Title Page
 Table of Contents
 List of Tables
 Biographical sketch

Title: Personality integration and the theory of open systems
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00098643/00001
 Material Information
Title: Personality integration and the theory of open systems a cross subcultural approach
Physical Description: viii, 115 leaves : ; 28 cm.
Language: English
Creator: Amerikaner, Martin John, 1950-
Publication Date: 1978
Copyright Date: 1978
Subject: Personality   ( lcsh )
System theory   ( lcsh )
Psychology thesis Ph. D   ( lcsh )
Dissertations, Academic -- Psychology -- UF   ( lcsh )
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
Thesis: Thesis--University of Florida.
Bibliography: Bibliography: leaves 109-113.
General Note: Typescript.
General Note: Vita.
Statement of Responsibility: by Martin Amerikaner.
 Record Information
Bibliographic ID: UF00098643
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.
Resource Identifier: alephbibnum - 000078340
oclc - 04893148
notis - AAJ3639


This item has the following downloads:

PDF ( 4 MBs ) ( PDF )

Table of Contents
    Title Page
        Page i
        Page ii
        Page iii
    Table of Contents
        Page iv
    List of Tables
        Page v
        Page vi
        Page vii
        Page viii
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
    Biographical sketch
        Page 114
        Page 115
        Page 116
        Page 117
        Page 118
Full Text








Many people provided support and help at various

stages of this dissertation, from the discouraging initial

contacts with countless military and seminary institutions,

through my long distance, cross-country finish.

My entire committee was understanding and helpful

well beyond the limits I would have allowed myself.

Dr. Franz Epting, my chairman, was consistently

supportive, and most importantly, available, especially

during my two week write-a-thon.

Dr. Robert Ziller, my co-chairman, provided enthusi-

asm, support, and stimulating conversation, both during the

dissertation project, and throughout my graduate student


Dr. Paul Schauble, my first counseling supervisor

and unofficial "training director," was an important teacher

and collaborator in many ways, and his use of imagery and

metaphor has had continuing growthful impact on my work as

a therapist.

Dr. Harry Grater was continually stimulating, sup-

portive, and challenging in all of our interactions, and I

thank him for each.

Though I worked less closely with Dr. Ellen Amatea,

she was consistently interested and enthusiastic about the

research, and her comments and questions were helpful and


Although many people helped with aspects of the

research, without Cindy Villis' caring and help, it would

probably still not be done, and I might still be in Murphys-


Thanks, too, to Chuck Landis, Director of the Coun-

seling Center at Southern Illinois University, for making

time available to collect the data, and to the administra-

tors and students at the institutions which did participate

in the study, for their time and sincere cooperation.

To Mom and to Robin, thanks for your continuing

love and interest. I would also like to express my appre-

ciation of the love and caring always shown to me by my

late father, Arthur Amerikaner.


ACKNOWLEDGMENTS . . . . . . . . . ii

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

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


I. INTRODUCTION . . . . . . . 1

General Systems Theory and
Personality Integration . . . . 7

Process Constructs . . . . . .. 11

Structural Constructs . . . . .. 15

Time Orientation Constructs . . . .. 18

Content Constructs . . . . . .. 22

Hypotheses . . . . . . . . 31

II. METHOD . . . . . . . . . 35

Subjects . . . . . .. . . . 35

Procedure and Instruments . . . .. 36

III. RESULTS . . . . . . . . . 50

IV. DISCUSSION . . . . ... . . . 79



B. TIME INVENTORY . . . . ... . . 102

C. BIERI REP TEST . . . . . ... 106

REFERENCES . . . . . . ... . . . 109

BIOGRAPHICAL SKETCH . . . . . . . .. 115



Table Page

1. Results of MANOVA: Personality Integration
X Groups X Allport-Vernon-Lindzey Scores 51

2. Results of MANOVA: Groups X Allport-Vernon-
Lindzey Scores . . . . . ... 52

3. Results of MANOVA: Groups X Personality
Integration X Independent Variables . .. 55

4. Means and Standard Deviations for Levels
of PI, Groups, and Total N on Dependent
Variables . . . . . . . . 57

5. T Tests for A Priori, Directional Hypotheses 58

6. ANOVA for Group X Personality Integration
X Complexity of Self-Concept . . ... 59

7. ANOVA for Group X Personality Integration
X FIC Scores . . . . . . . 61

8. ANOVA for Group X Personality Integration
X Ordination . . . . . . . 62

9. ANOVA for Group X Personality Integration
X Openness . . . . .... . . 63

10. ANOVA for Group X Personality Integration
X Self-Esteem . . . . . . . 65

11. ANOVA for Group X Personality Integration
X Experiential Inventory . . . ... 66

12. Contingency Table for Group X Personality
Integration X Time Dominance Categories 68

13. ANOVA for Group X Personality Integration
X PI Subscale (TSCS) . . . . ... 69

14. ANOVA for Group X Personality Integration
X Meaningfulness ..... . . . 71

Table Page

15. Results of MANOVA: Personality Integration
X Groups X Lines Test Variables . . .. 72

16. ANOVA for Group X Personality Integration
X GPA . . . . . . . . .. . 74

17. ANOVA for Group X Personality Integration
X Age . . . . . . . . .. . 75

18. ANOVA for Group X Personality Integration
X Group Tenure . . . . . ... 77

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



Martin Amerikaner

December 1978

Chairman: Franz Epting
Major Department: Psychology

The study of personal growth and high level person-

ality functioning has been a recent focus of study for

humanistically oriented psychologists. Current theoretical

perspectives and empirical work is reviewed from a framework

suggested by general systems theory. The focus is upon the-

oretical and empirical work with the construct of "personal-

ity integration." The generalizability of earlier results

is seen as constrained by limited population sampling and by

measurement shortcomings.

The present research was designed to study personal-

ity integration, conceptualized from an open system perspec-

tive, in three subcultures--military cadets, seminarians,

and college fraternity members, thereby extending research

to a variety of populations.

The study was designed to test hypotheses that highly

integrated subjects would differ from contrast subjects by

exhibiting greater cognitive differentiation and integration,

more complex self-concepts, greater openness to social expe-

riences, and time orientations characterized more by a future

orientation, and less by a past orientation. It was further

hypothesized that these differences would be independent of

value orientation and subcultural group affiliation.

The procedure involved the administration of the

Allport-Vernon-Lindzey Study of Values, the Bieri Role-

Construct REP Test, the Cottle Time Orientation Inventory,

and the Ziller Self-other Orientation Inventory to a sample

of 44 highly integrated and 46 contrast subjects, comprised

of military cadets, seminary students, and fraternity mem-

bers. These were selected on the basis of scores on Duncan's

Personality Integration Reputation Test administered to a

large sample (total N = 226) of subjects from the three


Results indicate support for the hypotheses that

personality integration is not dependent on value orientation

or subcultural identification, and moderate support for the

complexity and openness hypothesis. The time orientation

hypothesis was not supported. Results are discussed in terms

of the system theory position discussed earlier and their

implication for counseling.




Psychologists who have chosen personality growth and

development as their focus of study have often offered con-

ceptualizations of "healthy" or "ideal" functioning from

their particular theoretical framework. Such high level func-

tioning has been variously labeled "self-actualization"

(Goldstein, 1939; Maslow, 1968), "full functioning" (Rogers,

1961), "healthy personality" (Jourard, 1974), "mature per-

sonality" (Allport, 1961), "adequate personality" (Combs,

Richards, & Richards, 1976), and "optimal personality (Coan,

1974), among others. This list of terms is not to suggest

that all theoretical descriptions are equivalent except for

the chosen label, or even that the various theoretical posi-

tions converge in agreement upon the level of personality

functioning for given individuals. Rather, the list of

constructs simply points to the diverse, yet consistent

emphasis this area of study has received in psychology,

especially within the branch which, over the last 20 years,

has been called "humanistic" psychology.

Richness in conceptualizing, however, has been

accompanied by a relatively impoverished bank of research

methods and data. Maslow's pioneering efforts were based


upon interviews with, discrete observations of, and biograph-

ical information published about a small group of people

which fit his vision of self-actualized persons. Though valu-

able and acceptable as an initial approach to research, the

admittedly (Maslow, 1970) tentative and highly subjective

methods of subject selection and data collection needed devel-

opment if research in the larger population would be possible.

The remainder of this introduction will be focused

upon four issues. The first is a brief critique of the major

approach taken to the study of "self-actualization" over the

last 15 years. This will be followed by the presentation

and development of an alternative theoretical position which

is seen as useful for the understanding of personality growth

and for integrating the diverse theoretical positions men-

tioned earlier. The next section will describe a method

which has been developed by Seeman and his associates (Dun-

can, 1964, 1966; Seeman, 1966) to study the construct of

"personality integration" (Seeman, 1959), and the studies

having used this method will be reviewed. The relationship

of Seeman's theoretical position to the open systems theo-

retical approach to be developed here will then be discussed,

leading to a description of the rationale for the research

undertaken in this study and a statement of the specific

research hypotheses to be tested.

The most widely used approach to the study of healthy

personality functioning has been the utilization of


self-report questionnaires, and specifically, the Personality

Orientation Inventory (POI) published by Shostrum (1966).

Since its development there have been literally hundreds of

studies using it to both select samples of self-actualized

subjects and to measure change in level of actualization as a

function of some treatment or intervention strategy. As

Oakland and his colleagues (Oakland, Freed, Loukin, Davis, &

Camilleri, 1978) point out in their recent critique, "a close

look at this test is appropriate" (p. 76), since conclusions

drawn from this body of research rest, at least in part,

upon the validity of the instrument. The Oakland et al.

review points to serious problems with the POI, in areas

ranging from theoretical basis to test construction to reli-

ability and validity data. As an example of the theoretical

difficulties, Oakland et al. note that Maslow specifically

states that, by his criteria, only a small percentage of the

population would be categorized as self-actualized, and that

self-actualization does not occur in young, college age

people. Yet, POI scores for college samples routinely fall

in the self-actualized range, and are used as an independent

variable selection criterion. Problems they cite with test

construction include the ambiguous wording of some items and

the nonexclusivity of supposedly polar opposite item choices.

Additionally, the very structure of the test (forced-choice)

is theoretically questionable, since Maslow's (1968) concep-

tion of self-actualization includes heightened ability to

relinquish "either-or" thinking, to transcend and integrate

dichotomies, and to be flexible in making choices, depend-

ing on the current life context. Ironically, while using a

forced-choice format, one of the POI's own subscales is

designed to measure "flexibility in applying values" (Sho-

strum, 1966).

Oakland et al. (1978) also point to questions sur-

rounding the data which have accumulated as apparent support

for the validity of the POI. A central concern has been the

ability of subjects to "fake" high scores if they have

knowledge of the concepts and language used to describe self-

actualization, and as they note, Braun and LaFaro (1969)

demonstrated that college subjects exposed to relevant con-

cepts can raise their POI scores. Thus, it is at least ques-

tionable whether studies using a pre-post design showing

changes in POI scores following some intervention such as

therapy (Shostrum & Knapp, 1966) or marathon groups (Guinan

& Foulds, 1970) can be said to demonstrate that subjects

have become more self-actualized, or have learned the neces-

sary language to score higher on the instrument. Addition-

ally, both Oakland et al. (1978) and Kay, Lyons, Newman, and

Mankin (1978) point out that relatively poor test-retest

reliability data raises questions of interpretation of

results; when test scores change without any systematic

intervention by a researcher, it is difficult to attribute

post-intervention score changes to the intervention strategy

employed in the research. In sum, the problems surrounding

the use of the POI as a research and clinical instrument are

rather extensive and severe, and argue against its use in

research designed to study the process of personality growth

and development.

