Title: Some personality characteristics of frequent and infrequent visitors to a university infirmary
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00097976/00001
 Material Information
Title: Some personality characteristics of frequent and infrequent visitors to a university infirmary
Physical Description: viii, 158 leaves : ill. ; 28 cm.
Language: English
Creator: Wharton, Mary Charlotte Ince, 1931-
Publication Date: 1962
Copyright Date: 1962
Subject: Personality   ( 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 101-109.
Additional Physical Form: Also available on World Wide Web
General Note: Manuscript copy.
General Note: Vita.
 Record Information
Bibliographic ID: UF00097976
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 - 000568353
oclc - 13647178
notis - ACZ5079


This item has the following downloads:

somepersonalityc00wharrich ( PDF )

Full Text





June, 1962

Copyright by
MYary Charlotte WharEton


The writer wishes to express her appreciation
for the guidance and advice given her by the members

of her supervisory committees Dr. Sidney M. Jourard,

chairman, Dr. Albert K. Kurts, Dr. James C. Dixon,

Dr. Richard J. Anderson, and Dr. George R, Ba~rtlett.

She would also like to thank Drs. Ben Barger, Henry

C. Schumacher, and Samuel Wlright for their permission

to use the Student Infirmalry records, their help in
contacting subjects, and their most kind assistance

in giving her the opportunity to plan and execute this

study. She would also like to express her appreciation
to the students who cooperated as subjects in this




ACKNOWLEDGMENT~S. . . .. .. . .. iii

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

LIST OF FIGURES. . .. . .. .. .viii



The Psychosomatic Concept
Formulations of the Psychosomatic Process
HolismI and Field Theory
Pertinent Research
The Problem to be Investigated

II. METHOD. .... ..... ....... 32

Bypoth~EE~~EE e sesE~~~~EEE

III. RESULTS. . . . . . ,. .1

IV. DISCUSSION .. ... . ,. . .. 83

The Data and the Bypotheses
Psychological Imma~turity
The Conflict Situation and the Stress
It Produced
The Mlechanism or Process of Psychosomaatic
Sex Differences
Implications of the Findings for Theory
Unanswered Questions

V. SUMMARY. . . ,. . . .. 97

REFERENCES ** ** ** ** ** *. ,., 101


APPENDIX B ** ** ** *. ,.. 116

APPENDIX C ** ** ** *. ,. .. 125


Table Page

1. Religious Preference of High, Mean and
Low Users 37

2. Findings Pertaining to Hypothesis 1 . 45

3. Findings Pertaining to B~ypothesis 2 . 46

4. Findings Pertaining to B~ypothesis 3 .. 49

5. Findings Pertaining to HIypothesis 4 ,. .. 50

6. Findings Pertaining to H3ypothesis ( . . 1

7. Analysis of Variance of MMPI Scale 9 (Ma) $4

8. Kind of Discrepancy between Level I (MIMPI)
and Level II (ICL) Self-description ., 9

9.Level I and Level V Self-descriptions of
Female High, Mean and Low Users . .. 6;0

10. High, Mean and Low Subjects' Choice of
Name or Student Number. 62

11. Product-Moment Correlation Coefficients
between Morale-Loss and Self-Disclosure . 64

12. MMPI Profile Defined Levels of Self-develop-
ment for High, Mean and Low Users 67

13. Analysis of Variance of Famnilial Discord
Scale . . .. .. .. .. 69

14. ICL Description of Mother as Loving or
Hostile . . . . . . . . . 70

15. Infirmary Visits of High and Mean Users
During the First 30 Days of the Preshman
Year . . . . . . . . . 72

16. Analysis of Variance of Complaint/Visit
Ratio Scores of High and Me~an Users ,. .. 73

17. Rank Order of Types of Complaints for High
and Mean Infirmary Users ......... 74

Table Page

18. Number of Visits M~ade to Infirmary During
the Second School Year for all Subjects
Continuously Enrolled ... .. 76

1. Self-Ratings of General Bealth . ... 76
20. Analysis of Yariance of Total ACE Scores. 78

21. Analysis of Variance of Grade Point Average
at the End of the Freshman Year . .. 78

22. Sorority and Fraternity Membership at the
End of the Sophomore Year . ., ,. 81

23. Analysis of Variance of the Number of Extra-
Curricular Activities, Other than Sororities
and Fraternities* *~ ** *. . .. 81

24. Weekends at Home Corrected for Distance from
University to Hometown. 82

25. Contacts with Prospective Subjects and Their
Responses **********....... 117
26. Analysis of Variance of MP~I~ Morale Loss
Scale ** ** ** *. . .. 117

27. Analysis of Variance of MMPI Scale 1 (HTS) 118

28. Analysis of Variance of MMPI Scale 3 (HY) 118

29. Analysis of Variance of MMPI Scale 4 (PD).. 119

30. $-Ratios between Experimental Groups and
same Sex Freshmen for 11 MiMPI Scales. . 120

31. Analysis of Variance of Welsh's Internaliza-
tion Ratio ** ** *. ,. .. 121

32. Analysis of Variance of ICL Intensity Scores,
Level II Self ********* 122

33. Amlount of Discrepancy between Level I (MMPI)
and Level II (ICL) Self-description .... 123

Table Page

34. Analysis of Variance of Self-disclosure
Scores for Four Targetas Mother, Father,
Same Sex Friend, and Opposite Sex Friend. .. 124

3. MMPI Standard Scores. . . .. .. 126

36. Morale Loss, Familial Discord, Internaliza-
tion Rattiot and Self-development Scores . 129

37. ICL Intensity Scores, Level II-Self . 132

38. ICL Octant and Level I-Self (MM~PI) Scores . 135

39. Self-disclosure Scores. . . . . . 138

4O. Total Medical Visits Freshman Year, First 30
Days of Freshman Year, and Sophomore Year .. 141

4l. Type of Complaints Complaint/Visit Ratio,
and Self-ratings o General Health. 144

42. ACE Total Scores and Freshman Grade Point
Averages. .................. 148

43. Religious Preference, Name or Student
Number, and Sorority/Fraternity Membership. 151

44. Weekends at Home Corrected for Distance,
and Number of Extracurricular Activities. 155



Figure Page

1. Visits to infirma~ry by 477 freshmen .. 34

2. Visits to infirmary by 315 freshman men and
162 freshman women... 36




The Pssycosomatic Concept

"Psychosomlatic" is a new word in mtany vo-
cabularies but one which is becoming increasingly

fashionable. Perhaps because it refers to a concept

which outs across lines of discipline and profession,
one finds a bewildering confusion of meanings and
connotations connected with it in the literature. As

Hartman (1955, pp. 591-592) points out, this term has
been variously conceived to apply to a particular

medical philosophy, to a conjoint research approach,
or to the application of psychological principles
and methods in certain forms of somatic sickness.

Few have addressed themselves to the questions re-
garding basic views of human behavior which are
raised by the proliferation of data on "psychosomatics."
At least they have not gone beyond a consideration
of the narrow range of behavior called sickness
(Mead, 1953).

The relation between the mind and the body is

a question which has intrigued man for centuries.
In medicine and biology, great thinkers from Galen
to Bernard have discussed this relationship (Green,

19$'1; Foster, 1899). Looking backward, one may see
that they seemed to have amazing insights. However,
the facts are that their insights were not accepted

by their colleagues nor by those of many generations
to follow (Galdston, ~i194 IBkrtman, 1995; Kaplanz-ed

Kaplan, 1997). As outlined by Kruse (1951+), theories
of the etiology of disease developed along three

lines, each independent of the others and each con-
sidered to be a necessary and sufficient cause of

specific diseasess Only recently has it become evident
that noxious agents (germs), stress, and deprivation
are inextricably interwoven in their relationships to
each other and to the development of disease sytmptomns.
Kru~se discusses the history of the study of beriberi
to show that workers will hold tp treasured theories
of single causes of disease even in the face of facts

to the contrary. Today there is little doubt that
some kind of relationship exists between the soma
and the psyche in the etiology of sickness. There
are dramatic demonstrations of this relationship such
as that of Spitz (1954) who studied 91 infants in a

fouhdling home. Details of this study will be pre-
sented later. However, a quotation from the summaary
of this paper illustrates the point being made here.
We feel justified in claiming that the depri-
Vation of emotional supplies, at least in the
period of early infancy, is a destructive
stressor agent. It provokes the general
adaptation syndrome as surely as does nutri-
tional deprivation and it leads to the same
consequences (p. 131).

The general adaptation syndrome referred to is from

Seyle's work (Seyle, 1990).
A perusal of the literature reveals that many
of those who write about psychosomatics have not

accepted the unity of the orga~nism, somatic and

psychic, which is indicated by their data. Rather,
they work from the premis that there is a definable
subclass of illness which can be termed "psychosomatic"
and that this is different from other illnesses. Thus,
one finds lists of psychosomatic diseases such as that
of Halliday (1953) which are outdated almost as soon

as they appear. This is because other investigators
are continuously discovering psychological factors
in the etiology of illnesses previously thought to be
purely somatic. Or one sees studies such as that of
Waxenberg (1996) demonstrating that patients with a

"psychosomatic" ailment are no different psychologically

from those with a "somatic" one. At the same time

other investigators had placed his so-called "somatic"

ailment among the "psychosomatic" ones. In general,
what one finds is that the great majority of studies

are done from the single cause point of view with

emotionrs, attitudes, and other psychological variables

added as ancillary and minor factors. The present

study is based upon holistic, field theory principles.
Therefore, those studies whithiinvestigate illness in

general, or the behavior of the total organism under
conditions of stress, deprivation or invasion by noxious

agents, are the most germane and will be most carefully
reviewed. HKowever, a short summHeary of the mnain
theories about the process or mech~anis of psychosomatic

illness together with somne representative studies will

be presented.

Formulations of the Psycho~somlatic Process
A number of writers have proposed formulations

of how the psychosomatic process takes place. These

may be divided into three main types, each to be dis-
cussed in turn. This division of theories into three

types follows; the divisionse made by Kaplan and Kaplan

(1957) and Mendelson, fHusch, and Webber (1956).

