parent

interactions with highly active children. While problematic activity levels tend to

decrease between ages four and eight (Achenbach, 1978), some relationships inevitably

result in poorness of fit, derived from the discrepancies in the characteristics of the child

and environment. Temperament risk factors predisposing a child to a poor fit increase

the probability of behavior disorders and maladapted personality (Carey, 1989; Thomas

et al., 1968).

Inhibition and development. The temperament roots of shyness (i.e., social

inhibition) and its developmental implications have been explored (Eisenberg, Shepard,

Fabes, Murphy, & Guthrie, 1998). Internalized negative emotions at ages 4 through 6

are predictive of shyness at ages 6 through 8 and again at ages 8 through 10. Also, shy

children's' proneness toward internalizing negative emotion predicts low instrumental

coping (e.g., reluctance to seek support from teachers). These results are consistent with

earlier findings that show most children classified as extremely shy and emotionally

subdued (e.g., inhibited) at age 2 retain their classification through age 8 (Kagan, 1989;

1994b). Kagan (1994b) also suggested that about one-third of inhibited children are at

risk for social phobia and anxiety in adolescence and extreme introversion in adulthood.

Interestingly, extremely shy children report fears, such as speaking in class and

attending summer camp (Kagan, 1989). Such fears might developmentally precede the

characteristics of adult shyness (i.e., anxiety, social fear, loneliness, suspicion, and

resentment) described by Eisenberg, Fabes, and Murphy (1995).

NYLS dimension of temperament and development. The developmental change

and stability of the NYLS dimensions from infancy to adolescence have been reported

(Guerin & Gottfried, 1994). Developmental change during this period is dimension and

time specific. That is, more change occurs during the period prior to the entry of

kindergarten (i.e., age three through five) than during the period after entry (i.e., age

eight through twelve). More specifically, in contrast with three-year-olds, five-year-olds

are more rhythmic, adaptable, positive in mood, and persistent. A developmental

direction toward milder intensity of reaction and lower activity level also is found at age

five. Mean changes in middle childhood, though fewer in number, also are found. Older

children show a developmental direction toward less activity, intensity, and reactivity

and greater approachability. Interestingly, despite these developmental changes, cross-

sectional (i.e., time) stability was found along all nine dimensions as assessed by the

relative rank of children across time (Guerin & Gottfried, 1994).

Easy/Difficult temperament and development. Temperament researchers also

have focused on the developmental implications of children identified with easy or

difficult temperaments. Two related studies (Kom, 1984; Chess & Thomas, 1984)

reported that easy or difficult temperament at age three is associated with easy or

difficult temperament in young adulthood (i.e., age 18-22). That is, easy children tend to

remain easy in early adult life while difficult children tend to remain difficult. Chess

and Thomas' study also showed that infants with difficult temperament have greater

potential for an improper fit than do infants with other temperament characteristics.

Accordingly, difficult temperaments at age three predispose inadequate adjustment,

including behavioral and psychiatric disorders, in early adulthood (Chess & Thomas,


Similarly, the difficult temperament cluster is a predisposing factor for an

unsuitable fit in school settings (Keogh, 1989). Students described by their teachers as

low in persistence and adaptability, high in distractibility and activity, and high in

reactivity also are less likely to be described as teachable. Moreover, these students

reportedly have significantly less positive social interactions with their teachers (Keogh,


A transcultural replication of the above findings regarding the predictive power

of easy or difficult temperaments also has been reported (Maziade, Cote, Boudreault,

Thivierge, & Boutin, 1986). The most temperamentally easy and difficult 7 year-olds

from a large Quebec sample were targeted. At age 12, temperaments were reassessed.

Results were consistent with earlier findings that showed continuity of easy and difficult

styles. However, results also showed "parental behavior control may be a protective

factor against the risk associated with difficult temperament" (Maziade et al., 1985, p.

945). That is, superior functioning families characterize their difficult children as less

difficult at age 12, whereas dysfunctional families are at increased risk for

temperamentally difficult children with possible psychiatric disorders at age 12.

A related study explored relationships between extreme difficult temperament

and psychiatric disorder in children between ages 3 and 12 referred for psychiatric

service (Maziade, Caron, Cote, Boutin, & Thivierge, 1990). Results showed an over-

proportion of extremely difficult temperaments in the referred population. Clinical

problems among these children centered on externalized disorders (i.e., opposition,

conduct or attention-deficit) and specific developmental delays.

Origins of behavior problems and competencies. The temperament origins of

adolescent behavior problems and competencies were explored in an extensive

longitudinal study (Caspi, Henry, McGee, Moffitt, & Silva, 1995). Lack of control (i.e.,

difficult), sluggishness (i.e., slow-to-warm-up), and approach dimensions were

measured at ages 3, 5, 7, and 9 and compared with reports of behavior problems at ages

9, 11, 13, and 15. Findings showed that children characterized by a lack of control in

early childhood are more likely to exhibit antisocial behavior, conduct disorder,

hyperactivity, and attention problems in adolescence. Early childhood sluggishness in

girls also is associated with later attention problems. Finally, children characterized as

high on approach later are described by their parents as having more personal strengths

than those characterized by a lack of control and sluggishness in early childhood. The

researchers postulated that a possible explanation for these behavioral connections was

that extreme temperament characteristics in early childhood are early manifestations of

adolescent behavior problems. Person-environment interaction patterns that tend to

reinforce behaviors across the course of life also were identified as a possible link

between temperament and later behavior.

A related study (Caspi & Silva, 1995) explored the predictive relationship of

temperament qualities in early childhood and personality traits in young adulthood.

Results demonstrated that behavioral styles at age 3 are associated with distinct

personality differences at age 18. More specifically, preschoolers identified as

undercontrolled describe themselves as danger seeking and impulsive at 18. Inhibited

preschoolers later describe themselves as nonassertive, cautious, and submissive. Both

reserved and inhibited preschoolers score low on social potency in early adult self-

reports. Confident children describe themselves as impulsive at age 18. Finally, children

identified as well adjusted at age 3 describe themselves as average young adults at age

18. The preservation of behavioral styles across time was tentatively attributed to

genetic factors and the previously mentioned person-environment interaction patterns.

Temperament and the Big Five structure of personality. A chapter by Martin,

Wisenbaker, and Huttunen (1994) provides highly speculative links between NYLS

based temperaments in childhood and the Big Five structure of adult personality. The

Big Five are considered to provide a comprehensive taxonomy of personality

characteristics and have been labeled as follows: extraversion, agreeableness,

conscientiousness, neuroticism, and openness to experience (McCrae, Costa, & Yik,

1996). Extraversion refers to individual differences in preferences for social interaction

and lively activity. Agreeableness refers to individual differences in the degree to which

one displays selfless concern for others. Conscientiousness refers to individual

differences in organization and achievement. Neuroticism refers to individual

differences in proneness for experiencing unpleasant and disturbing emotions and

thoughts. Openness refers to individual differences in receptiveness to new ideas and

experiences (McCrae & Costa, 1990).

Assuming that temperament characteristics may be understood as early-

appearing personality traits, Martin et al. (1994) postulated that childhood activity is a

strong predictor of extraversion in adulthood. Pointing out that the extraversion cluster

of the Big Five structure of adult personality includes the components of activity,

energy, and vigor, Eaton (1994) also concluded that childhood activity predicts

extraversion in adulthood.

Social inhibition in childhood is speculatively related to both the extraversion

and neuroticism factors of the Big Five. The adult neuroticism factor also is thought to

be associated with the childhood temperament dimensions of adaptability and negative

emotionality. Emotionality in childhood also is conjectured to predict agreeableness in

adult personality. Finally, childhood task persistence is hypothesized to predict

adulthood conscientiousness. While these conjectures have little theoretical and

empirical foundation, Strelau (1998) noted that relationships between the Big Five and

temperament will become one of the more thoroughly investigated areas of personality

research in coming years.

Student Styles dimensions of temperament and development. A study of

students age 8 through 17 identified several developmental trends in the expression of

temperament (Thayer, 1996). First, extroversion increases with age. That is, students

age 12 through 17 display a stronger preference for extroversion than do students age 8

through 11. Interestingly, students age 12 through 15 display the greatest preference for

extroversion; preferences for extroversion level off and decrease after age 15.

Older students (i.e., age 16 and 17) also exhibit a higher preference for feeling

than do younger students. Additionally, students age 14 through 17 exhibit a higher

preference for flexible styles than do students age 8 through 11. Finally, students age

10 through 13 exhibit a higher preference for imaginative styles than do 8-year-old


MBTI and type development. The theory driving the MBTI is explicit with

regard to type development. "Type theory assumes that type does not change over the

life span" (Myers et al., 1998, p. 28). According to Myers and Myers (1980), children

are born with type predispositions and are motivated to exercise these preferences. As

children became more skillful and adept in their use of preferred functions, opposite


functions are relatively neglected. In adolescence and adulthood, individuals continue to

exercise and refine their use of the preferred functions. Finally, during midlife,

individuals are motivated to complete their personality by "gradually adding the

previously neglected tertiary and inferior functions to the sphere of operation....

however, this did not include a change in a person's type" (Myers et al., 1998, p. 28).