A second questionnaire approach to this area of

research should be mentioned here. Coan (1974, 1976) con-

structed a battery of instruments which were designed to

measure aspects of personality functioning thought to be com-

ponents of the "optimal personality." His factor analytic

work investigated the extent to which the data indicated

either an integrated, general dimension of "optimal person-

ality," or autonomous, individual and/or mutually exclusive

personality dimensions which would be incompatible with a

general definition of optimal functioning. Coan (1974)

interprets his results as supporting the latter view; that

is, that no general definition is empirically supported,

since the scores on his battery did not tend to covary.

There are,however, serious methodological shortcomings with

this research. As mentioned in the earlier POI discussion,

most theoretical descriptions of self-actualization suggest

that high level functioning is relatively "rare" and is not

to be expected in relatively young people. Coan, however,

collected his data from undergraduate psychology students,

with no independent measures of adequacy of personal func-

tioning. It is not surprising to find that the different

dimensions did not covary, since this integration of function-

ing would only be expected in a relatively small percentage

of the population. Coan's results, therefore, demonstrate

that in a college population, the dimensions he attempted to

measure do not systematically covary, but this says nothing

about the personality characteristics of independently

selected, highly integrated people. A second, and possibly

more serious criticism of Coan's research concerns the valid-

ity of his instruments. Most of the measures were newly

designed for this study, and yet no independent validational

studies or data are reported; that is, there is no evidence

reported which demonstrates that the instruments measure the

personality dimensions which are claimed (e.g., "openness

to experience," "personal control," "time orientation," etc.).

It appears that his approach was to assume the validity of

his measures, have a large number of students complete the

battery, and report the results of a factor analysis of this

uncertain mass of data. Conclusions and generalization from

such research are confusing and empirically unsupportable.

It is interesting to note, however, that in discussing his

research Coan (1974, 1976) continues to identify two major

dimensions which are important to personal growth. These he

calls being "open" and "ordered," and are conceptually simi-

lar to the "openness to exchange with the environment" and

"hierarchial organization," which will be discussed below as

characteristic of the open system functioning suggested as a

model of personality development.

The focus now shifts from a critique of earlier

methodological approaches to a description and discussion of

personality functioning based upon a model known as general

systems theory. The introduction concludes by reviewing a

research method which is both an alternative to the POI and

is based upon theory related to the open-systems theoretical

position developed here.

General Systems Theory and
Personality Integration

The presentation of systems theoretical constructs

is based upon the work of von Bertalanffy (1967, 1968). After

the major systems constructs are presented and defined, they

will be discussed in relation to current theories of healthy

personality functioning.

From a general systems theory perspective (von Berta-

lanffy, 1967, 1968), people are seen as functioning analo-

gously to other systems, and, optimally, the dynamic qualities

of open systems (as opposed to those of "closed" systems)

would be characteristic of them. In other words, to the

extent that the properties of a well-functioning open system

are applicable to an individual's personality functioning,

then that personality might be seen as a healthy one. Taken

further, a person's areas of "blockage" or nongrowth might

be seen as the ways in which his functioning is no longer

open and has become more like a closed system. The charac-

teristics of open systems will not be discussed as they


pertain to the major conceptual dimensions to be used in dis-

cussing the healthy personality constructs.

When conceptualizing system "process," the nature of

change is of primary interest.

1. An open system is defined as becoming more complex

over time. This process is further defined as follows:

Though remaining identifiable, unique, and integral through

time, the system develops from a more general and homogeneous

to a more specialized and heterogeneous condition. This is

called "progressive differentiation." One aspect of differ-

entiation is a process called "centralization," through which

the action of some components comes to exert greater influence

over the system than do others. This is more fully discussed

below, under "structure."

2. Change is continual, since exchange with the sur-

rounding environment never ceases.

3. Change within one component of the system is seen

as affecting the system as a whole. Structural relationships,

as discussed below, are important here, since some components

are seen as more influential than others.

4. System change is irreversible; one cannot reverse

the process and return to a prior condition.

5. The system may reach a state of development,

called a "steady state," where the system as a whole remains

relatively constant and integral, but there is continual

exchange of component material with the environment.

6. System movement towards this steady state, or

"dynamic equilibrium" (von Bertalanffy, 1968, pp. 131-32), is

marked by increasing self-regulation. This self-regulatory

function is aided by the use of positive and negative feed-

back. As described by Miller (1969) positive feedback is the

use of information by a system to increase distance from a

steady state, or in other words, to initiate change, whereas

negative feedback is information used to maintain or return

to a steady state. Feedback is called "internal" when the

feedback loop never passes outside the system boundaries

(e.g., temperature control in mammals); it is called "exter-

nal" when the loop passes beyond the boundaries of the system

as when an individual asks assistance from another in return-

ing to an equilibrium state ("Will you bring me a glass of

water?"). Whether internal or external, the use of feedback

is seen as important to two of the basic processes of all

organisms--both maintaining a state of relative dynamic

equilibrium and moving towards a more final steady state

(i.e., the process of growth).

The "structural" dimension is primarily concerned

with the ordered relationships between components of the

whole. A primary characteristic of a system is its develop-

ment of a hierarchical structure. Within this structure,

some few components are relatively dominant in their influ-

ence over system behavior, such that a change in one of these

(called a "leading part") leads to large change throughout

the system.


difficulties need not keep a system from ultimately reaching

its final state, and implicitly it is not necessary to go

back to or "undo" previous blockages in order to move forward;

indeed, the irreversibility of change described under "pro-

cess" suggests that this would be an impossible undertaking).

This brief presentation of some major systems theory

concepts develops from a conceptualization of man functioning

as an open system; a unique system, to be sure, with "compo-

nents" (e.g., abilities, physical structures) unlike those

of others, but, on his own level, a system nonetheless. From

this view, the healthy or optimal personality is one which is

most clearly describable by the characteristics of open sys-

tems discussed above, and nonoptimal or dysfunctional states

are those in which the system is "closing down" in some

respects. This conceptualization will now be developed by

moving to a discussion of some major constructs used by psy-

chological theorists in describing healthy personality,

especially those which define their position in relation to

the dimensions outlined above (process, structure, content,

time orientation), and examine how these might fit into the

systems theory framework.

Process Constructs

All theories of optimal personality must deal with

"how" or process questions; therefore, it is not surprising

that many constructs are used by theorists in their attempts

to describe the functioning of healthy persons over time.

Rogers (1961) is clear in his emphasis on the process dimen-

sion and points to the "willingness to experience oneself as

process" as a goal or therapeutic change. Several theorists

describe optimal process as including increasing self-

direction and autonomy and decreasing confluence with, and

dependence on the outside environment. Rogers (1961), for

example, describes the "fully functioning person" as one who

has developed an "internal locus of evaluation," such that

he, rather than aspects of his environment, is the source of

his valuing process. This lack of dependence on others

allows for occurrence of the "organismic valuing process,"

in which organismically experienced needs and satisfactions

come to be the source of values. Similarly, Perls (1969)

discusses maturation as a process of developing from environ-

mental support to self-support or "organismic self-regula-

tion." Maslow's (1968) description of the perceptual

processes of self-actualized people suggests a clear percep-

tion of other people as distinct from oneself, unique and

whole in themselves, and not existing solely to satisfy one's

own deficits. His description of the valuing process is one

in which values emerge from the growing organism, rather than

being simply accepted from the environment.

Another group of constructs used by theorists in the

field seems to emphasize the openness to exchange with the

environment that is characteristic of open systems process.

The term "authenticity," for example, is used by existential

philosophers and psychologists (Ellenberger, 1958; Gendlin,

1973; Jourard, 1974) to describe the process of free choice

of action, rather than automatic responding, in all life sit-

uations. Authentic action thus implies clear perception of

change in the world, openness to the acceptance of change,

and willingness to act in light of change. Jourard (1974)

suggests that "authentic being is a sign of healthy personal-

ity, and it is the means of achieving healthy personality

growth" (p. 168). Rogers' (1961) description of full func-

tioning includes the construct of "openness to experience,"

in which a person is fully aware of all organismic experi-

ences in all situations, without resorting to defensive

distortion or repression to block threatening messages from

the environment. Related to this view are the implications

of Kelly's (1955, 1963) metaphor of "man the scientist."

This process involves hypothesizing, observing, and revising

one's hypotheses. The "optimal scientist"--one who is most

fresh in his hypothesizing, most clear in his observations,

and most willing to drop habitual constructions and revise

them in light of new "data" or new situations. The poor

scientist clings to his hypotheses, selectively sees only

that which is confirming, and closes himself off to revision

and change.

The next sets of constructs are those involving the

process of becoming more complex and differentiated, yet

whole and integral. Regarding complexity, Kelly's (1955)


description of process is again relevant; the good scientist

is continually revising his hypotheses and reconstruing his

situations, and the creation of new constructions seems to

imply an increasingly complex and differentiated world view.

Rogers (1961), too, suggests that the fully functioning per-

son enjoys a "greater richness of life," involving a wider

range and greater variety of experiencing than "the con-

stricted living in which most of us find ourselves" (p. 195).

His construct of "congruence," also characteristic of

healthy functioning, involves the matching of experience,

awareness, and communication into a unified whole, so that

this greater variety in experience can be lived fully.

Another conception of becoming "healthy" or achieving "self-

realization" is that of Jung's (1968; Singer, 1973) process

of individuation. This lifelong process involves growing

awareness of the complexity of one's personality (e.g., the

several archetypical components) and integrating them into a

larger, more unified whole. Again, the focus is on increas-

ing differentiation as well as growthful integration.

The last set of process constructs to be presented

are those which describe the overall functioning of the healthy

individual. In the systems model presented above, it was

noted that open systems often reach a state known as "steady

state" or "dynamic equilibrium," in which the system main-

tains a relatively integral structure, though open exchange

with the environment remains continual. The optimal


personality, when fully matured, can be seen in similar terms:

Maslow's (1968) "self-actualized" person is described as hav-

ing all needs met and a relatively stable, biologically based

value system which allows for free perception of and inter-

action with the environment, yet maintenance of uniqueness

and autonomy. Rogers' (1961) "fully functioning person" has

developed to a point where all aspects of one's organismic

experience are available to awareness, suggesting a mature,

whole, and autonomous functioning, yet quite open to new sit-

uations and experiences. The "self-realized" individual in

Jung's (1968) theoretical system has matured such that his

life processes demonstrate a dynamic balance between polar

opposites, such as ins and outs, hero and victim, extraver-

sion and introversion; this mature person is described as

individuatedd" or autonomous from the collective, though, of

course, not closed off to interaction with others.

Structural Constructs

The basic structural characteristic of the system is

that of "hierarchy." Hierarchical structure suggests that

some components exist prior to, or are more fundamental or

basic to, the system than the others. A related psychologi-

cal position is Kelly's (1955) view of construct systems,

which are made up of relatively core and peripheral (super-

ordinate and subordinate) constructs. Following the descrip-

tion of hierarchical order given earlier, the core constructs

may be seen as holding central importance to the system as


they subsume the more subordinate constructs extending to the

periphery. It does not seem unreasonable to see core con-

structs as analogous to the "leading parts" described in the

earlier "structure" section. Though Kelly does not discuss

construct systems in this way, it may be that healthy or opti-

mal men have systems which have developed hierarchically,

with a relatively stable core structure; change would then

occur principally in the periphery, where less stress would

be exerted to the overall system per "unit change" than would

change in the core. The unhealthy system might be character-

ized by a nonhierarchical structure, such that most any

change or new event is likely to induce stress and shifting

throughout the entire system (i.e., all or most constructs

take on the role of "leading parts").

Other theoretical systems, though not written in quite

this way, can be seen as structuring personality functioning

around certain key elements. Ellis (1973) suggests that dif-

ficulties in living can be traced in the belief systems to a

few central "irrational beliefs." Implied here is that

healthy functioning develops from a core of "rational beliefs,"

the acquisition of which is the goal of his therapeutic pro-

cedures. Berne's (1964) position suggests that healthy or

nonhealthy functioning is basically an elaboration or "life

script" developed from a person's position in respect to the

fundamental dimensions of I'm OK (not OK); you're OK (not OK).