The first type of formnulation includes those
theories which posit that specific emotional conflicts

(Alexander, 1950), specific personality patterns
(Dunbar, 1938), or specific psychological traumata
(Greenaere, 1952) are followed by specific physiological

dysfunction. In general, the evidence does not support
such specificity. As Bartman (19TS) points out,

piecemeal correlation of specific psychodynamnics
with symPlrptoms of disease has produced data that are
contradictory and do not differentiate consistently
or reliably among patients with different diseases.
As both Ha~rtman and Mendelson f2 .21* point out,

specificity of relationships has too soften been a
conclusion based upon failure to look for relationships

other than those being hypothesized. Studies such as
those of Waxe~Xnberg (1995) and Buck and Hobbs (1959)

offer experimental evidence for the nonspecificity of

the psychosomatic process. This is not to say, of
course, that sickness and symptoms may not have

particular meanings to particular patients as ex-
pressions of feeling and attempts to communicate. A
number of single case analyses such as those of Hartz

(1956), Galdston (1956), and Kaplan and Gottschalk

(1958) offer evidence that this is true. There is a
mounting body of evidence which supports the notion
that certain individuals have idiosyncratic "body

language" (Ruesch, 1957) which they use as an avenue
of emotional expression and communication. Ma~lsro

and Shagass (1949) found, for exuample, that patients
who had a history of cardiovascular symaptoms or head
and neck pains tended to react to stress through the
cardiovascular or mlusculoskeletal systems even though

they were suffering no symptoms at the time. Fisher
and Cleveland (1959; 1996), in several carefully

designed studies, showed that the locus of the symnptom~s
developed in several different diseases was related

to the image which the person had of his body as

something~ which was surrounded by an inpenetrable
barrier (exterior sym~aptoas) or some~9thing easily pene-
trated (interior symr~ptoms). These classifications

were made on the basis of certain responses to the

Rorachach plates. Studies such as that of Richmaond
and Lustman (1995) suggest that such individual dif-

forences in tendencies toward body language mlay be

present very early in life. What seems unlikely, at

the present time, is Alexaunder's statement that a
particular somatic symaptom is always the physiological

consequence of specific sequences of spcific uncon-
scious conflicts (Alexanderi 1950), The "ulcer type"
seems to be an oversimplification of the facts.

Symptops as Regressipp
The second type of theory is that which states

that psychosomatic symaptoms are regressive phenomena

signifying a return to an earlier mode of behavior
(Ruesch, 1948; Alexander, 1960; Margolin, 1954; Szass,
1992). Kaplan and Kaplan criticize this point of view
on the basis that the phenomena being studied are

the product of complex learning and, therefore, a
regression hypothesis is not necessary. Experimentally,
there is very little evidence one way of the other.

Szass's hypothesis that parasymrpathetic activity is

paramount in psychosomatic illness is questionable,
according to Mendelson 1 L* (1956). Generally, the
only evidence offered for this point of view is
similar to that presented by Ruesch (1948) who showed
that people suffering from certain diseases tended to
behave in a manner which he judged to be immrature. B~ow-

ever, he stressed immr~aturity, a slightly different con-
cept from regression.
Psychoanalytic Views
The third type of theory represents an extension

of Freudian psychology into the psychosomatic field.

In this view a somatic symptom is a discharge through

the autonomic nervous system of psychic energy which

is "blocked" from discharge through motor channels.

In other words, a somatic symptom is like a psycho-

logical one in that both are expressions of repressed

impulses (Freud, 1938; Alexander, 1990; Ruesch, 1967;
Grinkser, & Spiegel, 1945). Grinker and Spiegel express
this point of view as follows

...when we speak of psychosomatic distur-
~banlces it is usually with reference to condi-
tions in which persistent or recurring emotion
is only recognized Z~ie 7 through those physi-
cal activities that normally accompany that
emotion, consciousness of the emotion in the
form of subjective feeling being absent. It
is a state of affairs in which nervous energy
is in part or wholly expressed through the
vegetative nervous gsstem because somle psycho-
logical barrier prevents the person from ex-
pressing the feeling at the conscious or beha-
Vioral level. The emotion is repress-ed and only
the lower-level visceral concomitants are ex-
pressed (Grinker, &e Spiegel, 191+T, p. 253).
The evidence for this theory is considerable,

perhaps because experimenters have frequently formu-
lated hypotheses derived from this theory. For example,
Ruesch and Bowman (1953) found that the rate of re-
covery from chronic disease was related to the

presence or absence of certain repressed emotional
conflicts. Mirsky, Thaler, Weiner, and Reiser (1956)

found the development of peptic ulcers in 107 army
inductees to be associated with, among other things,

"persistent oral dependency wishes in conflict writh
environmental sources of gratification and oral
hostility which is unexpressed" (p. 514). In their
review of the literature on cancer and personality,
LeShan and Worthington (1996) list inability to

express hostility as a factor consistently found
in patients who develop malignancies. Thaler, Weiner,
and Reiser (1957) found that patients ill with alters

or hypertension tended to project certain attitudes
and emotions into unstructured test stimuli in a

way which distinguished them from~ each other and
from control groups. Miller and Baruch (1956) com-

pared 100 children suffering froma asth8an to 60
nonallergic children. They found that the asthmratic
children were unable to express hostility to their

parents. This was not true of the control subjects.
Calden, Dupertius, Hokanson, and Lewis (1960) found
that the rate of recovery from tuberculosis for 316
male patients was related to their ability to openly

express feelings, especially aggression and ~hostility,
and that the "good, cooperative" patients tended to
be slower 1 recovering. Studying the rate of re-

cover from thyroidectomy, Ruesch (1947) found
patients slow to recover were significantly more
rigid, conscientious, and dutygroriented than were
those who recovered more rapidly. In a study of 22
women who had undergone surgery for cancer of the
cervix, Shrifte (1960) hyrpothesized that recurrence
of the tum~or would be related to underlying un-

pleasant feeling tone. Her hypothesis was not
supported. All subjects made dysphoric responses
on the Rorschach. However, the tendency to live
inwardly, as indicated by the Rorschach, did
distinguish between the two groups. Lieberman,
Stock, and Whittma~n (1969), Leary (1957), Raifrman
(1957), and Thaler 4kf Al* (1957) have all demaon-
stratedl differences between the conscious self-

perception of ulcer patients and the kinds of
feelings and attitudes which they project into
unstructured material. Feelings of dependency
and a need for intimacy were not expressed in overt
behavior but were expressed in fantasy behavior.
Two studies deserve special mention because

they represent a more direct test of the psycho-
analytic theory (Learmnonth, Ackerly, &e Kaplan, 1959;
Frankly, 1952). Moreover, since the subjects of


both studies were healthy, the results were not con-

taminated by the effects being sick may have upon

psychological variables. Both found a relationship
between being what Learmonth et iAl* call an "inner

expressed" and what Frankle calls "emotional intro-
version" and the tendency to use body language. In

the first study, 20 nursing students were given the

MM4PI and the Borschach. They then submitted to

three different types of stress situations. GSR

readings were made during the stress situations.

Results showed a significant positive relationship
between fluctuation of the GSR during stress and

high scores on indices of "restraint and curtailment
of unpleasant or prohibitive feelings and actions"

(p. 153). Using 75 normal graduate students, Frankle
found a significant relationship between a tendency

toward emotional introversion, as measured by the
Emotional Introver sion-Extroversion Scale of the

Minnesota T-S-E Inventory, and incidence of somatic

complaints checked on the Cornell Medical Index.

Many studies showed, however, that inhibition
of unacceptable affect per so is but one of the many

variables involved. For examples, Mirsky at Apl. (1956)

demonstrated the quite complex interrelationship


existing between serum pepsinogen, unexpressed
"oral hostility," and environmental stress in the
development of duodenal ulcers. Since these

variables were measured before the subjects developed
ulcers, one may place relatively high credence in
their conclusions. They says "Neither high rate

of gastric secretion or psychological findings alone
are responsible for ulcer development. Together,
however, these parameters constitute the essential

determinants in the precipitation of peptic ulcer
on exposure to social situations noxious to the

specific individual" (p. $14). The complexity of
variables and their interrelationships will be
discussed more fully below.

Holipm and Field Theory
The underlying finding which is repeated over
and over in the studies reviewed above is that the

human organisma is a unity, a whole. It is not at body
which is invaded by other organisms noxious to its

functioning, but a person who becomes ill. An
individual's attitudes toward himself and other

people, his relative rigidity or flexibility of self-
control and self-expression, the degree of maturity

of total development he has reached, idiosyncratic


patterns of functioning of body organs and systems,
learned patterns of communication and expression

through his body, and the degree and suddenness
of stress impinging upon him have all been shown

to be related to his state of health. That is,
whether he functions in a relatively well organized

manner and approaches a steady-state, or whether
he is relatively more poorly organized and closer

to entropy and death. A field theory of man as a

total system seems to be the best model to encompass

all the data now available. Other data, not included
in that discussed above as pertinent to the three

main theories of the psychosomatic process, would
seem to be outside the scope of the various theories

discussed and explainable most coherently on the
basis of a field theory. These include the growing

literature on the relationship between illness and -

stress (Seyle, 1990; Janis, 1958) Hinkile, 1959;

Holmes, 1956; Weisz, 195~7) Rezntedrfl, 195'7; Bradyl 195'8;
Rothstein, &e Cohen, 1958; Luby, Ware, Sent, &c Froh-

man, 1959; Spitz, 199i4). Also included are the studies
which demonstrate a significant temporal relationship
between subsequent development of illness and pre-

ceding loss of morale (Canter, 1960), feelings of


hopelessness and helplessness (Schmale, 1958),
object loss (LeShan, &r Waorthington, 1956; Greene,
1958), loss or threatened loss of parental love
(Miller, &r Baruch, 1956), feelings that life is a

painful duty rather than a challenge (Hinkle, 1968),
and depression (Lewin, 1959). A number of studies

suggest that whether a person is able to openly
express his feelings is not so important as is the
kind of dyadie or triadie interpersonal system he is
involved in with significant other persons. For

example, Greene and Miller (1958) found leukemia
developing in children who had had a certain kind of

relationship with their mothers, while, preceding
the child's illness, the mothers developed symptoms
of anxiety or depression. Thaler et AL* (1957)
demonstrated that ulcer and hypertensive patients
were constricted and rigid in their expression of
affect in interpersonal situations, particularly the
doctor-patient dyad. Miller and Baruch (1956),
Jessner, Lamont, Rollins, Whipple, and Prentice
(1955), and Long, Lfamont, Whipple, Bandler, Blom,
Burgin, and Jessner (1958) all demonstrated to a
significant degree the presence of a particular kind
of relationship between the asthmatic child and his


mother, This was a relationship in which the mother

rejects the child because she feels rejected by her

own parents, the child reacts with rage conflicted
with fear, and represses his rage. LeShan (1967)
found a particular life-long pattern of interpersonal

relationships which characterized 20 patients with

Hodgkin's disease as compared to normal controls.