Gender Perspectives

While gender is recognized as an important influence on child and adolescent

development and socialization (Macoby & Jacklin, 1974), males and females generally

are more similar than different with respect to childhood and adolescent temperament.

Nevertheless, gender differences in temperament dimensions exist albeit are somewhat

rare and often are restricted to differences in activity level (Kohnstamm, 1989; Persson-

Blennow & McNeil, 1981; Rutter, 1982; Prior, Smart, Sanson, & Oberklaid, 1993). As

Chess and Thomas (1984) stated, "the findings from the NYLS and from other centers

abroad are consistent in the absence of striking sex differences in temperament" (p. 94).

Gender differences in temperament are discussed below. The focus is on relevant to the

current investigation.

NYLS dimensions of temperament and gender. Macoby and Jacklin's (1974)

research regarding human activity level is a useful starting point. Their research showed

that boys are more active than girls after the first year of life. Activity is one of the few

dimensions along which gender differences have been documented consistently.

Maziade, Boutin, Cote, and Thivierge's (1986) extensive review of gender differences

in research utilizing the NYLS model of temperament found boys, ranging from ages 1

to 12, are more active in seven of eight studies (including three of their own). These

results are consistent with another major review (Eaton & Enns, 1986) of 127 studies

reporting on activity and activity-like dimensions (e.g. vigor in play, energy level,

restlessness). Studies included in the review consistently showed that males display

higher activity levels and activity-like dimensions. "Expressed in yet another way, the

average male subject (at the 50h percentile) is more active than 69% of the female

subjects" (Eaton & Enns, 1986, p. 22).

Few studies have found gender differences along other NYLS dimensions.

Differences, when found, were moderately congruent with other findings. For example,

males age 8 through 12 are reported to exhibit higher sensory threshold together with

less adaptability, predictability, and persistence than females (Hegvik, McDevitt, &

Carey, 1982). Another study found males at age 7 to be more negative in mood and

higher in sensory threshold than females (Maziade, Cote, Boudreault, Thivierge, &

Caperaa, 1984). Related studies found males ages 9 through 12 to be less predictable

and persistent and higher in sensory threshold than females (Maziade et al., 1985;

Maziade et al., 1986). Moreover, males at age 7 are more likely to be more difficult

than females (Maziade, Cote et al., 1984). These results contradict Kom's (1984)

finding that the number of males and females in the extreme quartiles for easy or

difficult temperament at ages 1 through 5 and during the teen and young adulthood

periods are approximately equal. Korn also reported that, during the teenage period,

females with easy temperament and males with difficult temperament are predictable

based on their temperament profiles from earlier ages.

Gender differences on adaptability were reported in a ten-year longitudinal study

of children ages 2 through 12 (Guerin & Gottfried, 1994). The findings are inconsistent

with others reported in NYLS-based research such as those described above and led

Guerin and Gottfried to conclude "that gender differences do not characterize the

construct of temperament with respect to the issues of developmental change and

stability during childhood" (p. 351).

CBQ dimensions of temperament and gender. Reports of gender differences in

temperament studies not employing questionnaires based on the NYLS model are even

more rare yet exist. Gender differences among 6 and 7 year-olds on the surgency and

effortful control scales of the Child Behavior Questionnaire (CBQ; Rothbart, 1988),

designed to measure Rothbart and Derryberry's construct of temperament, have been

reported (Ahadi, Rothbart, & Ye, 1993). On the surgency scale, boys are higher on

activity. On the effortful control scale, boys are lower on inhibitory control, low

intensity pleasure, and perceptual sensitivity. Similar findings were reported by

Eisenberg et al. (1998) in their investigation of the relationship between children's

shyness and regulation. Their CBQ results show boys at ages 7, 9, and 11 are viewed as

lower in behavioral regulation (i.e., self-control and inhibition control). Findings also

show boys at ages 7 and 11 are lower on attention shifting and attention focusing.

Although low attention shifting and focusing have been associated with adults' reports

of shyness (Eisenberg et al., 1995), no significant gender differences for shyness were

found at ages 7, 9, and 11.

Adolescent behavior problems and gender. In their exploration of the

temperament origins of child and adolescent behavior problems, Caspi et al. (1995)

found that boys high on approach in early childhood are less likely to experience

distress and anxiety in later years. Girls high on sluggishness in early childhood are

more likely to suffer from distress, anxiety, and attention problems in later years.

Student Styles dimensions of temperament and gender. Gender differences have

been reported on dimensions of temperament measured by the Student Styles

Questionnaire (Oakland et al., 1996). The greatest differences were found on the

thinking/feeling dimension. More specifically, 72% of females prefer a feeling to a

thinking style. On the other hand, 64% of males prefer a thinking to a feeling style.

Smaller gender differences were found on the organized/flexible dimension. Fifty-nine

percent of females, compared with 42% of males, prefer an organized style. Fewer

gender differences appeared on the extroverted/introverted and practical/imaginative

dimensions. Sixty-four percent of females and 66% of males prefer the extroverted

style. Similarly, 64% of females and 67% of males prefer the practical style.

A study of students ages 8 through 17 examined relationships between gender

and temperament dimensions as measured by the SSQ (Thayer, 1996). These results

are consistent with earlier observations. More specifically, in contrast to female

students, male students exhibit a higher preference for practical, thinking, and flexible

styles. Male and female students do not differ in their preferences for extroversion and


MBTI research and gender. The best current estimates of adult MBTI type

distributions in the U.S. (Hammer & Mitchell, 1996) found gender differences on all

dimensions. The most striking is on thinking/feeling. More specifically, 69% of males

prefer thinking to feeling. On the other hand, 61% of women prefer feeling to thinking.

Other gender differences were less sizable. Fifty-four percent of males and 53% of

females report a preference for extroversion. Sixty-four percent of males and 71% of

females report a preference for a practical style. Fifty-five percent of males and 61% of

females report a preference for an organized style. Despite inconsistencies with

previous estimates, Hammer and Mitchell suggest these norms provide a current and

close approximation to a nationally representative sample and recommend their use

whenever type data are needed.

Purposes of the Study

A large portion of the extant temperament research has focused on temperament

variation among infants and young children. Indeed, research with older children and

adolescence is virtually nonexistent. The primary purpose of this investigation is to

examine and describe possible changes and continuity in the prevalence rates of

temperament preferences displayed by U.S. male and female students ages 8 through


The Student Styles Questionnaire provides a synergy of dimensional and

categorical approaches to temperament measurement (Oakland, Glutting & Horton,

1996). It can yield categorical data in that one can sort individuals into one or another

classification group (e.g., extrovert or introvert). It can yield dimensional data in that

one can utilize the data to estimate the strength of a temperament trait an individual

possesses relative to others.

While norming the SSQ, careful consideration was taken to accurately record

the percentage of each of the temperament traits occurring in the general population.

Population prevalence rates were estimated based on an extensive review of related

research and instruments measuring related constructs such as the MBTI (Oakland et

al., 1996). This review showed that the base rate for an extroverted style is

approximately 65% and 35% for an introverted style. Similarly, the base rate for a

practical style is about 65% and 35% for an imaginative style. The base rates for

thinking-feeling and organized-flexible styles are approximately 50%. Support for these

base rates were found in the preliminary analyses of the SSQ data. Based on this

information, the authors of the SSQ developed a nontraditional, prevalence-based T-

score system in which score of T50 was set in such a manner that the percentage of

children scoring above score of T50 on each scale would match each dimension's

assumed prevalence rate in the general population (Oakland et al., 1996). For example,

65% of children in the standardization sample score above a score of T50 in the

prevalence-based T-score system and therefore are classified as extroverted. Thirty-five

percent of children in the sample score below a score of T50 and are classified as


The use of a nontraditional prevalence-based T-score system is not without

precedent. The Millon Index of Personality Styles (MIPS; Millon, Weiss, Millon, &

Davis, 1994), an instrument designed to measure personal styles of normally

functioning adults, also uses such a scaling procedure.

While the decision to use a prevalence-based T-score system is empirically

supported, the use of population prevalence rates obtained from research with adults

(e.g., MBTI, MIPS) may be insufficient for describing prevalence rates among children.

It is plausible that the prevalence of temperament traits in children are different than

prevalence rates found among adults.

Therefore, the current investigation is descriptive in nature and attempts to

delineate accurately the temperament prevalence rates among children by circumventing

prevalence-based T scores generated by the SSQ. The procedures used in this process

are discussed in the methods section.

Research Questions

This study addresses the following questions:

Question 1: What are the prevalence of extroversion-introversion styles and do they
change as a function of age and gender?

Question 2: What are the prevalence of practical-imaginative styles and do they change
as a function of age and gender?

Question 3: What are the prevalence of thinking-feeling styles and do they change as a
function of age and gender?

Question 4: What are the prevalence of organized-flexible styles and do they change as
a function of age and gender?