Adler (1964) considered a person's psychological processes,


or "style of life," to be structured around a central "final

goal"; the relative health of the life style is directly

related to the degree of social feeling inherent in the goal.

A more temporal hierarchy, in which certain events precede

and are fundamental to others,is evident in the theoretical

view of Maslow (1968). Using his motivational construct of

the "hierarchy of needs," self-actualization results from the

satisfaction of successively emerging deficit needs, which

are necessary precursors to the emergence of "Being" or

actualization needs. At the risk of stretching a point,

Jung's (1968) description of the psyche may also be seen as

hierarchical, with the collective psyche as fundamental, both

motivationally and temporally (or historically). The personal

unconscious and, in time, conscious sphere acquire greater

influence over the individual's personality system, as more

material from both collective and personal unconsciousness

emerges into consciousness and is "conquered" or integrated

through the individualization process. Though not described

hierarchically, Jourard's (1974) construct of the "self-

structure" places the authentic or "real" self in a key role,

with the health of the personality system being directly

related to the "positioning" or the other self-structures

(public self, self-concept, self-ideal) vis-a-vis the real

self (i.e., as the other structures become congruent with the

authentic self, the system becomes healthier).


Though the nature of their structural theories are in

many respects quite different, the theorists discussed above

share an emphasis on the relationship between components of

the personality system and especially the importance that key

or fundamental components extend over system functioning as a


Time Orientation Constructs

This dimension focuses upon time perspective, and

especially the view of the future. For several theorists,

the relative health of a personality system is directly

related to a person's orientation towards future events. It

will be remembered that systems were described as future-

oriented or goal-oriented; similarly, goal orientation is

prominent in several theoretical systems. The most explicit

example here is that of Adler (1964), whose description of

personality is clearly future-oriented and specifically focuses

upon goals as the key to understanding human functions. As

was mentioned earlier, Adler sees behavior as organized

around the attainment of goals, with one's "fictional final

goals" in particular as organizers of the life style. "Dis-

couragement" results from goals lacking in social feeling,

while health is attainable when one's goals are contributions

to and development of mankind.

For Jung (1968), too, personality function is defined

in primarily teleological terms. Both mankind in general,

and each person in particular, are developing towards

attainment of goal-states. For the individual, self-realiza-

tion is the goal, and one's life is organized around the

processes needed to attain it; the healthy personality is one

which achieves this goal.

Though Maslow (1968) does not include specifically

goal-oriented constructs in his theorizing, the state of self-

actualization is an implicit goal-state for all persons. His

is a biologically based purposiveness, in that as the organ-

ism's needs are satisfied, new ones emerge, and are satisfied,

continually, until the "Being" state is attained. Maslow's

insistence on a biological basis suggests that, given a

sufficiently rich environment, the organism will naturally

develop towards this more fully grown or healthy state.

Kelly's (1955, 1963) position is also future oriented.

His fundamental postulate suggests that all behavior derives

from the ways in which people "anticipate events." However,

he clearly rejects a teleological construction of personality

functioning (1963). Thus, the healthy person lives within

his world based on the most system-elaborative constructions

he can make, and revises his anticipations as needed, though

he is not motivated towards attainment of any particular

goals or state of potential being. Kelly does reject a goal

model of development, although his work suggests that at any

given moment a person's "goal" is to anticipate the future

accurately, and attainment or nonattainment of the goal is

reflected in the degree of revision necessary in the person's

constructing process. Again, though Kelly does not word his

theory in this way, it seems reasonable to suggest that a

long range goal of an individual is to become progressively

better at anticipating events, with "optimal man" the one who

is consistently effective at anticipating the wide range of

events in his life. The development of this anticipatory

ability is not clearly described by Kelly; the "choice corol-

lary" suggests that persons are continually making "elabora-

tive choices," which extend and define their systems. Yet, it

seems that some people are "better at" construing than others;

that is, their systems are more effective at anticipating

events than others--they are functioning more effectively,

in other words, than those who find their lives difficult,

discouraging, or full of despair. The systems perspective

underscores the importance of this issue and offers a con-

ceptual framework for understanding this developmental pro-

cess. It is the relatively open system which is in continual

exchange with the environment and progressively differentiat-

ing, whereas the closed system is not. It may be, then, that

as long as one's construct system is functioning in an open

way, it is operating elaboratively; that is, extending and

defining its capabilities to anticipate one's changing world.

The relatively closed, or "unhealthy," system is then marked

by continual nonelaborative choices, resulting in a constant

effort to construe the world with a limited system, and fail-

ure to extend and redefine in the face of failed anticipation.

It is as if one is "locked in" to a constricted and ineffec-

tive system of construing, and experiences an inability or

unwillingness to "open up" and elaborate or "reconstrue." At

any rate, even if the liberties taken with Kelly's theory are

held aside, it is clear that the time dimension, and espe-

cially one's orientation to the future, is central to the


Another theorist for whom the time dimension is

important is Perls (1969). However, his position regarding

the future seems inconsistent. One of the central themes of

gestalt therapy is that of living "in the here and now,"

involvement with future goals is "rehearsing," and "fantasy"

to be avoided, and only the present moment is to be experi-

enced. However, the theory suggests that all behavior is

oriented towards the future, in that it seeks "completion" or

"closure," and will move in that direction. From this per-

spective, gestalt formation (completion) is the "goal" of

behavior, and anticipation of the optimal ways of attaining

closure would seem to be healthy and growth promoting.

Though Perls' emphasis on bringing a person "out of his head"

and "into his senses" serves a useful therapeutic purpose,

the concurrent practice of idolizing the present seems an

unfortunate by-product. Gendlin's (1973) existential expe-

riential construct of "carrying forward" is revelant here.

Similar to the gestalt concept of completion, Gendlin sug-

gests that change involves the "carrying forward" of any

feeling through its bodily felt continuity to completion.

Gendlin explicitly recognizes, however, that the future is

important here, in that an authentic choice of one's future is

necessary to carry forward one's experienced present; an

inauthentic choice of the future simply blocks the carry for-

ward process.

In partial summary, then, it seems clear that the time

dimension is an important one to many theoretical views of

healthy function, and one's conceptualization of, orientation

towards, and mode of dealing with the future areof primary


Content Constructs

The dimension along which theories seem to differ the

most is that of defining "what" makes up the personality sys-

tem, or more precisely, what are the key elements or components

to healthy functioning. Happily enough, it is precisely here

that the systems framework has the least at stake. Being pri-

marily descriptive of process and structure, the labels chosen

to describe component parts are seen as less important than

the interactions over time of whatever components are focused


It is, of course, the hypothetical content constructs

employed which tend to give each theory its distinctive "fla-

vor" or identity. Ellis (1973), for example, considers

"beliefs" to be the components of primary concern; Kelly

focuses upon personal constructs. For Jung (1968), the

personality is made up of components such as archetypes, the

shadow and the self, each of which has specific, though com-

plex, functions in the system. Maslow (1967) sees values as

central and discusses the role played by different kinds of

values in personality development. For Jourard (1974), the

components of the self-structure (real self, ideal self, public

self, and self-concept) are important, and their "positioning"

in relation to one another is an important aspect of person-

ality function, while Adler (1964) focuses upon "life-goals."

In presenting this section, the attempt is certainly

not to dismiss or devalue the theoretical constructs pertain-

ing to content, or pretend that the differences between the-

ories are insignificant. Rather, the emphasis is to point out

that each discusses personality from a unique perspective,

and as a result each emphasizes different life components or

dimensions along which people's behavior may vary. Therefore,

the evaluation of personality on a dimension of relatively

high or low level functioning will, for each theorist, be in

terms of "his" components (i.e., the subsystems he chooses to

emphasize) and the theoretical role played by these competents

in the theorist's view of human process. As Kelly (1966)

points out, theories are built to account for different areas

of interest (i.e., have different ranges and foci of conven-

iences) and thus theorists would be expected to use different

constructs in building their positions. Further, theorists

differ in the degree to which their content categories should

empirically testable terms. For example, when terms such as

"boundaries," "feedback," and "differentiation" are applied

to human functioning, they are hypothesized to be meaningful

descriptors, and yet are not directly observable or measur-

able. However, to suggest that a theoretical portion has

scientific merit necessitates its being conducive to empiri-

cal research. Though the position developed here has not

been directly investigated up until now, there has been

research into relevant variables, from a closely related the-

oretical perspective.

A program of research by Seeman (1959) and his asso-

ciates has, for the past fifteen years, been investigating

the construct of "personality integration" (PI). Defined as

"a configuration of behavioral subsystems . that inter-

act in an adaptive and effective manner" (Thomas & Seeman,

1972, p. 154), and emphasizing the effective use of a maximum

amount of information (Seeman, 1959), the perspective is simi-

lar (but not identical), to the systems theory position out-

lined above. Though there are conceptual differences between

the approaches, the research undertaken by Seeman and his col-

leagues has produced evidence which is supportive of ideas

derived from systems theory.

Their data suggest that highly integrated persons

differ from less integrated contrast subjects in several sub-

systems of personal functioning. Cognitive processing of

highly integrated subjects is characterized as being both more

complex (Thomas & Seeman, 1971) and more efficient, as mea-

sured by GPAs, but high PI subjects are not more intelligent,

as measured by scores on the College Entrance Examination

Boards and the American Council of Education Test of Intelli-

gence (Duncan, 1966; Seeman, 1966). Interpersonally, highly

integrated persons demonstrate greater "environmental con-

tact," defined as involvement in a variety of activities

(Duncan, 1966; Seeman, 1966) and are more positively oriented

towards and less threatened by the social environment (Hearn

& Seeman, 1971). Based upon Rorschach protocals scored

according to Klopfer's method, highly integrated subjects

demonstrated more imaginative and constructive modes of think-

ing about people, as well as a capacity for more empathic

relationships with people (Thomas & Seeman, 1971). In the

affective domain, high PI subjects have been found to be more

comfortable with, and better able to express the affective

components of their lives, as well as demonstrating greater

variability in feeling states (Hearn & Seeman, 1971). Addi-

tionally, highly integrated men have demonstrated more inter-

nal loci of control and loci of evaluation (Duncan, 1964,

1966) but this finding has not been replicated in women (See-

man, 1966), which Seeman suggests may indicate a culturally

based "sex-linkage" to some behaviors and beliefs irrespec-

tive of level of integration.

The data presented here can be seen as conceptually

related to the description of high level functioning based


on systems theory, which would predict greater complexity of

component parts, more openness to the environment ("environ-

mental contact," "positive orientation to social environment"),

and more effective perception and usage of relevant informa-

tion ("effective perceptual styles," "intellectual efficiency

as opposed to intelligence"). The locus of control data, too,

suggest that integrated males perceive themselves as in suf-

ficient control over the events in their worlds to be potent

decision makers in forming goals and moving towards them.

It was mentioned earlier that systems theory recog-

nizes the interrelationship of different levels of systems,

such that from another perspective, individual systems can be

seen as components of a higher level system, and the adequacy

of functioning of the levels is interdependent. Interper-

sonally, this has been an emphasis of family theorists (e.g.,

Satir, 1967), and the implication of this position has been

receiving increasing research interest. In a study support-

ing the conceptual relationship between the functioning of

persons-as-systems and families-as-systems, Odom, Seeman,

and Newbrough (1971) found that children identified as highly

integrated (using a child-relevant nomination method similar

to the Personality Integration Reputation Test [PIRT] proce-

dure described below) came from families whose communication

patterns differ qualitatively and quantitatively from the

families of poorly integrated children. The families of

high PI children were characterized as exhibiting more

cooperation, more warmth, less dependence, more direct and

clearer communication, more flexibility in reaching decisions,

and more clearly defined roles than the contrast families of

poorly integrated children. These findings are also conver-

gent with the systems theory conceptualization of high level

functioning, and lend support to the position from the family

unit level of analysis.