The present study is an investigation of the
relationships; between illness and certain variables

of self-description, interpersonal relationships,
and emotional expressivity, as predicted by field
theory. A brief d~escription of the theoretical
model upon which it is based will be presented,

together with a discussion of how this model accounts
for the body of evidence discussed above. Much of

this discussion follows writings of Grinker (1953),

Bertalanffy (1950), Koffka (1935), Galdston (1954),
and Miller (1959). The human organism is seen as

an open system, behavior being the final outcome of
a field of forces, both "psychological" and "somatic."
Therefore, both ktinds of forces are always effective

although in varying proportions. According to this

view, the invasion of the body by bacteria might be
the necessary but not the sufficient cause of symptoms


of disease, there beint other factors, e~g., "psy-
chological" ones, which must also be present in the
field. The words "somatic" and "psychological"
m~erelyr designate two points of view which the ob-
server may take in viewing the total system, or they

may refer to two "foci of organization" (Grinker,
1953) among many such foci or subcenters of organiza-
In field theory what we call 'parts' are not
to be conceived as separate, independent,
randomly self-acting entities. The 'parts'
are what we have selected for observation
within a dynamic organized complex .. .
(Frank, 1951, p. 907).
Therefore the accepted difference between matter
and the mind, between body and psyche may be
considered not as essential but as a product
of our particular capacity to know (Seguin, 1950,
p. 30).

The relationship between these subsystems, e~g.,

psychological and somatic, is not one of cause and
effect chains. Such a concept of causation is
meaningless since any "cause" produces its "effect"

by acting through the whole.1 Such cause !and

1This is, of course, especially pertinent in
understanding and predicting behavior in molar terms.
Cause and effect chains mtay haves limited value as in
analyzing molecular samples of behavior out of con-
text of the total field in which they take place.


effect analysis must be replaced by an understanding
of transaction processes within the system and be-

tween the system and other, larger systems of which

it is a part (Ruesch, & Bateson, 1949). A question
such as, Is the production of "'somatic" disease

by mental "cause" possible?, becomes meaningless
since it implies a one-to-one cause and effect

relationship. Rather, one must ask what is the

nature of the relationship expressed in the total

organization of the system. It is rarely possible
to study an entire living system at once, although
this task is being approached by M~iller and his co-

workers (Mill1er, 1995). Therefore, certain foci of

organization may be chosen for study keeping in mind
that any conclusions about their ultimate effect

upon the total behavior of the system must be tenta-

Somatic illness may be seen as molar be-
havior which is one outcome of field forces. It

represents a tendency away from good organization or
Praenanz toward entropy or disorganization. There

are many forces which interact to produce the dynamic

condition called a steady-state, or homeostasis. This

steady-state encompasses not only physiological


processes, such as those described by Cannon (1932)
and others, but also processes of balance within the

psychological subsystems of the organism and between
the organism and other systems of which it is a part.
Stress is any variation in external conditions which
is too sudden or excessive for the system to handle

(Galdston, 19549 p. 12). Or defined from the point
of view of the organism, it is a stimulus which

brings about the general adaptation syndrome (Seyle,
1950). The ability of a living system to regain
its equilibrium or steady state is partially a

function of the ease of communication among its

subsystems (e~g., amount of repression), its ability
to handle the load of information coming from outside,
the amount of supplies coming from outside (de-

privation or excessiveness), and its ability to call
into play forces which restore equilibrium by
counter~mcting disrupting stress (defense mechannisms).

One way a system can handle stress is to regress to

a simpler, less differentiated level of organization.
This m3ode of adjustment is similar to the regression

theory discussed above. Failure to restore equilibriuma
leads to what Seyle has termed the third stage of

exhaustion and breakdowniof the general adaptation syn-


Spitz's study of infants in a foundling home
is a good illustration of system functioning under

conditions of psychological stress (Spitz, 1954).
He studied 91 infants who were separated from their
mothers at three months of age and placed in the
home where one nurse attended ten infants. Their

development over one year was compared to that of

123 infants cared for by their own mnothers. Al-

though the physical needs of these infants were
attended to, they received little affection or

attention from the nurses. Spitz describes them as

severely deprived of "emotional supplies." At first
the infants countered by being demanding and "weepy."

By the end of the second month of separation from
their mIothe~rs, they had begun to lose weight and to

become arrested in their psychomotor development.

Development in social ability was not arrested but

showed an absolute increase in these early stages,
an example of the s~ystemll's direct attempt to comnpen-

sate .fori deprivation. By the end of the third month

of separation, the infants refused contact with others,

had insomnia, presented rigid facial expressions, and
were liable to infections and intercurrent ailments.

By the fifth month of separation, development began


to regress precipitously and irreversibly. By the

end of the first year of separation, 26 per cent of
the children were dead and the majority of the total

sample had not learned to walk or talk.
Four variables which affect the ability of a

system to regain a steady state were listed above.
Each of these is represented in the literature
already discussed indicating that field theory

principles can account for the experimental evidence

on psychosomaitic illness. For exramlple, repression
of certain attitudes and emotions corresponds to the
amount of communication among subsystems. Demands
and obligations placed upon the individual fromr the

environme~knt corresponds to information coming from

the outside which the system must handle. Depriva-
tion of emotional supplies corresponds to the amount
of supplies coming into the system. Feelings of low

morale, hopelessness, and helplessness correspond to
failure of the system to regain equilibrium.

Pertinent Research
As mentioned above, the majority of studies
labeled "psychnosometion~ have studied patients with a

certain disease considered to be "psychosomatic" with
the intention of adding another variable to the


symptom syndrome. Thus, ulcer patients not only
have gastrointestinal pain but also repress tender

feelings. Since this study is concerned with illness

behavior in general (Parsons, 195S3) or general

susceptibility to illness (Hinkle, 1958), pertinent
studies are those concerned with illness in general

rather than with a specific disease.
Hinkle and his associates have done a number

of investigations of large groups of subjects analyz-

ing their health records over long periods of their
lifetimes. Their study of 1700 American working

women (Hinkle, 1959) is representative. All of

these women were employed in the samne factory; 1297
of them had completed one or more years of service

and had complete records. Most were second generation
Irish and Italians of lower middle class status

living in a narrowly circumscribed neighborhood of

New York City and subjected to very similar conditions

of climate, diet and living conditions. All had been

examained and declared healthy when first employed.

Three hundred and thirty-six of these women had a

greater than twenty year record of unbroken employ-
ment. Ninety-six of these were selected for intensive

study. Twenty-five per cent of these subjects had


52 per cent of all the episodes of disabling illness.
Wkhen they were ranked according to the number of

episodes of illness and also the number of different
kinds of illness, a correlation of .83 (p(.01) was
obtained. A similar correlation was obtained when

the number of episodes of illness was compared to
the number of organ systems involved and to the number

of etiologies of disease. There was also a tendency
for the subjects who had the greatest number of

episodes of illness to report the greatest number
of periods of mood-thought disturbance. Hinkle
draws the conclusion that these women differed in

their susceptibility to illness in general. Be also

states that this finding, of a general susceptibility

to illness, has occurred in every group he has studied

(Hinkle, & Wolff, 1957). A more intensive study of
individuals at either extreme of the distribution of
illness revealed no significant differences in social

class, health in childhood, health of parents and
siblings, or in exposure to infection or trauma in
childhood. Familial or inherited diseases were m~ore

frequent in the frequently ill group than in the
healthy Eroup, however. In all their various investi-

tations (Binkle, Christenson, Kane, Ostfeld, Thetford,


& Wrolff, 1968; Hinkle, 1969) these workers have found
a consistent picture of the style of life of those
who were frequently ill as compared to healthy sub-

3ects. Although they did not objectively have any
more difficult lives than did the control subjects,

frequently ill subjects perceived their lives as
demanding and frustrating and themselves as burdened

down with duties. They also tended to have conflicted
and unsatisfying interpersonal relationships and to

perceive their childhood relations with parents and
siblings as unsatisfactory. In general, the more

frequently ill subject tended to be more inner-
directed, more intensely involved in his or his

intimates' problems, more duty-oriented, and more
resentful and worried. In contrast, relatively
healthy subjects tended to be less intensely involved

with goals, responsibilities and duties, and more

apt to see life as exciting, interesting, and
challenging. They also tended to feel more satisfied
and less conflicted about their interpersonal relation-

Schmale (1958) interviewed 42 patients hos-

pitalized on a semiprivate ward who were suffering
from a variety of ailments. Twenty-nine patients

or relatives of patients reported that loss of a
loved person or valued goal followed by feelings of

hopelessness and helplessness immediately preceded
the onset of symptoms. Five patients reported
feelings of hopelessness and helplessness but no
loss of a loved person or goal.