Self-reports of temperament preferences displayed by U.S. male and female

students ages 8 through 17 were utilized. Data were obtained during the norming of the

Student Styles Questionnaire (SSQ). The standardization sample was designed to reflect

1990 U.S. Bureau of the Census data (Oakland et al., 1996) and was stratified along five

categories: age, gender, race/ethnicity, geographic region, and school type. The

participant pool of 7,902 students included a total of 10 age groups ranging from 8

through 17 years (sample sizes of 531, 648, 932, 1,174, 1,119, 1,162, 921, 639, 409, and

367 at ages 8, 9, 10, 11, 12, 13, 14, 15, 16, and 17, respectively) and was representative

of both male and female genders (sample sizes of 3,950 and 3,952 respectively). Four

racial/ethnic groups were represented, including 5,547 Whites, 1,194 African

Americans, and 868 Hispanics. The fourth group, labeled "Other", included 293

students whose racial identity was unknown. The subjects were drawn from four

geographic regions including the Northeast, North Central, South, and West. The

sample included students from public and private schools in 61 school districts from 29

U.S. states and Puerto Rico.


The SSQ is a 69 item temperament scale patterned after the Jungian constructs

popularized by Myers-Briggs Type Indicator. The SSQ may be administered to

individuals or groups.

The SSQ was administered to 7,902 students between ages of 8 through 17 (see

Appendix B for detailed administration directions). The paper and pencil test requires

approximately 20 to 30 minutes to complete. Each of its 69 forced-choice items has two

alternatives that fall at either extreme of one of the four primary dimensions (i.e.,

scales): extroversion (E) introversion (I), practical (P) imaginative (M), thinking (T)

- feeling (F), and organized (0) flexible (L). Among the 69 items, 23 are included on

the El scale, 16 on the PM scale, 10 on the TF scale, and 26 on the OL scale.

Additionally, 6 items provide information simultaneously for two scales.


Three methods can be used to interpret information yielded by the SSQ. The

first method focuses on each of the eight basic and dichotomous styles, the second

focuses on combinations of two styles, and the third focuses on combinations of four

styles. For purposes of the present investigation, interpretations based on the eight basic

styles are utilized and have been previously described.


The reliability of a measure refers to the degree to which its scores are free from

errors of measurement. Most theories of personality suggest that personality traits are

somewhat stable throughout development. Therefore, reliability estimates that examine

stability are most suitable for measures of temperament and personality (Kamphaus &

Frick, 1996). Thus, test-retest methods commonly are used for evaluating an

instrument's stability.

Test-retest reliability of the SSQ was examined by administering the SSQ to a

sub-sample of the 7, 902 students included in the standardization phase. One-hundred

thirty-seven ethnically diverse male and female students ranging from ages 8 through 17

completed the SSQ twice. The interval between the two SSQ completions was 7

months. Reliability coefficients for the EI, PM, TF, and OL dimensions are .80, .67, .70,

and .78 respectively. An average reliability coefficient of .74 is obtained when Fisher's

z transformation is applied. Results of these analyses suggest SSQ scores are

satisfactorily reliable. The reliability coefficients may be quite satisfactory given the

long period between the two administrations of the SSQ.


Validity refers to the degree to which a test measures what it purports to

measure. Validity constitutes the single most important psychometric characteristic of a

test (Messick, 1989). Careful consideration was taken to establish both the internal and

external validity of the SSQ throughout its development and standardization.

Internal validity. A series of factor analyses were used to establish the internal

validity of the SSQ. Initial factor analyses assessed the degree to which a pool of 245

original items loaded on the four basic scales hypothesized for the SSQ. Results of this

analysis yielded 100 items that loaded appreciably on one or more of the four factors.

These items were subsequently subjected to a second series of factor analyses during

the standardization phase of the SSQ. First, dichotomous item scores were forced into

the hypothesized four-factor solution using squared multiple correlations as the

communality estimate. The solution was evaluated using two criteria; each item show

an appreciable loading of 2 .20 on its hypothesized factor and < .10 of the items load on

more than one factor. Results of item factoring yielded 69 items that satisfied both

criteria and guided the formation of four to five item parcels used in the second

analysis. Items with similar factor loadings were grouped into 17 parcels and

correlations among item parcels were submitted to factoring. Each parcel showed an

appreciable loading on its hypothesized factor (i.e., 2 .40) and < 10% of the parcels

loaded on more than one factor. Finally, the factoring of the parcels was repeated 17

times, leaving one of the parcels opened during each factoring. Such analysis allowed

for direct examination of whether the individual items in each open parcel loaded on

their hypothesized factor. Results of all factor analyses met the internal validity

requirements specified by the substantive-construct model of instrument development

(Cronbach & Meehl, 1955).

Construct validity. Two studies examining the construct validity of the SSQ for

comparisons of racial-ethnic groups have been conducted since publication of the SSQ.

The first study (Stafford & Oakland, 1996b) used confirmatory factor analyses to

determine whether the 17 item parcels created during standardization of the SSQ are

statistically similar for African American, Hispanic American, and Anglo American

samples. Coefficients of congruence calculated for the three comparisons showed a high

degree of congruence (i.e., .90 to .99).

The second study (Stafford & Oakland, 1996a), similar in nature, examined the

construct validity of the SSQ for the comparison of African American, Hispanic

American, and Anglo American when age is considered a factor. Specifically, the study

used factor analyses to examine the factor structure of the SSQ across the three racial-

ethnic groups for children ages 8 though 10, 11 through 13, and 14 through 17.

Coefficients of congruence for the three comparisons were between .75 and .94, .87 and

.99, and .82 and .98 for the 8 to 10, 11 to 13, and 14 to 17 age groups respectively.

Results of these studies suggest the constructs of temperament appear to be similar

across racial-ethnic groups and the use of the SSQ with these groups is supported.

External validity. External validity of the SSQ was established using contrasted

groups, convergent validity, and divergent validity approaches. The contrasted groups

approach involved analysis of groups selected from the SSQ standardization sample

based on their vocational preferences, class preferences, leisure activity preferences,

special program involvement, and self-evaluation. Each groups' scores on the SSQ's

four scales were contrasted with scores for all other students in the sample. Students

with different interests in vocations, personal beliefs, and other personal qualities were

expected to perform differently on one or more of the SSQ scales. Analysis of variance

and Tukey post hoc comparison results (significance ofp < .01 and p < .05 respectively)

showed that styles indicated by the SSQ were related to personal preferences indicated

by students and generally confirm the theoretical foundation of the instrument.

Convergent validity of an instrument is established when the measure correlates

highly with constructs with which it is hypothesized to have a strong relationship. Two

studies were undertaken to assess the convergent validity of the SSQ. The first study

evaluated the relationship between the SSQ and the Values Inventory (VI; Oakland,

1990). The VI is a close-ended 10 item self-report instrument designed to measure the

importance of values such as helpfulness and loyalty in children's daily lives. A sample

of 134 ethnically diverse male and female students ranging in age from 8 to 17 years

was utilized for this study. Results of univariate comparisons indicate significant

correlations exist between SSQ dimensions and relevant VI counterparts. For example,

significant positive relationships were found between the helpfulness dimension of the

VI and SSQ dimensions of practical-imaginative (r = .28, p < .001) and organized-

flexible (r = .39, p < .001). This suggests that students who prefer a practical style are

more likely to value being helpful than those with an imaginative style. Similarly,

students who prefer an organized style to a flexible style also are more likely to value

being helpful. Additionally, significant negative relationships were found between the

SSQ dimension of thinking-feeling and the helpfulness (r = -.25, p < .01) and loyalty (r

= -.35, p < .001) dimensions of the VI. That is, students who prefer a feeling style to a

thinking style are more likely to value being helpful and loyal. Moreover, results of

multivariate comparisons between the SSQ and VI yielded significant correlations

(multiple R: helpfulness = .54, p < .001; multiple R: loyalty = .38, p < .001). The

appreciable correlations found in both the univariate and multivariate analyses indicate

that there is substantial convergent validity between children's values and temperament

qualities indicated by the SSQ.

A second study of convergent validity evaluated relationships between the SSQ

and the Myers-Briggs Type Indicator. The sample was composed of 99 White and

African American male and female students. This study was confined to older students

who were eligible to take both SSQ and MBTI, therefore students ranged in age from 12

through 17 years. Since these instruments purport to measure similar qualities, high

correlations between the SSQ scales and similar MBTI dimensions were expected.

Results of univariate comparisons showed appropriate convergent validity

across all four scales. More specifically, the SSQ extroverted-introverted, practical-

imaginative, thinking-feeling, and organized-flexible scales correlated appreciably with

their MBTI counterparts of extraversion (r = .62, p < .001) and introversion (r = -.65, p

< .001), sensing (r = .40, p < .001) and intuiting (r = -.29, p < .01), thinking (r = .36, p <

.001) and feeling (r = -.25, p < .05), and judging (r = .22, p < .05) and perceiving (r =

.26, p < .01). Additionally, canonical correlation analysis was conducted to better reveal

the nature of multivariate correlations between the SSQ and MBTI. Canonical analysis

showed a statistically significant set of relationships between the instruments (Wilks'

Lambda = 2.99; approximate F[4, 91] = 4.53; p < .001). Three out of four canonical

variates yielded relatively strong canonical correlations (i.e., .48, .23, and .20) and were

retained. The averaged square canonical correlation was used to assess the degree of

multivariate overlap between the SSQ and MBTI. Scores from the SSQ captured 30%

of the total variability in the MBTI and suggest that SSQ measures the same general

temperament types indicated by the MBTI. Results of the univariate and multivariate

analyses support the convergent validity of the SSQ and indicate that both instruments

assess important temperament characteristics.