These studies are an important beginning in the study

of personality integration as the construct is related to the

systems position developed here. However, there are limita-

tions to the conclusions that can be drawn from them when

developing a description of high level personality function-

ing, and several of the questions which arise from these

limitations form the basis and rationale for the present study

The limitations of the studies cited above center in

two areas. The first is that of population sampling; the

second is in the methodology employed and more specifically,

the instruments chosen to measure dependent variables. As

Wright, Bond, and Denison (1968) pointed out, all of the

research discussed above (except the family study by Odom et

al., 1971) have used undergraduate college students who lived

in fraternity or sorority houses, or were dormitory residents.

This is understandable, in that the PIRT method requires a

long term group relationship among people who know quite a lot

about many facets of each other's lives. Though these popu-

lations do fit the criteria, they form a rather limited sample

from which to speak about "personality integration" as a


general phenomenon. It seems important, therefore, to extend

the study of personality integration to people who differ

from these groups on several parameters, including a variety

of ages, socioeconomic classes, subcultures and even inter-

national cultures, in order to investigate the dimensions

which consistently differentiate highly integrated people from

others. Of course, no one study can incorporate all of these

variables, and for the purposes of this research, the cross

subcultural dimension emerges as an important focus. From

the open systems perspective discussed earlier, the issue can

be conceptualized as a need to study groups which differ in

"content" areas (values, beliefs, life-styles) to identify

the process variables which are characteristic of highly

integrated persons across subcultural groups. The present

study is designed to investigate this area.

The second area of limitation pointed to concerned

methodology, and specifically, instrumentation. Several of

the studies cited earlier used instruments with little valid-

ity data to support their use, such as the "environmental

contact" measure and the "locus of evaluation" instrument in

Duncan (1966) and Seeman (1966). In the Thomas and Seeman

(1971) study, it is not clear what, specifically, they were

measuring with their assessment of cognitive complexity; they

employed a simplified REP test which elicits from subjects

the ways in which groups of people are similar and different

from each other. The score is total number of discrimination

made by the subject, which makes no attempt to separate the

number of "words" from the number of true construct dimen-

sions (i.e., several words may have been different labels for

the same underlying construct dimensions). Thus, their

results are difficult to interpret, although they are theo-

retically sound, and are convergent with the work of Wexler

(1974), who found that high scores on the previously cri-

tiqued POI (Shostrum, 1966) were more cognitively complex

than lower scoring subjects, based upon a differentiation-

integration assessment of their verbal descriptions of emo-

tional experiences. The issue being focused upon here is the

need to continue with what Campbell and Fiske (1959) call a

"multitrait, multimethod" approach to the study of the com-

plex phenomenon of personality integration. In the present

study, several variables which from a systems theory perspec-

tive are hypothesized to be characteristics of personality

integration are studied in several subcultural population

settings. The study makes use of both (a) multiple methods

for assessing a single personality dimension, such as "com-

plexity"; and (b) the measurement of multiple "traits" by

using the instruments which measure the same or conceptually

related dimensions as those in earlier studies but are

methodological alternatives to the measures previously

employed. The rationale here is that the study of a con-

struct as broad as personality integration necessitates

divergent methods of measuring relevant personality

characteristics both within a single study and as part of a

program of theory testing research. For example, the selec-

tion of instruments for this study was guided, in part, by a

desire to employ methods requiring varying modes of response

(i.e., not using all forced choice, verbal questionnaires).

In sum, the present study is designed to further the study of

personality integration by (a) expanding the population base

of subcultures not previously studied; (b) continuing the

multitrait, multimethod approach of previous studies by

studying personality dimensions not previously examined (time

orientation) as well as using alternative instrumentation in

measuring similar personality dimensions; and (c) providing a

comprehensive theoretical basis for conceptualizing high level

personality functioning by employing a general systems theo-

retical framework.


The open systems perspective, as discussed earlier,

discriminates between "process" and "content" dimensions, sug-

gesting that the former is of central concern in understanding

high level functioning, while giving less weight to the lat-

ter. Therefore, the hypotheses which follow are based upon an

overriding hypothesis that personality integration can and

does occur within a broad range of subcultural groups which

differ from one another in value orientation and life style

("content dimensions"), and that highly integrated people dif-

fer predictably from less integrated peers along specifiable


"process dimensions." The content dimension investigated in

this study is operationally defined as value orientation, and

is measured by using the Allport-Vernon-Lindzey Study of

Values (1960). The process dimensions included (a) complexity

of the self-concept, (b) cognitive differentiation, (c) cogni-

tive integration, (d) boundary permeability (openness to

social experiences), (e) self-esteem, and (f) time orienta-

tion. The instruments used to measure these dimensions are,

respectively, Ziller's (1973) Complexity of Self-Concept

Scale; Bieri, Atkins, Briar, Cobeck, Miller, and Tripodi's

(1966) REP Test for b and c above; Ziller's (1973) Self-

Esteem Measure; and Cottle's (1976) Time Orientation Inven-

tory (which includes the Experiential Inventory, Circles Test,

and Lines Test). The justification for selecting these

instruments and supporting validity data is presented in the

Methods Chapter. The specific hypotheses to be tested in

this study are as follows:

Content Dimensions

1. There will be no difference between Personality

Integration (PI) and contrast subjects on the

Allport-Vernon-Lindzey (AVL) value profile scores.

2. There will be significant differences between sub-

cultural groups on the AVL value profile scores.

Process Dimensions

3. PI subjects, compared to contrast subjects, will

demonstrate greater complexity of the self-concept

by scoring higher on the Ziller complexity of

the self-concept instrument.

4. PI subjects, compared to contrast subjects, will

demonstrate greater cognitive differentiation by

having higher FIC scores on the Bieri grid.

5. PI subjects, compared to contrast subjects, will

demonstrate greater cognitive integration by hav-

ing higher ordination scores on the Bieri grid.

6. PI subjects, compared to contrast subjects, will

demonstrate more permeable social boundaries by

scoring higher on the Ziller "openness" instrument.

7. PI subjects, compared to contrast subjects, will

demonstrate greater self-esteem by scoring higher

on the Ziller self-esteem instrument.

Time Orientation

8. The PI subjects, compared to contrast subjects,

will demonstrate less orientation to the past by

having higher scores on the Cottle Experiential


9. The PI subjects, compared to contrast subjects,

will demonstrate more present and future orienta-

tions and less of a past orientation by more often

being categorized as present and/or future domi-

nant and less often past dominant on the Cottle

Circles Test.

PI Validity Check

10. PI subjects, as compared to contrast subjects,

will have higher scores on the Personality Inte-

gration subscale of the Tennessee Self-Concept


Other Variables

The following variables will also be investigated for

the relationships to the independent variables of PI level

and group membership, although they are not specifically

related to the theoretical position developed here, and thus

no directional hypotheses will be offered.

1. Meaningfulness of construct dimensions, as mea-

sured by the Bieri REP Test.

2. Temporal relatedness, as measured by the Circles


3. Scores on the categories of the Lines Test (His-

torical Past, Personal Past, Present, Personal

Future, Historical Future, Lifetime).

4. GPA (as reported by subjects).

5. Age.

6. Length of membership in group (Group Tenure).




The procedure for selecting the personality integra-

tion (PI) and contrast groups required relatively large

groups who knew each other well over an extended period of

time. For the theoretical reasons discussed above, it was

important that the selected groups could be assumed to differ

in value orientation (which would then be empirically inves-

tigated). Additionally, it was desirable for selected groups

to be equated in age and SEC. Based upon Seeman's (1966)

discussions of sex differences in PI, it was decided to limit

this investigation to male subjects, recognizing the limita-

tions that this approach places on conclusions and generali-

zations from the data. For these reasons, it was decided to

select subjects from groups of the following types: military

cadets, seminarians and college fraternities. Letters were

sent, and follow-up telephone calls were placed to the

administrators of many institutions which met these criteria,

and arrangements were completed with a military academy in

the southeastern part of the country, a Protestant seminary

in the northcentral region, and social fraternities at a pub-

lic midwestern university.


The final group of subjects was selected from larger

preexisting groups at each of the institutions by employing

the Personality Integration Reputation Test (PIRT). At the

military academy, this refers to members of companies with

the PIRT nomination data collected from a total of 175 volun-

teers, resulting in a PI group of N = 17 and contrast group

of N = 23. At the seminary, 75 volunteers from dormitory

units constituted nominating groups, resulting in a PI group

of N = 15 and contrast group of N = 11.

The initial fraternity group consisted of 76 volun-

teer participants from social fraternities, resulting in a

PI group of N = 12 and a contrast group of N = 12. In total,

226 subjects completed the PIRT nomination instrument, with

44 PI subjects and 46 contrast group subjects.

Procedure and Instruments

At each setting, the PIRT instrument (see below) was

administered to each large group at one session. The final

group of PI and contrast subjects, selected as described

below, completed the remaining instruments singly in a follow

up session. The selection of instruments was made based upon

two major criteria. The first was the degree to which the

personality dimensions they purported to measure were theo-

retically relevant to the hypotheses being tested. The second,

which was more a limiting factor than a legitimate "crite-

rion," was the element of time restriction. The subjects'

participation was completely voluntary in all three groups,


and thus no negative sanctions or rewards were available for

not returning to follow up sessions or staying to complete

all instruments. Therefore, it was decided that one hour was

all the time that could reasonably be asked of the subjects

in the follow up testing session. In one case, such a

restriction was necessary to receive permission from the

institutional administrators to arrange any of the data col-

lection. Therefore, it was necessary that the total time

for completing all dependent measures not exceed one hour,

which eliminated the selection of several alternative or addi-

tional instruments (e.g., Bieri, Atkins, Briar, Cobeck,

Miller, & Tripodi's [1966] "construct provided" form of the

REP test was used rather than a more time-consuming "con-

struct-elicited" form; the use of the PI subscale of the

Tennessee Self-Concept Scale [Fitts, 1964], rather than the

more widely used POI [Shostrum, 1966]). Though it could be

argued that fewer instruments should have been used or fewer

personality dimensions should have been investigated, it was

decided that in an exploratory study such as this, it was

justifiable to begin by examining the relationship of several

dimensions to personality integration, rather than doing a

more thorough investigation of one or two.


Personality Integration: The selection of "inte-

grated" and "contrast" subject groups followed the Personal-

ity Integration Reputation Test (PIRT) method described by

Duncan (1964, 1966) (see Appendix A). In this method, each

subject was presented with six components of personality

integration (based on the work of Jahoda, 1958), and for each

dimension, the subject was asked to nominate three group mem-

bers who most closely fit the description given (e.g., "Who

are the three persons who seem the most able to deal effec-

tively with everyday tensions and anxieties?"). The result-

ing highly skewed distribution allowed for selection of a

small group of "integrated" subjects; the contrast group was

randomly selected from the remaining group members receiving

at least one nomination. Thus, the contrast group was not at

an opposite "extreme" from the PI group, and was not likely

to be made upof "pathological" or dysfunctional group members.

Duncan (1966) has reported split half reliability coefficients

of .82, .78, and .85 for this instrument, and a test-retest

reliability coefficient of .88, with the second set of nomi-

nations resulting in an identical set of names for inclusion

in the high PI group. The series of studies reviewed above

evidences the extent to which PIRT has demonstrated its valid-

ity in college age populations. Additionally, a series of

studies by Wright (1966, 1967a, 1967b) provided further con-

struct validity, reliability data, and evidence for a single

"personality integration" factor, respectively, for the use

of PIRT.

As discussed earlier, both peer nomination and self-

report measures have been used in selecting groups for


research in this area. For the reasons presented in Chapter

I, PIRT was chosen in the study as the basis for selecting

"high" and "contrast" PI groups. It was decided, addition-

ally, to explore the relationship between PIRT and self-

report methods from the cross subcultural perspective devel-

oped here by administering the 25 item PI subscale of Fitts'

(1964) Tennessee Self-Concept Scale (TSCS) to all subjects.