Spitz's study of infants in a foundling home,
already discussed, demonstrated the broad effects of

emotional deprivation on perception, learning, psycho-
motor development, social development, and central
nervous system functioning (Spitz, 1954).
In a review of studies of recovery from
chronic disease and accidents, Ruesch and Bowman
(1953) found certain "character conflicts" which

characterized patients prone to invalidism regardless
of the specific ailment which they suffered. Male

patients were conflicted between needs for dependency
and independency. Female patients were conflicted
between desires for self-love and love of others.

Amon all groups of patients who were slow in re-
covering the overwhelming majority were law-abiding,
intelligent, from broken homes, and neurotic appearing
on the MMPI. From 100 per cent to 33-1/3 per cent in

a particular group were found to be persons who had


one parent who had fulfilled roles of both authority
and affection while the other parent was weak and

Three of the studies reviewed thus far in

this section suffer from a methodological fault.

They fail to establish the fact that certain psy-
chological states either preceded or accompanied the
development of illness. In each case subjects who
were already sick were questioned about their memories

of the past or the past was inferred from their

present psychological status. Since Spitz followed
his subjects from the time they entered the experi-
mental condition, and compared them to matched control

subjects, his findings deserve more credence. The

present study attempts to clarify time relationships
more carefully.

Several studies have experimentally invresti-

gated the nature of the relationship between attitudes
and affects and bodily expressivity in healthy subjects.

They are interesting because they elucidate the basic
phenomena without the contamination of possible effects
of sickness upon the dependent variables. Graham,
Stern, and Wcinokur (1958) hypothesized that there is

a specific relationship between the attitude a patient

develops toward the life situation disturbing him and

the symptoms he develops in response to it. They

were able to produce changes in skin temperature

characteristic of hives or Raynaud's disease by

suggesting to hytpnotized subjects they adopt attitudes

typical of patients suffering from these diseases.
In a more recent study (Stern, Winokur, &e Gra~heam,

1961) these experimenters have replicated their
findings adding to the evidence that the way a sub-

3ect perceives a situation confronting him and his
total reactions to it are lawfully related.

Following up on his earlier studies of the
relation between body image and the site of disease

(1956; 1956), Fisher discovered a relationship be-
tween differential GSR response on right and left

sides of the body and tw~o psychological variables,

v~ig, maatureness of body ima~ge and the stage reached
in defining sex roles (Fisher, 1959). In other words,

his work suggests that the whole person, including

his autonomaic nervous system, takes part in the pro-

cess of learning male from female, weak from strong,

Three studies already discussed should be

mentioned here as further evidence for a relationship

between attitudes and affects and bodily expressivity
in healthy subjects. These are the investigations of
Mallmo and Shalgass (1949), Learmonth 1~ AL. (1959),
and Frankle (1992).

Since each individual is part of the larger
social system, his relationship to that system
should affect his general functioning and his health.
Brief mention will be made of several studies which

substantiate this hypothesis. Rennie and Scrole

(1956) sent questionnaires to 172,000 adult subjects
asking them about the incidence of eleven psychosomatic

complaints in their lifetimes. They found a relation-

ship between certain types of complaints and social
status. Investigating an entire community, Koos

(1954) reported a relationship between inc~idebnce of
disabling illness and social class. Seeking medical

help when ill was also related to ~social class.
Abrahamson (1961) found that adolescent girls in
South Africa checked more somatic symptoms on the
Cornell Medical Index when they also indicated that
there was a discrepancy in either direction between
the traditionalism of the daughter and her mother. In

other words, conflict between mother and daughter in
their perception of social roles was associated with

more sGomatiC complaints on the part of the daughter.

In a study of college students' use of the infirmary

for upper respiratory complaints, Summerskill and
Darlington (1997) analyzed certain sociological

characteristics of high and low users, High users

were characterized as female, minority group members,

younger than other subjects, and from large cities
rather than small towns.

In summary, these studies substantiate the

feasibility of studying illness behavior or general

susceptibility to illness in relation to other molar

level variables of personality and interpersonal

The Problepl to be Investigated
This study investigates the relationship be-

tween some of the characteristics of system function-

ing discussed above and the number of episodes of
illness of three groups of college students. The

system characteristics studied are communication

among subsystems, stage of maturation (or regression),
and amount of communication between the system and

its interpersonal environment. Translated into

operational terms, it is predicted that subjects who


have frequent episodes of illness will be "inner

expressers" (Learmonth et AEl*, 19T9). They will have
restricted and constricted self-descriptions. They

will indicate that they have remained relatively

close to their parents rather than having developed

close relationships to their peers. And, they will

disclose little of themselves to other people in
their lives.

To avoid the problems inherent in asking a

subject to recall how many times he has been sick,
the number of visits made to a university infirmary

was used as an index to the num~ber of episodes of

illness suffered by the subjects. Such an index has

been termed "illness behavior" by Parsons (1953) to

differentiate between the state of physiological

malfunction and the response of the individual to

this malfunction. Working from the same point of

view, Mechanic and Volkart (1961) define illness
behavior as "the way in which symptoms are perceived,
evaluated, and acted upon by a person who recognizes

some pain, discomfort, or other signs of organic
malfunction" (p. 52). Clearly, visiting the univer-

sity infirmary is an indication that subjects have

adopted the sick role. The possibility that some


subjects may have experienced organic malfunction
without adopting the sick role is, of course, quite
likely. By using visits to the infirmlary as an

index to the number of episodes of illness, this
investigation is confined to reported illnesses

implying adoption of the sick role. Another problem
arises from the fact that the infirmary may be used

for other purposes than to seek medical treatment.

For this reason, care was taken not to include
visits for permits and excuses when deciding whether

a subject was a high or low user of th~e infirmary.

It is also likely that some subjects simply did not

use the infirmary but preferred private medical

care. As will be discussed later, no subjects were
included in the study who reported frequent visits to

private physicians although they infrequently visited
the infirmary.

Stress and invasion by noxious agents (Kruse,

1954) have been shown to be extremely important forces
associated with the etiology of illness. They will

not be handled directly in this study. However, it

can probably be assumed that students on a university

campus were exposed to about the samne number of
infectious agents and no subjects with a known history

of chronic disease were included. Certain stressful
and demanding situations occur throughout the year
and affect all students. These include the pressure
to make good grades examinations, and deadlines.
Two studies of student use of a university infirmary
suggest that the ordinary stresses of college life
are not especially important in determining use of
the infirmarry (Mlechanic, &e Volkarts 1961; Summershill,
& Darlington, 1957).



Three groups of subjects, twenty males and

twenty females in each group, were selected from

among those who entered the University of Florida
as freshmen in September, 1959, and completed the first

two years. These groups were designated: "BEigh," made

up of male students who made ten or more visits and
female students who made eight or more visits to the

student infirmary during the academic year September,

1959, to June, 1960; "Me~an," rmade up of students who
made three to four visits during the same period; and

"Lov," made up of students who made no visits or one
visit during that period. Specific individuals were

selected randomly from among those eligible for

inclusion in the study. The criterion for group

designation was based on a tabulation of the number
of visits mtade, over the same period of time, by a
randomly chosen ample of 477 students. These students
were members of the same class from which subjects



for this study were selected. Figure 1 shows the

shape of this distribution. The groups were equated
on certain variables to insure control of extraneous

factors. All subjects were 17, 18, or 19 years old

at the time of entering the university and were

unmarried then as well as when they were tested for

this study, at the end of their sophomore year. All

subjects had a general physical examination prior to
entering the university. None were found to have

significant physical defects or handicaps. Subjects
were questioned as to the number of visits to physicians

they had made during the academic year September, 1959,
to June, 1960, at some place other than the student
infirmary. It was found that in no case had a student

made more such visits than he did to the infirmary.

Since the number of visits to the student infirmary

was considered to be! an index to the number of episodes

of illness, visits for nonmnedical purposes were not

counted in assigning a subject to a group. Nonmedical
visits were defined as those in which the student did

not present a complaint or ask for treatment as in
visits for routine physical examinations, excuses and

permits, and prophylactic innoculations.
Separate analyses were mnade by sex since previous

\ a





El 6

s~~a~qng 30 ~aqmnN


work with this population indicated that there were
wide sex differences on many of the variables to be

investigated (Wrharton, Barger, & Scchumacher, 1961).

For example, the mean number of visits for males in
the randomly selected sample of 477 was about one and

one-half times that of the females. See Figure 2.

Previous research also indicated a significant differ-

ence in religious preference among high, mean, and low

users of the student infirmary. For this reason, in

all three groups, major religious denominations were

represented in the same proportions as were present
in the total student body. See Table 1. This was

done in order to control any extraneous variables,

other than illness, which might influence use of the
student infirmary.

Minnesota Multiphasic Personality Inventory

(MMPI)2 protocols were available on all subjects as a

2Test-retest reliability coefficients for the
various scales were mostly in the sixties and seventies
when college students were retested after intervals
of from one week to three months. They were in the
forties and fifties when intervals of from two to four
years were involved (Dahlstrom, & tWelsh, 1960, pp. 472-
473). Evidence for the validity of the particular
combinations of subscales used in this study will be
presented as they are discussed in the text.