Divergent validity of an instrument is established when the measure correlates

poorly with constructs with which it is hypothesized to be unrelated. Four studies were

undertaken by the authors to assess the divergent validity of the SSQ (Oakland,

Glutting, & Horton, 1996). These studies evaluated the degree to which students' scores

on the SSQ were influenced by their academic achievement as measured by the

California Achievement Test (CAT; CTB/McGraw-Hill, 1985) and Metropolitan


Achievement Test, Sixth Edition (MAT6; Prescott, Barlow, Hogan, & Farr, 1985). The

first study examined relationships between the SSQ and the CAT and included a sample

of 208 ethnically diverse male and female students ages 8 through 15. Univariate

correlations between the four SSQ scales and the CAT Total Reading, Total

Mathematics, Total Language, and Overall Battery scales were calculated with

significance for the correlations set at p < .05. Of the 16 correlations, 14 yielded non-

significant values. The highest univariate correlation was found between the organized-

flexible and Total Mathematics scales and was quite low (r = -.16). Multiple regression

analyses were conducted to obtain multivariate correlations between the instruments.

The highest multivariate correlation was obtained between the SSQ and the Total

Mathematics scale (r = .18) and also was low.

A second study replicated the first CAT study with an independent sample of

687 ethnically diverse, male and female students ages 7 through 17. Results of

univariate and multivariate analyses yielded correlations that were generally lower than

those reported in the first study. Similarly, results of a third study showed univariate and

multivariate analyses between the SSQ and MAT6 also were quite low.

A fourth study was undertaken to establish divergent validity between the SSQ

and children's intellectual ability as measured by the Wechsler Intelligence Scale for

Children Revised (WISC-R; Wechsler, 1974). The sample was composed of 86

ethnically diverse male and female students ages 8 through 17. Univariate correlations

between the four SSQ scales and the WISC-R Full Scale IQ, Verbal Scale IQ, and

Performance Scale IQ were calculated. All 12 correlations were non-significant (p >

.05). Similarly, none of the multivariate correlations between the instruments were

significant. Findings of this nature indicate substantial divergent validity for the SSQ

and demonstrate that students' academic achievement and intellectual ability are

unlikely to affect performance on the SSQ and interpretations derived from SSQ scores.

Data Analyses Procedures

Prior to statistical analyses, the percentage of children expressing a preference

on the eight basic SSQ styles was calculated. These values were obtained first by

analyzing individual responses to the 69 SSQ items. For example, among the 16 items

that assess the practical-imaginative style, students who select a larger number of

practical options were labeled practical. Conversely, students selecting a larger number

of imaginative options on such items were labeled imaginative. Students with no

preference (i.e., those that select 8 practical options and 8 imaginative options) were not

included in subsequent analyses on that scale. Finally, for each sample, the number of

students displaying a particular preference on each scale was divided by the total

number of students displaying any preference on the scale. To illustrate, let us assume

560 of 648 U.S. 9-year-olds sampled display a preference on the El scale. Among these,

302 students age display a preference for introversion. Thus, 53.9 % of 9-year-olds

would be labeled introverted and 46.1 % would be labeled extroverted. Percentages of

style preferences displayed by students at ages 8 through 17 are presented in Table 1 in

the results section.

A series of data analyses was used to describe the data. One study using logistic

regression procedures, a form of multiple regression that can be used with dichotomous

dependent variables, was undertaken. The study was designed to detect age, gender, and

age-by-gender interaction effects. Thus, this series of analyses used two independent


variables (i.e., age and gender) to describe developmental trends. Conceptually, this can

be thought of as a model that considers two regression equations, one for each gender,

displayed over the ten age groups.

Analysis was facilitated by visually examining line graphs, one for each of the

four temperament scales depicting one line each for the percentage of males and

females displaying preferences for a particular temperament style at each of the 10 ages.

Model fitting began with the development of a regression equation that included the

highest order polynomial term necessary for describing the developmental trends found

in either male's or female's temperament preferences on each of the four temperament

dimensions, thus ensuring that a sufficient number of age-by-gender interaction terms

were included in the model. Specifically, age-by-gender interaction terms

corresponding to each polynomial term included in the proposed model were added to

the data set and the regression equation. For example, for extroversion, if visual analysis

of the line depicting male preferences for extroversion shows one turn while the line

depicting female preferences shows two turns, then a cubic polynomial term would be

added to the data set and model, even though the line for males suggests the need for

only a quadratic term. Additionally, interaction terms corresponding to the linear,

quadratic, and cubic polynomial terms then would be added to the data set and

regression equation.

The resulting model was tested using the likelihood-ratio approach to statistical

testing to determine if its degree of complexity was necessary for describing the

features of the data. First the interaction terms were dropped from the model to

determine their utility. The difference between the likelihood statistics generated by the

full and reduced models being compared was computed. The resulting likelihood ratio

was compared to a chi-square critical value. For the analyses described in this and the

results section, the critical value for significance at the .05 level varied depending on the

number of interaction terms dropped from the full model. Comparisons between two

models yielding likelihood ratios greater than the critical value suggest that the full

model with interaction terms should be retained and the analysis discontinued.

Conversely, likelihood ratios less than the critical value suggest the complex model

should be disregarded in favor of the reduced model without interaction terms.

If the reduced model described the data more accurately, analyses continued by

working backward, removing the highest order term from the model and comparing the

resulting two models to determine if a more accurate model for describing the data

could be obtained.

Thus, continuing with the example, the cubic term would be dropped from the

model to produce a quadratic model and the resulting two models would be compared.

If the cubic model was determined to describe the features of the data more accurately

than the quadratic model, then the cubic model would be retained and the analysis

discontinued. However, if the quadratic model more accurately described the data, then

cubic model would be discarded in favor of the quadratic model. Finally, the quadratic

model was compared to the linear model by dropping the quadratic term.

The models were compared using the likelihood-ratio approach to statistical

testing. First, the difference between the likelihood statistics generated by the two

models being compared was computed. The resulting likelihood ratio then was

compared to a chi-square critical value. For the analyses described in this and the results


section (i.e., without interaction terms), the critical value for significance at the .05 level

with one degree of freedom is 3.84. Thus, comparisons between two models yielding

likelihood ratios greater than 3.84 suggest that the more complex model should be

retained. Conversely, likelihood ratios less than 3.84 suggest the complex model should

be disregarded in favor of the more simple model.

While logistic regression are useful primarily for describing the major features

of the data (i.e., trends and effects), chi-square analyses may further elucidate the

significance of possible differences in temperament preferences displayed as a function

of age and gender. Therefore, a final series of analyses employed chi-square tests

conducted at each age level. A significance level of.05 was set for all analyses. Thus, 2-

values equal to or less than .05 indicated the groups are statistically different on a

temperament dimension. A summary of the research questions, sample sizes, and

significant relationships for chi-square tests are provided in Tables 2 through 5 of the

results section.


Question 1: Does the prevalence of extroversion-introversion styles change as a

function of age and gender?

Logistic regression for the extroversion-introversion dimension. A complex

model including polynomial terms up to a quadratic and quadratic interaction terms first

was modeled. Interaction terms then were dropped from the model. Differences

between the likelihood statistics generated by the full and reduced models indicate that

the full or complex model should be retained:

10789.025 10771.127 = 17.89; 2dfcv = 5.99; p < .05.

Thus, logistic regression procedures indicate an age by gender interaction effect.

Graphically (see Figure 1), the effect may be represented best by two non-parallel lines

with intersecting concave curves, one for each gender, with both female and males

preferences for an extroverted style increasing between ages 8 and 13 and decreasing


Chi-square analyses for age on the extroversion-introversion dimension. On the

El dimension, differences were not significant between ages 8 and 9 (X2 =.183, p >

.05). Preferences for extroversion increased between ages 9 and 10 (X2 = .002, p < .05).

Differences were not significant between ages 10 and 11 (x2 = .105, p > .05).

Preferences for extroversion increased between ages 11 and 12 (x2 = .003, p < .05) and


12 and 13 (X2 = .000, p < .05). Preferences for extroversion decreased between ages 13

and 14 (X2 = .000, p < .05). Differences were not significant between ages 14 and 15

(x2 = .395, p > .05) and 15 and 16 (x2 = .469, p > .05). Preferences for extroversion

decreased between ages 16 and 17 (X2 = .006, p < .05).

Chi-square analyses for gender on the extroversion-introversion dimension. A

higher percentage of 8-year-old and 9-year-old males preferred the extroverted style

compared to 8-year-old (X2 = .000, p < .05) and 9-year-old (x2 = .000, p < .05) females.