This subscale is empirically based, representing the items

which best differentiated "a group of subjects judged, by out-

side criteria, to have a better-than-average level of adjust-

ment" (Radford, Thompson, & Fitts, 1971, p. 45). Fitts

(1964; Fitts, Adams, Radford, Richard, Thomas, Thomas, &

Thompson, 1971) reports no further validity data for using

scores on the PI subscale as a criterion for selecting high

PI subjects, although he reports (1971) several studies

demonstrating the high PI subjects chosen by other criteria

have "healthier" scores on many of the TSCS scales than con-

trol groups (e.g., total P [positive] scores, indicating a

more favorable self-concept, and lower total conflicts scores,

reflecting less internal conflict or contradiction in their

self-concepts). In the only other study reported which used

the PI subscale by itself (without the remaining items on the

TSCS), Dunn's study of college students (1969, reported in

Fitts et al., 1971) reported a significant intercorrelation

between the PI subscale and two other self-report measures

(Shostrum [1966] POI, and Barron's [1963] ego strength scale,


the PI subscale demonstrated little correlation to sociomet-

ric ratings [two versions of PIRT]).

Cognitive Complexity: Two instruments were chosen to

measure different dimensions of cognitive complexity. In the

Bieri et al. (1966) REP Test subjects were provided with

eight construct dimensions, and asked to rate each of eight

persons who fit provided role-descriptions on each construct

dimension (see Appendix C). The resulting 8 x 8 grids were

then individually analyzed to produce scores on dimensions of

cognitive differentiation, integration, and meaningfulness,

using a procedure developed by Landfield (1977b). The major

criterion for selecting the provided form of the REP Test was

time, in that this form takes 15 minutes or less to complete,

whereas the elicited form developed by Kelly (1955) takes

considerably longer. Though there is controversy in the

field, Bieri et al. (1966) present evidence based on the

research of Kieferle and Sechrest (1961) and Tripodi and

Bieri (1963) that distribution of complexity scores obtained

by provided and elicited forms from the same populations are

not significantly different, supporting the contention that

the provided form is an acceptable alternative to the more

common elicited form for the purposes of this study.

Ziller's (1973) complexity instrument was used as a

measure of complexity of the self-concept, a more delimited

area of complexity than the generalized complexity dimensions

assessed by the Bieri procedure. Ziller's instrument is a

109 item adjective checklist (Ziller, 1973). The instru-

ment has demonstrated acceptable split half reliability

(r = .92 in an adolescent population [Ziller, 1973], and test-

retest reliability of r = .72 [Ziller, 1973]). The validation

of the measure has included studies which demonstrate its

independence from other personality dimensions (self-esteem,

intelligence) and a significant positive correlation to self-

described complexity and complexity of photographic self-por-

traits. Also demonstrated has been a significant relationship

to degree of social interaction, with physically handicapped

and terminally ill patients having less complex self-concepts

than matched controls (see Ziller, Martell, & Morrison, 1977,

for a more complete discussion of the validation process for

this instrument).

Self-esteem, openness: These dimensions were

assessed by using selected parts of Ziller's (1973) Self-

other Orientation Inventory, a group of primarily nonverbal

items which are purported to measure self-esteem, and open-

ness to social experiences and interaction. For the self-

esteem scale, Ziller (1973) reports split half reliability

coefficients ranging from .80 to .89 for student and adoles-

cent groups. Validity data include demonstrations that

sociometric "stars" have higher self-esteem than sociometric

"isolates" (Ziller, 1973) and normal (nonclient) groups had

higher self-esteem than neuropsychiatric patients (Ziller,


The "openness" items have been developed more

recently and have less supportive construct validity data.

The only reported study thus far demonstrates that high scores

on the "openness" items include significantly more people in

a photographic self-description task than do low openness

scores (Ziller & Smith, 1978).

Value Orientation: In this study, the "content" dimen-

sion discussed in Chapter I is operationally defined as value

orientation, and more specifically, the profile of scores

obtained on the Allport-Vernon-Lindzey Study of Values (1960).

This instrument is a 45 item inventory, providing an ipsative

profile of scores on six value dimensions developed from the

work of Spranger. It is a widely used instrument for counsel-

ing and research purposes, with supporting validity presented

by the authors (1970) demonstrating its ability to discrimi-

nate between varied occupations, to predict chosen occupation

in 5 to 15 years, and to predictably change as a result of

life experiences such as type of education.

Time Orientation: The instruments used to assess

time orientation were selected from the battery of tests

developed by Cottle (1976) in his exploratory study of how

subjects perceived themselves in relation to a range of tem-

poral dimensions. The instruments selected for the present

study were designed to assess the relative importance or

impact on current functions that subjects attribute to their

conceptions of past, present, and future events respectively.

Specifically, the instruments included (a) the Experiential

Inventory, which measures the frequency with which important

life events reported by subjects occur in different "time

zones" (e.g., "distant past," "near future"); (b) the Circle

Test, which uses the positioning and size of circles repre-

senting the past, present, and future time zone as a spatial,

nonverbal assessment of time orientation which can be analyzed

to produce scores on the dimension of Temporal Zone Dominance

and Temporal Relatedness; and (c) the Lines Test, in which

time zones are linearly defined by subjects larger or smaller

units of a continuous line segment representing "time-as-a-

whole" (see Appendix B). These instruments have not been

extensively used for research purposes, and thus supporting

reliability and validity data are still minimal. The results

presented by Cottle (1976) include a discussion of the dif-

ferent patterns of time orientation exhibited by his subjects

and the relationship of scores on one section of the inven-

tory to scores on other parts. There is, however, no presen-

tation of data that demonstrates construct validity of the

instruments. However, the dimensions which Cottle's instru-

ments attempt to assess approximate the theoretically

relevant time dimensions discussed earlier, and the instru-

ments took relatively little time for subjects to complete.

In the absence of available alternative instrumentation,

therefore, it was decided to include these measures as an

interesting preliminary exploration of the relationship


between personality integration and time perception and ori-


Other Data: A biographical information sheet was also

given to all subjects. This recorded data on age, class in

school, and duration of group membership.


The several instruments were scored according to the

following procedures:

Personality Integration Reputation Test: Following

the procedure described by Duncan (1964), the scoring of PIRT

is a straightforward count of the total number of times each

individual's name is nominated, summed across the six items.

The resulting distribution is highly skewed, with a small

number of individuals receiving a large number of the nomina-

tions. These persons were identified by simple observation

of breaks in the distribution of nominations.

The number of subjects above the break in each group

is partially dependent upon the size of the nominating group

itself, and thus the larger cadet group had more members

receiving a large number of nominations. This is reflected

in the larger number of both PI and contrast subjects within

the cadet group, compared to the seminary and fraternity


Allport-Vernon-Lindzey (AVL) Study of values:

Detailed instruction for scoring are provided with the

instruments. Items on Part I ask subjects to distribute 3

points, 2 alternative indications personal preference. For

example, choice "a" might get 2 points, while "b" gets 1 point

or a 3-0 spread might be chosen. Each choice is designed to

reflect one of the six value dimensions. Part II provides

four alternatives, which are rank ordered by the respondent,

with 4 points assigned to the most preferred alternative,

and 1 point to the least preferred. Scores are basically the

sum of points assigned for each value across all items. The

final score, therefore, is six numbers, corresponding to the

total number of points given to each of the six value dimen-

sions. No alternatives were made in the standard scoring

procedures, resulting in an ipsative profile of scores for

each subject on the six values.

Self-other Orientation Method: Scoring procedures

for completing of the self-concept, self-esteem, and openness

followed Ziller's (1973) method. The complexity scores for

each subjectwere the sum of items check on adjective check-

list. Openness was scored by summing the number of circles

connected to "self" on each of the six openness items of the


Self-esteem was scored by counting, from right to

left, for each of the six esteem items in the instrument, the

number of the circle marked "yourself." The extreme right

circle was scored "1," the second circle "2," etc. The scores

from each item were summed, yielding a total score for each


Time Orientation: The time instrument was scored

following Cottle's (1976) method. For Experiential Inventory

Scores, an arithmetic mean was computed, based upon the num-

bers assigned by subjects to each of their listed experiences

with higher means indicating subjects' time perspective as

relatively more future oriented. The Circles Test was ana-

lyzed to yield scores for each subject on temporal dominance

and on temporal relatedness. Temporal dominance was scored

by a procedure where subjects were categorized as past, pre-

sent, or future dominant if the corresponding circle was

drawn significantly larger than the other two circles. If no

circle was significantly larger, the subject's category was

"none." All classifications were based upon 100% agreement

between two trained raters. Temporal relatedness was scored

by assigning points based on the degree of overlap between

the circles drawn representing past, present, and future.

Completely "atomized" or separated circles received a score

of zero. Two points were scored each time a circle was drawn

touching another circle, while four points were assigned each

overlap, and six points for a complete engulfmentt" or

encirclement of one circle by another (i.e., if one circle

was drawn totally within another one). The points were then

summed yielding the final relatedness score. The Lines Test

was scored by simply measuring, in centimeters, the length

of the line segments demarcated by each subject. The dis-

tance between the left edge of the line and the marking for

"Birth" was the score for Historical Past, while Personal

Past was the distance between that point and the "Past-Present

Boundary." From these to the "Present-Future Boundary" was

the Present, and Personal Future extended from that boundary

to the mark for "Your Death." Historical Future was the dis-

tance from this latter marking to the right edge of the line

segment. The "Lifetime" score was the total distance in cen-

timeters, from "Birth" to "Death."

Bieri REP Grid: The scoring procedure for the REP

Grid followed the procedures discussed by Landfield (1977a),

and made use of the FIC-Ordination computer program developed

by him (1977b). The program yields scores for FIC, Ordina-

tion, and Meaningfulness. Each REP Test Grid was analyzed by

matching each construct dimension (each row) with each other

construct dimension, and noting each occurrence of agreement

in usage of constructs when applied to the same role-title.

The fewer the occurrences of overlap, the greater the func-

tional independence of the two construct dimensions; con-

versely, more frequent overlaps indicate more functionally

interdependent constructs. Following Landfield's (1977a) 70%

criterion, in this study the cut off point for the independent-

dependent criterion was six agreements out of the total of

eight ratings made for each construct. The same procedure is

applied to the columns of the grid ("People"), to yield FIC

scores for "Constructs" and "People." Those were summed,

yielding a "total" FIC score for each subject.

Whereas FIC scores reflect the differentiation of a

person's system, the ordination measure was developed (Land-

field & Barr, 1976) to assess the degree of hierarchical

relationship within an individual's set of constructs (i.e.,

the degree of integration within the system). There are

several assumptions which underly this measure. Constructs

are assumed to be hierarchically related, such that systems

contain constructs which are superordinate to other, subor-

dinate, constructs. These superordinate constructs are

considered more meaningful, relative to the subordinate con-

structs for the individual. The ordination measure is

designed to examine the relative superordinancy within the

individual's set of constructs, assuming that relatively high

ordination scores indicate a relatively high degree of sys-

tem integration. The assumption underlying the use of this

instrument as a measure of ordination is that a subject's

use of a relatively superordinate construct will be reflected

by greater ability to make finer discrimination between

levels of extremity along that construct dimension. There-

fore, when subjects make use of relatively superordinate con-

structs, they will demonstrate this by applying these

constructs across a wider range of levels on the REP Test.

The procedure for obtaining ordination scores is some-

what more complex than that for the FIC scores, and is quoted


from Landfield (1977a). To compute the Ordination score on a

given construct:

. if the person has used four different levels of
extremeness--for example, for scale points 0, 2, 3, 5,
a score of 4 is multiplied by the difference between
his highest and lowest ratings (5) and the ordination
score is 20. If we want to obtain the ordination level
of a particular (person) . we observe how he has
rated that person across his . constructs. Again,
the number of levels used is multiplied by the high-low
rating difference . a correction factor for exces-
sive use of a particular rating is explained by Leitner,
Landfield and Barr (1975). (pp. 153-154)

Final Ordination scores are a combination of the average

"Construct" ordination scores and "Person" ordination scores.