10 0


f \lr\


I he:





O in\ O s1O



Males N=6Q_ Females N=60
High Mean Low High Mean Low

Methodist 4 4 4 G 4 4

Baptist 4 4 4 33 3
Catholic 3 3 3 3 3 3
Bebrew 2 2 1~2 3 3 3
Other Major Prot.* 4 4 4
Other Minor Prot.** 1 1 1 1 1 1
No Preference 2 2 2 11 1

*Includes Congregational, Episcopal, Lutheran,
Presbyterian, and Desciples of Christ.
**Includes Seventh Day Aldventist, Christian
Science, Church of Christ, Quakers, Latter Day Saints,
Unitarian, Universalist, Advent Christian, and Others.

result of university-wide administration of this
instrument in connection with the initial processing

of freshmen. Also available for each subject were

such information as home town, religious preference,
number and type of complaints, scholastic aptitude
test scores, and grade point average. These data vere
current as of the end of the freshman year.
Each subject was administered the Leary Inter-

personal Cheeklist (ICL), the Jourard Self-Disclosure
Questionnaire, and an information sheet requesting
certain identifying information. (See Appendix A.)
These tests were administered near the end of the

sophomore year, or, in the case of about four subjects
in each group, during the first semester of the junior

year. Subjects were tested individually for the most

part, but occasionally in small groups.
The ICL3 is a 128 item checklist; in filling

3Test-retest reliability coefficients for 77
obese women retested after an interval of two weeks
averaged .78 for octant reliability (Leary, 1957,
p. 461). The validity of the circular grid model of
personality is demonstrated in the regular decrease
in octant correlations as one progresses around the
circle (Leary, 1957, p. 462), Foa (1961) has dis-
cussed the general validity of this circular grid
model of interpersonal behavior. He~ found a "strong
convergence of thinking and results" (p. 341) in the
work done independently by several investigators in
studies done from different research traditions and
with different types of groups. They all arrived
at the s~ame main axes used by Leary, although different
synonyms were used which corresponded to Leary's names,
"Dom"' and "Lov." Leary (1957, pp. 388-389) reports
multi-level personality measures of 42 ulcer patients
in which Level I scores stressed responsibility and
strength. Level II scores stressed aggression and
independence. Level III scores stressed passivity
and dependency. This analysis attests to the validity
of the scale for psychosomatic subjects,


it out the subject is asked to indicate whether each

trait of interpersonal behavior is true or false of

himself and other target-persons. The 128 traits

represent eight types or classes of behavior which are
arranged consecutively around a circle or grid in

Leary's system so that those octants opposite each
other are less alike than those near each other

(Leary, 1957). Protocols are scored in terms of the

two main axes of the grid, love-hostility and dominance-

submission. A subject's final score is a single number

representing that position on the grid which reflects
the relative weight of his score in each of the eight

octants. The intensity, on a four-point scale, of the
items selected may also be measured. See Leary (1997)

for a description of the empirical development and

validation of ICL intensity scores. Scores are in

standard form so that octant scores may be used in

comparing different target-persons. Discrepancy tables
which indicate the amount of discrepancy between any
two scores have been prepared by Leary. Ble has also

indicated a method of translating MMelPI scores into

his system and for computing the discrepancy between

these and self-descriptions on the ICL. This is an

example of his multi-level model of personality in


which MMPI scores measure Level I or the Public Com-

munication Level, while the Checklist self-description
measures Level II or the Level of Private Communication.

Discrepancy between these levels is an operational
definition of self-deception or failure to see oneself
as others see one.

The Self-disclosure Questionnaire4 (Jourard, &c

Lasakow, 1998) is comprised of forty statements about
several different aspects of the self. The subject is

asked to indicate, on a three-point scale, the degree to
which he has directly confided information about these

aspects of himself to four "target-persons'": mother,
father, same sex friend, and opposite sex friend. The

subject writes his responses on an answer sheet. For
each subject, the sumn of scores for each target-

person as well as the sum of all scores are

4Forty-four male medical students and 43 female
nursing students were retested after one year. Re-
liability coefficients for total disclosure scores were
.62 and .61 respectively (Jourard, S.M. Personal
communication. Ma~y, 1962). Validity of the scale as
an indicator of closeness to another person is shown by
the report that mzarried subjects disclosed less to parents
and same sex friend than did unmaa~rried subjects. More-
over, there was mBore disclosure to spouse for these
subjects than to any other target-person. Total dis-
closure for married and unmarried subjects wars the same,
however. Correlations at the .01 level.of significance
between liking a parent (Parent-Cathexis) and self-
disclosure to parents were also reported by the samea
authors; (Jourard, & Lasako, 1958).


counted yielding a target score as well as an over-all
score. Statistical properties of these instruments
as well as more details of validity and, reliability may

be found in (Leary, 19S7; LaForge, &e Suezek, 1966) and

(Jourard, & Lasaksow, 1958).


Subjects were selected at random from the
university census according to the criteria already

mentioned. In addition, only those subjects whose
MMPI profiles were valid according to hathaway and

McKinley's criteria (Hatha~way, & MlcKinley, 1951,

p. 23) were used. All potential subjects were con-
tacted by letter and those not responding were tele-

phoned until chough had agreed to participate.
Table 25, in Appendix B, gives details of the number
of students contacted and methods used.

Subjects were asked to take part in a research

project investigating interpersonal relations of
college students. Although the research was identified

with the student health service, subjects were not
made aware of the relationship between their selection

and their frequency of infirmcary usage. This was done

to avoid giving them the set that use or nonuse of the

infirmary made thema suitable subjects for psychological


research, a procedure thought to be wise both from

the point of view of public relations of the student
health department and also because it was felt variables
should be studied without the contamination of the

subjects' notions of how they might be! related. As
mentioned above, all but about four subjects in each

group were tested during their sophomore year. The
remainder were tested at the beginning of their junior

year. Instructions were those which are standard
with the two main instruments usetd If either parent

was deceased, or parents; were divorced or separated

prior to the subject's adolescence, he was asked to
describe a parent-substitute. In two such eases, one

male and one female, subjects were permitted to omit

descriptions of their fathers because of inability

to determine a suitable parent-substitute.

Confidentiality was assured the subjects and

they were permitted to use either their nam~es or just
their student numbers, a umxrber similar to a serial
number assigned to each student when he enters the


Any spontaneous commesnts made by a subject were
recorded by the examiner after the subject had left.

These were used in a qualitative analysis of what


Leary calls Level I behavior, or how a subject appeared
to an observer.


The hypotheses of this study may be stated as

1. At a time immediately preceding the period

during which visits to the infirmary are counted,

frequently ill subjects will have lower morale, as
indicated by the morale loss scale, than will in-

frequently ill subjects.
2. Subjects who are frequently ill are people
who will not express their feelings directly and openly

but who will tend to express them somatically, as

compared to infrequently ill subjects.

3. Frequently ill subjects will not attribute
to themselves certain modes of interpersonal behavior,

in particular negative ones, but will rather make a
self-description which is restricted, when compared
to infrequently ill subjects.

4. Frequently ill subjects will have signifi-

cantly fewer intimate interpersonal relationships, as

indicated by their disclosure of themselves to others,
when compared to infrequently ill subjects.

'.The main target for self-disclosure of

frequently ill subjects will be parents rather than
peers, especially the opposite sex peer. The opposite
will be true for infrequently ill subjects. This is

considered to be a measure of dependency upon parents
or emancipation from them.


Tables 2 through 6 summarize the main findings

in relationship to the five hypotheses. Tables 26

through 34, in Appendix B, show results of statistical
tests underlying these findings.



Hyp. 1. Low morale will precede period of infirmary
use for high users but not for m~ean and low users.

Test: Morale loss seale

GosuD Predicted. Found. SiRnifaicance

High + 6.93 N.S.
Mean -6.75~
Low -6.90


Hfyp. 2. High usrsr will be "inner expressers" while
mean and low users will not.



1811: MMIPI Scale 1 (Hs)
Frour Predieted j

High +
Mean 0
Low -

Tls1: MMPI Scale 3 (hy)
High +
Mean o

Low -

TeSst MMPPI Scale G (Pd)

Righ +
Mean 0

Low -

Terts Com arison to samen sex









60.28 p *(.025


94. 88
freshme~n on MMPI Scale 1



p 410


13.40 rs 12.17

12.05 vs 12.17
12.00 vs 12.17


TABLE 3--Continued

_ __ ___



Sig rnif ic ane e




13.90 vs 13.24

13.95 vs 13.24

13.05 vs 13.24

same sex freshmen on


MMPI Scale 3







MMEPI Scale 4

Low -

Test: Comparison to


High +

Mean 0

Low -


High +

Mean 0

Low -

Test: Com arison to

20.80 vs

19.40 vs

19.75 vs




22.85 vs 22.20

23.10 vs 22.20

21.4+0 vs 22.20

same sex freshmen on





22.65 vs 22.99

19.80 vs 22.59

21.50 vs 22.59






TABLE 3--Cont inued

GSrous Prledicted Fou~nd. Sinnificance


High + 23.85 vs 22.15 p (.10
Mean 0 22.35 vs 22.15 N, .S

Low 20.30 vs 22.15 p 1.05

Test Welsh's Internalization Ratio (MMEPI)
High + *92 N, s.
Mean 0 ~J.98




Hyp, 3. High users will not attribute to themselves
certain categories of interpersonal behavior to a
greater extent than will mnean and low users. They
will especially repress negative feelings.

IR&&:ICL Intensity Scores

Pred icted







Hi gh


agist Level I vs Level II Discrepancy Scores (ICL)



($ above Mian)












H~yp. b. High users will hav~e lower self-disclosure
scores than will mrean sand low users.

Tegst Jourard Self-disclosure Questionnaire
GrouD Predicte6d Found Significance

Righ -166.78 N. S.
Mean 0 166.60

Low + 166.53



Hyp. 5. Self-disclosure will be higher to parents
than to peers for high users. It will be higher to
peers than to parents for mean and low users.

'estt Jourard Self-disclosure Questionnaire




40.1 l

Mo. +

Mo. O

Moi9. -

Fa. +

Fa. O

Fa. -

























Note--SSF = Same sex friend
OSF' = Opposite sex friend

Schmaale's (1998) report that patients he

interviewed felt "helpless and hopeless" preceding

their illness raises the question of whether low

morale may predispose one to frequent illnesses. The

first hypothesis of this study is concerned with that

question. Canter (1960) developed a "morale loss"
scale from MMPI scores which distinguished among

subjects who had made recent suicide attempts, non-
suicidal patients, and normals.5 Mbore pertinent to
this study, Imaboden, Canter, Cluiff and Trevor (1969)
showed that the morale loss scale successfully pre-

dicted speed of recovery from brucelosis in two groups

of adult patients. In the present study, Canter's

morale loss scale was scored fromn the MMIPI protocols

obtained from subjects when they entered the univer-

sity as freshmen. Since the criterion for selection
as a subject was the number of visits madae to the
infirmary during the academic year following the

administration of the MMPI, morale loss scores should
be higher for those students who subsequently became

high users of the infirmary than for the other two

6Canter offers no information on the reliability
of the scale.


groups. Table 2 shows that this hypothesis was not

supported. Table 26 in Appendix B shows the results
of the analysis of variance performed on these scores.