Gender differences were not significant for 10-year-olds (X2 = .351, p> .05), 11-year-

olds (X2 = .733, p > .05) and 12-year-olds (X2 = .918, p > .05). A lower percentage of

13-year-old males preferred the extroverted style compared to 13-year-old females (X2

= .019, p < .05). Gender differences were not significant for 14-year-olds (x2 = .599, p

> .05), 15-year-olds (x2 = .601, p > .05) and 16-year-olds (X2 = .138, p > .05). A higher

percentage of 17-year-old males preferred the extroverted style compared to 17-year-old

females (X2 = .000, p < .05).

Question 2: Does the prevalence of practical-imaginative styles change as a function of

age and gender?

Logistic regression for the practical-imaginative dimension. A complex model

including polynomial terms up to a quartic and quartic interaction terms first was

modeled. Interaction terms then were dropped from the model. Differences between the

likelihood statistics generated by the full and reduced models indicate that the full or

complex model should be rejected in favor of the reduced model:

9239.520 9231.161 = 8.359; 4df cv = 9.49; p > .05.

The reduced model, or quartic without interaction terms, then was compared to one

without a quartic term (i.e., cubic model). Differences between the likelihood statistics

indicate that the quartic term is unnecessary for describing the data:

9239.754 9239.520 = .234, Idfcv = 3.84; p > .05.

Next, the cubic term was dropped from the data. Differences between the likelihood

statistics indicate that cubic model is necessary for describing the features of the data:

9256.062 9239.754 = 16.308; Idf cv = 3.84; p < .05.

Thus, logistic regression procedures do not indicate an age by gender interaction.

Graphically (see Figure 2), preferences may be represented best by one line showing a

significant increase in a preference for an imaginative style between age 8 and 10. A

significant return toward a balanced style is found between age 10 and 15. Preferences

for an imaginative style increase again between the age of 15 and 17.

Chi-square analyses for age on the practical-imaginative dimension. Preferences

for a practical style decreased between ages 8 and 9 (X2 = .000, p < .05) and 9 and 10

(2 = .023, p < .05). Preferences for a practical style increased between ages 10 and 11

(X2 = .022, p < .05). Differences were not significant between ages 11 and 12 (x2

.778, p > .05) and 12 and 13 (x2 = .634, p > .05). Preferences for a practical style

increased between ages 13 and 14 (X2 = .013, p < .05). Differences were not significant

between ages 14 and 15 (x2 = .452, p > .05), 15 and 16 (2 = .079, p > .05), and 16 and

17 (x2 =.543, > .05).

Chi-square analyses for gender on the practical-imaginative dimension. Gender

differences in preferences for the practical/imaginative style were not significant for 8-

year-olds (x2 = .257, p> .05). A higher percentage of 9-year-old males preferred the

practical style compared to 9-year-old females (X2 = .043, p < .05). Gender differences

were not significant for 10-year-olds (X2 = .147, p > .05), 11-year-olds (x2 = .108, p >

.05), and 12-year-olds (X2 = .054, p > .05). A higher percentage of 13-year-old males

preferred the practical style compared to 13-year-old females (x2 =.000, p < .05).

Gender differences were not significant for 14-year-olds (Q2 = .156, p > .05) and 15-

year-olds (X2 = .595, p > .05). A higher percentage of 16-year-old males preferred the

practical style compared to 16-year-old females (X2 = .018, p < .05). A lower

percentage of 17-year-old males preferred the practical style compared to 17-year-old

females (x2 = .012, p <.05).

Question 3: Does the prevalence of thinking-feeling styles change as a function of age

and gender?

Logistic regression for the thinking-feeling dimension. A complex model

including polynomial terms up to a quadratic and quadratic interaction terms first was

modeled. Interaction terms then were dropped from the model. Differences between the

likelihood statistics generated by the full and reduced models indicate that the full or

complex model should be retained:

7831.020 7821.722 = 9.298; 2df cv = 5.99; p < .05.

Thus, logistic regression procedures indicate an age by gender interaction effect.

Graphically (see Figure 3), the effect may be represented best by two declining non-

parallel lines, one for each gender, showing females' preference for a thinking style

declining more rapidly over time than males.

Chi-square analyses for age on the thinking-feeling dimension. Preferences for a

thinking style decreased between ages 8 and 9 (X2 = .027, E < .05). Differences were

not significant between ages 9 and 10 (x2 = .892, p > .05), 10 and 11 (x2 = .560, p >

.05), 11 and 12 (X2 = .986, p > .05), 12 and 13 (X2 = .406, 2 > .05), 13 and 14 (x2 =

.745, p > .05), 14 and 15 (X2 = .248, p > .05), 15 and 16 (X2 = .096, p > .05), and 16 and

17 (2 = .204, p > .05).

Chi-square analyses for gender on the thinking-feeling dimension. Significant

gender differences were found at all age levels. A higher percentage of males at all ages

preferred the thinking style compared to same age females at ages 8 (2 = .000, p <

.05), 9 (x2 = .000, p < .05), 10 (x2 = .000, p < .05), 11 (X2 = .000, p < .05), 12 (2 =

.000, p < .05), 13 (2 = .000, < .05), 14 (X2 = .000, p < .05), 15 (2 = .000, p < .05),

16 (X2 = .000, p < .05), and 17 (X2 = .000, p < .05).

Question 4: Does the prevalence of organized-flexible styles change as a function of age

and gender?

Logistic regression for the organized-flexible dimension. A complex model

including polynomial terms up to a quadratic and quadratic interaction terms first was

modeled. Interaction terms then were dropped from the model. Differences between the

likelihood statistics generated by the full and reduced models indicate that the full or

complex model should be retained:

8381.165 8375.072 = 6.093; 2dfcv = 5.99; p < .05.

Thus, logistic regression procedures indicate an age by gender interaction effect.

Graphically (see Figure 4), the effect may be represented best by two declining non-

parallel lines, one for each gender, showing males' preference for an organized style

declining more rapidly than females between age 8 and 15. Girls consistently show

stronger preferences for an organized style. A significant return toward an organized

style is found between age 15 and 17.

Chi-square analyses for age on the organized-flexible dimension. Preferences for

an organized style decreased between ages 8 and 9 (x2 = .000, p < .05). Differences

were not significant between ages 9 and 10 (x2 = .096, p > .05). Preferences for

organized style decreased between ages 10 and 11 (X2 = .000, p < .05), 11 and 12 (2 =

.000, p <.05), 12 and 13 (x2 = .002, p < .05), 13 and 14 (x2 = .001, p <.05), and 14 and

15 (x2 = .008, p < .05). Differences were not significant between ages 15 and 16 (2 =

.549, p > .05). Preferences for an organized style increased between ages 16 and 17 (x2

=.003, p < .05).

Chi-square analyses for gender on the organized-flexible dimension. Significant

gender differences were found at all age levels. A lower percentage of males at all ages

preferred an organized style compared to same age females at ages 8 (x2 = .004, p <

.05), 9 (X2 = .000, p <.05), 10 (x2 = .000, p < .05), 11 (2 = .000, p <.05), 12 (2 =

.000, p < .05), 13 (x2 = .000, p < .05), 14 (x2 = .000, p < .05), 15 (x2 = .000, p < .05),

16 (x2 = .000, p < .05), and 17 (x2 = .000, p > .05).

Table 1

Preference frequency percentages by age and gender
(N) E I P M T F O L

Male 8
Female 8
Total 8
Male 9
Female 9
Total 9
Male 10
Female 10
Total 10
Male 11
Female 11
Total 11
Male 12
Female 12
Total 12
Male 13
Female 13
Total 13
Male 14
Female 14
Total 14
Male 15
Female 15
Total 15
Male 16
Female 16
Total 16
Male 17
Female 17
Total 17

(265) 49
(266) 38
(531) 44
(324) 52
(324) 40
(648) 46
(466) 52
(466) 50
(932) 51
(587) 54
(587) 53
(1174) 54
(560) 58
(559) 58
(1119) 58
(581) 62
(581) 66
(1162) 64
(460) 57
(461) 59
(921) 58
(319) 57
(320) 56
(639) 56
(204) 52
(205) 57
(409) 54
(184) 54
(183) 40
(367) 47
(3950) 56
(3952) 54
(7902) 55

45 42 58 52 48 70 30


Note. N = Sample Size; E = Extroverted; I = Introverted; P = Practical; M = Imaginative; F
= Feeling; O = Organized; L = Flexible

Table 2

Summary of chi-square results for extroversion/introversion dimension
Age comparisons

8-year-olds to 9-year-olds
9-year-olds to 10-year-olds0
I0-year-olds to 11-year-olds
11-year-olds to 12-year-olds *
12-year-olds to 13-year-olds *
13-year-olds to 14-year-olds *
14-year-olds to 15-year-olds
15-year-olds to 16-year-olds
16-year-olds0 to 17-year-olds

Gender comparisons

8-year-old males0 to 8-year-old females
9-year-old males0 to 9-year-old females
10-year-old males to 10-year-old females
11-year-old males to 11-year-old females
12-year-old males to 12-year-old females
13-year-old males to 13-year-old females0
14-year-old males to 14-year-old females
15-year-old males to 15-year-old females
16-year-old males to 16-year-old females
17-year-old males0 tol7-year-old females

Note. denotes p < .05. denotes the group that is higher on Extroversion.