Meaningfulness scores are the sum of the extremity of

ratings (distance from zero, or midpoint) for the construct

and person dimensions. See Landfield and Barr (1976), for a

more complete description of the computer-program designed

to analyze REP Grids and produce the scores discussed here.



This chapter will present the results of the analyses

performed on data collected for this study. The data per-

taining to each hypothesis will be presented in order, fol-

lowed by a presentation of the data analysis for variables

about which no hypotheses were generated.

Hyp. 1: There will be no differences between per-

sonality integration (PI) and contrast

subjects on the Allport-Vernon-Lindzey (AVL)

Value Profile scores.

Hyp. 2: There will be significant differences

between Groups on the AVL Value Profile


Table 1 presents the results of MANOVA analysis per-

formed on the AVL data. The results showed a significant

main effect for groups, which provides support for Hypothesis

1, and no main effect for PI, which supports Hypothesis 2.

There was no significant Group X PI interaction effect. To

examine how the three groups differed from one another on the

AVL, a second set of MANOVA analyses was performed, comparing

each pair of groups on the set of AVL variables. These

Table 1

Results of MANOVA: Personality Integration X Groups X
Allport-Vernon-Lindzey Scores

Test Hypothesis Pillai's Trace Statistic F (d.f.) E

No overall effect for V = .536 4.82 (12,156) p = .0001

No overall effect for V = .070 0.98 (6,78) p = .444
PI (6,78)

No overall effect for V = .112 0.78 (6,156) p = .668
Group X PI interaction

Note: Results for MANOVA tables (1-3) are presented in null hypothesis
testing form.

analyses served, essentially as a MANOVA analogue to the a

posteriori tests in univariate analysis of variable proce-


Table 2 presents the results of these analyses. The

results show significant differences between Groups 1 and 3,

and 2 and 3, with no significant differences between Groups

1 and 2.

No further analyses of AVL scores, such as exploring

on which specific values the groups differed, were performed

due to the ipsative nature of the instrument. Subjects'

scores on the several values are mathematically interdepen-

dent, and thus it was decided that value scores should only

be analyzed as sets rather than as separate variables.

However, a verbal description of each Group's profile

provides an interesting reflection of the Group's value

orientation. The rank ordering of the AVL's six values for

each Group are as follows: highest scoring values are pre-

sented first, with the Group mean score for that value in


Seminarians: Religious (52.4), Social (43.8),

Aesthetic (38.8), Political (36.2), Theoretical (35.8), Eco-

nomic (33.0).

Cadets: Political (44.5), Economic (43.6), Religious

(41.0), Theoretical (38.1), Aesthetic (37.5), Social (35.2).

Fraternities: Political (43.6), Economic (43.3),

Social (39.2), Theoretical (38.8), Religious (38.2), Aes-

thetic (36.5).

Table 2

Results of MANOVA: Groups X Allport-Vernon-Lindzey Scores
(Pairwise Comparisons)

Test Hypothesis Pillais Trace Statistic F (d.f.)

No overall effect for V = .128 1.37 (6,56) p = .243
Group (Group 1 and 2)*

No overall effect for V = .564 9.26 (6,43) P = .0001
Group (Group 1 and 3)

No overall effect for V = .485 11.11 (6,59) q = .0001
Group (Group 2 and 3)

*Note: Group 1 =
Group 2 =
Group 3 =

Military Cadets


When a relatively large number of dependent measures

are used in a study and a series of ANOVA analyses are per-

formed, the risk of Type 1 errors rapidly increases. There-

fore, to hold down this risk a MANOVA analysis was performed

upon the dependent variables as a group. Table 3 presents

the results of this analysis. The univariate ANOVA analyses

reported below are based upon this MANOVA; that is, the df

and sums of squares for each variable in the tables that fol-

low were calculated within the MANOVA procedure, rather than

in separate ANOVA's for each variable. A comment on the

multivariate statistic reported in this and other MANOVA

tables is relevant here. Olsen (1976) reports that Pillai's

Trace (V) is the statistic of choice when there is doubt as

to the homogeneity of variance for each dependent variable;

it is the most robust of the several multivariable statistics

available, and the least likely to inflate the numbers of

Type I errors.

The results show main effects for Group and for PI,

with no Group X PI interaction. These results both justify

the ensuing analyses to determine on which variable there are

significant differences between the levels of PI and between

Groups, and suggest that any PI X Group interactions which

are found in the univariate ANOVA should be considered as


Table 3

Results of MANOVA: Groups X Personality
Independent Variables

Integration X

Test Hypotheses Pillai's Trace Statistic F (d.f.) E

No overall effect for V = .636 2.75 (27,130) p = .0002

No overall effect for V = .460 4.97 (11,64) p = .0001

No overall effect for V = .266 0.91 (27,130) p = .586
Group X PI interaction

Note: Dependent variables include: Complexity of Self-Concept, Openness,
Self-Esteem, Personality Integration (TSCS), FIC, Ordination, Mean-
ingfulness,Experiential Inventory, Age, Group Tenure, GPA.

Table 4 presents the means and standard deviations

for each of the dependent variables, categorized by Group and

by level of PI, as well as the total sample.

Table 5 presents the results of t tests performed on

the dependent variable scores for which directional, a priori

hypotheses had been made. The results indicate that the

complexity of the self-concept variable produced the only

scores on which the PI and contrast subjects differed in the

predicted direction. The Openness and PI subscale (TSCS)

scores approached significance (p < .07), while all others

were clearly nonsignificant differences.

Hyp. 3: PI subjects, compared to contrast subjects,

will demonstrate greater complexity of the

self-concept by scoring higher on the Ziller

complexity of the self-concept instrument.

Table 6 presents the results of the ANOVA analysis

performed on the complexity of self-concept scores. The

results indicate a significant difference between levels of

PI in the predicted direction, giving support to Hypothesis

3. No significant differences between Groups and no signifi-

cant Group X PI interaction were found.

Hyp. 4: PI subjects, compared to contrast subjects,

will demonstrate greater cognitive differ-

entiation by having higher FIC scores on the

Bieri grid.

Table 4

Means and Standard Deviations for Levels of PI, Group,
and Total N on Dependent Variables

PI Contrast Fraternities Cadets Seminarians Total

11.67 4.09 42 10.09 3.33 46 11.50 4.30 24 9.60 3.59 40 12.25 2.91 24

46.63 12.83 43

5.82 2.62 44

32.18 8.45 44

242,75 47.82 44

45.19 13.04 43

22.14 7.86 43

38.93 17.46 46

5.85 3.31 46

34.05 7.44 46

232.02 41.98 46

42.59 14.57 46

22.78 5.48 46

44.96 21.42 24

5.63 2.87 24

32.03 8.46 24

215.00 51.33 24

42.67 13.96 24

24.63 5.70 24

40.62 12.21 40

5.77 2.81 40

33.19 8.09 40

253.80 37.99 40

45.43 14.79 40

22.43 6.80 40

43.68 14.81 25

6.11 3.40 26

34.06 7.45 40

232.38 39.94 26

42.44 12.33 25

20.48 7.05 25

10.84 3.78 88

42.65 15.79 89

5.83 2.98 90

33.13 7.96 90

237.27 44.99 90

43.84 13.83 89

22.47 6.70 89

2.61 0.55 41 2.62 0.61 46 2.49 0.46 23 2.77 0.66 39 2.49 0.51 25 2.62 0.58 87

20.71 1.40 41 19.91 1.20 44 20.68 1.32 22 20.08 0.94 40 20.30 1.87 35 20.29 1.35 85

PI Subscale

Complexity of








Group Tenure


27.56 12.57 88

2.72 0.67 88


32.21 11.03 42 23.30 12. 4 46 28.00 15.90 24 29.30 9.32 40 24.20 13.43 24

2.97 0.56 42 2.50 0.69 46 2.68 0.65 24 2.48 0.57 40 3.15 0.66 24

Table 5

T Tests for A Priori, Directional Hypotheses

Variable T df (One-Tailed)

Complexity of Self-Concept 2.35 88 .02

FIC 0.00 88 n.s.

Ordination -0.98 88 n.s.

PI Subscale 1.58 86 .07 (Approx.)

Self-Esteem -0.73 87 n.s.

Openness 1.48 87 .07 (Approx.)

Experiential Inventory -0.39 84 n.s.

Table 6

ANOVA for Group X Personality Integration
X Complexity of Self-Concept

Source df SS F PR>F

Group 2 153.95 0.32 .73

PI 1 1333.13 5.54 .02

Group X PI 2 0.23 0.00 .995

Error 74 17817.74

Total 79 19387.20

Table 7 presents the results of the ANOVA performed

on the FIC data. The results indicate no significant main

effects for Groups, or for PI. There was no significant

Group X PI interaction. Hypothesis 4 is therefore not sup-

ported by these data.

Hyp. 5: PI subjects, compared to contrast subjects,

will demonstrate greater cognitive integra-

tion by having higher ordinative scores on

the Bieri grid.

Table 8 presents the results of ANOVA performed on

the ordination data. The results indicate no significant

main effects for Groups or for PI. There was no significant

Group X PI interaction. Hypothesis 5 is therefore not sup-

ported by these data.

Hyp. 6: PI subjects, compared to contrast subjects,

will demonstrate more permeable social

boundaries by scoring higher on the Ziller

"openness" instrument.

Table 9 presents the results of the ANOVA performed

on the openness data. The results indicate no significant

main effect for PI or for Groups and no Group X PI interac-

tion. As reported in Table 5, however, the PI means do differ

in the predicted direction (t = 1.48) which approaches sig-

nificance (p < .07). Therefore, Hypothesis 6 receives some

moderate support from these data.

Table 7

ANOVA for Group X Personality Integration
X FIC Scores

Source df SS F PR>F

Grouo 2 2.998 0.15 .86

PI 1 0.023 0.00 .96

Group X PI 2 15.514 0.77 .47

Error 74 747.809

Total 79 766.388

Table 8

ANOVA for Group X Personality Integration
X Ordination

Source df SS F PR>F

Group 2 51.04 0.40 .67

PI 1 61.34 0.96 .33

Group X PI 2 51.33 0.40 .67

Error 74 4749.52

Total 79 4909.84

Table 9

ANOVA for Group X Personality Integration
X Openness

Source df SS F PR>F

Group 2 259.40 0.65 .52

PI 1 439.34 2.19 .14

Group X PI 2 141.32 0.35 .70

Error 74 14873.66

Total 79 15672.19

Hyp. 7: PI subjects, compared to contrast subjects,

will demonstrate greater self-esteem by

scoring higher on the Ziller Self-Esteem


The results of the ANOVA performed on the self-esteem

data are presented in Table 10. The results indicate no main

effects for PI or Group, with the significant interaction

(p < .05) seen as statistically spurious, based on the insig-

nificant Group X PI effect in the MANOVA discussed earlier.

Therefore, Hypothesis 7 receives no support from these data.

Hyp. 8: The PI subjects, compared to contrast sub-

jects, will demonstrate less orientation to

the past by having higher scores on the

Cottle Experiential Inventory.

Table 11 presents the results of the ANOVA performed

on the Experiential Inventory data. The results indicate no

support for Hypothesis 8, with no main effect for either

Group or PI; the Group X PI interaction was also insignifi-


Hyp. 9: The PI subjects, compared to contrast sub-

jects, will demonstrate more present and

future orientations and less of a past orien-

tation by more often being categorized as

present and/or future dominant and less often

past dominant on the Cottle Circles test.

Table 10

ANOVA For Group X Personality
X Self-Esteem


Source df SS F PR>F

Group 2 156.815 1.78 .18

PI 1 23.966 0.54 .46

Group X PI 2 270.901 3.07 .05*

Error 74 3259.809

Total 79 3713.550

*Note: Overall
was not
seen as


X PI interaction effect
this effect is therefore

Table 11

ANOVA for Group X Personality Integration
X Experiential Inventory

Source df SS F PR>F

Group 2 1.397 2.02 .14

PI 1 0.052 0.15 .70

Group X PI 2 0.491 0.71 .49

Error 74 25.526

Total 79 27.434

The results of chi-square analysis of the Time Domi-

nance categorizations, based on the Circles Test, are pre-

sented in Table 12. The overall X2 was insignificant, as

were the X2's for each Group analyzed individually. The data

give no support to Hypothesis 9.