It will be noted that, although the scale is named

"morale loss," it is considered throughout this work

to be a measure of low morale only since it does not

measure change in morale level.

Outward vs Ipward Expreyssiop of Feelings
The second hypothesis of this study predicted

that high infirmary users would indicate through

their test behavior that they tended to express their

feelings inwardly rather than outward to their physical

and interpersonal environments. Two tests were made

of this hypothesis, both using MMPI data. If high
use subjects were "inner expressers" (Learmonth gi gi.,

1959), they should have scored significantly higher on
the three MYMPI scales found by these investigators to

discriminate among such subjects. Table 3 shows that

this was not the case and the hypothesis was not

supported. Tables 27, 28, and 29 show the results of
the analyses of variance performed with these scores.

Scores on all the other MMPI scales, except

F, L, and Q, were also subjected to analysis of
variance. F, L, and Q were omitted since protocols

of high scorers on these scales were considered to be
invalid (Bathaway, &e McKinley, 1991) and were not
included in the study. Only scale 9 (Ma) distinguished

among levels of infirmary use. Table 7 shows the
analysis of scale 9 (M-a).


Source of Variance df MS F

Sex 193.63*9
Levels of Use 2 .533.85 5.45*
Interaction 2 103.89 1.06
Within Cells 114 97.91
Total 119

*p <.01

An alternative test of this hypothesis involved

comparing each of the six experimental groups to the
appropriate sex group of the entire freshman class of
which these subjects were members. Freshman men or
women were used as a finite population to which the

experimental groups were compared.6 Table 3 shows

Although the experimental subjects were also
part of the freshman class, it was felt that the N's

that none of the "inner expressed" scales, except

scale 4 (Pd) distinguished these subjects from their

classmates. Table 30 shows the results of 1 tests

performed on this data. It also shows that scale 9

(Ma), as well as scale 4 (Pd), distinguished subjects
from freshmen in general. In both cases, high users

were higher than freshmen as a whole on these scales
and mean and low users lower. Table 30 also indicates

that male high users were higher than freshmen men

on scale 1 (Hs) at the .05 level of confidence using
a one-tailed test. Since the direction of the

difference was predicted by the hypothesis, a one-

tailed test is appropriate. However, considering the

finding that there was no significant difference among

the various levels of use nor any significant inter-

action effect, as shown in Table 27, this finding should
be evaluated with caution.

Welsh (1992) has developed a more direct measure
of internalization-externalization from the MMPI and

offers convincing validation evidence for it.? If high

were so large that the results were not materially
affected by this. Freshman men numbered 2171 and
women 1027.

7The index discriminated among several diagnostic
groups in the predicted direction. For example, de-

users of the infirmary internalize affect, overcon-

trolling any outward expression, an analysis of
internalization ratio scores should distinguish among

levels of use. Table 3 shows that this was not the

case. The analysis of variance of these scores is

reported in Table 31.
The evidence clearly indicates, then, that high

use subjects did not internalize their feelings but

rather that they tended to show poor control over

impulse expression. Parenthetically, it may be noted
that scale 4 (Pd) does not seema to be related to

internalization of affect, as stated by Learmonth at al.


Hypothesis 3 states that high use subjects will
not attribute to themselves certain categories of

interpersonal behavior to the extent that mean and low

users will, indicating poor self-knowledge and repres-
sion of certain interpersonal behavior tendencies.

Table 4 shows that this was not the case. This hypo-

thesis was tested using intensity scores on the ICL.

pressed patients had higher scores than did manic
patients. The scale also successfully predicted re-
sponse to psychotherapy. Welsh reports no reliability
data, however the reliability of the index would be
closely related to that of the scales composing it.
These are Hs, D, Pt, Hy, Pd, and Ma.

Table 32 shows the analysis of variance performed on

these scores. Comparison of data in Table 32 with
those of Table 4 shows that all trends were in a

direction opposite to that hypothesized.

Another way of approaching hypothesis 3 is to

ascertain whether high users tended to have inaccurate

self-concepts, reflecting suppression or repression

of interpersonal behavior perceived in them by others.

An operational definition of this is the discrepancy
between Level I (self as seen by others) and Level II

(self as seen by oneself) on the ICL. Level I is
obtained from MMPI scores with the rationale that

these scores represent symptoms which the subject is

communicating to a hypothetical interviewer." Level II

is measured by the Interpersonal Cheeklist self-

description. Table 4 shows the results of this test

of the hypothesis indicating that there were no

differences among the levels of use in amount of

discrepancy. Table 33 shows the Chi Square analysis

of these data. Discrepancy can be expressed in terms
of direction as well as amount. Table 8 shows that

8Leary states these scores satisfactorily
predict actual ratipgs of a patient's behavior by
an interviewer (Leaay, 1957, p. 108),


high female users described themselves as more hostile
than they appeared at Level I more frequently than did
the other females. This is certainly no indication
of repression of negative feeling and behavior toward

others. They can be said, however, to have had self-
concepts which were inaccurate when compared to their
Level I descriptions. Mean females tended to suppress
or repress hostile feelings and behavior in their

self-descriptions. No differences were found for

males. Discrepancies in the dominant-passive direction
were not significant for either sex. Median tests

were also performed for each sex comparing Level I

self, Level II self, and Level V self (ideal self)
as to whether they fell on the dominant-passive or
loving-hostile side of the circle. None of these

Chi Squares reached the .05 level of significance.

However, high females were more frequently dominant

at Level I and meran users were more frequently passive
at Level V than were the other females, the latter Chi

Square reaching the .10 level of significance. Level
I data could not be tested with Chi Square because the

expected frequencies were too small. Table 9 shows these
data. In general, then, there was a tendency for high
females to describe themselves and to appear to others

Direction of Discrepancy High M~ean Low
N=20 N=20 N=20

Chi Square =2.25, df=2,
P <. 50




Level II More Loving

Level II Less Loving

No Discrepancy

Chi Square not appropriate

Level II M~ore Dominant

Level II Less Dominant

No Discrepancy

Chi Square not appropriate


Level II Mlore Loving

Level II Less Loving

Chi Square = 6.62, af=-2,
p <.05

Level II More Dominant

Level II Less Dominant

ICL LEvel Domn Pass Lov Host




Level I

16 4

Chi Square
not appro-

Chi Square
not appro-

9 11
14 6

Level V
Ideal Self

7 13

Chi Square=
4.46, af=2,
p <.20

Chi Square =
4.83, df=2,
p <.10

where N=19q

Note--N=20 in each group except Low Femasle
for Level I.


as dominant and hostile and for mean females to

aspire to the usual feminine role of passivity.


Hypothesis 4 predicted that high users would
be lower in self-disclosure than would mean and low

users. Self-disclosure was considered to be an indi-

cation of the subjects' closeness to important figures

in their lives. Table 5 shows that this hypothesis was

not confirmed. Table 34 shows the analysis of variance

of these scores. As check on this finding, it was
considered that the opportunity to write one's name on

the questionnaire or to omit it and use only the
student number was a measure of self-disclosure to

the experimenter. Table 10 shows that there were no

significant differences among the groups on this
measure. Intensity scores on the ICL Level II self-

description may likewise be interpreted as communication

about oneself to another (the experimenter). As dis-

cussed above, no significant differences were obtained
here either (Table 4).

Self-Disclo~sure to Parents and Peers

Hypothesis $ predicted that high users would
be closer to parents than to peers while mean and low

Itales Females
H' M L Hi M L

Namae 14 10 13 11 10 9

Student Number 6 10 7 o 10 1

Chi Square = Chi Square =
1.83, df=2, .40, df=2,
p <. 50 p 4.70

Note--N=20 in each sex-use group.

users would be more emaancipated from parents and more
closely related to peers. Differences in self-dis-
closure among targets for the three levels would then

be predicted. As can be seen from Table 6, no such
differences were found. Table 34 shows the analysis
of variance of these scores.

It was felt, however, that differences among

groups light be obscured by the large standard devia-
tions obtained in the self-disclosure data so that a

correlational measure might reveal some pattern of
self-disclosure not discernible in the analysis of
variance. Therefore, morale loss scores were correlated

with self-disclosure scores to the various targets. The





rationale for selecting the morale loss scale
rather than some other intrapsychic measure such as
the ICL was that the correlations between intrapsychic

mood state and self-disclosure would reveal something

about the qualitative nature of interpersonal relation-

ships. In other words, close relationships to others
when one is high in morale are psychologically different

from close relationships when morale is low. Table 11
shows; these correlations. For all mean and low use

groups, except the female mean users, correlations were
consistently negative. Since a high score on the morale

loss scale means low morale, negative correlation
coefficients indicate that sthCdents who were low in

morale at the beginning of their freshman year were,

on the whole, not close to ta~iei~tlpersson at the end
of their sophomore year. This relationship was

significant for mean and low males when closeness to

parents was considered. With the exception of one

negative correlation coefficient, high users had a
consistent pattern of positive correlation indicating

that low morale predicted closeness to target-persons

a year later. Since none of these correlation co-
efficients reached statistical significance, conclusions

must be tentative. However, the pattern of positive


Males Fem~ales
Target H M L H M L

Mother .15~ -.43*t -.50* -.12 .14 -.34
Father .25 -.51* -.46* .27 -.08 -.06
Same Sex Friend .16 -.38 -.21 .31 -.04 -.22

Opp. Sex Friend .19 -.17 -.46* .33 -.23 -.14

*tp (.05
Note--N=20 in each sex-use group.

correlation does suggest that these subjects were from
a different population than the me~an and low users.

Schaeter (19$9) has termed the phenomenon of turning to
close interpersonal relations when anxious or low in
morale as "anxiety affiliation." This term characterizes

high users but not mecan and low users suggesting that
their relations to their parents may have been of a
different quality from those of the mean and low
users. The significant negative correlation between

morale loss and self-disclosure to girlfriend, seen
for low maales, suggests that these subjects may have

made more progress in interpersonal maturity, as dis-
cussed by Sullivan (1953), than did other subjects.