Table 3

Summary of chi-square results for practical/imaginative dimension
Age comparisons

8-year-olds0 to 9-year-olds*
9-year-olds0 to 10-year-olds
10-year-olds to 11-year-olds0
11-year-olds to 12-year-olds
12-year-olds to 13-year-olds
13-year-olds to 14-year-olds0
14-year-olds to 15-year-olds
15-year-olds to 16-year-olds
16-year-olds to 17-year-olds

Gender comparisons

8-year-old males to 8-year-old females
9-year-old males0 to 9-year-old females
10-year-old males to 10-year-old females
11-year-old males to 11-year-old females
12-year-old males to 12-year-old females
13-year-old males0 to 13-year-old females
14-year-old males to 14-year-old females
15-year-old males to 15-year-old females
16-year-old males0 to 16-year-old females
17-year-old males to 17-year-old females *

Note. denotes p < .05. o denotes the group that is higher on Practical.

Table 4

Summary of chi-square results for thinking/feeling dimension
Age comparisons

8-year-olds0 to 9-year-olds
9-year-olds to 10-year-olds
10-year-olds to 11-year-olds
11-year-olds to 12-year-olds
12-year-olds to 13-year-olds
13-year-olds to 14-year-olds
14-year-olds to 15-year-olds
15-year-olds to 16-year-olds
16-year-olds to 17-year-olds

Gender comparisons

8-year-old males0 to 8-year-old females
9-year-old males0 to 9-year-old females
10-year-old males0 to 10-year-old females
11-year-old males0 to 11-year-old females
12-year-old males0 to 12-year-old females
13-year-old males0 to 13-year-old females
14-year-old males0 to 14-year-old females
15-year-old males0 to 15-year-old females
16-year-old males0 to 16-year-old females
17-year-old males0 to 17-year-old females

Note. denotes p < .05. o denotes the group that is higher on Thinking.

Table 5

Summary of chi-square results for organized/flexible dimension
Age comparisons
8-year-olds0 to 9-year-olds
9-year-olds to 10-year-olds
10-year-olds0 to 11-year-olds *
11-year-olds0 to 12-year-olds *
12-year-olds0 to 13-year-olds *
13-year-olds0 to 14-year-olds *
14-year-olds0 to 15-year-olds *
15-year-olds to 16-year-olds
16-year-olds to 17-year-olds0
Gender comparisons:
8-year-old males to 8-year-old females0 *
9-year-old males to 9-year-old females0 *
10-year-old males to 10-year-old females0 *
11-year-old males to 11-year-old females0 *
12-year-old males to 12-year-old females0 *
13-year-old males to 13-year-old females0 *
14-year-old males to 14-year-old females0 *
15-year-old males to 15-year-old females0 *
16-year-old males to 16-year-old females0 *
17-year-old males to 17-year-old females0 *

Note. denotes p < .05. o denotes the group that is higher on Organized.




C C 40


| >
o 309

o 20 1

0 10


-- Male


-T . . T T ... r r


- --- I


Figure 1. Percentage ofchildren who display an extroverted style across ten age groups.
Age differences are signified by solid lines. Gender differences are signified by an asterisk.

--- Female
--- Total




E 20




-r---- -------i-- -F--~ ~~1~~ -- I


Figure 2. Percentage of children who display a practical style across ten age groups.
Age differences are signified by solid lines. Gender differences are signified by an asterisk.

T- I

-- Male
-*- Female

--- "---

80 1

70 i

D- 60



"- 30




16 17


Figure 3. Percentage of children who display a thinking style across ten age groups.
Age differences are signified by solid lines. Gender differences are signified by an asterisk.


-- ------f

--- Female


a ,












20 -



* *

16 17

Figure 4. Percentage of children who display an organized style cross ten age groups.
Age differences are signified by solid lines. Gender differences are signified by an asterisk.

* *

8 9

* *

10 11 12 13 14 15


Preferences for an Extroverted or Introverted Style

Preferences for extrovertion and introversion are impacted by age and gender.

Graphically, the effect may be represented best by two non-parallel lines with

intersecting concave curves, one for each gender, with preferences for an extroverted

style increasing between age 8 and 13 and decreasing thereafter for both females and

males. Gender differences in El preferences are evident at only four of the ten age

levels; 8, 9, 13, and 17. Thus, the most notable finding may be that initial preferences

for a somewhat balanced perspective for extroversion and introversion seen in early

childhood gives way to a significant shift toward a more extroverted style in later

childhood, asymptoting during the early teens. A preference for an extroverted style

peaks at age 13 and returns toward a more balanced preference for extroversion and

introversion at later ages.

From a classic psychoanalytic viewpoint, the psychosocial goal of adolescence

is the severance of emotional attachment and dependencies from parents (Freud, 1958).

Detachment makes possible mature autonomy and later attachment to extrafamilial

objects. From a Freudian perspective, this process leads to repudiation of parents and an

orientation toward peers.

While childhood and adolescence may be marked by a shift away from the

influence of parents toward greater peer influence (Ambert, 2001), psychoanalytic

viewpoints for explaining this phenomenon have been downplayed over the past half



century due to a lack of empirical support for the theories as well as the development of

alternative interpretations. For example, adolescents may turn away from the family and

toward peers for companionship because peers are seen to be more socially and

emotionally supportive than parents (Furman & Buhrmester, 1992).

Another alternative suggests that adolescents turn to outside social networks and

activities because parents encourage them to do so. According to typical Western

definitions of adjustment, children are considered to be well functioning and normal

when they are eager to invest their energy in active and independent exploration of their

immediate physical and social environment (Matas, Arend, & Sroufe, 1978). Thus,

Western families generally expect children to become physically and psychologically

separated from parents (Steinberg & Silverberg 1986) and encourage children to acquire

a repertoire of social skills, experiment with interpersonal relationships, assume new

behaviors, and engage in new experiences (Higgins & Parsons, 1983). Moreover, the

role of the parents is not necessarily replaced by a peer system during this period;

rather, the child may move quite comfortably between the two worlds (Brown, Mounts,

Lamborn, & Steinberg, 1993). In other words, adolescence may represent another stage

in life, one in which understanding of family has been established and understanding of

peers becomes better rooted.

Regardless of ideology, each of the models described above suggests that an

expanding social network marks the periods of childhood and adolescence. Both the

number of peers and time spent interacting with peers increase during childhood and

adolescence. Some report a linear increase in daily social contact with peers between

early childhood and middle childhood (Feiring & Lewis, 1989). At age 3, children

reportedly have an average of 22 people in their networks beyond their immediate


family members. Although peer network size did not change at age 6, a 70% increase in

number of peers was observed between ages 6 and 9; by age 9 the average number of

peers increased to 39 people. The trend toward increasing peer contact continues into

late childhood as preadolescents spend more time with friends and develop a larger

number friends (Bee, 1995; Feiring & Lewis, 1991). By adolescence, more time is spent

interacting with peers than with other social partners (e.g., parents, teachers). Teenagers

generally spend more than half of their waking hours with their peers (Berndt, 1999;

Csikszentmihalyi & Larson, 1984). Notably, this period also is marked by an increase in

cross-gender interactions. This may be a result of the discovery of one's sexuality and

burgeoning interest in romantic relationships (Berndt, 1999, Crockett, Losoff, &

Petersen, 1984).

The proclivity toward larger social networks and more frequent social

interactions through childhood and adolescence may be adaptive. Erikson (1968)

believed that self-concept is forged by exploring a variety of identities, a process that

requires shifting one's crowd membership several times in order to sample a range of

provisional identities. Sullivan (1953) also emphasized a critical role played by peers in

one's self development. A central theme of his theory is that peer interactions provide

opportunities for validation of self-worth and a unique context for exploration and

development of personal strengths. Given the value of peer relationships during

adolescence, some previously introverted children prefer extroverted qualities and thus

seek more peer orientation and friendships with peers with whom they can interact on

an equal footing.

Sullivan further proposed that peer relations may be a primary source from

which arise critical characteristics of the mature personality. Indeed, peer interactions

are conducive to the development of a host of adaptive social behaviors including an

understanding of fairness, proclivities toward sharing and kindness, mastery of

symbolic expression, perspective-taking and communication skills, and skills in

negotiating and managing conflict (Damon & Phelps, 1989; LeMare & Rubin, 1987;

Parker, Rubin, Price, & DeRosier, 1995).

Peer interactions also are predictive of later emotional and social adjustment.

More specifically, the sociability of children ages 7 through 12 is predictive of behavior

and social competencies 7 years later. Elementary school students who were relatively

socially isolated are more likely to display less suitable social skills and various kinds of

psychological disturbance as adolescents (Morrison & Masten, 1991). In addition,

withdrawn children are less likely to independently generate solutions to hypothetical

social problems and display more rigidity in generative alternative solutions to

problems (Rubin, 1982).

The possible restructuring of one's peer network to include a larger number of

peers and more frequent interactions with them also may provide a functional value in

school settings. As students transition from elementary to middle schools, they typically

move from self-contained classrooms to larger, constantly shifting classes with many

classmates they have not seen before (Brown, 1989). To the extent that peers provide

mutual support and guidance (Moran & Eckenrode, 1991), increased interaction with

peers in different groups may help them navigate these changes. Indeed, for many,

being a member of a group is more important in early and middle adolescence (i.e.,

seventh through tenth grades) than at prior or later ages (Furman, 1989).