Hyp. 10: PI subjects, as compared to contrast sub-

jects, will have higher scores on the

Personality Integration subscale of the

Tennessee Self-Concept Scale.

The results of the ANOVA performed on the Personality

Integration subscale of the TSCS are presented in Table 13.

The results indicate no significant main effect for Group,

although the effect approaches significance (p < .06). A

Duncan's Multiple Range Test performed on the data indicates

a direction of difference wherein the Cadet Group (X = 9.60)

scored lower than the Seminary Group (X = 12.25) and the

Fraternity Group (7 = 11.5), who did not differ significantly

from one another df = 74, MSE = 13.44).

The PI main effect did not reach significance, as

reported in Table 5 the one-tailed t tests for significance

between the PI and contrast group means, which differed in the

predicted direction, approached significance, reaching approx-

imately the .07 level.

The results, therefore, give some minimal support to

Hypothesis 10.

Table 12

Contingency Table for Group X
Time Dominance

Personality Integration X

Past Present Future None
Fr. Cad. Sem. Fr. Cad. Sem. Fr. Cad. Sem. Fr. Cad. Sem. Total

1 6 6 6 4 5 6 0 1 1 2 5 2 44

Total 18 Total 15 Total 2 Total 9
2 5 14 3 2 3 3 1 2 2 3 4 3 45

Total 22 Total 8 Total 5 Total 10

Total 40 23 7 19 N = 89

Note: 1 = PI Group; 2 = Contrast
1. PI X Time Dominance Categories, X2 = 4.83, df = 3, p = .20
2. PI X Time Dominance Categories X Group
For Fraternity Group: X2 = 2.95, df = 3, = .40
For Cadet Group X = 3.32, df = 3, = .35
For Seminary Group: X = 1.96, df = 3, p = .58

Table 13

ANOVA for Group X Personality Integration
X PI Subscale (TSCS)

Source df SS F PR>F

Group 2 77.093 2.87 .06

PI 1 33.810 2.51 .12

Group X PI 2 25.939 0.96 .39

Error 74 994.834

Total 79 1140.487

Tables 14-18 present the results of analyses for the

variables for which no theoretical derived hypotheses were

made. These variables included Meaningfulness (Table 14),

scores on the categories of the Lines Test (Table 15), GPA's

(Table 16), Age (Table 17), and Group Tenure (Table 18).

The results of ANOVA performed on the meaningfulness

data are presented in Table 14. The results indicate signifi-

cant mean effects for PI and for Groups. The Group X PI

interaction was insignificant. A posteriori analysis showed

that the PI group mean was higher than the contrast group.

Duncan's Multiple Range Test showed that the Cadet group

scored higher than both the Fraternity group and the Semi-

nary group; the latter two groups did not significantly

differ from one another (o(= .05, PF = 74, MS = 1631.41).

Table 15 presents the results of MANOVA analysis per-

formed on the scores for the categories of the Lines test.

The results show a main effect for Groups, with no signifi-

cant mean effect for PI. The Group X PI interaction was not

significant. Inspection of the ANOVAs for the separate

scores shows a significant Group effect on two scores: Life

Time (f = 3.43; p < .04) and Historical Future (f = 3.53;

p < .03). For the Lifetime scores, a posteriori analysis

(Duncan's Multiple Range Test) showed that the means for the

three groups were rank ordered as follows: Cadets >

Fraternities > Seminary, with the Cadet > Seminary comparison

as the only significant difference (( = .05, df = 84,

Table 14

ANOVA for Group X Personality Integration
X Meaningfulness

Source df SS F PR>F

Group 2 22032.281 6.75 .002

PI 1 7799.945 4.78 .03

Group X PI 2 2627.156 0.81 .45

Error 74 120724.027

Total 79 151910.750

Table 15

Results of MANOVA: Personality Integration X Groups X
Lines Test Variables

Test Hypotheses Pillai's Trace Statistic F (d.f.) Prob > F

No overall effect for
Group V = .240 1.82(12, 160) .05

No overall effect for
PI V = .036 0.49 79) .81

No overall effect for
Group X PI interaction V = .145 1.04(12, 160) .42

MS = 6.845). For the Historical Future scores, a posteriori

analysis (Duncan's Multiple Range Test) showed that the Group

means were rank ordered as follows: Seminary > Fraternity >

Cadets, with the Seminary > Cadets comparison as the only

significant difference, (c4= .05, df = 84, MS = 7.855). The

results indicate that the groups differed on the Lines Test,

with Seminary students, as compared to Cadets,recording sig-

nificantly shorter Lifetime line segments, and significantly

longer Historical Future segments.

Table 16 presents the results of ANOVA performed on

the GPA data. The results indicate significant mean effects

for Group and for PI. The Group X PI interaction was not


A posteriori analysis indicates that the PI subjects

reported significantly higher GPA's than the Contrast sub-

jects. For the Group data, Duncan's Multiple Range Test

showed that the Group mean GPA's were rank ordered: Semi-

nary > Fraternities > Cadets, with the Seminary mean GPA

significantly higher than the other two, which did not differ

significantly from one another (- = .05, df = 74, MS = 0.352).

It is likely that these results reflect institutional differ-

ences in grading policy, rather than "real" differences in

scholastic ability, between groups. Therefore, it is diffi-

cult to attach much significance to this finding.

The results of ANOVA performed on the Age data are

presented in Table 17. Results indicate a significant main

Table 16

ANOVA for Group X Personality Integration

Source df SS F PR>F

Group 2 5.147 7.31 .001

PI 1 3.369 9.57 .003

Group X PI 2 0.265 0.38 .68

Error 74 26.055

Total 79 35.522

Table 17

ANOVA for Group X Personality Integration
X Age

Source df SS F PR>F

Group 2 4.354 1.25 .29

PI 1 13.860 7.98 .006

Group X PI 2 0.972 0.28 .76

Error 74 128.532

Total 79 148.800

effect for PI, while neither the Group main effect, nor the

Group X PI interaction was significant. Examination of the

PI and Contrast means indicates that the PI subjects were

older than the Contrast subjects.

Table 18 presents the results of the ANOVA performed

on the Length of Group Tenure data. Results indicate a

significant main effect for PI, with nonsignificant Group

and Group X PI interaction effects. Inspection of the data

shows that the PI subjects have been members of their groups

longer than the contrast subjects.

Upon discovering the strong relationship between PI

and the GPA, Group Tenure and Age variables, which had not

been directly predicted, it was decided to use these vari-

ables as covariate in a "post-hoc" series of ANOCOVA's, to

investigate what effect these variables may have had on the

significant impact on the main effects reported in Tables 5

and 13, respectively. ANOCOVA analysis, with age, month,

and GPA as covariates of PI X Group for Complexity of Self-

Concept resulted in a significant main effect for PI

(F = 5.06, p .03) with neither the Group nor Group X PI

effect being significant. ANOCOVA for PI X Group for Mean-

ingfulness resulted in significant main effects for Group

(F = 4.52, p .006) and for PI (F = 6.64, p .002) with

the PI Group interaction remaining insignificant.

Scores for Temporal Relatedness, as measured on the

Circles Test, provide data which are ordinal in nature.

Table 18

ANOVA for Group X Personality Integration
X Group Tenure

Source df SS F PR>F

Group 2 308.489 1.26 .29

PI 1 2443.009 19.93 .001

Group X PI 2 438.095 1.79 .17

Error 74 9072.597

Total 79 12155.950


In order to investigate the relationship of scores on this

variable with the PI dimension, a point-biserial Spearman

rank order correlation coefficient (rho) was computed; the

results indicate no significant correlation between the two

dimensions (rs = .06, n = 90). Level of personality inte-

gration is apparently unrelated to the degree to which the

"spheres" of the time dimension (past, present, future) are

seen atomistically or interrelatedly as measured on the

Circles Test.



From the theoretical position developed in the intro-

duction, several hypotheses were generated and tested

empirically. Taken as a set, the hypotheses proposed that

the term "personality integration" is a meaningful and use-

ful theoretical construct. More specifically, it was sug-

gested that people could be identified as currently

functioning at different levels of integration and that

these levels would be characterized by predictable differ-

ences along personal and interpersonal experience-processing

dimensions related to the person-as-open-system model

described earlier. Further, it was hypothesized that these

differences between levels of personality integration would

be relatively consistent across groups of people who dif-

fered in the "content" of their life choices (i.e., the

differences would be independent of differences in value

orientation, life style, and life goals). This discussion

will focus upon the nature of support and confirmation of

these hypotheses presented by the present data, as well as

pointing to unsupported aspects of the position. Methodo-

logical strengths and shortcomings will then be examined,


followed by brief presentation of questions and issues to be

explored by future conceptualizing and research.

Possibly the most striking outcome of this study,

providing broad support for both personality integration as

a construct, and the Personality Integration Reputation Test

(PIRT) as a method, is the consistency of results across sub-

cultural groups. Subjects identified as highly integrated

did not differ from one another as a function of group mem-

bership; this is represented statistically by the group-PI

(personality integration) interaction not having significant

impact on the results, although there were multivariate and

univariate differences, distinguishing between subcultural

groups and between levels of personality integration. Fol-

lowing the open system model developed earlier, the process

which characterizes personality growth and integration can

occur in a variety of subcultural contexts, to the extent

that the groups do not exert forces which require closed-

system behavior from its members and/or the individual mem-

bers cannot successfully modify or eliminate such growth

stopping forces. Similarly, of course, less integrated

persons can also be identified in a variety of subcultural

contexts, and the functioning of these persons can be

expected to demonstrate less open--and more relatively

closed--system characteristics.

In the present study subjects were selected from

groups which differed from each other along several content

dimensions. The Allport-Vernon-Lindzey (AVL) data, for

example, demonstrate differences in value orientation. The

education and training provided at the three institutions

reflect differences in the career goals of its members, with

the seminary preparing its students for ministerial and

teaching work, the military academy having a strong scien-

tific engineering component to its curriculum, while the

fraternity members, enrolled in a large public university

were enrolled in major programs ranging from agriculture to

prelaw. Similarly, the content of life-style patterns was

noticeably different from group to group. Cadets wore uni-

forms, lived within a relatively structured time schedule,

and had parade training grounds in the center of their cam-

pus. The seminarians attended daily services, and lived in

an environment rich in religious symbolism and reminders,

with the Chapel in a central location on campus.

Fraternity functions included rival participation in

athletic competition, and inter- and intra-group social

activities, and they were housed in physical structures with

a large living-meeting-socializing area at their centers.

Certainly the physical/social environments, and concomitant

value orientations differed between groups, yet the highly

integrated subjects differed from the contrast subjects con-

sistently, across groups, on several relatively content free,

"process" dimensions. These included complexity of the self-

concept, GPA, Age, and Group Tenure, and, to a less clear-

cut degree, openness.

This group of variables lends itself to examination

as a set, with a pattern emerging which is consistent with

the open system perspective. The first and last variables

(complexity and openness) pertain directly to the theoret-

ically derived hypotheses, and within the limitations of

methodology discussed below, provide support for the rela-

tively greater "complexity" and "openness" attributed to

healthily functioning, open person/systems, compared to less

integrated and relatively closed person/systems. The GPA

and Group Tenure results suggest that along several dimen-

sions, similar to what Adler (1964) called life-tasks, the

PI group was highly successful at attaining its goals. In

work or career tasks, the higher GPA suggests successful

movement towards short term (course work) and long term

(career) goals. As noted earlier, Duncan (1964) ascribes

the higher GPAs found in PI subjects to more efficient usage

or information, rather than to a higher level of intelli-

gence, and his GPA/intelligence data support this sugges-

tion. This is consistent with the position developed here,

which suggests that personality integration is associated

with increased openness to awareness of, and ability to pro-

cess goal-relevant information from other, higher order

systems. The interpersonal success of the PI subjects is

pointed to by the Group Tenure data. Most simply stated, they

have chosen to remain in systems within which their style of

functioning has proven successful for meeting their needs


and goals. Certainly there are many potential motivators for

continued membership, which are not necessarily mutually

exclusive, including fear of leaving, and an unwillingness

to resist group maintenance forces.