It may be seen that their closeness to girlfriend

had a significant relationship to earlier morale.
These findings together with high users'

elevation on scales 9 (Ma) and 4 (Pd) of the MMPI are

congruent with a notion of immature acting-out of feelings.
Sarbin (19l52) has considered such poor control

over impulse expression to be characteristic of what
he calls the "Primitive-Construed Self" said to be a

stage in the development of the self which takes
place about half way through the process of psycho-

logical maturation. At this stage of development, the
child has learned to distinguish between patterns of
stimuli received from humans and stimuli received

from objects in the environment but he has not yet

learned to delay response to these stimuli. In later

stages of development, the stage of the "Introjecting-
Extrojecting Self" and the stage of the "Social Self,"

the child progressively learns to use language to

differentiate and strengthen cognitions, to communicate

them to others, and to discriminate not only discrete
acts of others but organized acts or roles. He also

introjects these perceived roles, identifying with others
in his interpersonal environment. As a test of this

"epistomogenic" theory of self-development, Mathews,

H~ardwyok, and Sarbin (1953) predicted the behavior
of groups of college students selected by means of
MIMPI profiles to represent the three stages of

personality development outlined above. Responses
to several cognitive tasks generally supported the
hypotheses and the validity of the MMPI as a method
for classifying subjects. MIMPI profiles of the sub-

jects of this study were examined and classified into
the same three levels of self-development. These were

S3 The Primitive-Construed Self, defined by a profile
in which either scale 4 (Pd) or Scale 9 (Mla) was
highest and at least one standard deviation above the

mean with the other being elevated also; Sb The Int~ro-
jecting-Extrojecting Self, defined by a profile in
which the highest score is for either scale 3 (EBy),
scale 7 (Pt), scale 1 (Hs), or scale 2 (D) with at
least two scores in the profile one standard deviation

above the mean; Sg The Social Self, defined by a
profile in which all scores are within one standard
deviation of the mean, one score being permitted to
deviate but not so far as two standard deviations from

the mean. Following Sarbin, scale 5 (Mf) was omitted
from the analysis. Table 12 shows the results of

assigning these designations to subjects of this study.


Use Level Primitive- Introjecting- Social No Classi-
N=4O Construed Extrojecting Self fiction
Self-Sz Self-Sq Sy

High 13 4 12 11
Mean 8 8 15 9
Low 2 5' 26 7

Chi Square = 17.06, df=6, p(.01

It can be seen that there was a clear progression
from most immature to most mature self-development as

one progresses from high to mean to low infirmary use.
Clearly, infirmary use is related to levels of maturity
in self- or personality development, as defined by
Other evidence about the families of these

subjects and their perceptions of parents corroborates
the findings discussed above of anxiety affiliation

and emotional immaturity of high use subjects, as
compared to the other groups. Sarbin and his coworkers
state that self development on an immature level is
the result of trauma or conflict which retards or


inhibits the development of later substructures or

foci of self-organization (Mathews et gi., 1953).

Table 13 shows the results of an analysis of variance

performed on "Famnilial Discord" scores derived from
the MIMPI (Harris, & Lingoes, 1960). Strong differences

between levels were found in the predicted direction,

i.e., high users reported more familial discord.
Further evidence of trauma and conflict is seen in

the Chi Square analysis performed on descriptions of
mother on the ICL. Differences among the females

were not significant, but male high users' descriptions
of mother fell more frequently on the hostile side of
the circle than did those of low and mean males.

Table 14 shows these data. This finding should be

compared to the anxiety affiliation evidence in Table
11. Another finding may be mentioned. While the

numrbers are too small to analyze statistically, the

direction of differences agrees with the picture being

developed here. "Lov"' and "Dom" scores for mother,
the standard scores from which the final summnary

score is plotted, were examined for combinations of

high dominance and low love, a combination previous
researchers have emphasized as especially significant

in the development of personality (Ruessh, 1948). All


Source of Variance df MS F

Sex 1 8.53 2.14
Levels of Use 2 16.36 4.10*
Interaction 2 .41 .10
Within Cells 114 3.99
Total 119

*Sp 4.02 5



Level of Use Loving Hostile


High 11 9
Menn 18 2

Low 16 4

v.=rr Chi Square = 6.71, df=2,
p <.05

High 14 6
Mean 14 6

Low 12 8

Chi Square = .80, df=2,

*One male omitted Mother
Note--N=~20 in each sex-use group except Male High
where N=19.


subjects whose Dom and Lov scores for mother were
both at least one standard deviation above the mean

for all subjects of their sex were counted. Six high

users fell into this category, no mean users did, and

two low users did. Neither ICL descriptions nor

Dom and Lov score combinations distinguished among

levels of use when descriptions of father considered.

Paittern~s of Inf~irmesryv Use

High users of the student infirmary can be
distinguished from low and mean users in qualitative

characteristics as well as quantitative ones. For

example, they began their visits almost immediately
upon arriving at the university, making an average
of more than two visits the first thirty days of

school, while mean users made an average of less than

one. Table 15 shows the analysis of these data. Since

visits during the first thirty days are included in

the total, one would expect the relationship between

being a high user of the infirmary and a high user
during the first thirty days to be positive, However,

one would not necessarily predict such large differences

among the groups in infirmary usage during the first
thirty days of school.


Level of Use5 Above Mdn Mdn & Below

High 29 11
Mean 16 24

Low 3 37
Chi Square = 35.21, df=2
p<'. 001

Note--N=4O in each group.

Table 16 shows the analysis of variance of

ratio scores arrived at from dividing each subject's

total number of kinds of complaints by his total number

of visits. Low users were omitted fromn this analysis

since they either made no visits and had no complaints

or made one visit for one complaint. Obviously, the
variance of their scores was not homogeneous with that
of high and mean users' SeoyaEib It may be seen that

high users had fewer kinds of complaints in relation

to their nimber of visits.?'Edghh~users did not, however,

have an overall different ranking of types of complaints.
See Table 17.



Source of Variance df MS F

Sex 1 .11 3.67
Levels of Use 1 1.81 60.33*
Interaction 1 0.00 0

Within Cells 76 .03

Total 79

*p <'.001. Mean higher than High.


Type of Complaint High Mean

Accidents and Injuries
DRI* and Somethroat

Gastrointestinal Symptoms

Skin Rash, Warts, etc.

Excuses and Permits

Tension Symptoms
Obesity and Glandular

Feminine Complaints
Chronic Ailments
(Present before University)
Allergy and Asthma













9 12.5
7.5 4
Rho** = .55, p ..Of

*URI = Upper respiratory infection.
**Corrected for ties.

One important question is whether visits to
the student infirmary during the freshman year re-

present a stable characteristic of these subjcts'
behavior or an ephemeral effect. Two answers to this

question are offered. First, no subject indicated
that he visited a private physician any number of times
which was not commensurate with his frequency of

visits to the student infirmary. That is, no low user
visited a private physician more than once, no mean
user visited one more than three times, and all sub-

jects who made four or more visits to private physicians
were also high infirmary users. Secondly, records
of visits made to the infirmary during the second year

of school were analyzed. The results of this analysis
are seen in Table 18. It is obvious that high users
maintained their status while mean and low users were

relatively indistinguishable. Moreover, high users
considered themselves to be sick as analysis of the

subjects' ratings of their own general health shows.
See Table 19. H~igh users were like other students

in the kinds of complaints they brought to the infirmary

then, but they came early and continued to come at least
through two years.


Level of Use Above fMan Maln &e Below 10 + Visits

Righ 32 7 28P
Mean 19 21 Of

Low 9 30- Ok
Chi Square = 28.68, df=2, p (.001

Note--High N=39
Mean N=lcO
Low NE39


Level of Use Excellent Good + Fair

High 12 28
Mean 24 16
Low 26 14

Chi Square =e 11.48, df=2, p C.01

N~ote--Nr4O in each group.

The Subjiects _as Students

Findings discussed above have indicated that

high users of the student infirmary were immature,

poorly controlled individuals who were unsuccessful
in emancipating themselves from their parents and in

forming strong relationships with their peers. These

individuals were also characterized by greater familial

discord and conflicted hostile-dependent feelings

about parents. How well they did in academic achieve-

ment, in social participation, and in being active in
university extracurricular life would be interesting

to know. If they were, in fact, ill prepared psycho-

logically for the demands of university life, this
should be manifest in their record at the end of the

freshman year. This is, in fact, what was found.

Tables 20 and 21 show analyses of variance of ACE total

scores and grade point averages at the end of the

freshman year. Because approximately four subjects

in each group were not tested until the beginning of

their Sunior year, ji ratios between these subjects

and the others in their group tested in their sopho-

more year were computed. This was done because of

the possibility that students still enrolled by the

junior year could have been those wpho earned unusually

Source of Variance df MS F

Sex 1 1710.75 6.41*
Levels of Use 2 424.11 1.59
Interaction 2 39.47.1
Within Cells 114 266.90
Total 119

*p <.05. Males higher than females.


Source of Yariance df MS F

Sex 1 .01 .02
Levels of Use 2 2.28 4.74*
Interaction 2 .02 .04
Within Cells 108 .48

Total 113



high grade point averages as freshmen. The only

group for which any differences reached the .05 level
of significance were the male mean users. Only the

seventeen individuals tested during the sophomore year
were included in the analysis of variance. Grade

point averages were not available for two high
females and one low female. Therefore, the total N

of Table 21 was 114. High, mean and low users did not
differ in their aptitude for university work, as mea-

sured by the AGfE, although the mean for males (186.4i0)
was significantly higher than that for femlel~1s (175'.08).