Research with infrahumans also suggests that social contact is essential to

normal development (Harlow, Dodsworth, & Harlow, (1965). For example, in social


animals, the effects of social isolation, diminished social contact, and modified patterns

of social contact resulted in less adequate social development, pathological affective

states, inappropriate sexual behavior, and inadequate development of parenting skills.

Results of the current investigation show a return toward a more introverted

style after age 13. Data from other studies of children and youth that may help

corroborate this finding could not be located. However, adult data address this issue.

Renorming data from the MBTI suggest that a trend toward an introverted style may

continue into early adulthood (Hammer & Mitchell, 1996). Adults displayed a greater

preference for an introverted style than do children less than 18 years of age.

Additional research supports the existence of a trend toward a decrease in a

preference for an extroverted style as adults mature. For example, sociability declines

for college freshman reassessed after 14 years (Mortimer, Finch, & Kumka, 1982).

Similarly, research with the Five Factor model of personality also shows that

extraversion, a measure of quantity and intensity of interpersonal interaction, decreases

in adulthood (Costa & McCrae, 1994). These results were replicated in several large

cross-national investigations, thus suggesting that adult personality development, in

part, is due to biologically based maturation (Costa et al., 2000; McCrae et al., 1999).

Theoretical interpretations also support empirical findings that show a return

toward an introverted style. Older adolescents may begin to realize that impending high

school graduation will bring an end to a social world that has helped define their lives

for years (Berndt, 1999). Having negotiated the intricacies of school life and developed

a sense of identity and social competence (Ambert, 2001), being part of a peer group

may become less critical (Furman, 1989). Thus, late adolescence may be marked by a


shift toward finding and building abiding relationships with a few close friends-a mark

of an introvert (Epstein, 1989).

Interestingly, Dunphy (1963) may have captured developmental preferences

along the El dimension in his comprehensive effort to describe structural

transformations in peer groups across adolescence. From observations of several groups

of teenagers, supplemented with interview and diary data, Dunphy identified five stages

of peer group structure. Stage one occurs in early adolescence and is characterized by

unisexual cliques consisting of four to six close friends. In stage two, opposite gender

cliques begin to draw together as a result of the budding heterosexual interests. In stage

three, heterosexual cliques form, and a crowd begins to develop. In stage four, roughly

between the ages of 13 and 15, the fully developed crowd consisting of a close

association of heterosexual cliques emerges. In stage five, toward the end of

adolescence, crowds disintegrate into relatively isolated cliques reminiscent of early

adolescence, except that the groups are heterosexual.

This discussion seemingly couches the El dimension in terms of one's peer

network size and quantity of time spent interacting with peers. While these certainly are

central tenants of an extroverted and introverted style, this may appear on the surface to

be a narrow conceptualization of the El dimension. No doubt, qualities associated with

the El dimension extend beyond social-oriented concepts. However, when considering

that peers often are the ones with whom children discover new enterprises and interests,

assume new behaviors and engage in new experiences, talk and learn about themselves

and others, and develop answers to questions of "who am I?" and "what will I

become?" (Asher, Parkhurst, Hymel, & Williams, 1990; Bagwell, Newcomb, &

Burkowski, 1998; Berk, 1994; Brown, 1989; Feiring & Lewis, 1989; Higgins &


Parsons, 1983), then the extent of the peer network and time spent interacting with peers

may be viewed as a bell weather for assessing preferences on the El dimension.

Preferences for a Practical or Imaginative Style

Gender appears to be a non-pervasive influence on forming preferences for a

practical or imaginative style. Compared to girls, boys are more likely to display

preferences for a practical style at three ages (i.e., 9, 13, and 16) and for an imaginative

style at one age (i.e., 17).

Several relatively short developmental trends are evident in the data. More

specifically, a significant shift from a somewhat balanced perspective toward a more

imaginative style is found between age 8 and 10. A gradual yet significant return toward

a more balanced style occurs between age 10 and 15. Finally, another shift toward an

imaginative style is found between age 15 and 17. While these developmental trends are

intriguing, their significance can be minimized in that comparatively few differences in

PM preferences are found across each of the ten age levels. Therefore, the most notable

finding may be that children and adolescents display a somewhat consistent preference,

albeit at times mild, for an imaginative style.

These results are incongruent with MBTI renorming data that show U.S. adults

prefer a sensing (i.e., practical) style (Hammer & Mitchell, 1996). However, the

renorming results also show that adults displayed greater preferences for a sensing (i.e.,

practical) style than do children less than 18 years of age. Thus, the results of the

current investigation together with the MBTI renorming study suggest a significant shift

from an imaginative style toward a practical style occurs across late adolescence, young

adulthood and into adulthood.

Unfortunately, there appears to be a dearth in related theory and research

addressing the notion of such a developmental trend. However, indications that late

adolescence may be a stage during which young people begin to think less

imaginatively and more realistically about the paths that their futures might take

(Nurmi, 1989) and display a preoccupation with economic well being and societal

expectations (Jessor, 1983) suggest that a shift toward a more practical style in young

adulthood is plausible.

Preferences for a Thinking or Feeling Style

Preferences for thinking and feeling styles are impacted by gender and to a

lesser extent age. Graphically, the effect may be represented best by two declining non-

parallel lines, one for each gender, showing females' preference for a thinking style

declining more rapidly over time than that for males. The prevalence of a thinking style

among males generally is stable between ages of 8 and 17. In contrast, age level

differences in the direction of an increasingly stronger preference for a feeling style are

evident among females. More specifically, among females, 58 percent show a

preference for a feeling style at age 8 while approximately 81 percent show a preference

for a feeling style at age 17.

The most notable finding is that males consistently display a strong preference

for a thinking style while females consistently display preferences for a feeling style.

These results are congruent with cross-national findings that show males also display

stronger preferences for a thinking style than do females among Australian and Chinese

adolescents (Bassett, 2001; Oakland, Faulkner, & Bassett, in press). Moreover, these

results also are consistent with the renorming data from the MBTI that show adult males

prefer a thinking style whereas adult females prefer a feeling style (Hammer &

Mitchell, 1996).

Some evidence suggests biological factors may explain these preferences.

Genetic coding, with males displaying greater relative strength and lung capacity, may

make them better suited for the competitive struggle to obtain resources necessary for

survival. In contrast, females, with childbearing and milk-production capacities, may be

predisposed to nurturing roles (Whyte, 1998). Brain differences also may help explain

these gender differences. Women's better verbal expression of emotions may be

explained by the fact that the right hemisphere of the female brain, which controls

emotions, is better connected to the left side of the brain, which controls verbal

expression, than is the case for men (Moir & Jessel, 1989). In contrast, females greater

reliance on their feelings may have helped develop this hemispheric difference.

Differences in emotional reactions of men and women also have been attributed to

differential hormonal responses (Fischer, 1998).

However, most sociobiologists are likely to acknowledge that environmental

factors also may help explain these behavioral differences among males and females

(Whyte, 1998). Indeed, psychoanalytic and socialization perspectives provide an

interesting forum for explaining these gender preferences. According to Freud, the

establishment of gender identity is a major developmental milestone in early childhood.

By age 3, almost all children identify themselves as either a boy or girl (Pogrebin, 1980:

Thompson, 1975) and begin to develop organized beliefs regarding qualities that

constitute femaleness and maleness (Bem, 1981). In Western cultures, masculinity is

assigned attributes of competitiveness and self-reliance. Femininity is assigned

attributes of emotional expressiveness, sensitivity, and nurturance (Ruble, 1983). These

and other gender-related beliefs may strengthen as children evaluate their adequacy

based on conformity to culturally defined role requirements regarding the way they

should think and behave (Ber, 1983). The development of this dichotomous

arrangement in gender-related beliefs and behavior also is evident in Western countries

with more liberal attitudes (Goldman & Goldman, 1982).

Research in child-rearing practices suggests that preferences for a thinking or

feeling style may be learned. According to a feminist version of psychoanalysis,

mothers relate to their sons and daughters differently (Chodorow, 1999). While fusing

identities with their daughters, mothers relate to their sons as being separate and

distinct. Consequently, young males and females often approach relationships

differently; girls are more inclined to adopt styles marked by connectedness while boys

are likely to prefer independence and autonomy.

Mothers talk more about emotions and a greater variety of emotions with

daughters than sons. Moreover, girls talk more about emotion and about a greater

variety of emotions than do boys. Additionally, girls initiate more emotion-related

discussions than do boys (Kuebli, Butler, & Fivush, 1995). Similarly, adolescent boys

have fewer intimate relationships than do girls (Berndt, 1999; Youniss & Smollar,


Social learning and behavioral theories suggest that gender-related personality

styles develop as result of social reinforcement (Bandura & Walters, 1963). Parents

encourage children to engage in activities considered sex appropriate (Lytton &

Romney, 1991). In Western cultures, girls generally receive praise and attention for

their interest in feminine toys or activities while boys are positively reinforced for their

interest in masculine toys or activities (Fagot & Hagan, 1991; Lytton & Romney, 1991).