However, the possible prominence of the social suc-

cess hypothesis for the group of subjects is supported by

the very nature of the PIRT nomination procedure. These are

the people seen by their peers as being the most interper-

sonally valued members of the group. This collective per-

ception forms the basis for supportive and powerful feedback

to the PI subjects that their interpersonal behavior is seen

within the group context, as highly consistent with success-

ful interpersonal goal attainment. In sum, a pattern of

increased complexity, openness to and efficient use of infor-

mation from the larger social systems) is associated with

successful life goal attainment by the individual person/

system from within the given group context. As long as the

individual's goals remain congruent with the group's values,

and its level of interpersonal functioning remains consis-

tent, then the powerful positive feedback system is main-

tained and developed, and the person continues his relationship

with the group.

The relationship between "meaningfulness" of construct

usage as measured by the Bieri, Atkins, Briar, Cobeck, Miller,

and Tripodi's (1966) REP Test and PI was not predicted, and

is worthy of brief, though admittedly speculative discussion

pertaining to both methodology and conceptualization of the

findings. The methodological issue relates to the difficul-

ties of interpreting the results of the Bieri grid data. The

variables which were predicted to distinguish between PI

levels, but did not, were scores on FIC (differentiation) and

ordination (integration), whereas, meaningfulness did.

An important point to focus upon here is the method

of analyzing grids which was employed. Landfield's (1977a)

procedure was developed to analyze how individuals made use

of their own (elicited) constructs. That is, he was examin-

ing a pattern of construing based upon constructs which had

emerged from the subject's own system, and was thus highly

individualized. In this study, however, the subjects were

provided with a set of construct dimensions to then apply to

people they knew. Regardless of how close the provided con-

structs are to their own personal dimensions, a case can be

made that the meaning of the task is qualitatively different

in the two situations. The "provided" format seems, in

essence, to examine how subjects apply a standard set of

stimuli rather than the personal constructs which emerge as

most appropriate to the individual for the set of role titles,

as in the "elicited" format. With the provided format, FIC

scores would then reflect the degree of independent usage of

dimensions which may or may not have personal relevance to

the subjects. Similarly, ordination scores may, in this for-

mat, reflect the degree to which those standard stimuli are


an integrated "subsystem" of the subject's own construct sys-

tem. Therefore, there would be no a priori reason to expect

the resulting score to accurately represent the integration

within the subject's entire personal construct system, or to

expect PI subjects to differ from contrast subjects on these

scores. The alternative explanation, that PI and contrast

groups simply do not differ on three complexity dimensions

must also be considered, but the previously reported work of

Thomas and Seeman (1971), Wexler (1974), and the other com-

plexity data from the present study would argue against such

a position. Returning now to the meaningfulness data, an

alternative explanation for the resulting difference in

extremity ratings is possible. It may be that these scores,

which represent the usage of extreme ratings in applying the

provided constructs to people, may be the best indication

from the grid data of differences in construing style between

the two levels of PI. High scores here may reflect greater

willingness to risk nontentative certainty in evaluating per-

sonal relationship along somebody else's dimensions. In other

words, these scores may indicate more efficient and confident

decision making and have little to do with the meaningfulness

or centrality of these constructs to the subjects' own sys-


It is also possible for extremity scores to represent

a "response set," rather than a series of meaningful choices.

It may be this factor which accounts for the higher mean

scores for the Cadet Group compared to the other two. That

there may be different factors which account for differences

between groups on one hand, and for difference between PI

level on the other, seems supported by the lack of statisti-

cal interaction between the main effects. In other words,

the difference between the PI levels was consistent across

groups, as was the difference between the Cadet and other

groups consistent across PI level, and thus the findings need

to be accounted for separately. This, of course, does not

rule out the possibility that a common principle accounts

for both difference, but does lend credence to the discussion

of separate explanations for the two findings.

The Time Orientation data revealed no differences

between PI and contrast subjects. Three alternative explana-

tions for the findings, which focus upon the hypothesis, the

instruments, and the conceptualizations of the problem,

respectively, will be discussed.

The first, which is the most direct extrapolation from

the data, is that there are no differences in orientation

towards time dimensions which are related to level of PI

(i.e., the hypotheses were simply incorrect). Accepting this

alternative, however, is also an implicit acceptance that the

instruments used were adequately measuring what they were

supposed to measure. As described earlier, the validation

process for these devices has been minimal, and their usage

in the study was exploratory, and in a sense, an attempt to

provide further validation for the instruments. That the

devices did not discriminate between the PI levels as pre-

dicted leaves the validity question open. The hypothesis may

still be viable, but the instruments were a poor empirical

test; the instruments may have been adequate, but the hypothe-

sis simply not supported. Of course, both the instruments

and hypotheses may have been poorly constructed; the only

clear result is that the hypotheses, as tested, were not sup-


Another perspective on these data is also possible,

and is possibly the most plausible. It may be that the

hypotheses, as worded and tested, were poorly conceptualized

in relation to the systems theoretical position. As dis-

cussed earlier, open systems were described as goal oriented;

that is, they move towards goal achievement rather than func-

tioning purposelessly. Then, well-functioning systems, rela-

tive to poorly functioning system, might be expected to (a)

have explicitly defined short and long term goals, (b) have

explicit plans or strategies for achieving these goals, and

(c) be behaving in accordance with these plans. Thus,

although the goal directed behavior described here is, in a

sense, future oriented, it is not necessarily tied to beliefs

about time, or to the relative importance attributed to the

past, present, or future time zones by subjects. Thus, for

example, a person may demonstrate an active goal-orientation

while attributing a great deal of value to his personal and/or


cultural past as the source of his creative ideas or success-

ful behavior. On the Cottle (1976) instruments this person

would be likely to have a relatively low experimental inven-

tory score, he might be categorized as Past Dominant, and he

might draw long personal and/or historical past line segments,

relative to his future segments. His scores on the instru-

ments would be an accurate representation of his beliefs

about these time dimensions, yet would not reflect his

activity future-oriented and goal relevant behavior.

Indirect support for this position comes from two

data sources. The first is from the time instruments them-

selves, specifically the Lines Test. Though not predicted,

there were differences between Groups, with the Seminarians

having the shortest Lifelines and longest Historical futures

compared to the other groups. These data make intuitive

sense when considered as indicative of beliefs about the rela-

tive importance of time after death, where the Seminarians

would be expected to have more clearly defined and highly

valued beliefs about the extended time after one's personal

death compared to the other groups. This finding can then be

seen as supportive (though in an admittedly post-hoc fashion)

of the proposition that the Time Instruments are measuring

beliefs about time, rather than providing an indication of

the degree to which current behavior is oriented towards

explicating and achieving future goals. The second source

is the GPA data discussed earlier. To the extent that GPA's

reflect degree of success in short and long term goal rele-

vant behavior, the PI subjects compared to the contrast sub-

jects were clearly more successful than contrast subjects,

though they were not necessarily more intelligent (see dis-

cussion above), nor did they hold consistently different

beliefs about the relative importance of the future, past, or

present, as measured by the Time Instruments. In sum, then,

the results derived from the Time Instruments do not support

the hypotheses as worded. Though no conclusive explanations

can be offered, the need for more sophisticated conceptuali-

zation and operationalization of Time- and goal-related

hypotheses relevant to open systems theory is suggested as

important consideration for future work.

The data concerned with Hypotheses 7 and 10 (the

self-esteem and PI subscale instruments) remain to be dis-

cussed. Neither of these hypotheses were supported by the

data. The self-esteem results are particularly surprising,

in that the PI subjects, in comparison to the contrast group,

are both perceived as more successful interpersonally by

their peers (demonstrated by their PIRT nominations) and are

experiencing greater academic success, as shown in GPA's.

Given these "ingredients," it makes theoretical sense that

the PI subjects would demonstrate higher self-esteem, as they

have other studies (Fitts, Adams, Radford, Richard, Thomas,

Thomas, & Thomspon, 1971). Such counter-intuitive results

suggest four considerations. The first, which is

theoretically unlikely, is that self-esteem is not related

to personality integration, and the data are accurate repre-

sentations of this independence. The second possibility is

that the PIRT instrument is not a good instrument for select-

ing groups of subjects differing on level of PI; the construct

validity data cited earlier, as well as the other results of

this study, however, provide support for the PIRT method. A

third possibility would be that the instrument does not ade-

quately measure the dimension, and thus the results would not

be expected to differentiate PI levels. The validity data

for the esteem instrument presented earlier, however, suggest

that a simple "no validity" disclaimer is not sufficient. The

fourth consideration, which seems more likely, is that the PI

and contrast groups did not differ enough in degree of per-

sonality integration for this self-esteem instrument to dis-

criminate between the groups. As mentioned earlier, the

contrast group members were not an "extreme" group, since all

had at least one PIRT nomination. The esteem instrument's

validity data, however, were based on relatively extreme

groups (e.g., sociometric "stars" and "isolates"), and extreme

situations (e.g., political winners and losers) (Ziller,

1973), and it may simply not be sensitive to differences

between nonextreme groups such as these. An empirical

exploration of this issue would be possible by employing the

Ziller instrument along with other self-esteem instruments

which have been used to measure subtle differences or changes

in self-esteem, such as the complete TSCS, and/or a "self-

ideal" Q-sort procedure, to see if the results from alterna-

tive instruments do indicate significant differences as

predicted, and if the Ziller instrument remains consistent in

not detecting differences.

The PI subscale of the TSCS was employed as a "valid-

ity check" under the hypothesis that PI subjects, selected

by the peer nomination PIRT method, would score higher than

contrast subjects on the written scale format of the TSCS.

The results provide marginal support, with the differences

between levels of PI being in the predicted direction, and

approaching, but not reaching, the designated level of sig-


Again, there may be more than one reason why the

difference between PI levels was not as strong as had been

predicted. First, the effect of utilizing the subscale items

separated from the entire TSCS is an unknown factor, which

was necessitated by the time limitation discussed earlier.

Second, and probably more influential, is the difference

between the subject selection method used in this study and

the procedures used in previous studies utilizing both the

TSCS and PIRT. These have tended to base their selection of

high PI subjects on nominations from more total subjects

divided into more subgroups than the present study. For

example Seeman (1966) used 695 initial subjects from 16 sub-

groups (dormitories, sororities), resulting in a sample of

23 PI and 20 contrast subjects. Thus, his selection of PI

subjects focused upon, roughly, the two subjects from each

subgroup receiving the most nominations. This was consider-

ably more selective than the present study, in which more

subjects from each of the considerably smaller number of

groups (three) were included in the PI sample.

Additionally, Seeman's (1966) contrast group was

selected via a random selection process of all 695 subjects,

whereas the present study selected contrast subjects only

from those receiving at least one nomination. This latter

procedure is considerably more restrictive, effectively

excluding the very poorly integrated subjects from the sam-

ple, whereas Seeman's study may well have included relatively

poorly functioning people in his contrast group. In other

words, the selection process in this study worked against

finding PI-Contrast group differences by the systematic

inclusion and exclusion, respectively, of subjects who might

have influenced the degree of difference between group means.

This argument, of course, holds for the other dependent vari-

ables too, and provides additional impact for the differences

which were found between levels of PI, since the selection

process was working against, rather than towards, finding

significant differences.

Exploratory studies of this type, which make use of

a variety of instruments in several population groups while

investigating abstract constructs such as personality

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