They did differ in their execution of this work, as
measured by grade point averages at the end of the
freshman year. Mean users did the best, low users

were next, and high users had the lowest grade point

averages. (They were 2.49, 2.46, and 2.06, respec-
tively). Social participation was indexed by sorority

and fraternity membership at the end of the sophomore

year when subjects were tested. This is a better
index than would be sorority and fraternity membership
at the end of the freshmra~n year since few students join

these organizations during their first year at school.

In Table 22 these data show clearly that high users were

less frequently members of sororities and fraternities

than were mean and low users. Subjects' answers to the

question about extracurricular activities they were
engaged in were analyzed in terms of the number of
activities they named in addition to sorority or

fraternity membership. Table 23 shows these data

indicating there were no significant differences among

the groups in number of extracurricular activities.

Of particular interest, however, are two kinds of
extracurricular activities which were found to par-

ticularly characterize the high meale users. They are

barbell club and pre-medical club, both of which in-

volve an interest and emphasis on the body. Seven high

male users named these activities while only one low
male user named the barbell club and one mean female

user named the pre-medical club. One final facet of

university life investigated is the number of times a

subject went home for the weekend. Subjects' answers
to this item on the questionnaire were corrected for the

distance between the university and the students' home.

Thus, if he lived more than one hundred miles away, but
not as far as three hundred milest his number of week-

ends at home was multiplied by two. Table 24 shows

these data -ndica~ting that there were no significant dif-

ferences amtong the groups in going home for the weekend.

Level of Use Member Nonmemiber

High 9 31
Mean 17 23

Low 20 20

Chi Square = 7.40, df=2, p(.Of

N~ote--Nr4O in each group.


Source of Variance df MS F

Sex 1 4.41 3.87

Levels of Use 2 1.36 1.19

Interaction 2 1.01 .89
Within Cells 114 1.14

Total 119


Source of Variance df MIS F

Sex 1 20.51 1.32

Levels of Use 2 .77 .05

Interaction 2 1.26 .08

Within Cells 106 14.84

Total 111

Note--Only those students living away from home were





The Data and the Hy~potheses

The high infirmary user was predicted to be a

student low in morale at a time just preceding his

first year at the university and the period during

which his visits to the infirmary were counted, a

person who would internalize his feelings rather than

express them outwardly, a person who would have an
inaccurate, restricted self-concept, a person who would

reveal little of himself to others, and a person who
would show evidence of psychological immaturity in

that he would not be as emancipated from his parents

and not as far along in development of interpersonal

ties with his opposite sex peers as others. The

findings did not support any but the last mentioned

prediction. In fact, the picture which emerged was

quite different from that which was hypothesized.
Morale state was seen to have no relationship to later
use of the infirmary. High use subjects did not tend

to internalize their feelings, but rather were seen


to have poor control over direct and immediate acting
out. Neither Welsh's Internalization-Ratio nor the

MMPI scales selected by Learmonth and his coworkrers

showed these subjects to be "internalizers." In fact,
the only MMPI scales which distinguished among the level

of infirmary use groups and distinguished these groups
from same sex freshmen in general were scale 9 (Ma)

and scale 4 (Pd), chosen by Sarbin to define the imma~ture

stage of self-development characterized by acting-out
and termed by him the "Primitive-Construed Self." Self-

descriptions of high users were not lower in intensity
of ICL items selected; rather, any trends discovered
were in the opposite direction. There wats evidence

that female high users did not see themselves as they

appeared to others but, rather than omitting negative
features from their self-descriptions, they accented
them. There were no significant differences at all,
neither for level of use ncrfor sex, in self-disclosure
scores. In general then, findings were that all high
users were psychologically immaEture and poor in their
ability to control impulse expression. High female

users also had a rather negative self-concept.

PsYcho~loaicea} Immaturity
A clue to the etiology of this state of affairs

may be found in the general notion of psychological
immaturity. Most personality theorists agree that

failure to progress in the process of psychological

maturity can often be attributed to conflict, traumats

or frustration at some earlier stage. Both Sarbin

(1952) and Sullivan (1993) emphasize the importance
of interpersonal relations as the raw material from

which the self is developed and both would consider
conflict or trauma in a child's relations with his

parents to be the 122 RM&~i gan for failure of his

personality to develop and mature as expected. The
combination of dependency upon parents together with

negative feelings toward them would constitute such a

conflict. The positive correlations between "morale-

loss" and self-disclosure to parents and the fact that

high users went home for the weekend equally as often

as did other subjects, in spite of the discord they

felt to be present there, attest to the dependency

side of the conflict for these subjects. Their de-

scriptions of mother as hostile or hostile-dominant
and their high scores on the familial discord scale

show the negative feelings associated with home and

parents. The significant negative correlation be-
tween the morale loss scale and self-disclosure to


girlfriend for the male low users further suggests
that there is some difference in maturity of inter-

personal relationships across the levels of use.
These boys were more concerned with what Sullivan

(1953) terms the main task of the late adolescent

period, the establishment of relationships with peers
of the opposite sex and emancipation from parents.
This is part of the overall problem which also includes
finding a satisfactory solution to the problem of a

choice of vocation, the successful execution of steps

in that direction and, finally, the establishment
of one's own family of procreiation. Evidence pre-

sented here indicates that, in comparison to the other

groups, high users were not only unsuccessful in the
interpersonal tasks facing them, but that they did poorly
in the academic and social tasks undertaken by univer-

sity students, 4Tj&*, low grade point averages and
failure to be included in sororities and fraternities.

The Confli~c~t Situation and the Stress It Produced~
The fact that these subjects were university
students, and students who managed to remain for at

least two years, is of paran~ount importance for this

establishes the other facet of the stressful, con-
flicted situation in which they found themselves.

Although they were ill-prepared for emancipation from
home and parents, they undertook the role of university

student, a role which calls for a certain degree of
self-control, maturity and ability to meet demands for

satisfactory academic performance. Stress has been

emphasized by many investigators as a factor in the
etiology of illness. The stress imposed upon a

relatively poorly controlled, immature system when

demands for attainment and autonomy are put upon it,

as seems to be true of these high users, definitely
follows the definition offered earlier of variation

in external conditions which is too sudden or too

excessive for the system to handle (G~aldston, 1954,

p. 12).

The M~echapfsm gr ~Propess of Psychosomatie JSllpess

The mechanism or process by which somatic
illness was related to this state of affairs is not

known. The findings do show that high users were not

just students who happened to develop some serious,
chronic disease which necessitated an inordinate number

of visits to the infirmary. They did not enter the

university with poor health, either; in fact all were

declared healthy by examining physicians prior to
matriculation at the university. The high rate of


infirmary use during the first thirty days at the
university mIay be interpreted in a number of ways.
It may have been that a pattern of "illness behavior"

Gas defined by Parsons and discussed above) was already
established by these subjects before they became students
or it may have been that the stress of university life
was felt, in anticipation, even before they actually were

enrolled. One possible explanation lies in the uses
to which the infirmary may have been put in satisfying
the needs of these students. The infirmary is free

to all students and thus is readily available as a

target for impulsive, acting-out when anxiety becoess

high. Coming to the infirmary is, in itself, a form
of acting-out behavior. In addition, the infirmary
offers a socially acceptable gratification of dependency
needs, needs which these students may have generalized

from their dependency upon their parents. Finally,
the choice of somatic complaints and infirmary visits
rather than some other form of acting-out such as

extracurricular activities in which high users did
not differ, may be related to the stress under which

these students found themselves. Self-ratings of

general health clearly showed that high users considered
themselves to be sick. There seems to be no reason for


assuming that their purpose in visiting the infirmary

was any other than to secure medical care, and the
infirmary records clearly indicate that they were

treated and considered to be ill by the attending


Sex Differences
Sex differences found in various parts of this

study warrant some comment. First, the typical boy
in the sample of 477 freshmen made nearly twice as

many visits to the infirmary as did the typical girl.
This finding is in contradiction to those reported
for older subjects in which women are said to greatly
outnumber men in their visits to physicians' offices

(Standish, Bennett, White, &c Powe~rs, 1955, p. 5). Our
data suggest that this relationship may shift by the

sophomore year when approximately twice as many
female high users than male high users made ten or
more visits (370 vs 20k), and about the same difference
was found between the sexes when eight or more visits
was used as the out-off point (580 vs 25 ). None of
the the mean and low users of either sex madea as many
as eight visits in the second year. The reason for
this shift in proportion of the sexes who remained

high users of the infirmary is not known. However,

it is interesting to note that while high users of

both sexes showed anxiety affiliation, the high

females presented a self-description on the ICL which

was negative while the high males described their
mothers negatively. NJo significant sex differences

were found in familial discord scores although the

level of use effect was significant. A tentative

explanation may be that the females had internalized
more of the conflict and discord while the males

tended to attribute hostility more to other significant

persons in their lives. It may be that negative
feelings about oneself were more stable across two

years than were patterns of interpersonal relation-

ships. The fact that correlations between morale
loss and self disclosure reached significance only

for the males also suggests that interpersonal re-

lations were more closely related to intrapsychic

state (i~e., morale) for thema than for females. The

notes ma~ded by the experimenter of subjects' spontaneous
comments support this interpretation in that f-~emaEle

high usrs'r~freq uently appeared to be anxious, hostile
and suspicions while ma~le high users were more co-

operative, but dependent and ready to either stress

the perfection of their parents or to reveal problems

in their relationships with them.

Imolications~of the.Findinns..Cor Theory
The implications of the findings of this study

for the various theories of the psychosomatic process
will be discussed in turn.

The specifieity theory

Findings presented here have no direct bearing

on the specificity theory, although the Fact that high
users had no specific types of illness which distinguished

them while they did seem to have certain common person-

ality traits and emotional conflicts not shared by
the other subjects, suggests that these data do not

support such a theory.

The regressilon theory
The regression theory calls for evidence of

a return, under stress, to an earlier mode of behavior.
Our findings do indicate that high users behaved in

an immature manner, but the question of whether this

represents regression, as such, is not answered. The
correl~tions between low morale as freshmen and self-

disclosure as sophomores seems to show a failure to

progress in the maturation process rather than a

regression. Ruesch's description of the "infantile

personality," which he calls the "core problem of

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