Children also develop gender-related focal concerns (i.e., specific sets of values,

interests, and problems central to peer culture) (Adler, Kless, & Adler, 1992). Boys'

focal concerns often center on competition, self-reliance, and a culture of coolness and

detachment. In contrast, girls' focal concerns often center on compliance and

conformity, intimacy, emotional expression, and a culture of romantic love.

Adult females tend to be more expressive, emotion oriented, and nurturing while

adult males tend to be more self-reliant, emotionally detached, and competitive. One

large study of college students from 30 countries found that men consistently are

described as adventurous, strong, dominant, assertive, task-oriented, aggressive,

enterprising, and independent. In contrast, women are consistently described as

sensitive, gentle, dependent, emotional, sentimental, weak, submissive, and people

oriented (Williams & Best, 1990). Women's social relationships are more dyadic and

intimate while men's friendships tend to focus on activities of an emotionally detached

nature such as sports or the completion of a task (Parker & de Vries, 1993; Wright,

1982). Women are more empathetic to other people (Eisenberg & Lennon, 1983), are

more willing to discuss their emotional experiences with others (Dindia & Allen, 1992),

and do so more overtly and with more confidence (Fischer, 1998). In short, results

appear to suggest that women emphasize social and emotional interdependence with

other people, whereas men emphasize independence.

Gender differences in public policy attitudes and work values also are evident.

For example, from a public policy perspective, women are more supportive of programs

that enhance people's well-being such as education, health programs, and social welfare

(Eagly, 1987; Schlozman, Burs, Verba, & Donahue, 1995; Shapiro & Mahajan, 1986).

With regard to employment attitudes, women are more likely to attach importance to

jobs that are worthwhile to society and help others (Beutel & Marini, 1995; Marini,


Confirmatory evidence indicating the presence of gender differences in a

preference for a thinking or feeling orientation also can be found in research on

leadership styles. Managers generally are concerned with two aspects of their work. The

first, task accomplishment, concerns organizing activities to perform assigned tasks.

The second, maintenance of interpersonal relationships, focuses on tending to the

morale and welfare of the people in a work setting. Women are more concerned than

men with both the maintenance of interpersonal relationships and task accomplishment

(Eagly & Johnson, 1990).

Scholarship on democratic versus autocratic style further elucidates gender

differences in the leadership orientations of managers. Democratic leadership is marked

by a participative style that encourages the contributions of subordinates in the decision-

making process. In contrast, autocratic leadership, is marked by a directive style that

discourages subordinates from participating in the decision making process. Once again,

women display a more democratic leadership style. More specifically, women are more

likely to use the collaborative style whereas men are more likely to employ a directive


Gender differences in preferences for a transactional or transformational style

also have been investigated (Powell & Graves, 2003). A transactional orientation is

characterized by command and control leadership whereas a transformational

orientation is characterized by shared power and information and enhancement of others

self-worth (Rosener, 1990). Females are more likely to favor transformational

leadership styles than their male counterparts.

Literature on the development of moral judgment also appears to underscore

differences males and females display in their preferences for a thinking or feeling style.

According to Kohlberg (1969), each of the six stages of moral judgment builds on the

reasoning of the preceding stage. Stages 3 and 4 are of particular interest for the

purposes of this study. At stage 3, moral understanding is based on the individual's

desire to promote social harmony. At stage 4, moral judgment encompasses a larger

perspective including societal laws and justice. Kohlberg reported that the average stage

of moral development is 4 for men, and 3 for women (Kohlberg & Kramer, 1969).

Gilligan (1982) is fundamentally dissatisfied with Kohlberg's interpretations

suggesting that the "goodness" of women (e.g., their care for and sensitivity to the

needs of others) leads them to be classified at only stage 3 of moral development. She

argues that Kohlberg places too much emphasis on a justice orientation and too little

emphasis on a caring orientation. In her view, multiple moral orientations are possible

and an ethic of care is different, but no less a valid basis for moral judgment compared

to an ethic of justice. Despite these differences, both Kohlberg and Gilligan agree that,

when making moral judgments, men are more likely to consider principles of fairness

while women are more likely to consider compassion.

At least one other observation regarding gender related preferences for a

thinking/feeling style is noteworthy. When considering U.S. heritage, the notion that

males display a preference for a thinking style is not entirely unexpected. At the time of

its formation, the framers of the U.S. constitution, all men, created a form of

government that stressed uncommon ideals of justice and equality (de Tocqueville,

1835/1984, Potter, 1970).

Preferences for an Organized or Flexible Style

A developmental trend in preferences for an organized-flexible style is evident.

More specifically, a significant shift from an initially strong preference for an organized

style toward a more balanced style perspective is evident between the age of 8 and 15.

However, a significant return to a preference for an organized style is observed between

the age of 15 and 17.

One observation is especially noteworthy. At every age, the majority of students

show a preference for an organized style. These findings are consistent with those from

the renorming data of the MBTI that show the majority of U.S. adults prefer a judging

(i.e., organized) style (Hammer & Mitchell, 1996).

One may reference culture-based factors to help explain a preference for an

organized style. From a historical perspective, evidence suggests that a preference for

an organized style, characterized by a strong work ethic and achievement, is valued in

U.S. society. Almost from the start, settlers in America believed that hard work and

education pay off. For example, John Wesley, who founded the Methodist church,

exalted entrepreneurial success and urged his followers "to gain all they can, and to

save all they can; that is, in effect, to grow rich" (Weber, 1904/1958, p. 175). The

American dream has been described as the "pursuit of individual salvation through hard

work, thrift, and competitive struggle" (Whyte, 1956, p. 4).

The achievement orientation also was evident in the frontiers of the West. While

land was inexpensive, the lack of a division of labor forced people to be self-reliant

(Potter, 1970). Francis J. Grund, an Austrian born journalist in the United States, noted

"there is probably no people on earth with whom business constitutes pleasure, and

industry amusement, in an equal degree with the inhabitants of the United States of

America" (Grund, 1837/1971, p. 1).

A strong work ethic remains a central feature of American character. A

dominant view held among Americans today is that leisure time ought to be used in a

productive manner (Yankelovich, 1994) and the United States consistently ranks as one

of the most task-oriented nations in the world (Trompenaars, & Hampden-Turner,

1997). Americans work longer hours than their European and Japanese counterparts

(Economist, 1994a; Bell & Freeman, 1994) and, in contrast to Europe, where status is

attributed by birth, gender or age, Americans are more likely to approve merit-based

reward systems (Trompenaars & Hampden-Turner, 1997).

An achievement orientation also is reflected in the value American's place on

education for success. In contrast to European culture, where welfare and other public

policies are more widely accepted as means for helping less affluent, Americans place

greater emphasis on achieving social mobility through investment in education (Lipset,

1996). The United States spends more of its gross domestic product on education than

do countries in the European Union or Japan (Economist, 1994b) and the United States

historically leads the world in the proportion of young people attending elementary

school, high school, and college (Lipset, 1996). The proportion of 20- to 24-year-olds in

higher education in the United States is almost double that of Japan and the most

affluent European countries (Hampden-Turner & Trompenaars, 1993).

Child-rearing practices may play a role in preferences for an organized style. In

Western cultures, parents admire children when they achieve by their own effort

(Hofstede, 1980; Triandis, 1995). Such ideology likely perpetuates an achievement

orientation from one generation to the next.


As noted previously, adolescents display a significant and gradual shift toward a

preference for a more flexible style between the age of 8 and 15. Similar results have

been reported in cross-national research. Australian and Chinese adolescents

increasingly show a stronger preference for a flexible style between the age of 9 and 15

(Bassett, 2001; Oakland, Faulkner, & Bassett, in press).

Developmental and psychoanalytic perspectives may be useful in explaining this

trend. During preadolescence, when an organized style is most pronounced, children are

most likely to accept parental views because they are viewed as most knowledgeable

and dependable (Kaplan, 1991; Selman, 1980; Smollar & Youniss, 1989; Weissman,

Cohen, Boxer, & Cohler, 1989). Deference to authority during this stage may be

adaptive. An incentive for children to learn to behave in ways that ensure parents'

affection is strong (Sullivan, 1953). Children may experience feelings of love and

acceptance when parental values are internalized and, consequently, work to correct

tendencies or impulses that would incur disapproval or abandonment (Sullivan, 1953).

However, beginning in middle childhood, changes occur in children's

susceptibility to parental influence (Bemdt, 1979). Whether prompted by parents'

attitudes suggesting that they become more autonomous in functioning (Baltes &

Silverberg, 1994; Baumrind, 1971; Maccoby & Martin, 1983; Steinberg & Silverberg,

1986), by psychosocial goals of detaching from parental ties (Erikson, 1968; Freud,

1958; Sullivan, 1953), or the onset of the realization of the fallibility of parents

(Steinberg & Silverberg, 1986), increasing opposition to adult authority marks the

period between childhood and adolescence (Kearney, 1998). Indeed, middle

adolescence has been characterized as a time of acting out and experimentation with

different forms of subjectivity (Kearney, 1998). Adolescents often characterize their

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs