University of Southern California Dissertations and Theses

Prediction Of Overt Behaviors In Hospitalized Psychiatric Patients

(USC Thesis Other)

Prediction Of Overt Behaviors In Hospitalized Psychiatric Patients

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content T his d isse rta tio n has been 64— 11,256
m icro film ed ex a ctly as receiv ed
LEONARD, C a lista V erne, 1 919-
PREDICTION OF OVERT BEHAVIORS IN
HOSPITALIZED PSYCHIATRIC PATIENTS.
U n iv ersity of Southern C aliforn ia, P h .D ., 1964
P sych ology, clin ic a l
University Microfilms, Inc.. Ann Arbor, Michigan
Copyright by
Calista Verne Leonard
1964
PREDICTION OF OVERT BEHAVIORS
IN HOSPITALIZED PSYCHIATRIC PATIENTS
by
Calista Verne Leonard
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(Psychology)
June 1964
UNIVERSITY O F SOUTHERN CALIFORNIA
THE GRADUATE SCHOOL
UNIVERSITY PARK
LOS ANGELES, CALIFORNIA 9 0 0 0 7
This dissertation, written by
under the direction of h£.?....Dissertation Com
mittee, and approved by all its members, has
been presented to and accepted by the Graduate
School, in partial fulfillment of requirements
for the degree of
D O C T O R OF P H IL O S O P H Y
Dean
Date. June....L9.6.4..
DISSERTATION COMMITTEE
Chairman
To my husband, Ernest, who has sustained
me throughout these many long months of endeavor
with love, encouragement and constant help with
the many details and decisions involved.
ii
ACKNOWLEDGMENTS
I have always read acknowledgments with a great
deal of interest, knowing how little one person can accom
plish alone and wondering who the individuals and groups
were who gave of their time and skills to make the research
project possible.
To my dissertation committee at the University of
Southern California I wish to express heartfelt gratitude.
Dr. C. L. Winder, Chairman, has continued to provide prac
tical guidance and encouragement both in person and via
long distance. Dr. J. P. Guilford has been helpful not
only in his keen understanding of the theoretical aspects
of the research but also in keeping a watchful eye on the
statistical methods. Dr. Maurice D. Van Arsdol has been of
helpful assistance throughout. Dr. Edwin S. Shneidman has
been a constant source of personal encouragement and sup
port which has proven invaluable.
The dedication of the Veterans Administration to
veteran welfare is amply demonstrated in their approval
of this research project, No. P 20-63. I wish to express
special appreciation to Dr. Harry M. Grayson, VA sponsor,
who kindly shared his professional skills and the services
iii
of the Psychology Department at Brentwood Hospital in in
numerable ways. My thanks also to Dr. Charles W. Acker
who acted as VA statistical consultant, and to Mr. Sam Mor-
ford not only for many hours of statistical computation but
for his personal interest and technical skill.
To Miss Barbara Cooper, who administered all the
questionnaires, I wish to add very special thanks for her
enthusiasm for this project. Her rapport with patients and
warm understanding of their problems and feelings stimu
lated patient response and suggestions for improving the
questionnaire. I am also most grateful to the nurses and
nurses aides of Building 257 who found time in their busy
schedules to rate patient behavior with such care and sen
sitivity. I wish as well to thank the other professional
staff members of Building 257 without whose support and
encouragement the project could not have been accomplished.
To Dr. Shneidman and Dr. Norman L. Farberow, Co
principal Investigators of the Central Research Unit, a 7 A
project devoted to the study of suicidal behavior, I am
grateful both for their friendship and scientific knowledge
as well as for the stimulating ideas emerging from their
important research efforts.
TABLE OF CONTENTS
CHAPTER Page
I. THE PROBLEM..................... 1
II. SURVEY OF LITERATURE............ 8
III. METHOD........................... 28
Subjects
Hospital Setting
Prediction of Behavior from
Self-report Questionnaire
Staff Ratings of Patient
Behavior
Tabulations of Nurses Notes
in Clinical Charts
Analysis of Data
IV. RESULTS AND DISCUSSION.......... 62
V. IMPLICATIONS FOR USE OF THE
QUESTIONNAIRE............... 85
VI. SUMMARY......................... 98
REFERENCES............................... 104
APPENDIX................................. 114
v
LIST OF TABLES
TABLE Page
1 Census Characteristics of Subjects in Study 30
2 Census Characteristics of Subjects in Study 31
3 Census Characteristics of Subjects not
Included in Study because Testing was
not Obtainable........................ 35
4 Census Characteristics of Subjects not
Included in Study because Testing was
not Obtainable......................... 36
5 Length of Hospitalization and Hospital
Status at Time of Behavior Ratings 47
6 Reliability Study of Behavior Rating Form . 51
7 Correlations between each Prediction of
Behavior and its Corresponding Item
for Staff Rating of Behavior.......... 66
8 Correlations between Predictions of
Behavior and Staff Ratings of Behavior 67
9 Relationship between Implementing Predictions
and Extent of Nurses Notes in Clinical
Chart.................................. 70
10 Rotated Factor Matrix: Predictions of
Behavior............................... 73
11 Rotated Factor Matrix: Staff Rating of
Behavior............................... 74
12 Rotated Factor Matrix: Predictions and
Staff Ratings Combined................ 75
13 Relationship between Criterion Keys and
Extent of Nurses Notes................ 79
14 Comparison of Criterion Key 2 with Notations
of Excessive use of Alcohol in Chart... 81
15 Three-way Comparison of Predictions, Staff
Ratings and Extent of Nurses Notes 83
vi
LIST OF FORMS
INCLUDED IN APPENDIX
FORM Page
Self-Report Questionnaire............ 115
Description of Criterion Keys........ 124
Criterion Keys....................... 125
Instructions for Scoring Self-Report
Questionnaire.................... 131
Prediction Work Sheet. ............. 132
Prediction of Behavior Form.......... 133
Staff Rating of Behavior Form........ 134
vii
CHAPTER I
THE PROBLEM
The purpose of this study was to develop a self-
report questionnaire which could be used to predict overt
behavior patterns of hospitalized psychiatric patients.
Assessment of behavioral potential is a problem of
particular urgency in psychiatric hospitals where judgments
about the care and treatment of newly admitted patients
must be made quickly and too frequently on very limited in
formation. The large psychiatric hospital with minimal
staff and heavy influx of admissions may of necessity be
forced to process new patients in the same manner regard
less of individual differences and needs while they wait
their turn for individual processing. This period of time
may be unfortunately long. What happens to the patient in
these crucial first days after admission may, however,
critically influence the subsequent course of his illness.
Goffman (1961) has pointed out the drastic loss of
identity which faces the newly admitted patient who is
stripped of status, belongings and normal communication
channels. The importance of treating the patient as an
individual is emphasized by many others including Sivadon
(1957), Pine and Levinson (1961) and Giedt (1961). This
humanistic approach has been put into action at a California
state hospital where the admission procedure is completed in
one hour with full emphasis on immediate cognizance of the
individuality and varying needs of each new patient (May,
Garrett, Nash and Jordan, 1962).
In an effort to provide some easing of this problem,
the current study has two practical goals: (1) To provide
the hospital staff with an efficient means for assessment of
individual differences in terms of three behavioral patterns
of major importance from the standpoint of patient needs and
hospital management. These are dependency orientation,
activity level and the manner in which energies are chan
neled, and implementing behavior or the apparent need to
control the environment as opposed to acceptance of the en
vironment as it is found.'*' (2) To provide the newly admit
ted patient with an opportunity for direct communication of
his feelings about his illness, his need for hospitalization
and his expectations concerning his care and treatment.
Important as these practical considerations are,
however, there are theoretical issues in the present study
Although the focus of this study was not upon
suicidal behavior, interest in these behavioral patterns
stems directly from suicide research projects which indica
ted their importance, particularly that of implementing
behavior, in the prediction of suicide (Shneidman, Farberow
and Leonard, IS62, and Farberow, Shneidman and Leonard,
1963).
3
which are of equal importance. The major purpose is to
determine whether a screening device (of the self-report
type) given to all newly admitted psychiatric patients will
predict later behaviors. This is based on several assump
tions which may be formally stated as hypotheses:
Hypothesis A
There are enduring personality traits expressed in
overt behaviors which are not totally obscured during epi
sodes of psychiatric disturbance.
One implication would be that it is possible to take
a series of consecutive psychiatric admissions and assess
the personality traits under study without regard to psychi
atric label and on this basis to make meaningful assertions
about the individuals. Behavior apparent on admission, pre
sumably during the height of illness, will be observable at
a later period when some improvement should have occurred.
Thus there are customary ways of reacting which remain rela
tively stable regardless of the type of psychiatric illness,
and the illness itself is not of sufficient importance that
it obscures the customary patterns of behavior.
Hypothesis B
Dependency orientation, activity level and imple
menting behavior represent traits which are measurable,
predictable and have important effects upon the hospital
course of the psychiatric patient.
Dependency Orientation
The trait of dependency presumably arises from the
human infant*s helpless period during which others provide
for its physical and social well-being. As the child
develops it becomes increasingly able to rely upon its own
resources and independent behavior becomes possible.
Whether the adult still relies heavily upon others, re
jects such dependence severely, or is able to move flexibly
between these extremes in appropriate and adaptive behavior
is largely a question of individual and cultural variation.
The question posed in this study is whether these
orientations can be measured and predicted in the psychi
atric patient. Such behaviors are of considerable impor
tance in the study and prediction of human reactions in
general but particularly so in the emotionally disturbed
population where extremes of orientations are commonly en
countered and present difficult problems in treatment and
management.
Activity Level
Individual differences in general activity level are
commonly recognized as well as the variations in the manner
in which this energy is released. Personality theorists
such as Guilford (1959) and Duffy (1962) report considerable
experimental evidence in support of this trait.
Some individuals appear to function at a high ener
gy level; others in more phlegmatic fashion. Some perform
at a consistent rate within the over-all level; others per
form in inconsistent bursts of activity which may or may
not be adaptive and accepted by others.
In addition to variations in general activity level
there are great differences as well in the manner in which
the energy is released or channeled and here the question
of psychiatric illness is of particular interest. To what
extent does such illness disrupt normal patterns of activ
ity? Is the illness merely a symptomatic extension of mal
adaptive activity patterns and other predisposing factors?
Is it possible from answers to a self-report questionnaire
given newly admitted psychiatric patients to predict the
activity or energy level and the manner in which it is chan
neled at a later period of hospitalization when there should
have been some lessening of psychiatric disturbances? These
are questions of primary interest in this study.
Implementing Behavior
This is a term for an important behavioral pattern
which emerged from studies of suicide in both psychiatric
and non-psychiatric populations (Shneidman, Farberow and
Leonard, 1962, and Farberow, Shneidman and Leonard, 1963).
It is wilful, needful and controlling behavior which may be
subtle or blatant.
This behavior indicates a strong need to control or
change the environment as opposed to an ability to accept
environmental conditions somewhat flexibly without an undue
injection of personal needs and wishes. Implementing
behavior is related to but not synonymous with autonomous
behavior. It appears rather to reflect a conflict surround
ing strong needs for both autonomy and dependence. This
conflict, which is normally worked through in very early
childhood, has never been effectively resolved and has re
sulted in continuing problems centering about need for con
trol and fear of control. Thus excessive dependency on
others to provide behavioral controls, excessive refusal to
submit to controls, or a vacillation between these extremes
is reflected in the behavior labeled "Implementing."
Aside from the question of suicide, this behavior
is of major importance from the standpoint of patient needs
and hospital management problems. Further study of the be
havior seems urgently needed and in particular an efficient
means of assessing this behavior upon admission seems desir
able. Another question of interest was an assessment of the
degree to which this behavior would be present in an average
group of patients. How many patients would evidence this
behavior? Would it appear to adhere to certain diagnostic
categories or would it seem relatively independent of
7
diagnostic category as had been the case in the suicide
research referred to above?
These intriguing and important questions formed the
impetus for inclusion of this behavior pattern in the pre
sent study. The concept of implementing is described more
fully in Chapter II together with examples of the empirical
basis upon which it refits.
Hypothesis C
It is possible to develop a self-report question
naire for newly admitted psychiatric patients which can be
objectively scored and used to predict overt behaviors.
Although results differ, many investigators have
found that this type of test is useful for predictive pur
poses (Sandler, 1954; Hillson and Worchel, 1957; Dahlstrom
and Welsh, 1960; Birkman, 1961; and Semeneoff, 1962).
Unquestionably self-report forms have a practical
advantage in terms of administration and scoring since they
need not be administered individually and scoring may be
done clerically. This results in considerable economy in
terms of professional time and number of patients reached
through psychological testing.
CHAPTER II
SURVEY OF LITERATURE
The review of literature relevant to this study will
be considered in terms of the three hypotheses presented.
Hypothesis A
There are enduring personality traits expressed in
overt behaviors which are not totally obscured during epi
sodes of psychiatric disturbance.
Allport (1937), Cattell (1955), Guilford (1959),
Eysenck (1960) and many other leading personality theorists
attribute enduring structure to personality and recognize
the presence of stable traits. Allport's description per
haps most accurately defines these traits as determining
tendencies, long-range mental sets, or dispositions of
readiness to behave in certain ways. In discussing traits
and resulting behavior, Guilford states,
. . . to those who are reluctant to think in terms of
causes, let us say that all we need to mean is that be
havior can be partially accounted for or predicted by
the personal properties called traits. It is not nec
essary to assume that traits by themselves can bring
about behavior or even account for its properties.
Behavior does not occur without temporary instigating
forces provided by drives or motives on the one hand
and by external stimulation on the other. Behavior is
a joint product of temporary organic conditions, the
situation and personality traits, all operating to
gether at the moment. (1959, p. 38).
8
9
A great many studies have been conducted, particu
larly through factor analysis, to locate major personality
traits. In such a study Jenkins (1962) assembled items
from all personality tests existing in 1946 together with
additional items from laymen and other professional persons
to arrive at a total of 128 primary traits plus two super
factors. Other investigators such as Cattell (1955),
Guilford (1959), Comrey (1962) and Eysenck (1960) arrive at
smaller numbers of primary traits.
Relatively few longitudinal studies of the stability
of traits have been conducted. Those reported by Strong
(1951), Smith, M. (1952), Owens (1953), Kelly (1955), Kagan
and Moss (1960), and Bronson (1962) indicate that there are
sex differences and differences in terms of the trait under
study. For example, Bronson (1962) was able to predict ex
tent of "involvement with people" and "competence and mas
tery" from observation of behavior of boys before the age
of three although he was not successful in doing so with
girls. Kagan and Moss (1960) could predict dependency
traits for women but not for men. Kelly (1955) feels that
some changes occur in personality during adulthood although
perhaps minor.
Let us return for the moment to the question of
psychiatric illness. What effect does this have on basic
personality traits? Certainly, in the studies on suicide
10
which formed the impetus for the present study, there were
indications that certain personality traits important in
suicidal behavior were relatively independent of diagnostic
category. Indeed, it is apparent from direct observation
of patients as well as from studies of the literature that
similar traits occur among widely differing diagnostic cate
gories. For example, there are over-dependent schizophrenic,
over-dependent ulcer and over-dependent neurotic patients
just as one may find hostile, independent schizophrenic,
ulcer or neurotic patients. As Numberger states,
. . . one of the elementary concepts (for the analysis
of symptoms) is that the human organism whether adapted
or maladapted can only exploit behavioral mechanisms
which have become a part of his familiar repertoire
acquired during the course of development. (1961, p.33)
Lorr (1955) has found many personality traits which
are descriptive of psychiatric patients and which are rela
tively independent of the formal diagnostic classification
now in use by most psychiatric institutions. He expresses
the feeling of many other investigators when he reports his
dissatisfaction with current psychiatric classification
(Ash, 1949; Mehlman, 1952; and Wood, Rakusin and Morse,
1962).
King states flatly, "The inadequacy of the present
neuropsychiatric diagnostic categories is practically
common knowledge" (1954, p. 383). He suggests using the
approaches of Wittenbom and Lorr in treating diagnosis as
11
a multi-dimensional approach not only in regard to psycho-
pathological dimensions but also in terms of relevant per
sonality variables.
Foulds feels also that personality measures are
important and are more stable than diagnostic measures,
stating,
. . . a personality trait is taken to be a relatively
consistent and enduring . . . personality variable.
Symptoms and signs of mental illness are distinguish
able from traits in that they are neither universal nor
necessarily enduring. Indeed they indicate a break in
the normal continuity of behavior. (1961, p. 269)
Hospital treatment and procedures may modify dis
turbing symptoms which form the basis for psychiatric diag
nosis but strong and enduring ways of responding and rela
ting to others are less amenable to change than are such
symptoms as delirium tremens, hallucinations and ideas of
reference. It is these enduring personality traits which
require modification if the patient is to prevent recur
rence of symptoms of maladjustment which form the basis for
psychiatric classification.
Hypothesis B
Dependency orientation, activity level, and imple
menting behavior represent traits which are measurable,
predictable and have important effects upon the hospital
course of the psychiatric patient.
12
Dependency Orientation
The dimension of "dependence-independence" has long
been considered an important personality trait. Dependency
has been defined by Gordon (1953) as a security system
oriented toward external sources of support (people or ob
jects) while mature independence is defined as a security
system which "evaluates and integrates realistically both
internal and external sources of support, thus making for
flexibility." Kagan and Mussen define dependency as
. . the need for emotional or authoritative support
in most situations, difficulty in making independent
decisions and taking on responsibilities and dread of
loneliness" (1956, p. 30).
These definitions represent the concept of depen
dency as it is used in this study. Of special interest
here was the expression of dependency needs, particularly
extremes of dependence or independence or the presence of
conflict in this area. For example, the very dependent
patient who is constantly at the nursing station makes it
difficult for the staff to meet his needs and yet to appor
tion time fairly among other patients.
On the other hand, the patient who refuses any sort
of care or treatment and is extremely resistant to receiv
ing help from others manifests an extreme of independence
which presents great difficulty to the staff in treating
13
his illness and in keeping him comfortable within the
social milieu of the hospital. The patient who vacillates
between over-dependence and rejection of dependency is in
the throes of a dependency conflict which leaves him no
satisfactory means of meeting his needs and can be a real
trial to the psychiatric staff charged with treating his
illness.
The presence of a consistent unified trait of
dependence is reported in experimental studies as well as
in clinical observation. This trait is included in the
Dynamic Factors Opinion Survey, an inventory of needs and
interests (Guilford, Christensen, and Bond, 1954).
Particularly interesting are studies of dependence
such as those by Gordon (1953), Witkin, Karp and Goodenough
(1959), and Elliott (1961), in which subjects who have been
rated in terms of extent of dependency by some criteria
(such as clinical judgment or test response) are given
tasks such as finding a simple figure which has been em
bedded among more complex figures or judging verticality by
kinesthetic cues while resisting visual cues. Here the
subject must resist external cues and rely on his own in
ternal cues in making judgments, and the subjects who have
been rated as dependent by other criteria do indeed show
difficulty in relying upon their own internal cues.
This quality is tapped in further studies such as
those of Caruth (1959) and Weiss and Emmerich (1962) in
14
which the subject is required to perform tasks in which he
must take into account judgments or social pressure from
other persons present. Again, the dependent person tends
to rely on external cues while the independent person is
able to depend upon internal cues.
There is much overlapping between studies of depen
dence and those of social conformity with fairly general
agreement by such investigators as Couch and Keniston (196L)
that the dependent person is an "agree-erM or one who ac
quiesces to the judgments of others regardless of the stim
ulus presented. That conformity or the tendency toward
answering in a socially conforming manner is only part of
the determinants at work, however, is indicated by Fordyce
and Lamphere's study (1960) in which independence/depen
dence was clearly differentiated from social desirability.
Other studies which have important implications for
the treatment of psychiatric patients are those by Cairns
(1961), Goldman (1961) and Cairns and Lewis (1962) which
indicate that not only is the trait of dependence measur
able but rates of conditioning and response to reinforce
ment are decidedly affected by the trait. Response to
psychotherapy is also reported as being affected by the
trait (Blyth, 1961).
That the trait is of major importance in develop
ment is seen in studies of the influence of dependence on
15
learning (Flanders, Anderson and Amidon, 1961) and in the
development of various forms of pathology including delin
quency (Ward, 1958, and Bandura and Walters, 1958), alcohol
ism (Munt, 1960) and psychosis (Downing, 1959, and McCord,
McCord and Verden, 1962).
Studies of suicide reveal the great importance of
dependency needs and frustration of these needs in the
suicidal patient (Shneidman, Farberow and Leonard, 1962,
and Farberow, Shneidman and Leonard, 1963). Any hospital
that has tried to move an over-dependent patient out of the
hospital or to keep an extremely independent patient in the
hospital has had practical experience with the treatment
problems inherent in dependency orientation. There is the
risk of disturbance in the patient, including in some the
possibility of suicidal action.
Activity Level
Duffy (1962) has devoted a recent volume to discus
sing this trait and presenting evidence in support of its
existence. In her words,
. . . individuals differ to a marked degree in the ex
tent of activation in the same situation; these differ
ences in activation tend to persist and to characterize
the individual; and in general, the individual who re
sponds with a high degree of activation in one situa
tion is likely, as compared with other individuals to
respond with a high degree of activation in other situ
ations also. Justification is thus given for conceiving
of some individuals as being excitable or responsive,
and for conceiving of other individuals as being phleg
matic or unresponsive. (1962 , p. 227)
16
She cites much experimental evidence with dogs,
rats and mice as well as some evidence with humans. She
also notes that there are age'and sex differences as well
as individual differences in humans. Not included in this
volume are two more recent studies which support the state
ment that activity level is related to age and intelligence
(Bastendorf, 1961, and Cromwell, Palk and Foshee, 1961).
Guilford reports that a factor of "general activity"
has been verified several times and is measured in the
Guilford-Zimmerman Temperament Survey (1959, p. 184).
The variation in activity level from individual to
individual is also complicated by the fact that there are
great differences in the manner in which the energy level
is released or channeled. It may be in stable productive
fashion or in impulsive bursts of activity which may or may
not be productive. The problem of destructive bursts of
unchanneled energy is particularly acute among a psychia
tric population and among a suicidal population.
In this study the focus was not only upon the ac
tivity or energy level, but also upon the manner in which
activity was channeled, whether held in with rigid control,
released readily, or released in impulsive bursts. Guil
ford (1959, p. 351) reports factors related to the discharge
of energy, such as an impulsion factor with impulsion mean
ing the rate of initiation of an act. He also describes a
17
temperament factor of "impulsiveness vs. deliberateness" or
a tendency to react promptly without thinking (1959,
p. 412).
Twain's study (1957) of behavioral control, parti
cularly impulsivity, suggests that impulsivity is not a
unitary trait but may be a combination of factors from
physiological to motivational. Sanford, Webster and Freed
man (1957) in a study of college women and psychological
test results relate the impulsivity factor isolated to a
general disturbance in adjustment, and note also the pos
sible importance of physiological factors.
Frosch and Wortis (1954) have attempted to classify
impulse disorders as symptom and character dimensions, with
such disorders as perversions, sadism and homosexuality
considered symptoms and psychopathological and organic
states considered characterological.
Child development studies and studies of other cul
tures suggest that inadequate inner control systems and
resulting difficulty with impulse control may stem from
deviations in early childhood training, particularly about
the individuation phase. Erikson (1950) and Parsons (1961)
describe this phase as occurring around the second year of
life with the rapid development of motor and verbal ability
and the resulting capacity for independence. At this time
the child must develop an inner control system which is
18
adequate for his own needs and the demands of his particular
culture. He must learn what limits he is expected to set
for himself and what may be expected from the environment.
This process is based upon identification with adequate
environmental figures who represent appropriate cultural
requirements. Enduring dependency conflicts, difficulty
with impulse control and general maladaptation appear to
stem from inadequate resolution of the individuation strug
gle.
Specific studies may be cited in general support of
these statements. Rexford (1963) points out that it is
during the second year of life that the development of im
pulse controls is of greatest importance and reports sup
portive evidence. Ritchie (1961) feels that difficulties
in impulse control among New Guinea tribes studied are
traceable to trauma during the individuation phase of child
hood. Bandura and Walters(1958) feel that impulse control
difficulty arises in our culture as a result of rejection
of dependency needs by parents and poor parental identifi
cation as a result. Goldman-Eisler (1951) relates impul
sivity in part to disturbances of patterns of early weanirg.
In turning to the question of suicidal behavior,
problems of impulse control are frequently seen. Durkheim
(1897) in his early actuarial study of suicide and its
sociological implications laid the cause squarely in terms
19
of breakdown of control systems but felt that this stemmed
from lack of external controls rather than from unsatisfac
tory inner control systems in the individual.
The question of the extent of difficulty with im
pulse control seen in a general psychiatric population
should also be touched upon. Giedt (1961) found this dif
ficulty characterized less than half of the 225 patients he
studied in a psychiatric hospital.
Implementing Behavior
This behavior pattern became increasingly evident
in the previously referred to studies of suicide conducted
by Shneidman, Farberow and Leonard. This pattern was not
associated with any particular diagnostic group and was
apparent in both psychiatric and non-psychiatric populations.
The pattern was first labeled "Dependent-Dissatisfied" and
later "Implementing Behavior" in an attempt at more precise
identification and description of the behavior.
Implementing behavior appears to reflect an unusual
need to control or change the environment as opposed to an
ability to accept environmental conditions without unusual
imposition of personal needs. Perhaps the closest synonyms
are needfulness and wilfulness although these terms do not
fully represent the concept.
Empirically the trait is evidenced in behavioral
notes in hospital charts, such as the following verbatim
20
notations from Subject 22 in the present study:
. . .refused to take liquid Mellaril. Agreed to take
pill.
Requesting medication for loose stool.
Very generous; offering nurse position with Mayo Clinic.
Bothering other patients.
Refused oral medication.
Talking of chartering a plane and going to Las Vegas.
Smoking in seclusion; set mattress on fire.
Complaining of soreness in left upper arm.
Helped sweep floor at bedtime. Had large BM in bed.
Foot soaks to lesion on right heel. Difficult to keep
shoes on him.
Claims liquid makes him sick. Given pill and checked
to be sure he had swallowed it.
States he has had several loose stools today and this
evening. Instructed to show same to nurses aide.
None noted.
Threatening to elope. Claims he's only waiting till
Saturday when his check comes.
Wanting to soak his foot (lesion on heel).
Requesting cough medication.
Awake since 12:30 a.m. making and remaking bed.
This record is quite representative of implementing
behavior. Notations similar to the above are found in
abundance. This patient averaged one or more such notes
daily for his hospitalization period. In contrast nearly
half the subjects in the present study averaged one or less
such notes in a ten-day period. Thus the behavior is not
typical of a general psychiatric population and was found
in the present study in approximately 20 per cent of the
subjects. The fact that this behavior is also found in
non-psychiatric populations in which suicide occurs lends
support to the belief that it represents a personality
trait or cluster of traits rather than a symptom of psychia
tric disturbance.
21
Since the term "Implementing" was introduced spe
cifically for the suicide studies previously cited, rele
vant literature must necessarily stem from the concepts
most closely related to that of implementing behavior.
Another look at the empirical evidence above is
indicated. First there are special requests for medication
coupled with refusals to accept certain kinds of treatments
or restrictions. There are active involvements with envi
ronmental figures. Difficulties with control are apparent.
This was a patient who wanted to set the terms himself but
at the same time appealed to others for assistance. Un
questionably he was a difficult, dissatisfied and demanding
patient whose behavior would arouse some resentment or ir
ritation in other people around him. An inadequate inner
control system seems evident.
Both needfulness and wilfulness are reflected in
these behavioral notes. There is a need to change things
in some way and to become actively involved in the envi
ronment. Thus both activity and dependency are part of
the pattern. The most closely related traits which are
found in the literature are those of dominance, autonomy
and manipulative behavior, which are discussed below. None
of these, however, seems quite to fit the trait or perhaps
cluster of traits which is reflected in implementing be
havioral notes.
22
Guilford (1959) reports traits of ascendance, self-
assertion and dominance and also delineates a factor of
"need for attention" as well as a factor of "self-reliance
vs. dependence." Thurstone (1938) refers to an ascendant or
dominant score and Gough's California Psychological Inven
tory (1957) has a dominance score. Cronbach (1960) cautions
against reliance on trait names, however, and feels that
they are a source of confusion. He suggests that the only
useful way to discuss these traits is by specific reference
to the investigator and the measures used (1960, p. 468).
The drive for autonomy or power is cited in develop
mental literature by Adler (1927), Hendrick(1943), Erikson
(1950), Veroff (1957), White (1959), Murphy (1960), and
Phillips (1961), and there appears to be considerable agree
ment that there is a power motivation or drive. A search of
the experimental literature with humans reveals relatively
little although the concepts are frequently discussed.
Veroff (1957) attempted to develop and validate a projective
measure of power motivation. Bronson (1962) rated children
under the age of three for competence and mastery, tested
them again at ages nine to ten and one-half and found that
the trait was predictable for boys but not for girls.
Angyal (1941) refers to the distinction between
autonomy and homonomy and defines homonomy as a tendency to
be in accord with forces from the outside which impinge
23
upon the person. Certainly in this sense, homonomy is des
criptive of non-implementing behavior.
From the empirical evidence presented in the nurses
notes above it is apparent that the term autonomy does not
quite represent the behavior although there is a heavy
overlay of this tendency. Although the patient was cer
tainly insistent upon setting the terms himself, his behav
ior also suggests that he wants and is directly asking for
attention and nurturance from others. One might well ex
pect the autonomous person to be primarily goal oriented
and relatively uninterested in unnecessary involvement with
others. The implementing behavior presented appears to re
flect a disruption of the autonomous drive in which there
is neither acceptance of outer controls nor goal-directed
autonomy. Rather there appears to be a conflict over con
trol, with a vacillation between the need to control or
dominate and the need to be nurtured.
Implementing behavior bears a resemblance to manipu
lative behavior in that the person is maneuvering and at
tempting to make changes in the environment. There appears
to be a distinction again in terms of goal orientation.
The manipulator or "operator" has goals and doesn't hesi
tate to use devious means to attain them. We may not ap
prove of his methods but his purpose is relatively clearcut
and adaptive from a self-centered point of view. He wants
24
something, is willing to connive to get it and seems to
enjoy it when he gets it. Implementing behavior seems to
reflect instead a diffuse general dissatisfaction and dis
comfort, with attempts to adjust the environment in order
to reduce discomfort. Manipulative behavior is referred to
frequently in description of clinical behavior but there is
a lack of direct study of the trait or attempts to isolate
its components.
Although the concept of implementing is difficult
to define semantically, the agreement among raters of the
trait in studies of suicidal behavior is quite high. There
may be objections to the label or to the interpretation of
underlying dynamics but there is very strong agreement con
cerning presence or absence of the characteristic. This
agreement is present whether based upon a study of a case
history or upon personal observation.
In the present study, staff raters had very little
difficulty in rating implementing items and expressed the
feeling that the behavior was important. Although the
label was new, the concept appeared quite familiar to the
raters.
Because of the major importance of this trait in
general interpersonal relationships and in suicidal behav
ior, further study of the behavior and clarification of its
complex components is warranted.
25
Hypothesis C
It is possible to develop a self-report question
naire for newly admitted psychiatric patients which can be
objectively scored and used to predict overt behaviors.
Ellis in his 1946 review of the literature to that
time reached the conclusion that group administered paper
and pencil questionnaires are of dubious value in distin
guishing between groups of adjusted and maladjusted individ
uals and that they are of much less value in the diagnosis
of individual adjustment or personality traits. Nevertheless
the vast literature, particularly on the Minnesota Multi-
phasic Personality Inventory (MMPI) which has accumulated
since that time contraindicates this position and is pre
sented in Dahlstron and Welsh1s 1960 MMPI Handbook.
For the most part this literature reports attempts
at discrimination of diagnostic groups or extent of malad
justment on the basis of response patterns. Many, however,
are attempts at predicting responses to therapies of various
types, academic achievement, need for hospitalization, suc
cess in the armed services, success in teaching or in medi
cal school, juvenile delinquency, alcoholism, irregular dis
charge from a hospital, social participation and an infinite
variety of behaviors.
Much i n t e r e s t h a s b e e n d i r e c t e d to w a rd i n v e s t i g a t i o n
of the idea that perhaps paper and pencil self-report
questionnaires actually measure only two or three specific
behavior patterns such as acquiescence (Wahler, 1961), or a
tendency to give socially desirable responses (Edwards,
1953; Kenny, 1956; Rosen, 1956; Klett, 1957; Wahler, 1961;
and Edwards and Diers, 1962), or to overgeneralize
(Jackson and Messick, 1958) or perhaps simply to claim symp
toms (De Soto and Kuethe, 1959). For the most part, however,
these investigators have not used self-report questionnaires
to predict that subjects will evidence these particular
behaviors in other areas besides test responses.
In spite of criticisms of self-report question
naires, successful use of such tests has been widely re
ported by such proponents as Dahlstrom and Welsh for the
MMPI (1960) and Sandler for the Tavistock Self-Assessment
Inventory (1954). Certainly there are practical advantages
to the use of such tests. They may be group administered
and clerically scored, thus saving valuable staff time re
quired for administration and scoring of individual tests,
particularly projective tests. In addition the time-con
suming nature of individual tests prohibits their use with
all but a select few patients. In spite of these difficul
ties however, projective tests have been used in efforts to
predict such behaviors as hostility, assaultiveness or
elopement tendencies (Gluck, 1955; Vernier, Whiting and
Meltzer, 1955; and Smith and Coleman, 1956). Such attempts
27
are sparse and somewhat uncertain as to results in contrast
to the numerous efforts at prediction with self-report
questionnaires and the many optimistic results cited (Dahl-
,strom and Welsh, 1960).
In terms of the behavior patterns in this study,
Diamond (1957) believes that activity level and social com
pliance are two dimensions tapped by the MMPI, but does not
refer to specific scales.
Navran (1954), Gordon (1953) and Warn (1958) report
dependency scales for the instrument and Gough, McClosky
and Meehl (1951) and Leary (1957) report dominance scales.
Thus it would appear that this type of test may be useful
in assessing the presence of these and related patterns.
CHAPTER III
METHOD
This study focused upon an attempt to predict overt
behaviors in a series of consecutive admissions to the
Brentwood Psychiatric Hospital of the Veterans Administra
tion at Los Angeles, California. Prediction was made on the
basis of patient responses to a self-report questionnaire
included in the admission psychological test battery.
A pilot study was first conducted during which the
questionnaire, scoring keys for use in prediction of behav
ior and a scale for staff rating of behavior were developed.
No subjects used in the pilot study were included in the
study proper. Practical problems of conducting the study
within this hospital setting were also worked through.
Upon completion of the pilot study the study proper
was begun using the following four-step procedure: (1) pre
diction of behavior from the self-report questionnaire, (2)
staff ratings of patient behavior, (3) tabulations of nurses
notes in clinical charts, and (4) analysis of data.
Before proceeding with a description of these steps,
a discussion of the subjects and the setting in which the
study took place should provide background information
which will make the methodology more meaningful.
28
29
Subjects
The 98 subjects consisted of all male patients ad
mitted to Building 257 at Brentwood Hospital during the 84
day period from January 21, 1963 to April 14, 1963,^ and
for whom testing was obtainable. All new admissions, read
missions and transfers from other buildings were included
except for those admissions which had been used previously
O
in the pilot study.
Census characteristics of the 98 subjects are repro
duced in Tables 1 and 2. Crumpton and Wine (1962) made a
study of the characteristics of Brentwood Hospital patients
in 1948 and 1958. Using this study as a basis for compari
son with the current 1963 study, the following similarities
may be seen.
There was an aging tendency evident, with the aver
age age of patients 34.7 in 1948, 37.3 in 1958 and 37.9 in
1963. As would be expected, there was a decreasing number
of veterans from the World War II period and the years
^-Building 257 was selected because it is fairly
representative of general Brentwood admissions and because
it receives the heaviest flow of new admissions, accounting
for nearly 40 per cent of all admissions either by direct
admission or by transfer from the reception building
(Building 258) after a brief period.
^Some subjects were admitted and separated again
more than once during this 84-day period but were included
only for the first admission period.
30
TABLE 1
CENSUS CHARACTERISTICS
OF SUBJECTS IN STUDY
N “ 98
Race_______________________per cent
Caucasian 85
Negro 12
Other 3
Religious Preference______per cent
Protestant 65
Catholic 26
Other 9
Marital Status per cent
Single 33
Married 33
Divorced 30
Other 4
TABLE 2
CENSUS CHARACTERISTICS
OF SUBJECTS IN STUDY
N - 98
Diagnosis_________________ per cent
Schizophrenia 58
Anxiety Reaction 12
Depressive Reaction 10
Passive-Aggressive
Personality 6
Other 13
Subtypes of Schizophrenia (among
patients diagnosed schizophrenic)
per cent
Paranoid 49
Undifferentiated 40
Other 10
32
preceding World War II and an increase in veterans of the
Korean crisis and the post-Korean years.
No change was noted in religious preference between
1948, 1958 and 1963, with a little over 60 per cent expres
sing a Protestant preference and nearly 30 per cent a Cath
olic preference. A rising tendency in the number of divorced
patients was noted, with 14 per cent in 1948, 19 per cent
in 1958 and 30 per cent in 1963. A racial difference was
noted with 85 per cent Caucasian and 12 per cent Negro in
the 1963 population contrasted with 93 per cent Caucasian
and 7 per cent Negro in the 1948 and 1958 populations.
No other characteristics are directly comparable
with the Crumpton and Wine study but the consistencies
noted are sufficient to suggest that the population in the
present study is fairly representative of the general
Brentwood Hospital psychiatric population.
Approximately 80 per cent of the patients in the
present study had had previous psychiatric hospitalizations.
A study of the diagnoses given during the previous hospital
izations revealed that nearly half of these patients had
been given diagnoses in another and quite different diag
nostic category. For example, a diagnosis of Anxiety
Reaction might have been previously carried by a patient
currently diagnosed Schizophrenic. This points up the
difficulty in categorizing emotional disturbance.
33
The four most frequent current diagnoses were
Schizophrenia of various types (58 per cent), Anxiety
Reaction (12 per cent), Depressive Reaction (10 per cent)
and Passive-Aggressive Personality (6 per cent). Of those
patients diagnosed Schizophrenic. 49 per cent were Paranoid
Schizophrenic and 40 per cent were Schizophrenic Undiffer
entiated, as noted in Table 2.
The problem of elopements and alcoholic excesses
seemed quite prevalent in this population, with a history
of elopement or unauthorized separation from the hospital
in nearly 40 per cent of the patients and a history in the
clinical chart of excessive use of alcohol in over 50 per
cent of these men.
The hypothetical or modal patient of the 1963 study
may be described briefly as a 38 year old Caucasian Protes
tant veteran of World War II with a diagnosis of Schizophre
nia.
During the 84-day period of the study 55 additional
patients could not be included because testing was not
available. Thus a total of 153 patients were potential sub
jects but only 98 were included in the study. Testing was
not available for the following reasons: 15 refused test
ing of any kind; 10 did not finish their questionnaires or
had so many changes or omissions that they were unscorable;
20 were discharged too quickly for testing or rating to
occur; 5 were too ill and 5 were lost for miscellaneous
34
reasons. The census characteristics of the 55 patients who
could not be included in the study together with reasons
for their exclusion are given in Tables 3 and 4.
Hospital Setting
The study was conducted at Brentwood Psychiatric
Hospital, formerly known as the Veterans Administration
Neuropsychiatric Hospital of Los Angeles. This hospital
has a population of approximately two thousand patients
with varying neuropsychiatric diagnoses. A psychological
test program for new admissions has been in operation since
1947 (Grayson, 1950) and afforded an excellent opportunity
for the addition of the self-report questionnaire. The
number of admissions is also large enough that it was pos
sible to obtain a sufficient number of subjects before any
major changes in procedure or personnel could introduce
extraneous variables.
Justification for the use of a single building in
the study is based on the fact that this eliminated some
undesirable variables such as differences in staff proce
dures in the various buildings and the necessity for using
additional teams of staff raters in each building. In
addition a very large percentage of all new admissions were
assigned to the building used in the study, Building 257.
TABLE 3
CENSUS CHARACTERISTICS OF SUBJECTS NOT INCLUDED
IN STUDY BECAUSE TESTING WAS NOT OBTAINABLE
N ■ 55
Reason for Previous Hos- History of
lack of testing _____ Diagnosis_______ pitalization Elopement
Schiz
Pa.
Schiz
Undiff
Schiz
Other
Anx
React
Depr
React
Other Yes No Yes No
Unscorable
Tests
N-10
3 1 3 2 0 1 4 6 5 5
Refused to
Take Tests
N-15
5 2 3 2 0 3 6 9 8 7
Left Hospital
Too Quickly for
Testing or
Rating to Occur
N-20
2 1 6 2 4 5 10 10 15 5
Too 111, or
Lost for Mis
cellaneous
Reasons
N-10
0 3 4 1 0 2 4 6 4 6
TABLE 4
CENSUS CHARACTERISTICS OF SUBJECTS NOT INCLUDED
IN STUDY BECAUSE TESTING WAS NOT OBTAINABLE
N - 55
Reason for
Age Marital Status
Religious
Preference
Under 25
25 -29
30
-34-
35
•39
40
-44
45
-49
50 "
Plus
M S D Other Prot Cath Other
Unscorable
Tests
N-10
0 2 5 0 3 0 0 3 6 1 0 5 5 0
Refused to
Take Tests
N-15
0 0 4 3 3 1 4 6 5 3 1 8 5 2
Left Hospital
Too Quickly for
Testing or
Rating ,to Occur
N-20
0 4 3 4 4 4 1 9 5 4 2 14 4 2
Too 111, or Lost
for Miscellaneous
Reasons 1
N-10
0 0 5 2 2 0 6 1 2 1 7 2 1
w
ON
37
Step 1. Prediction of Behavior
from Self-Report Questionnaire
In this step questionnaires were administered to a
series of consecutive admissions. The forms were objective
ly scored through the use of criterion keys which provided
a basis for prediction of behavior in terms of the patterns
under study: dependency orientation, activity level and
manner in which energies would be channeled or released,
and implementing behavior.
a. Development of Self-Report Questionaire
The questionnaire was a 135 item True/False form
which each newly admitted patient filled out as part of the
regular admission psychological test battery.
Patient reaction to the questionnaire, which was en
titled "Letting You Know How I Feel" was carefully studied
and used as a basis, for refinement of items since a major
purpose of the form was to provide an opportunity for the
patient to communicate his feelings about his illness and
hospitalization and his reactions to his care and treat
ment. The form is reproduced in the Appendix.
All questionnaires were administered by one exam
iner who gave the regular test battery during the period of
the study. The questionnaire was included as part of the
battery with no special instructions beyond those given on
the face of the questionnaire. Questions about the test
38
were answered on an individual basis and in the same man
ner that questions were answered about other tests in the
battery.
b. Derivation of Items for the Self-report Questionnaire
Discussion of the derivation of items for the ques
tionnaire is important in order to clarify the rationale
behind the items. The initial items emerged directly from
studies of hospitalized schizophrenic suicides and reflec
ted behaviors which were important not only in the predic
tion of suicide but were also important in terms of general
patient welfare and hospital management. These behaviors
represented basic differences in dependency orientation,
activity level and needs for personal involvement and con
trol over the hospital environment (Shneidman, Farberow
and Leonard, 1962). Three suicidal patterns were distin
guished: Dependent-Satisfied, Dependent-Dissatisfied, and
Unaccepting. Active involvement in the hospital course was
present in all three types although manifested in quite
different forms. One non-suicidal type of patient was dis
tinguished and labeled "Marginal." This patient was de
scribed as quiet and accepting of the hospital environment.
While the current study does not focus upon suicide,
the behaviors under study are those which appeared particu
larly important in assessing potential suicidal behavior
and interest in these patterns arose directly from suicide
39
research. The four patterns distinguished in the study of
suicide among schizophrenics are also included in the pres
ent study as four of the criterion keys used as a basis for
prediction of behavior.
The fact that these behaviors are important, not
only in the evaluation of suicidal potential but are of
major importance in terms of patient needs and management
problems, provided much of the impetus for the present
research. Both an efficient means of assessing these be
haviors early in hospitalization and an actual evaluation
of the prevalence of these varying patterns in a consecu
tive series of admissions seemed urgently needed.
Initial questionnaire items, then, were direct a
priori attempts at translating these behavioral patterns
into statements concerning the patient's feelings and re
actions toward his illness and hospitalization and eventual
return to outside life.
For example, the following items were included for
the patient who overtly rejected dependency (the "Unaccept-
ing" patient in the schizophrenic suicide research):
Item 4. "I'm here because I was forced to come here."
Item 5. "There's nothing wrong with me that I can't
handle myself.
On the other hand, the following items were intended
for the patient who admits his dependency and clings to the
hospital for help (the "Dependent-Satisfied" patient in the
schizophrenic suicide research):
40
Item 10. "The hospital knows more about what I need than
I do. It's beyond me."
Item 12. "I am anxious to cooperate with whatever treat
ment they give me and am grateful for getting
it.
Examples of items intended to tap differences in
activity level and manner of channeling energy were the
following:
Item 22. "I stay out of other people's way and never
cause any trouble."
Item 28. "There are times when I just say, 'The Hell
with everything' and cut loose."
Item 46. "I am restless and uneasy unless I am working
hard."
Item 106. "I have always been impulsive; it's just my
nature."
Items intended to express differences in the need
for personal control were the following:
Item 63. "I manage my own affairs in the way I like."
Item 11. "I have a pretty good idea what I need and
would like to talk to the doctor as soon as
possible."
Item 21. "When other people decide what you're supposed
to do you might as well go along with it."
Item 66. "I don't take orders from other people."
Items predominantly reflected the three behaviors
under study but many other items were included which were
based upon hypotheses that were not the focus of this study.
For example, behavioral rigidity and reactions to changes
in the state of consciousness were included in item content
as well as a series of items concerned with feelings and
reactions concerning the use of alcohol.
Many neutral items were also retained since the
questionnaire was intended as an acceptable channel for
41
patient communication. Neutral items were those answered
in the same direction by almost all patients and therefore
presumably "popular" or non-threatening items. An example
is Item 75, "People need help from others now and then and
I'm no exception," to which almost all patients answered
"True."
One can not be certain, of course, that a patient
will answer in the direction consistent with his actual be
havior, but initial items were intended to provide direct
opportunity for commication of feelings. Although many
such direct items did indeed prove efficient discriminators
of behavioral differences, it was not necessary that they do
so since scoring was done not on the basis of item content
but by the method of criterion keying.
c. Scoring the Questionnaire
In the method of criterion keying the truthfulness
of the patient's responses is not at issue since prediction
is based upon locating similar response patterns to the
questionnaire (regardless of item content) which are char
acteristic of patients who also share similar behavioral
patterns which are of interest. This is the method which
was used in formulating the scoring for the Minnesota
Multiphasic Personality Inventory (Dahlstrom & Welsh, 1960).
This approach has been described by Berg and Bass
(1961) in their discussion of the "Deviation Hypothesis."
42
They state: "The important thing is that the deviant group
has some validly ascertainable characteristics in a criti
cal area which sets them apart behaviorally from other
groups" (p. 343). Thus if all patients who answered "True"
to Item 33 on the present questionnaire shared some behav
ioral characteristic which distinguished them from all
patients who answered "False" to the item, this item could
be considered a means of predicting the behavioral charac
teristic regardless of content of the item.
This assessment technique has been variously called
the "Nominated Group," "Criterion Group" or "Criterion
Keying" method, with the following description by Meehl
justifying the general approach:
. . . tCriterion KeyingJ consists simply in the explicit
denial that we accept a self-rating as a feeble surro
gate for a behavioral sample and substitutes the asser
tion that a "self-rating" constitutes an intrinsically
interesting and significant bit of behavior, the non
test correlates of which must be discovered by empirical
means. (1962, p. 8)
The non-test correlates in this case were overt
behavior patterns which were determined by empirical means:
through direct observation of patient behavior on the wards,
through interview with staff members, and through analysis
of clinical charts. Again it should be emphasized that no
subjects used in the pilot study were used in the study
proper in order that criterion keys might be independently
established.
43
In the pilot study the questionnaire was given to
many criterion groups, that is, patients who were judged as
sharing certain behavioral similarities. Patients who
appeared to fit the patterns in the study of schizophrenic
suicides were included: Dependent-Satisfied, Dependent-
Dissatisfied, and Unaccepting (the three suicidal types),
as well as Marginal (the non-suicidal type). The question
of suicide was bypassed in selecting these groups; rather,
they were selected on the basis of sharing variations in
the behavior patterns under study which also had been noted
in the above suicide types. Other behavioral groups were
given the questionnaire as well: assaultive patients, com
plaining patients, phlegmatic patients, chronic alcoholics,
character disorders, process schizophrenics and many others.
When response patterns had been isolated for the
groups which appeared most clearly distinguishable in terms
of the behavioral patterns under study, scoring keys were
established for each of these patterns. During the pilot
study these scoring keys or criterion keys were used on new
groups of patients and revised until they seemed maximally
effective in distinguishing the behavioral types of inter
est in this study.
Ten criterion keys were adopted for scoring pur
poses in the final study. Two were double-scored; that is,
scored in one direction if the score was high and in an
other direction if the score was low.
44
Each of the criterion keys consists of a series of
from 8 to 27 items from the self-report questionnaire which
must be answered in a given direction. If the number of
items answered in this direction reaches a specified criti
cal number or cut-off point the key is considered scorable.
If this point is not reached the key is considered unscor-
able and is omitted. A cut-off point is provided for each
criterion key.
Each key is also provided with numerical values to
be assigned to each of the 12 items on the "Prediction of
Behavior" form. If a key is scorable, these values are
merely entered upon the prediction form. Assigned values
for each of the 12 behavioral items are averaged to obtain
a single score which is the final basis for prediction of
behavior. In the present study patients averaged three to
four scorable keys out of the possible total. All criteri
on keys, with cut-off points and values to be assigned to
the prediction of behavior for each of the 12 items, are
reproduced in the appendix together with detailed instruc
tions for following this procedure. Additional character
istics which describe the criterion groups may also be
found in the appendix.
The "Prediction of Behavior" form is a duplicate of
the form used for staff ratings of behavior and will be
discussed fully in the next section. Briefly, it consists
r
45
of 12 Items representing different ways of expressing the
three behavioral patterns under study. Four items are sub
sumed under each of the three patterns, making the total of
twelve. Each item is rated on a 1 to 7 point scale in terms
of the extent to which each behavior might be expected to
occur in the patient.
Step 2. Staff Ratings
of Patient Behavior
Staff ratings of actual patient behavior were se
lected as validation measures for the behavior predicted
from the self-report questionnaire.
a. Selection and Training of Raters
Nurses and nurses aides were selected to rate
patient behavior. They were found to be the only hospital
staff members who were fully acquainted with all patients
including those very quiet and retiring patients who do not
get the complete attention that more aggressive patients
receive. Further, experimental studies suggest that rat
ings by such staff members may be equal to and sometimes
superior to those given by other professional staff members
(Stilson, Mason, Gynther and Gertz, 1958, and Lasky, Hover,
Smith, Bostian, Duffendock and Nord, 1959).
Since Building 257, in which the study took place,
consists of two floors, each with a relatively independent
staff, it was necessary to train two sets of raters. In
46
addition, one rater from the evening shift was included.
Preliminary group training sessions were held with staff
members on each floor in which instructions for use of the
form were given and practice ratings were made to assess
competence in using the form and to provide a basis for
further discussion of the mechanics of its usage.
Staff members were most helpful in the development
of the rating form, both in improvement of item content and
wording and in streamlining the form.
b. Obtaining the Staff Ratings
Ratings were done independently by raters, a task
greatly simplified by the fact that the staff lacked both
time and proximity for inter-rater influence.
Ratings were made as soon as possible after a pa
tient had been separated from the building whether through
authorized separation (trial visit, leave of absence,
transfer to another building, or regular discharge) or un
authorized separation (elopement, irregular discharge,
absence without leave, or absence against medical advice).
In the case of patients who had not separated from
the building after a period of a month or more, ratings
were made while the patient was still in the building.
Table 5 presents the length of hospitalization at
the time of staff ratings as well as the patient's status
at the time of staff ratings (whether separated or still
47
TABLE 5
LENGTH OF HOSPITALIZATION AND
HOSPITAL STATUS AT TIME OF BEHAVIOR EATINGS
Hospital Status
Length of Patient
Hospital- Still in Authorized Unauthorized
ization Hospital Separation* Separation**
Number of Days
in Hospital
1- 9 0 2 0
10-19 0 5 5
20-29 1 5 3
30-39 0 5 8
40-49 3 8 2
50-59 3 3 3
60-69 8 4 2
70-79 10 3 2
80 plus 10 3 0
Total 35 38 25
*Authorized Separation: Trial Visit, Leave of Absence,
Transfer to Another Building, or Regular Discharge.
**Unauthorized Separation: Elopement, Irregular Dis
charge, Absence without Leave, Absence against
Medical Advice.
48
hospitalized). It is interesting to note that there does
not seem to be much difference between the authorized
and unauthorized separations in terms of number of days
spent in the hospital.
Eight raters did 94 per cent of the ratings on B
Floor and eight raters did 91 per cent of the ratings on A
Floor. Three of the sixteen were nurses; all from the B
Floor, day shift. A total of 344 ratings was secured with
an average of 3.5 ratings for each subject. Three patients
were rated by just two staff members and three patients
were rated by five staff members.
Ratings for each individual rater were transformed
into standard scores in order to reduce inter-rater varia
bility since some raters tended to rate consistently on the
high or low side. Standard scores for each subject on each
item were combined and averaged in order to obtain a single
numerical rating for each subject on each item.
c. Development of Staff Rating Form
This was done in both rational and empirical
fashion.^ A tentative form was designed to assess the be
havior under study and was subsequently refined and revised
^Many psychiatric rating forms now in existence
were studied hopefully but were found to include so many
variables extraneous to this study that drastic revision
would have been necessary. It seemed more efficient to
develop a form tailored to the present study.
49
until it seemed maximally efficient both in assessing the
behavioral patterns under study and in securing the ratings
with a minimum of disruption to the busy hospital staff.
The final scale consisted of 12 items representing
different ways of expressing the behavior patterns under
study. Four items are subsumed under each of the three
patterns, making the total of twelve. Each item was rated
on a 1 to 7 point scale in terms of the extent to which the
staff member had observed the behavior in the patient.
Provision was made on the rating form for the rater
to enter the degree of confidence he felt in assessing each
patient. In actual practice this did not prove particular
ly helpful although staff members seemed reassured by the
fact that this evaluation was available if they chose to use
it. Most raters either did not mark a level of confidence
or consistently marked one level, such as "Moderately Con
fident."
Provision was also made on the rating form for
those few patients whose behavior was so changable that it
was simply impossible to assign a single rating. This pro
vision was a "4x” rating placed just below the "4" on the
scale. The "4" indicated the midpoint between 1 and 7.
Very few raters resorted to the "4x" rating and no consis
tent pattern emerged in terms of items given this rating or
patients for whom the rating was felt to be necessary.
50
When a "4xM rating was given it was computed as a rating of
”4" in the final results.
Two revisions of the rating form were made during
the pilot study before it was felt to be adequate in terms
of content and minimal demands for staff rating time. A
reliability study was then conducted with six staff members
rating each of twelve patients on the form. Kendall’s Co
efficient of Concordance W was calculated and reached sig
nificance on ten of the twelve items beyond the .05 level
so it was felt that the form was now usable. The results
of the reliability study are given in Table 6.
The same form was used for both staff ratings of
behavior and as the form for entering predictions of be
havior obtained from the self-report questionnaire. Thus
the same form was used for two completely different and
independent measures. The duplication in form was felt to
be desirable since it eliminated unwanted variables which
would have been introduced had different forms been used.
The final form is reproduced in the appendix.
d. Item Content of Staff Rating Form
Although the Staff Rating Form is reproduced in the
appendix, items will be given verbatim here in order that
the rationale behind the items and staff reactions may be
more meaningfully presented.
TABLE 6
RELIABILITY STUDY
OF BEHAVIOR RATING FORM
Item W* Probability
1 .48
.01
2 .64
.01
3 .51
.01
4 .45 .01
5 .32 .01
6 .33 .01
7 .29 .05
8 .61 .01
9 .30 .05
10 .32 .01
11 .24 .10
12 .15 ns**
*Kendall's Coefficient of Concordance
**ns - not significant
52
Implementing Behavior (Items 1, 2, 3, and 4).—
These items are intended to express various ways in which a
need to control or change the environment may be manifested.
Item 1. "Needs a lot of attention and reassurance."
Item 2. "Complains a lot; needs to be noticed."
Item 3. "Demanding, insistent, resistive to routine;
wants things his way."
Item 4. "Anxious to please; goes out of his way to be
helpful."
Item 1 is intended to tap that behavior which is
controlling even though couched in positive manner. An
overlay of dependency is inherent in the item but such be
havior is also used to manipulate environmental conditions.
Staff members were instructed to consider the item
entirely from the patient's standpoint. If he appeared to
be definitely seeking attention or reassurance (or perhaps
hopefully waiting for it) he would be rated on the high
side; if, on the other hand, he neither appeared to be
seeking much interaction nor even particularly hoping for
it, he would be rated on the low side. Raters seemed to
have no difficulty with the item, although there was a
tendency to rate all patients toward the high side.
Item 2 was intended to measure overt attempts at
manipulation and Item 3 outright wilfulness. Staff members
had little difficulty rating these items and their predic
tive value was very good.
Item 4 proved to be a consistently poor item in the
study. It had been retained largely because it had seemed
n
53
promising in the pilot study for some patients in whom this
behavior appeared to represent a positive attempt to manipu
late and control through "good behavior" which might bring
special consideration in return.
Dependency Orientation (Items 5, 6, 7, and 8).--
These four items are intended to reflect different orienta
tions toward dependency:
Item 5. "Admits he is very dependent upon others."
Item 6. "Insists very strongly that he is independent
and can get along on his own."
Item 7. "Does whatever others tell him; seldom making
suggestions of his own."
Item 8. "Makes demands on others but somehow their
attempts to help him meet with failure (rate
hospital behavior primarily)."
Staff members were instructed to rate Item 5 entire
ly upon overt admission of dependency on the part of the
patient and to ignore their own feelings about his actual
dependency needs.
Many raters struggled with this task, particularly
when the patient was actually quite dependent but they had
difficulty recalling actual admissions of dependency. Un
doubtedly a cultural factor operates here since dependency
in men is generally considered less admirable than indepen
dence and overt admissions therefore less frequently were
apparent.
Item 5 proved to be unrelated to other dependency
items and emerged as a separate factor in the factor analy
tic study reported in the chapter on results.
[...........
54
On the other hand, Items 6 and 8 proved to be high
ly correlated in a positive direction and to Item 7 in a
negative direction.
Item 6 was intended to reflect a rejection of depen
dency needs regardless of the patient's actual dependence
or independence, and again staff members were rating on the
basis of the patient's expressed attitude. Raters had very
little apparent difficulty with this item and on occasion
expressed amusement with a comment such as the following:
"Patient A says, 'I don't need Nobody,1 but he's always
hanging around wanting this and wanting that."
Item 7 was intended to tap the range of actual be
havior regardless of attitude and it thus contains a strong
component of activity. It is phrased in a passive direction
and therefore shows a negative correlation with active
items. Staff members seemed to have no difficulty rating
the item.
With Item 8 the attempt was to reflect an underly
ing dependency conflict in which the person both insisted
that dependency needs be met but subverted these attempts
in some fashion, apparently because of anxiety concerning
the need for dependency. Raters appeared to enjoy this
item and some reported anecdotes such as the following:
"That really fits Patient B. I remember when he insis
ted that I get him an appointment with the chiropodist
and kept at me about it. Then when his appointment
slip came through he refused to go."
55
Activity Level (Items 9. 10, 11, and 12).— These
items were intended to assess both activity or energy level
and manner in which these energies are channeled or released
Item 9. "Has such a high energy level that it is a real
problem to hold himself in check."
Item 10. "Lets his impulses go freely (doesn't hold them
in. "
Item 11. "Holds his impulses in but has to watch himself
rigidly in order to do it."
Item 12. "Holds his impulses in for awhile and then cuts
loose."
Staff members expressed more difficulty in rating
these items than they did with the implementing and depend
ency items. In spite of this difficulty, however, Items 9,
10 and 12 were highly correlated both as predictions of be
havior and as ratings of behavior.
Item 9 was intended to assess the over-all activity
or energy level. Staff members had no apparent difficulty
in rating quiet patients but were perplexed by some active
patients. For example, one rater commented,
"Patient C has a high energy level all right but he
doesn't make any attempt to hold it in check."
The qualifying phrase, "it is a real problem to hold
himself in check" was included to sharpen awareness of those
patients whose general activity level was such that it might
be troublesome regardless of how control was or was not
exerted.
In Item 10 the attempt was to assess the manner in
which activity or energy level was channeled. In contrast
to Item 9, raters had no apparent difficulty with rating
56
patients who had high activity levels but puzzled about how
to rate the very passive patients who did indeed seem to re
lease their impulses rather freely but who did not seem
particularly troubled by impulses in the first place.
Item 11 was intended to provide a measure of the
type of control in which tremendous effort must be exerted
to maintain control over built-up energies. It was included
in part because of the importance of the characteristic in
sudden unexpected assaultive behavior, but did not prove
successful in predicting this behavior in the present study.
Item 12 was included to assess a problem with im
pulse control (especially noticeable among some alcoholics)
in which a repetitive, cyclic pattern was evident, with
quiet, rigidly restrained behavior followed by abrupt and
seemingly complete release of inhibitions and energies.
Staff members seemed to enjoy rating the item and
had very little difficulty applying it in this setting. It
correlated highly with both predictions and ratings of Items
2, 3, 6, 8, 9, and 10 (and with Item 7 in a negative direc
tion).
Step 3. Tabulations of
Nurses Notes In Clinical Charts
In order that this study need not place complete
reliance upon staff ratings to validate the predictions of
behavior made from the self-report questionnaire, an
57
additional step was included as a validation measure which
was quite independent of staff ratings.
This step consisted of comparing predictions of
implementing behavior with tabulations of nurses notes in
the clinical charts. It was predicted that those patients
whose questionnaires indicated they would be low in imple
menting behavior would have very few behavioral notations
of any kind in their clinical charts. These would be the
patients who accepted the hospital environment as they
found it without any unusual need to impose their own in
dividual requirements on others and for whom therefore it
would not be necessary to note unusual behavior in the
chart.
The value of this validation measure is demonstrated
by the fact that three of the four items used for prediction
of implementing behavior were significantly correlated with
extent of nurses notes, as reported in the chapter on
results.
Tabulation of nurses notes included all special re
quests or orders for medication which were not part of the
medication prescribed as part of the patient's routine
treatment. For example, daily charts of tranquilizing medi
cation were not included but a special request for aspirin
or an additional dosage of tranquilizer was included.
All special orders for medication were tallied
whether or not it was certain the order came at the patient's
58
instigation. Notes of polio innoculations were not included
nor were notations of routine cold medications for the four
or five day period which was standard ward treatment for
colds but was entered under special medication notation.
All behavioral notations were counted in the tabu
lation except for unquestionably passive observations such
as "usual quiet day," "quiet night," ' ‘ usual quiet behavior,"
"seems improved," or "never has anything to say." If there
was any suggestion of nonpassivity, however, the notation
was included.
Each notation tallied was given a value of one,
regardless of content, since a qualitative analysis was not
attempted. Thus a notation of "requesting aspirin" and a
notation "brought in by Security Officers in combative con
dition" each had the same value. A qualitative analysis
did not seem warranted since by and large these special
notations adhered to certain patients regardless of content.
When tabulation was completed a ratio of number of
nurses notes to days in the hospital was computed for each
patient. The number of days in the hospital was calculated
as the time from admission to the time of staff rating of
behavior. This ratio was then used to compare extent of
behavioral notes in the clinical chart with prediction of
implementing behavior.
Predictions of implementing behavior were obtained
from the self-report questionnaire by use of the criterion
keys. Since this was done in exactly the same manner as
were predictions of implementing behavior described in the
preceding sections, the predictions in both cases are, of
course, identical.
Step 4. Analysis of Data
1. Effectiveness of prediction of behavior as
determined by staff ratings of behavior was evaluated by
means of a Pearson Product-Moment Correlation Coefficient.
Correlations were obtained not only between each prediction
and its corresponding behavioral rating but also between
each single prediction and each of the twelve behavioral
ratings. This second step was taken in order to determine
the independence of items and also the effectiveness of
prediction not only on its matching rating item but on other
rating items as well.
Both ratings and predictions were transformed into
standard scores in order to reduce inter-rater variability
as much as possible. Rating scores were combined and aver
aged in order to provide a single score for each patient on
each item.
2. Effectiveness of predictions of implementing be
havior as determined by tabulation of nurses notes in the
clinical chart was evaluated by means of chi-square analysis
Ratio of nurses notes to length of hospitalization was con
trasted with implementing predictions for each of the four
60
implementing items alone. The four items were also combined
and averaged to perform an over-all chi-square analysis.
Three-fold tables were used in each case, with four degrees
of freedom.
3. Three factor analyses were made between inter
correlations to determine whether similar factor patterns
emerged to account for the behavioral dimensions in the
prediction scores alone, rating scores alone, and predic
tions and ratings combined. The Bi-MD 03 Program of the
Health Sciences Computing Facility at the University of
California was used for this purpose.^ This program in
volves an orthogonal rotation using a verimax criterion.
Although this tends to produce one major factor in the solu
tion this factor did not appear to be an artifact since this
large factor was also amply evident in each of the correla
tion matrices involved.
4. Comparison of individual criterion keys with ex
tent of nurses notes was made by means of chi-square analy
sis, using two by two contingency tables in each case, with
one degree of freedom.
5. Comparison of the criterion key concerned with
alcoholic excess was made with notations of excessive
^Much appreciation is due the Health Sciences Com
puting Facility at the University of California at Los
Angeles for their generous help with this part of the
research project, and to Dr. James 0. Palmer, for making
this service available as faculty sponsor.
alcoholic consumption in the clinical chart. These compar
isons were also by means of chi-square analysis, using a
two by two contingency table with one decree of freedom.
6. A three-way comparison was made between individ
uals who scored high in terms of predictions and ratings of
active, involved behavior as well as a high number of nurses
notes, and individuals who scored low on these three meas
ures. This was a qualitative rather than a quantitative
analysis although the results are presented in tabular form
in the chapter on results.
CHAPTER IV
RESULTS AND DISCUSSION
Results of this study indicate that it is possible
to take a series of psychiatric hospital admissions without
regard to diagnostic category and on the basis of patient
responses to a self-report questionnaire predict the behav
iors under study beyond chance expectancy.
These behaviors: dependency orientation, activity
level and the manner in which energies are channeled, and
implementing behavior, are observable and reliably ratable
by staff members with minimal training. These behaviors
are also reflected in the quantity of nurses notes in the
clinical charts.
The Hypotheses
The three hypotheses presented in the opening chap
ter may now be viewed in the light of results obtained.
Hypothesis A. There are enduring personality traits
expressed in overt behaviors which are not totally
obscured during episodes of psychiatric disturbance
The findings in this study support this hypothesis
although necessarily in limited fashion. The behaviors
were predictable over a time lapse. Although this time
lapse was in no case longer than three months, the fact
62
supported the belief that these were not transient behaviors
but enduring personality traits. That they are not obscured
during episodes of psychiatric disturbance is seen both in
the fact that prediction was possible without regard to
diagnosis and that behaviors predictable at time of admis
sion, during the presumably acute phase of the illness, were
apparent later in hospitalization when the acute phase
should have lessened.
Hypothesis B. Dependency orientation, activity level
and implementing behavior represent traits which are
measurable, predictable, and have important effects
upon the hospital course oi: the psychiatric patient
These behaviors were both measurable and predictable
and it was found that prediction of behavior could be made
from patient responses to the self-report questionnaire and
later observed by the staff and seen in the nurses notes.
These behaviors do not emerge as separate traits,
however, and appear highly interrelated. This important
interrelationship is considered in a separate section.
That these behaviors have important effects upon
the hospital course of the psychiatric patient is evident
in the significant relationship between these potential
behaviors as measured in responses to the questionnaire and
the extent of nurses notes in the clinical chart. Such
notes indicate that special attention, time and care must be
given to patients who score high on this behavioral complex.
Implications for hospital treatment of patients showing
these varying needs are discussed in a separate chapter.
Hypothesis C. It is possible to develop a self-report
questionnaire for newly admitted psychiatric patients which
can be objectively scored and used to predict overt behaviors
The questionnaire and its objective scoring system
proved successful in predicting overt behavior. Although
higher correlations than were obtained would be desirable
in using the instrument in a practical setting, the results
were nevertheless statistically significant and encouraging
from the standpoint of the difficulties encountered in
assessing complex human behavior in a natural setting where
so many variables can not be controlled.
Analysis of Data
Using the six steps listed under "Analysis of Data"
in the chapter on method, the following results may be
presented and discussed:
1. Correlation between Predictions of Behavior and Staff
Ratings of Behavior
A Pearsonian product-moment correlation coefficient
was computed between the twelve predictions of behavior
based on patient responses to the self-report questionnaire
and ratings of these twelve behaviors by staff members at
a later period.
Nine of the twelve correlations between each pre
diction and its corresponding behavioral rating were
65
significant at the .01 level or beyond and two were signif
icant at the .05 level. The nine highest correlations
ranged from .27 to .56 as indicated in Table 7.
In computing the correlations it was possible to
reduce individual rater variation to some extent through
the transformation of raw behavioral rating scores into
standard scores before pooling the ratings for each individ
ual subject.
Additional correlations were computed between each
of the twelve prediction item scores and all of the twelve
behavioral rating scores.
As can be seen in Table 8, seven of the prediction
items (Items 2, 3, 6, 8, 9, 10 and 12) could predict each
or any of the same behavioral items at the .01 level of
significance or beyond. Items 1 and 11 were nearly this
highly correlated as was Item 7 in a negative direction.
Thus all but two of the 12 items proved interrelated both
in terms of predictions and behavioral ratings.
The two unrelated items were Items 4 and 5. Item 4
proved of no value even in predicting its own matching be
havioral rating. Item 5 (an overt admission of dependency)
seemed independent of the other items and predicted its own
matching behavioral rating at the .01 level as can be seen
from Table 7.
Because of the close interrelationship between most
of the items, the factor analytic studies reported below
66
TABLE 7
CORRELATIONS BETWEEN EACH PREDICTION
OF BEHAVIOR ITEM AND ITS CORRESPONDING
ITEM FOR STAFF RATING OF BEHAVIOR
N = 98
Behavior Item Correlation* Significance
Pattern________Number________r______________ Level____
Implementing
Behavior 1 .36 .01
2 .44 .01
3 .56 .01
4 .12 ns**
Dependency
Orientation 5 .27 .01
6 .37 .01
7 .22 .05
8 .40 .01
Activity
Level 9 .47 .01
10 .53 .01
11 .23 .05
12 .45 .01
*Pearson product-moment coefficient of correlation
**Not Significant
STAFF RATINGS O F BEHAVIOR ITEMS
r
67
TABLE 8
CORRELATIONS* BETWEEN PREDICTIONS OF
BEHAVIOR AND STAFF RATINGS OF BEHAVIOR
(Items 4 and 5 have been excluded
because of lack of significant
correlations with other items)
N = 98
PREDICTIONS OF BEHAVIOR ITEMS
1 2 3 6 7 8 9 10 11 12
1 .36 .29 .32
** **
.28 .30 .30 .27
2 .46 .44 .44 .33 -.34 .42 .45 .44 .31 .41
3 .54 . 60 .56 .48 -.50 .59 .60 .58 .35 .55
6 .30 .33 .32 .37 -.32 .34 .38 .33 .31
7
** ** -** -**
8 .35 .43 .38 .35 -.32 .40 .43 .38 .34
9 .48 .50 .49 .38 -.35 .43 .47 .45 .35 .43
10 .51 .49 .52 .43 -.45 .52 .54 .53 .35 .50
11
**
.31
** **
12 .45 .53 .48 .40 -.37 .46 .44 .46 .31 .45
*Pearson product-moment coefficient of correlation
r was computed. All correlations which reached
the .01 level are given. Correlations are
positive unless preceded by a minus sign.
**Correlation reached the .05 level. Correlations
below the .05 level are omitted.
68
are particularly needed in order to present the structure
of the relationship in simpler form.
2o Correlations between Predictions of Implementing
Behavior and Extent of Nurses Notes
Successful results are also apparent in the chi-
square analysis of the relationship between prediction of
implementing behavior and extent of nurses notes. It must
be added, however, that the high correlations between pre
dictions of implementing, dependency orientation and activ
ity level indicate an over-all relationship rather than one
which adheres exclusively to implementing predictions.
Predictions of implementing behavior were obtained
from the self-report questionnaires and are therefore iden
tical to those used in the predictions of implementing which
were compared with staff ratings of such behavior. Raw
scores were converted to standard scores for each of the
four implementing items. These scores were averaged to com
pute the chi-square value for combined implementing scores.
Extent of nurses notes in the clinical chart was
determined by tabulating behavioral notations, computing
the length of hospital stay (as described on pages 57 and
58) and deriving a ratio of behavioral notes per day.
Interestingly enough a simple count of nurses notes in the
clinical record was sufficient to bring out the relation
ship without an intensive qualitative analysis of entries.
By and large patients who made special requests for care,
medication, or other services were also the patients who
manifested behaviors of many kinds sufficiently unusual to
warrant inclusion in the nurses notes.
Behavioral notes were given the following three
fold division in computing chi-square values:
Extent of Nurses Notes:
High....... .300 and above
Medium....... 100 to . 299
Low........ under . 100
Implementing behavior in standard scores was also
given a three-fold division:
Prediction of Implementing:
High....... +1.00 and above
Medium..... -1.00 to +.99
Low......... below -1.00
Four chi-square values were computed; one for each
of the four implementing items and extent of nurses notes.
One chi-square value was computed for the four implementing
items combined and extent of nurses notes. The results
of these five ohi-square analyses are given in Table 9.
From this table it can be seen that high scores on the firs t
three implementing items were significantly correlated with
a high number of nurses notes. The fourth item was of
doubtful significance, again indicating the questionable
value of this item in the study.
In spite of the weakness of Item 4, however, the
chi-square for the four items combined reached the .01 level
of significance as indicated in Table 9.
70
TABLE 9
RELATIONSHIP BETWEEN IMPLEMENTING
PREDICTIONS AND EXTENT
OF NURSES NOTES IN CLINICAL CHART
Implementing
Item
Chi-square
Value
Significance
Level*
1 11.30 .03
2 17.74 .01
3 15.84 .01
4 6.58 .20
Items 1, 2,
3, and 4
Combined
20.72 .01
degrees of freedom = 4
^Direction of significance in all cases was a
relationship between high implementing predictions
and high number of nurses notes.
71
3. Factor Analyses of Predictions and Staff Ratings
Three factor analytic studies were done to deter
mine the structure of prediction scores alone, staff
rating scores alone, and the structure of the two combined.
The program used was the performance of a principal compo
nent solution in an orthogonal rotation. Communalities
were estimated from a squared multiple correlation coef- '
ficient.
The structures of the factor matrices are given
in Tables 10, 11 and 12. An inspection of the factor
structure of prediction scores alone and staff rating
scores alone (Tables 10 and 11) reveals one consistent pri
mary factor with loadings on Items 1, 2, 3, 8, 9, 10 and 12.
This factor appears to represent active, needful involve
ment with the environment vs. quiet, non-involved acceptance
of environmental conditions. Although there are differences
in the factor structures in Tables 10 and 11, they appear
to share this first factor. A direct quotation of these
items involved may provide background for the interpreta
tion given:
Item 1. "Needs a lot of attention and reassurance."
Item 2. "Complains a lot; needs to be noticed."
Item 3. "Demanding, insistent, resistive to routine;
wants things his way."
Item 8. f , Makes demands on others but somehow their
attempts to help him meet with failure (rate
hospital behavior primarily)."
Item 9. "Has such a high energy level that it is a
real problem to hold himself in check."
72
Item 10. "Lets his impulses go freely (doesn’t hold
them in)."
Item 12. "Holds his impulses in for awhile and then
cuts loose."
In this combination of items there is a quality of
activity and needfulness if scored toward the high side and
certainly this type of patient would require and demand
different treatment in the hospital than would the patient
who was scored toward the low side on this set of items.
Most of the variance in the three studies appeared to
be accounted for by two factors. Suggestions of other
factors were present but loadings were small and results
were not clear-cut enough to do more than speculate.
The second factor appeared to represent a need to re
late positively to the environment vs. lack of this need.
The dimension of activity appeared to be absent in the
factor, however. Most of the variance appeared to be ac
counted for by Item 5, an overt admission of dependency;
Item 4, the need to please or be helpful; and Item 7, a pas
sive compliant attitude. A tendency toward negative corre
lation with Items 3 and 6 was also present in the factor.
Th£ factor seems to represent a cooperative attitude with
passive dependency vs. indifferent lack of cooperation.
73
TABLE 10
ROTATED FACTOR MATRIX
PREDICTIONS OF BEHAVIOR
Rotated Factors
Variable
I II III IV
1 -.90 -.23 .30 .03
2 -.81 .07 .37 -.13
3 -.70 .24 .53 .00
4 .17 -.37 -.30 .69
5 -.02
0^
00
•
1
-.09 .24
6 -.69 .42 .43 -.09
7 .40 -.26 -.79 .26
8 -.54 .14 .77 -.21
9 -.74 .01 .60 -.17
10 -.73
00
0
•
1
.64 -.10
11 -.44 -.39 .53 .45
12 -.53 .06 .82 -.08
Eigenvalues:
7.94
1.91
.44
.36
r
74
TABLE 11
ROTATED FACTOR MATRIX
STAFF RATING OF BEHAVIOR
Variable I
Rotated Factors
II III IV V
1 .69 -.21 .07 -.16
i
•
o
o
2 .78 .07 -.03 -.31 .37
3 .59 .31 .11 -.50 .36
4 -.02 -.51 -.04 .20 -.03
5 .12 -.50 .08 .01 -.20
6 .22 .29 .12 -.19 .59
7 -.13 -.38 -.16 .27 -.13
8 .46 .28 .17 -.31 .39
9 .40 .08 .22 -.53 .39
10 .38 .31 .02 -.71 .20
11
C M
O
•
.01 .49 -.07 .05
12 .27 .22 .36 -.59 .11
Eigenvalues
4.70
.96
.44
.29
75
TABLE 12
ROTATED FACTOR MATRIX
PREDICTIONS AND STAFF RATINGS COMBINED
Variables
Prediction
Values I II III
Rotated Factors
(Decimal Points Omitted)
IV V VI VII VIII IX
1 -77 31 -30 02 -11 37 -09 -02 -08
2 -77 34 00 -03 -18 30 09 06 -05
3 -80 30 17 04 -25 09 -03 01 -09
4 41 -07 -39 13 -07 03 -66 01 -04
5 10 -03 -91 14 04 03 -14 02 -02
6 -80 20 35 -10 -09 14 01 -06 -09
7 85 -18 -24 00 01 21 -16 03 -02
8 -90 26 10 -03 -07 -15 14 -00 -01
9 -92 28 -04 -10 -01 06 08 -06 -06
10 -94 25 -12 -04 -06 05 01 -01 -06
11 -59 17 -46 07 -18 -15 -42 04 -10
12 -92 22 01 -01 -13 -17 02 -01 -04
Rating
Values
1 -18 42 -04 28 -01 02 -07 06 -54
2 24 76 -08 04 10 -01 02 -17 -38
3 -38 82 -02 -19 -02 00 04 -02 -14
4 -03 -23 03 52 04 -05 -08 -16 -04
5 05 -05 -18 56 -08 04 02 18 -08
6 -20 53 11 -23 -08 02 04 -40 08
7 09 -31 -06 31 14 02 -08 -00 01
8 -18 74 04 -19 -12 -04 21 -04 -10
9 -28 73 01 00 -14 04 -09 -01 03
10 -37 69 00 -23 08 06 -12 18 -01
11 -16 07 00 01 -55 01 -03 -00 00
12 -30 57 05 -16 -30 08 -04 26 02
Eigenvalues:
10.51
2.30
2.19
.85
.60
.48
.40
.33
Some discussion of the fact that the three behavior
patterns under study were not independent is of interest.
Not only did these patterns appear related in predictions
and staff ratings but these same relational tendencies were
apparent in the extent of nurses notes in the clinical chart.
The active, needful, involved patient had indeed a large
number of nurses notes in his chart.
Some of the reason for this relationship unquestion
ably lies in item content. Some of the overlapping in con
tent has already been pointed out in previous sections.
Implementing behavior, for example, undoubtedly contains
strong elements of activity and dependency. It is quite
possible that a poorly developed inner control system ren
ders the "implementing" individual dependent upon arranging
and maneuvering outer controls because of weakness of his
own inner controls, yet fearful of these very controls be
cause of his vulnerability. Active, needful involvement
with the environment may represent defenses against this
deficient inner control system and a resulting inability
to accept environmental conditions with some flexibility.
Investigators in this and other cultures have
pointed out the relationship between difficulty with im
pulse control in adults and early childhood experiences
which presumably interfere with the development of an
adequate inner control system (Rexford,1963; Ritchie, 1961;
Iga, 1961; Bandura and Walters, 1958; and Ward, 1958).
The overlapping between activity/passivity and
field dependence/independence has been pointed out by
League and Jackson (1961) who feel that individuals who
are analytical and independent of the context in their per
ceptual performance can be characterized also as active as
reflected in measures of personality.
Speaking more broadly Duffy (1962) suggests that
general activity level affects performance in general, im
plying that the person with a high activity level will mani
fest his personality traits more strenuously than will the
person with a low activity level. Guilford (1964) has
stated that he feels the activity/passivity dimension may
act as a catalyst. When a high activity level exists, be
haviors will be more manifest than when passivity predomi
nates .
The complex interrelationship between the three
behavior patterns under study seems attributable both to the
measures used and to the fact that it may simply not be pos
sible to isolate some behavioral traits in "pure" or inde
pendent form.
4. Comparison of Criterion Keys with Extent of Nurses Notes
This comparison was made by chi-square analysis
using two by two contingency tables with one degree of free
dom. Presence or absence of each criterion key was contras
ted with a high ratio of nurses notes (.300 and above) and
78
a low ratio of nurses notes (below .300). This ratio was
obtained as described on page 57. It will be remembered
that criterion keys were considered present and scorable if
the patient answered a certain minimum number of items in
the given direction. The key was considered unscorable
and was omitted if this cut-off point was not reached.
As seen in Table 13 four of the criterion keys cor
related at a statistically significant level with extent of
nurses notes. Keys 6 (Unaccepting) and 7 (Dependent-Dis
satisfied) correlated with a high ratio of nurses notes
while Keys 1 (Quiet) and 10b (Low Energy) correlated with
a low ratio of nurses notes. More detailed description
of these criterion keys may be found in the appendix.
Again the direction of the relationship coincided
■ with that seen in the predictions and ratings of behavior.
5. Comparison of Criterion Key 2 (Quiet Patient, Problem
with Alcohol) with Notations in the Clinical Chart of:
Problems with Alcoholic Excess
In developing Criterion Key 2 labeled "Quiet Patient
Problem with Alcohol" the criterion group in the pilot study
consisted of men who seemed to be able to get along fairly
well so long as they could remain sober. These were the
men referred to on the ward as "Alkies" and whose primary
difficulty seemed to be an inability to stay away from al
cohol. This was in contrast to the patients whose difficul
ty with alcohol was accompanied by general anti-social or
79
TABLE 13
RELATIONSHIP BETWEEN CRITERION KEYS*
AND EXTENT OF NURSES NOTES
(Chi-squares were computed,"
using a two-fold table with
one degree of freedom)
No.
Criterion Key
Label
Chi-
Square
Significance
Level
Direction of
Relationship
1 Quiet Patient 4.69 .05 Few Nurses Notes
2 Quiet, Problem
with Alcohol
ns
3 Marginal 3.11 .10 Few Nurses Notes
4 Aggressive ns
5 Hostile ns
6 Unaccepting 6.48 .02 Many Nurses Notes
7 Dependent-
Dissatisfied 7.04 .01 Many Nurses Notes
8 Dependent-
Satisfied ns
9a
9b
Conformist
Non-conformist
ns
ns
10a
10b
High Energy
Low Energy 7.30
ns
.01 Few Nurses Notes
^Labels for criterion keys are included for general
description but do not necessarily reflect accurately
the behavior suggested and must not be considered
to do so.
**ns * not significant
80
difficult behavior in which alcohol seemed but one behavior
in a general repertoire of aggressive rebellion.
The men who were used to develop the criterion key
were not included in the study proper, of course, and no
attempt is made to classify the men in this study as alco
holics if they happened to have a high score on this cri
terion key. Nevertheless a significant number of the sub
jects in the study who scored on this scale had a history
of excessive use of alcohol in their clinical chart.
In searching the clinical charts for notations of
problems with alcohol, these notations were considered
indicative: arrests for drunkenness or drunk driving, an
affiliation with an Alcoholics Anonymous organization, a
history of delirium tremens, or actual notations of problems
with excessive alcoholic intake. Patients were omitted if
notations were not clear-cut.
A chi-square analysis of results was made using a
two-fold table in which presence or absence of a history of
difficulty with use of alcohol was contrasted with presence
or absence of Criterion Key 2 which was labeled, "Quiet,
Problem with Alcohol.1' The significance level reached .05
as seen in Table 14.
6. Three-way Comparison between Predictions. Staff Ratings
o£ Behavior and Extent of Nurses Notes
This three-way comparison of each subject seemed of
interest in order to evaluate the number of false positives
81
TABLE 14
COMPARISON OF CRITERION KEY 2
(QUIET, PROBLEM WITH ALCOHOL)
WITH NOTATIONS OF EXCESSIVE USE
OF ALCOHOL IN CLINICAL CHART*
Excessive Use
of Alcohol
Indicated in
Clinical Chart
No Indication
of Excessive
Use of Alcohol
in Clinical
Chart
Criterion Key
No. 2 Present
Criterion Key
No. 2 Absent
N=25 N=22
(fe = 20.1)**
(fe = 26.9)
N = 16
N-33
(fe - 20.9) (fe - 28.1)
Total N = 96
Chi-square Value = 3.88
Significance Level .05
Chi-square was computed using a two-fold table with one
degree of freedom.
fe = Frequency Expected if Chance Alone were Operating.
82
included in the study. To simplify the task, Item 3 was
used as representative of other items since its correlation
was the highest generally with all other prediction and
staff rating items.
Each individual was allocated in terms of: high,
medium or low prediction of behavior; high, medium or low
staff rating of behavior; and high, medium or low number of
nurses notes in the clinical chart. Values which formed the
basis for rating high, medium or low on these behaviors are
given in Table 15.
Again the directions were consistent as can be seen
from Table 15. The patient who was rated low in terms of
his need for active involvement with the environment was
also predicted to be low in this behavior and there was in
deed a scarcity of nurses notes in his clinical chart.
False positives were examined individually and
occasional explanations were found for the reversal of the
expected direction. One patient, for example, was predic
ted and rated high in terms of need for active involvement
with his hospital environment yet had almost no nurses
notes. During this particular period he was being extreme
ly careful in his ward behavior because of previous diffi
culties which had resulted in the placing of sanctions
against him. Notes from his previous hospitalizations
were plentifully filled with notations entered by nurses.
TABLE 15
THREE-WAY COMPARISON OF PREDICTIONS, *
STAFF RATINGS AND EXTENT OF NURSES NOTES
N - 96
Hi-Rate
Hi-Pred
Hi-Rate
Med-Pred
Med-Rate
Hi-Pred
Med-Rate
Med-Pred
Med-Rate
Lo-Pred
Lo-Rate
Med-Pred
Lo-Rate
Lo-Pred
Hi-Rate
Lo-Pred
Lo-Rate
Hi-Pred
High
Nurses
Notes
16 4 3 3 2 2 3 0 1
Medium
Nurses
Notes
0 1 0 2 4 1 2 1 0
Low
Nurses
Notes
4
1 5 4 13 6 14 1 3
Basis for Assigning Predictions
and Staff Ratings (Standard Scores)
High » 55 and above
Medium ■ 45-54
Low = below 45
Not all clinical charts were available for tabulation of nurses notes
and the number is therefore less than 98.
Basis for Assigning Extent
of Nurses Notes (Ratios)
High ■ *300 and above
Medium » .200 to .299
Low ■ below .200
00
00
84
Another false positive was a very quiet patient in
terms of predictions and staff ratings who nevertheless had
an unusually high number of nurses notes. This appeared
due to an unusual physical condition which required frequent
special medication. The notes of medication were not char
ted in the routine way and thus under the method used in the
present study were necessarily included in the tabulation of
nurses notes.
CHAPTER V
IMPLICATIONS FOR USE OF
THE QUESTIONNAIRE
What might this study mean to the psychiatric
hospital? How useful is the questionnaire in its present
form? How good a screening device is it? If it is of some
use in general screening, what suggestions may be given for
differential treatment of patients? How did patients react
to the questionnaire?
Since interest in the behaviors under study arose
from suicide research, does the questionnaire have useful
ness in predicting suicidal behavior?
These issues will be discussed as forthrightly as
possible. It is clearly recognized, however, that this
discussion is based primarily upon the results of a single
study of one particular type of population. As in any type
of research effort, caution must always be observed in
depending upon statements based upon unreplicated studies
and in applying results to other populations. At the same
time, results and implications of the results of a research
should be shared and discussed as frankly as possible.
From the results of the present study the question
naire appears useful as a general screening device for the
85
related behaviors under study. Use of the questionnaire
will be discussed first, followed by suggestions for general
hospital application and a discussion of its potential merit
in predicting suicidal behavior.
Use of the Questionnaire
as~~a General Screening bevice
The following estimate of predictive value of the
questionnaire may be made on the basis of present results.
Out of 150 consecutive admissions, 125 would remain
in the hospital long enough to take part in the psychologi
cal testing program. One hundred of these patients would
produce scorable questionnaires. Of this group of 100, 25
would be predicted to be active, needful, involved patients
and 20 would indeed prove to be so while 5 would not. Ap
proximately 50 patients would be predicted to be quiet,
accepting, routine patients and would prove to be so with
perhaps 6 exceptions. The remaining patients would fall in
an intermediate range both in terms of predicted and actual
behavior, with a general tendency for behavior to be on the
quiet side although there would be interest in treatment
provided by the hospital.
The following groups would then be expected among
patients taking the questionnaire as part of the admission
psychological test battery in a psychiatric hospital setting
with clerical scoring of the test bringing out these general
87
groupings for the convenience of professional staff members
charged with their immediate treatment:
1. 20% predicted to be active, needful, involved
patients.
2. 35% predicted to be quiet, accepting, relatively
uninvolved patients.
3. 25% moderately active and needful.
4. 20% unscorable either through changes, omissions or
lack of cooperation which would render the question
naire unusable.
Suggestions for
General Hospital Use
With the understanding at the outset that the fol
lowing speculations go beyond the data in the present study,
suggestions may be made for differential treatment of the
above four groups.
A. No change in routine hospital procedure seems
called for in the case of Groups 2 and 4. It seems likely
that these patients prefer the anonymity and distance pro
vided by routine treatment where they may as one nurses
aide aptly put it, "blend in with the woodwork and get by
unnoticed."
In a study of 225 psychiatric admissions by Giedt
(1961} approximately four different populations were iso
lated for which he felt different treatments should be
utilized. Although the Giedt populations do not coincide
88
entirely with the groups in the present study, there are
many similarities. Giedt describes a quiet, accepting,
uninvolved patient type accounting for nearly 50% of the
subjects in his study. His suggestions for treatment were
somewhat similar to those given here for Groups 2 and 4 in
the present study and would seem to center about milieu and
supportive therapy which forms a major part of current hos
pital treatment.
B. No special suggestions will be made for Group 3.
The patients in this group showed characteristics of both
the active and the uninvolved patients. Unless further
refinement of the screening device provides evidence to the
contrary, normal hospital procedures seem appropriate for
this group.
C. It is to the active, involved, needful patients
in Group 1 that we would address special attention. These
generalizations may be made about this group:
1. They demand, and like the demanding child, get spe
cial attention whether or not the staff wishes to
devote this much attention to them. Unless the
staff has the time or sensitivity to understand
this needfulness the attention may not necessarily
be therapeutic.
2. Their activity and need for involvement is a time-
consuming imposition upon staff time and is made
necessarily at the expense of time which could be
89
devoted to other patients when patients are inter
mingled regardless of individual variations in needs.
3. These needs seem an integral part of their illness
and should be considered as such in spite of the
irritating form these manifestations often take in
the Group 1 patient.
Very little research has been done concerning the
type of treatment most effective for this difficult type of
patient. The writer has in the past several years, however,
been involved in research concerning the identification and
description of this type of patient and presents this sug
gestion for treatment.
Place these patients in small special groups resem
bling a family unit. This would seem profitable from sev
eral standpoints. First, they would get the attention they
require— and manage to get anyway. Second, they will get
this individual attention early in the hospital course when
the therapeutic value may offset possible negative reaction
to routine hospital treatment. Third, although this would
necessitate an undue amount of staff time for the small
groups, this time would be extracted anyway by these demand
ing patients if placed in routine group settings. Placing
them in small groups should relieve stress on hospital
personnel and other patients which results from attempts to
force Group 1 patients into the general hospital routine.
90
This special treatment for Group 1 patients should
not be viewed as an attempt to pamper them but rather as a
recognition of a particular form of illness and an attempt
to relieve their discomfort and the stress they inflict
upon others about them.
There is a rationale for placing them into small
groups resembling small family units. It seems likely that
their active, needful, involved behavior reflects acute
dependency anxiety and conflict. They both demand attention
and nurturance yet fear control and submission. This prob
lem probably stems from an unresolved individuation conflict
arising in early childhood and perhaps within a small group
they may be able to work through some more satisfactory
resolution of the conflict.
In such a group with supportive staff members (pa
rental surrogates) they can both receive individual atten
tion and yet be relieved of their fear of complete control
and the need for submission. In addition the presence of
other patients with similar conflicts may help with their
recognition of the common problem which besets them.
Giedt (1961) has suggested the value of forming small ho
mogeneous groups in which meaningful social comparison and
group communication can occur.
Some experimental studies support the belief that
variations in dependency orientation have important bear
ings on patient response to therapy (Blyth, 1961) as well
91
as response to learning under conditions of reward or pun
ishment (Cairns, 1961; Goldman, 1961; and Cairns and Lewis,
1962). Leary (1957) in his interpersonal approach to per
sonality emphasizes the importance of differences in depend
ency orientations. He describes personality types in which
this is an integral part of behavior and plays an important
function not only in "pulling" responses from others but
also in the process of psychotherapy. He lends much weight
also to the activity/passivity dimension in his approach.
Whether isolation of the active, needful and in
volved patients into small groups resembling small family
units would result in a lessening of their difficult behav
ior would need to be determined experimentally. That it
would ease tension among staff members and other patients,
however, which would indirectly have therapeutic resultants
for the patients,seems quite probable.
Implications of the Questionnaire
Concerning Suicidal Behavior
Although the focus of this study was not upon
suicide, some discussion of the subject is indicated since
the behavioral patterns under study derived from research
on suicide and the question of a relationship between these
patterns and possible suicidal action is intriguing and
pervasive. One very pertinent question may be asked. Did
any suicides or suicide attempts occur during the period
when these patients were being studied?
92
No suicide attempts were known to have occurred
during this period. One patient did become seriously sui
cidal, depressed and agitated after admission. His behavior
became assaultive and he was noted as striking himself, ex
pressing suicidal preoccupation, guilt and fear of loss of
control. His questionnaire had been filled out before the
onset of this behavior and the predictions of behavior co
incided clearly with his later active, needful and involved
behavior. Staff ratings of behavior and nurses notes also
coincided but this would, of course, be expected since these
ratings and notations occurred after onset of the disturb
ance.
Two deaths occurred during the period of the study.
One patient with a convulsive disorder suddenly struck his
head against a wall, with death occurring within a few hours.
The death was not considered a suicide by hospital author
ities and was officially listed as "death from seizure."
A second patient in the study suddenly struck an
other patient severely on the head while both were engaged
in an outdoor activity. The patient who was struck died
from the injury and the death was considered a homicide.
Both the man who was listed as a death from a sei
zure and the homicidal patient were men who had been rated
by the staff as quiet, untroublesome patients from whom
outbursts would not be expected. Predictions of behavior
for both men were that they would prove quiet, uninvolved
93
patients. Nurses notes were sparse until the episodes in
volving these serious outbursts. Certainly, in the case of
these two patients the predictions from the self-report
questionnaire were ineffective for this type of behavior
and staff knowledge of patient behavior did not prepare
them for the outbursts.
What of behaviors related to suicide, such as de
pression and a history of suicidal attempts or threats? Did
they bear a relationship to the behavioral complex under
study?
Depression is commonly reported as an accompaniment
of suicide. In the present study 30 patients had a history
of some type of depression (some carried a diagnosis of
Depressive Reaction) but it was not possible to assess with
much confidence how serious these depressions were. No
differences were seen in the patients with and without a
history of depression and the patients in the study who
were or were not active, involved and needful.
A suicidal history consisting of reported threats,
attempts or ideation was present in the clinical charts of
35 of the 98 patients. Again no relationship was seen
between the presence or absence of such a history and the
presence or absence of active, needful behavior. This was
true in terms of predictions, staff ratings and extent of
nurses notes.
94
This apparent lack of relationship probably results
from two factors. First, not all suicides have a history
of attempts, threats or suicidal ideation although it is
present often enough that such indications must be taken
with great seriousness. Second, a common difficulty en
countered in assessing a "suicidal history" is the obscure
question of motivation. When a patient reports that he has
felt suicidal or that he has made a suicide attempt, this
report must be duly noted in the record. The question of
his actual intent is, of course, one which is not open to
observation. Even the report of a suicide attempt may not
be verified. An example of obviously questionable intent
in reporting suicidal behavior is seen in the fact that some
patients are aware that a threat of suicide may gain them
admission to a hospital where other efforts might fail.
The difficulties of classifying suicidal behavior,
various suggested classifications, and discussion of the
problems of methodology in conducting research on suicidal
behavior are abundantly presented in publications by Shneid-
man and Farberow (1957, 1961, 1962). These issues will not
be touched upon here except to comment briefly that even in
an apparently unequivocal suicide, motivation remains ob
scure. Attempts at reconstructing the factors which led to
this sad event must be based upon indirect evidence such as
reports by survivors or written records which are available.
Controlled collection of data is quite difficult when an
95
event is infrequent, unexpected, and from a population of
uncertain characteristics.
One promising approach in suicide research is the
study of behaviors, such as depression or suicidal threats
and attempts, which have been found to be present in many
suicide histories. This is the approach taken in this study
of active, needful, involved behavior which has also been
found to be present in the records of many suicides. Fol
low-up studies of the populations which do or do not have
these behaviors may then be undertaken to determine the
occurrence of future suicidal behavior. A follow-up study
of the population in the present study was not possible
because of time limitations but might well be undertaken as
an important future research effort.
Aside from the question of potential suicidal be
havior, additional study of treatment methods which might
prove effective in modifying the behavior patterns in this
study seems urgently needed. The difficult, demanding
patient is not only uncomfortable himself and in need of
assistance but creates difficulties for other patients and
for hospital staff members. As Leary (1957) has pointed
out in his interpersonal approach to personality, mutual
interaction (in this case between the active, needful
patient and the hospital staff members and other patients)
must be recognized and taken into consideration in treat
ment j>rocesses.
96
Implications for Patient
Use of Questionnaire
Since a major purpose of this study was to provide
a self-report form which would give newly admitted patients
the opportunity for immediate communication of their feel
ings and reactions about their new situation, some discus
sion of patient reaction is indicated.
Patient reaction to the questionnaire was varied as
is patient reaction to psychological testing in general.
By and large the reaction was positive. Many patients felt
a real need to communicate their feelings and reactions con
cerning the immediate concerns of their illness, treatment
and hospitalization. Some asked that their physicians be
given a copy of the questionnaire and added special com
ments which they felt would be helpful in their treatment.
Other patients used the form to vent anger and re
sentment or to defend various actions and behaviors. Some
answered in extremely cautious manner and expressed fear
that answers might be used against them. Others expressed
appreciation of particular items, stating, for example,
"That is certainly true for me," or, "That says just what I
would like to tell them."
On occasion patients from other buildings requested
permission to fill out the questionnaire and asked that
their physicians receive a copy. These requests were
97
honored as a matter of courtesy although the questionnaires
were not used in the study.
In summary, the questionnaire provided an oppor
tunity for communication which was welcomed by many patients
who wished to express their feelings about their new situ
ation. Even when the feelings were angry or hostile, the
need to communicate was strong and the opportunity seemed
therapeutic. Since the content of items directly concerned
immediate problems of illness and hospitalization, the form
appeared to contain validity for the patients.
CHAPTER VI
SUMMARY
The purpose of this study was to develop a self-
report questionnaire which could be used as a basis for
immediate assessment of individual differences among newly
admitted psychiatric patients and to predict certain overt
behaviors in the hospital which have importance both in
terms of patient needs and hospital management.
Three behavior patterns were predicted: dependency
orientation, activity level and the manner in which energies
are channeled, and implementing behavior or the apparent
need to control the environment as opposed to acceptance
of the environment as it is found.
An additional purpose was to develop the screening
instrument in a form which would provide the newly admitted
patient with an immediate opportunity for communicating his
feelings concern ing his illness and need for hospitaliza
tion and his reactions to his hospital course and treatment.
The study was undertaken in the Veterans* Adminis
tration Brentwood Psychiatric Hospital at Los Angeles,
California. The 98 male subjects consisted of all consecu
tive admissions to a selected psychiatric unit, irrespec
tive of diagnosis, for whom it was possible to obtain
psychological testing.
98
99
A pilot study was first undertaken in which the
self-report questionnaire, scoring criteria for predictions
of behavior, and a rating scale for staff rating of patient
behavior were developed. No subjects used in the pilot
study were included in the study proper.
The final self-report questionnaire was a 135 item
true-false form with each item concerned with direct ex
pression of the varying attitudes of patients toward their
illness and hospitalization and other immediate concerns.
The questionnaire was administered to the 98 patients in
the study as part of the regular admission psychological
test battery, with one examiner giving all the tests during
this period.
Predictions of behavior were made on the basis of
patient responses to the questionnaire. Ten criterion keys
developed during the pilot study were used to score the
questionnaires and provide the basis for prediction of
behavior. These predictions were entered upon a form which
was identical to the form used for staff ratings of these
behaviors.
The form which was used for both predictions and
staff ratings of behavior consisted of four statements or
items for each of the three behavioral patterns under study,
making a total of twelve items. Each item was rated on a
one to seven point scale indicating the extent to which
the behavior was predicted or was observed by staff members.
100
As soon as a patient had taken the self-report
questionnaire it was scored by means of the criterion keys
and the resulting predictions were entered on the form for
prediction of behavior. Staff members had no knowledge of
these predictions.
After the patient had been hospitalized for some
time staff members rated his behavior as they had observed
it on the ward,using a blank rating form. These ratings
were secured as soon as a patient was separated from the
psychiatric unit, regardless of the reason for separation.
If a patient had not been separated after a period of a
month or more, ratings were obtained while the patient was
still on the unit. Nurses and nurses aides performed the
ratings independently, with an average of 3.5 ratings for
e ach subj ec t.
When all predictions and staff ratings of behavior
had been obtained it was possible to determine how well
they coincided. In addition to this validation step, an
additional measure was undertaken which was quite independ
ent of staff ratings. This step was to compare predictions
with nurses notes in the clinical charts. Nurses notes
were simply counted in order to determine the extent of
notations present in each chart. Prediction of implementing
behavior could then be compared with extent of nurses notes
present in the clinical record.
101
Statistically significant correlations were found
between all but one of the twelve predictions of behavior
and its equivalent staff rating. These correlations ranged
between .25 and .56 which is encouragingly high for a study
such as this in which complex human behavior is assessed in
an uncontrolled environment. A statistically significant
relationship was also found between predictions of imple
menting behavior and extent of nurses notes.
The three behavior patterns under study: dependency
orientation, implementing behavior and activity level, did
not emerge as separate traits. Rather a consistent and per
vasive relationship was apparent, with very high correla
tions between the twelve predictions and high correlations
between most of the staff ratings. This relationship was
true of predictions, ratings, and extent of nurses notes.
A factor analytic study was undertaken to determine
the structure of this interrelationship. One major factor
emerged which was interpreted as an active, needful involve
ment with the environment vs. a quiet, accepting lack of
involvement. Approximately 20 per cent of the subjects be
longed clearly in the active group and approximately 40 per
cent in the quiet group. The intermediate group tended to
distribute somewhat toward the quiet side of the pattern.
The statistical results as well as the subjective
reaction of hospital patients and staff members indicated
102
the importance of the behavioral complex under study. The
three behavior patterns did not emerge as separate traits
but rather as related components of the complex which has
been labeled "active, needful and involved vs. quiet,
accepting and uninvolved behavior." Further investigation
of this behavioral complex seems indicated in view of the
special needs and problems which beset the active, demanding
patient who chafes under delays and controls and needs
special attention and consideration. This is in marked
contrast to the quiet patient who fits into the hospital
milieu with comparative ease.
On the basis of results from the present study the
self-report questionnaire appeared useful as a screening
device for determining the presence of the behavioral com
plex described. It also appeared useful as a means of pro
viding the newly admitted patient with an opportunity for
immediate communication of his feelings and reactions con
cerning his new situation.
Implications for hospital and patient use of the
questionnaire were discussed in the preceding chapter. In
addition suicidal implications were touched upon since
interest in the behaviors under study stemmed from suicide
research which indicated their importance.
The possibility that this behavioral complex is im
portant in evaluating suicidal potential suggests that a
follow-up study of the patients in this study should prove
103
of considerable Interest and might well be undertaken as a
future research project.
In research, studies must be replicated before the
results reported may be accepted with confidence. Many
promising and encouraging studies such as the present one
have not held up when repeated with other populations or at
other times. Further study of the self-report questionnaire
is necessary not only to confirm the present findings but to
refine the content and scoring system should the question
naire continue to prove useful as a general screening
device.
REFERENCES
REFERENCES
Adler, A. Understanding human nature. New York:
G reenberg, 1927.
Allport, G. W. Personality: a psychological inter
pretation. New York: Holt, 1937.
Angyal, A. Foundations for a science of personality.
New York” Harvard Univ. Press, 1941.
Ash, P. The reliability of psychiatric diagnoses.
J. abnorm. soc. Psychol., 1949, 44, 272-277.
'Bandura, A. & Walters, R. H. Dependency conflicts in
aggressive delinquents. J. soc. Issues, 1958, 14,
52-65.------------------- -------------- —
Bastendorf, W. L. Activation level as measured by
palmar conductance and intelligence in children.
Dissert. Abstr., 1961, 21 (10), 3156.
Berg, I. A. & Bass, B. M. Conformity and deviation.
New York: Harper, 1961.
Birkman, R. W. Development of a personality test using
social and self-perception inventories. Dissert.
Abstr., 1961, 22 (6), 2061-2062.
Blyth, D. D. Dependence, independence and other factors
related to veterans reaction to an offer of psycho
therapy. Dissert. Abstr. , 1961, _20 (7-8), 2899.
Bronson, G. Critical periods in human development.
Brit. J. med. Psychol., 1962, 35 (2), 127-133.
Cairns, R. B. The influence of dependency inhibition on
the effectiveness of social reinforcement. J. Pers.,
1961, 29, 466-488.
Cairns, R. B. & Lewis, M. Dependency and the reinforce
ment value of a verbal stimulus. J. consult. PsychoL
1962, 26, 1-8.
Caruth, Elaine G. The relationship of dependency to
verbal learning without awareness. Unpublished doc
toral dissertation. Univ. of Southern Calif., 1959.
105
106
Cattell, R. B. The principal replicated factors discov
ered in objective personality tests. J. abnorm.
soc. Psychol., 50, 291-314.
Cohen, S., Leonard, Calista V., Farberow, N. L. &
Shneidman, E. S. Tranquilizers and suicide in the
schizophrenic patient. Accepted for publication,
Arch, gen. Psychiat., 1964.
Comrey, A. L. A study of 35 personality dimensions.
Educ. psychol. Measmt., 1962, 22 (3), 543-522.
Couch, A. 6 c Keniston, K. Agreeing response set and
social desirability. J. abnorm. soc. Psychol.,
1961, 62, 175-179. ----
Cromwell, R. L., Palk, B. E. 6c Foshee, J. G. Studies
in activity level: The relationships among eyelid
conditioning, intelligence, activity level and age.
Amer. J. ment. Defic., 1961, _65, 744-748.
Cronbach, L. J. Essentials of psychological testing.
(2d ed.) New York: Harper, i960.
Crumpton, Evelyn 6c Wine, D. B. Hie veteran NP patient
ten years ago and now. Psychiat. Quart., 1962, 36,
301-311. —
Dahlstrom, W. G. 6c Welsh, G. S. An MMPI Handbook.
Minnesota: Univ. of Minn. Press, I960.
DeSoto, C. B. 6 c Kuethe, J. L. Ihe set to claim undesir*
able symptoms in personality inventories. J. con-
sult. Psychol., 1959, 23 (6), 496-500.
Diamond, S. Personality and temperament. New York:
Harper, 1957.
Downing, J. Chronic mental hospital dependency as a
character defense. Psychiat. Quart., 1959, 32,
489-499.
Duffy, Elizabeth. Activation and behavior. New York:
Wiley 6c Sons, 1962.
Durkheim, E. Suicide: A study in sociology. Tr. by
Spaulding, J. A. 6e Simpson, G. A. Illinois:
Glencoe Free Press, 1951.
107
Edwards, A. L. The relationship between the judged
desirability of a trait and the probability that
the trait will be endorsed. J. appl. Psychol.,
1953, 37, 90-93. ----------------
Edwards, A. L. & Diers, C. J. Social desirability and
the factorial interpretation of the MMPI. Educ.
psychol. Measmt., 1962, 22, 501-509.
Elliott, R. Interrelationships among measures of field
dependence, ability and personality traits.
J. abnorm. soc. Psychol., 1961, 63, 27-36.
Ellis, A. The validity of personality questionnaires.
Psychol. Bull., 1946, 43, 385-440.
Erikson, E. H. Childhood and society. New York:
W. W. Norton, 1950.
Eysenck, H. J. Experiments in personality. London:Paul,
1960.
Farberow, N. L. & Shneidman, E. S. The cry for help.
New York: McGraw-Hill, 1961.
Farberow, N. L., Shneidman, E. S. & Leonard, Calista V.
Suicide among general medical and surgical hospital
patients with malignant neoplasms. Department of
Medicine and Surgery, Medical Bulletin No. 9.
Veterans Aministration, Washington, D. C., 1963.
Flanders, N., Anderson, J. P. & Amidon, E. J. Measuring
dependence proneness in the classroom. Educ.
psychol. Measmt.. 1961, 21, 575-587.
Foulds, G. A. The relative stability of personality
measures compared with diagnostic measures.
J. ment. Sci., 1959, 105, 783.
Foulds, G. A. Personality traits and neurotic symptoms
and signs. Brit. J. med. Psychol., 1961, 34, 263-
270.
Fourdyce, W. E. & Lamphere, A. V. The control of social
desirability in a structured Q-sort of dependency-
independence. Educ. psychol. Measmt., 1960, 20, 1.
Frosch, J. & Wortis, S. B. A contribution to the nosol
ogy of the impulse disorders. A.M.A. J. Psychiat.,
1954, 111, 132-138.
108
Giedt, F. H. Psychological characteristics of psychiat
ric admissions and implications for treatment.
J. nerv. ment. Pis., 1961, 132, 324-334.
Gluck, M. R. Rorschach content and hostile behavior.
J. consult. Psychol.. 1955, JL9, 475-478.
Goffman, E. Asylums. Chicago: Aldine, 1961.
Goldman, A. R. The effects of dependency and dependency
anxiety on schizophrenic1s rate of learning under
conditions of social reward and punishment.
Dissert. Abstr., 1961, 22 (6), 2063.
Goldman-Eisler, F. The problem of "orality" and its
origin in early childhood. J. ment. Sci. , 1951,
97, 765-782.
Gordon, B. An experimental study of dependence-independ
ence in a social and laboratory setting. Unpublished
doctoral dissertation. Univ. of Southern Calif.,
1953.
Gough, H. G. Manual, California Psychological Inventory.
Consulting Psychologists Press, 1957.
Gough, H. G. , McClosky, H. & Meehl, P. E. A personality
scale for dominance. J. abnorm. soc. Psychol.,
1951, 46, 360-366.
Grayson, H. M. Psychological admissions testing program
and manual. Los Angeles U. S. Veterans Administra
tion Neuropsychiatric Hospital, 1950.
Guilford, J. P. Psychometric methods. New York:
McGraw-Hill, 1954.
Guilford, J. P., Christensen, P. R., & Bond, N. A. The
DF Opinion Survey. Manual of instructions and inter
pretations. California: Sheridan Supply Co., 1954.
Guilford, J. P. Fundamental statistics in psychology
and education! New York: Mc-Graw Hill, 1^56.
Guilford, J. P. Personality. New York: Mc-Graw Hill,
1959.
Guilford, J. P., Christensen, P. R., Taaffee, G. &
Wilson, R. C. Ratings should be scrutinized. Educ.
psychol. Measmt. , 1962, J22, 439-447.
109
Heider, F. The psychology of interpersonal relations.
New York: Wiley & Sons, 1958.
Hendrik, I. The discussion of the "instinct to master."
Psychoanal. Quart.. 1943, 12, 311-329.
Hillson, J. S. & Worchel, P. Self concept and defensive
behavior in the maladiusted. J. consult. Psychol.,
1957, 21, 83-88.
Iga, M. Cultural factors in suicide of Japanese youth
with focus on personality. Sociol. soc. Res.,
1961, 46, No. 1.
Jackson, D. N. 6c Messick, S. Content and style in
personality assessment. Psychol. Bull., 1958, 55,
243-252.
Jenkins, T. N. Measurement of the primary factors of
the total personality. J. Psychol., 1962, 54,
417-442.
Kagan, J. 6c Mussen, P. H. Dependency themes on the TAT
and group conformity. J. consult. Psychol., 1956,
20, 29-32.
Kagan, J. 6c Moss, H. A. The stability of passive and
dependent behavior from childhood through adulthood.
Child Develpm., 1960, 31, 577-591.
Kantor, R. E., Wollner, J. M. 6c Winder, C. L. Process
and reactive schizophrenia. J. consult. Psychol.,
1953, 17> 157-162.
Kelly, E. L. Consistency of adult personality. Amer.
Psychologist, 1955, 10, 659-681.
Kenny, D. T. The influence of social desirability on
discrepancy measures between real self and ideal
self. J. consult. Psychol., 1956, ^0, 315-319.
King, G. F. Research with neuropsychiatric samples.
J. Psychol., 1954, 38, 383-387.
Klett, C. J. The social desirability stereotype in a
hospital population. J. consult. Psychol., 1957.
21, 419-421.
110
Lasky, J. J., Hover, G. L., Smith, P. A., Bostian, D. W.,
Duffendock, S. C. & Nord, C. L. Post-hospital ad
justment as predicted by psychiatric patients and by
their staff. J. consult. Psychol., 1959, 23, 213-
218.
League, Betty J. & Jackson, D. N. Activity and passivity
as correlates of field-independence. Percept, mot.
Skills, 1961, 12, 291-298.
Leary, T. Interpersonal diagnosis of personality.
New Yorlc: Ronald Press, 1957.
Lorr, M. & Rubinstein, E. A. Factors descriptive of
psychiatric out-patients. J. abnorm. soc. Psychol.,
1955, 51, 514-522.
Lorr, M. , O'Connor, J. P., & Stafford, J. W. Confirma
tion of nine psychotic symptom patterns. J. clin.
Psychol., 1957, 252-257.
May, P. R., Garrett, F. H., Nash, L. R. & Jordan, B.
A one-hour admissions procedure. Ment. Hosp., 1962,
13, 462-465.
McCord, W., McCord, J., & Verden, P. Familial and be
havioral correlates of dependency in male children.
Child Develpm., 1962, 33 (2), 313-326.
Meehl, P. E. The dynamics of "structured" personality
tests. In Contemporary research in personality.
Ed. by Sarason, I. (L New York: Van Nostrand, 1962.
Mehlman, B. The reliability of psychiatric diagnosis.
J. abnorm. soc. Psychol., 1952, 47, 577-578.
Munt, Janet S. Fear of dependency. Soc. Wk., 1960, 5,
27-32. “
Murphy, Lois B. Pride and its relation to narcissism,
autonomy, and identity. Bull. Menninger Clin.,
1960, 24, 136-143.
Navran, L. A rationally derived MMPI scale to measure
dependence. J. consult. Psychol., 1954, _18, 192.
Numberger, J. I, Diagnostic signs and symptoms of
emotional disorders. Psychol. Abstr., 1963. 37, 129.
Ill
Owens, W. A., Jr. Age and mental abilities: A longi
tudinal study. Genet. Psychol. Monogr., 1953, 48,
3—54.
Parsons, T. Social structure and personality develop
ment. In Studying personality cross culturally.
Ed. by Kaplan, B. Illinois: 1961, Row, Peterson.
Phillips, R. H. Environmental mastery and personality
development. Arch, gen. Psychiat.. 1961, Jj, 146-150.
Pine, F., & Levinson, D. J. A sociopsychological con
ception of patienthood. Int. J. soc. Psychiat.,
1961, 2, 106-122.
Rexford, E. N. A developmental concept of the problems
of acting out. J. Child Psychiat., 1963, 2, No. 1.
Ritchie, J. Childhood in Rakau: The first five years
of life. In Studying personality cross culturally.
Ed. by Kaplan, B. Illinois: Row, Peterson, 1961.
Rosen, E. Self-appraisal, personal desirability, and
perceived social desirability of personality traits.
J. abnorm. soc. Psychol., 1956, 52, 151-159.
Sandler, J. Studies of psychopathology using a self-
assessment inventory. Brit. J. med. Psychol., 1954,
27, 142-145.
Sanford, N., Webster, H., & Freedman, M. Impulse expres
sion as a variable of personality. Psychol. Monogr.,
1957, No. 440, 71.
Secord, P. F., & Beckman, C. W. Personality theory and
the problem of stability change in individual be
havior: An interpersonal approach. Psychol. Rev.,
1961, 68 (1), 21-32.
Semeonoff, B. Self-description as an instrument in per
sonality assessment. Brit. J. med. Psychol., 1962,
35, (2), 165-175.
Shneidman, E. S., & Farberow, N. L. Clues to suicide.
New York: McGraw-Hill, 1957.
Shneidman, E. S., & Farberow, N. L. Suicide--the prob
lem and its magnitude. Department of Medicine and
Surgery. Medical Bulletin No, 7. Veterans Administra
tion, Washington, D. C., 1961.
112
Shneidman, E. S. , Farberow, N. L. , & Leonard, Calista V.
Suicide— evaluation and treatment of suicidal risk
among schizophrenic patients in psychiatric hospitals
Department of Medicine and Surgery, Medical Bulletin
No. 8. Veterans Administration, Washington. D. C.,
1*62.
Shneidman, E. S., Farberow, N. L. & Leonard, Calista V.
Facts about suicide. U. S. Bureau of Public Health
Publication, No. 872. Bethesda, Maryland, 1961.
Sivadon, P. D. Techniques of sociotherapy. Psychiatry,
1957, 20, 205-210.
Smith, J. R. & Coleman, J. C. The relationship between
manifestations of hostility in projective tests and
overt behavior. J. proj. Tech., 1956, 20, 326-334.
Smith, Madorah E. A comparison of certain personality
traits as rated in the same individuals in childhood
and fifty years later. Child Develpm. , 1952, 23,
161-180.
Strong, E. K., Jr. Permanence of interest scores over
22 years. J. appl. Psychol., 1951, _35, 85-91.
Stilson, D. W., Mason, D. J., Gynther, M. D. 6e Gertz, B.
Comparability and reliabilities of behavior rating
scales. J. consult. Psychol., 1958, 22!, 213-216.
Thurstone, L. L. Primary mental abilities. Psychometr.
Monogr., 1938, No. 1.
Twain, D. C. Factor analysis of a particular aspect of
behavioral control: Impulsivity. J. clin. Psychol.,
1957, 13, 133-136.
Vernier, C. M., Whiting, J. F. 6c Meltzer, M. L. Dif
ferential prediction of a specific behavior from
three proiective techniques. J. consult. Psychol.,
1955, 19, 175-182.
Veroff, J. Development and validation of a projective
measure of power motivation. J. abnorm. soc.
Psychol., 1957, 5&, 1-8.
Wahler, H. J. Response styles in clinical and non-
clinical groups. J. consult. Psychol., 1961, 25,
533-539. “
113
Ward, J. L. Homosexuality and the instutionalized
delinquent. Psychiat. Quart. Suppl., 1958, £2, 301-
Wam, L. J. A comparative investigation of dependency
in epilepsy, paraplegia, and tuberculosis. Unpub
lished doctoral dissertation, Univ. of California
at Los Angeles, 1958.
Weiss, P., & Emmerich, W. Dependency fantasy and group
conformity in ulcer patients. J. consult. Psychol.,
1962, 26, 61-64.
White, R. W. Motivation reconsidered: The concept of
competence. Psychol. Rev., 1959, 66, 297-333.
Winder, C. L., Ahmad, F. Z., Bandura, A., & Raul, C.
Dependency of patients, psychotherapists responses
and aspects of psychotherapy. J. consult. Psychol.,
1962, 26, 129-134.
Wirkin, H. A. , Karp, S. A., & Goodenough, D. R. Depen
dence in alcoholics. Ouart. J. Stud. Alcohol,
1959, 20, 493-504.
Wood, E. C., Rakusin, J. M. & Morse, E. Interpersonal
aspects of psychiatric hospitalization. Arch, gen.
Psychiat. , 1962, j>, 39-46.
APPENDIX
SELF-REPORT QUESTIONNAIRE
Name
Date_
Ward”
LETTING YOU KNOW HCW I FEEL
(Read each statement and decide whether it is true as
applied to you or false as applied to you. Circle T if it
is true; circle F it it is false. It is important to mark
every statement you can so that we can better understand
you. If it is absolutely impossible to decide how to mark
a statement, however, simply leave it out and go on to the
next one).
T F 1. I'm here at the hospital because everyone feels
it's a good place to get myself straightened out.
T F 2. I came here because I was so tense and nervous
at home I couldn't handle myself.
T F 3. I'm not just sure why I'm here.
T F 4. I'm here because I was forced to come here.
T F 5. Ihere's nothing wrong with me that I can't
handle myself.
T F 6. I know what's wrong with me but don't know what
to do about it.
T F 7. I'm here to find out what's bothering me. If I
knew what it was I wouldn't be here.
T F 8. I'm here because they say I need to be but I'm
not sure just what's wrong.
T F 9. I insist on having a voice in what happens to me
here. That's my right.
T F 10. The hospital knows more about what I need than I
do. It s beyond me.
T F 11. I have a pretty good idea what I need and would
like to talk to the doctor as soon as possible.
115
T F 12.
T F 13.
T F 14.
T F 15.
T F 16.
T F 17.
T F 18.
T F 19.
T F 20.
T F 21.
T F 22.
T F 23.
T F 24.
T F 25.
T F 26.
T F 27.
T F 28.
116
I am anxious to cooperate with whatever kind of
treatment they give me and am grateful for
getting it.
I don't hesitate to ask for what I need.
You're better off to take what comes along than
to keep asking for things.
I can get what I need but I have to work very
hard at it because other people are so self-cen
tered as a rule.
I can usually get what I need by cooperating
with other people--then they cooperate with you.
I'm satisfied with the way I've been treated hesa
As far as I'm concerned it is a waste of time
for me to fill this form out.
I can only wait just so long for something then
I do something about it no matter what anybody
says.
I let other people speak to me first— then I
answer.
When other people decide what you're supposed to
do you might as well go along with it.
I stay out of other people's way and never cause
any trouble.
When I don't like the way things are done I say
something about it even if it gets me in trouble.
Somehow I always seem to get involved with other
people one way or another.
It is easy for me to make decisions.
If you act without thinking it just gets you
into more trouble so I just forget it and go my
own way.
Usually I cool off before I do anything drastic.
There are times when I just say "The hell with
everything" and cut loose.
T F 29.
T F 30.
T F 31.
T F 32.
T F 33.
T F 34.
T F 35.
T F 36.
T F 37.
T F 38.
T F 39.
T F 40.
T F 41.
T F 42.
T F 43.
X f 44.
T F 45.
117
When I get upset all I want Is something to take
so I won't have to think about anything at all.
When I get upset I do something about it even if
it isn't the right thing.
When I get upset I get help from someone else.
I very seldom get upset but when I do I can
handle things myself.
I would be all right if I just had some way to
blot eve^thing out when I feel very badly.
Although I try not to be a bother I seem to need
a tremendous amount of attention and reassurance
from others.
My home life is in bad shape. I've done every
thing to please my wife but she's still not
satisfied.
I've tried marriage but it has never worked out
for me.
I've never been married.
My wife never gives me any trouble; she does
what's expected of her.
All my life I've had trouble getting any coopera
tion from other people.
I've wanted a lot more from life than I've gotten
but I'm used to it by now.
I know what I want but most of the time I have to
accept what comes my way.
All my life I’ve worked very hard trying to help
others and make them happy.
I'm used to having my own way and am not satis
fied unless I can set the terms in my life.
I can't seem to accept the troubles that come my
way like other people can.
I have always started out by working very hard at
something but then I would just lose interest
completely in it.
T F 46.
T F 47.
T F 48.
T F 49.
T F 50.
T F 51.
T F 52.
T F 53.
T F 54.
T F 55.
T F 56.
T F 57.
T F 58.
T F 59.
T F 60.
T F 61.
T F 62.
T F 63.
T F 64.
118
I'm restless and uneasy unless I'm working hard.
When I was little my folks let me have my own way
most of the time.
When I was little I was anxious to please every
body so they would like me.
I found out when I was a child that it was best to
say nothing and go my own way.
My folks didn't pay much attention to me when I
was small.
I hardly had any folks; I raised myself for the
most part.
My folks always had very high ambitions for me.
I always thought that if I could be a success some
day then my folks would really appreciate me.
I was pampered as a child.
Things were pretty tough for me when I was little.
I had about the same childhood as most kids, I
guess.
I was very close to my mother when I was a child.
I wanted to be exactly like my Dad when I was
little.
I was so independent when I was little that all I
wanted was just to be my own boss some day.
I could always get my folks to do what I wanted;
they couldn't do much with me.
I don't remember much about when I was a child.
Somehow I never seem to manage things as well as
other people.
I manage my own affairs in the way I like.
I can manage to get other people to do things
right if I work at it long enough.
T F 65.
T F 66.
T F 67.
T F 68.
T F 69.
T F 70.
T F 71.
T F 72.
T F 73.
T F 74.
T F 75.
T F 76.
T F 77.
T F 78.
T F 79.
T F 80.
T F 81.
T F 82.
119
There is no point in trying to manage things; you
might as well take what comes along.
I don't take orders from other people.
Orders are fine but there should always be excep
tions for people who are exceptional.
If you don't follow orders you're in for trouble.
I always follow orders because that makes things
run smoother for everybody.
All my life I've done just what other people
wanted but it isn't working anymore.
I've never asked for help from anybody.
I know very well that I annoy people around me but
I can't seem to help myself.
Other people could help me if they would only try
to understand me.
All my life I've been trying to help other people
--now I need help myself.
People need help from others now and then and I'm
no exception.
I am unable to control myself and need someone to
help me.
If I could just get control of myself I'd be all
right.
Somehow I seem to be restless and dissatisfied all
the time.
When things get too much for me I start drinking.
When things get too much for me I feel like I'm
going to blow my stack.
I'm so tense and nervous at home I can't even
sleep at night.
I am quieter here in the hospital with something
to make me sleep.
T F 83.
T F 84.
T F 85.
T F 86.
T F 87.
T F 88.
T F 89.
T F 90.
T F 91.
T F 92.
T F 93.
T F 94.
T F 95.
T F 96.
T F 97.
T F 98.
T F 99.
T F 100.
120
I'm not upset. All I want is my freedom--not
medication.
I'm tense and nervous at home and here too. I've
always been that way and I seem to be worse now.
I've tried alcohol when I get tense and nervous
but it doesn't help much.
I never drink.
When I get tense and nervous drinking relaxes me
although I know it's no solution to my troubles.
I drink because I like it, not because I'm tense
and nervous.
Now that I've stopped drinking I know everything
will work out.
If I can't make a go of it myself I don't want
any interference from anybody.
No matter what I do I can't get anybody to really
like me.
The one thing I can't live without is my wife.
The one thing I can't stand is other people
bothering me.
I'd give up hope if my wife left me.
I'd give up hope if I didn't have the hospital to
turn to.
If I had no friends to help me I would give up
hope.
There are some things I simply will not accept in
life.
If I could only control myself I'd be o.k.
If other people would only cooperate I'd get
along all right.
I try to think that I will get well but sometimes
I feel very depressed and discouraged.
121
T F 101.
T F 102.
T F 103.
T F 104.
T F 105.
T F 106.
T F 107.
T F 108.
T F 109.
T F 110.
T F 111.
T F 112.
T F 113.
T F 114.
When I get tense, walking up and down helps some.
I wouldn't feel so tense if I were allowed to
leave here.
When I get tense I can't sit quietly, I've got
to do something.
There have been times in the past when I've
acted pretty badly but I've got myself under
control now.
Although I try not to be, I'm very tense as soon
as I leave the hospital and I would appreciate
having a supply of medication to take with me.
One of the reasons why I feel better in the hos
pital is that I can control my impulses better
here.
I have always been impulsive; it's just my
nature.
I was never impulsive until recently; lately I
can't seem to handle myself.
The medication I'm getting doesn't seem to help
much. I feel badly most of the time.
I don't need help from the hospital because there
really isn't anything wrong with me.
This hospital is o.k. I suppose. I don't pay
much attention to hospitals.
The hospital is a real help to me when I am
tense and nervous.
There are some people on the hospital staff who
are especially understanding and sympathetic
with people who are ill.
Some of the staff members here think they own
the place. This hospital is provided for treat
ment of patients and when a patient asks for
something they should at least listen to him.
Everybody likes to be noticed by other people and
I really appreciate it when a staff member takes
an interest in me.
T F 115.
T F 116.
T F 117.
T F 118.
T F 119.
T F 120.
T F 121.
T F 122.
T F 123.
T F 124.
T F 125.
T F 126.
T F 127.
T F 128.
T F 129.
122
I'm much sicker than the doctors and nurses
realize.
The hospital staff really doesn't seem to under
stand what kind of medication I need. I need a
different kind of treatment from what I've been
getting.
Although I take my medication I would rather not
have to take it because I really don't need it.
I felt much better before I began taking my
medication. It seems to make me worse.
I don't usually say anything even if I don't feel
very good because they usually give you some
medication anyway.
If I had my way I wouldn't take any medication.
I usually feel better when I'm under medication.
I know I should be able to get along well away
from the hospital but I still am afraid some
times that I won't be able to handle my feelings.
I'm not well enough to leave the hospital but I
get so sick of it here that I want to leave
anyway.
I have never wanted to elope from the hospital.
It has been a good place to stay while I have
been getting myself straightened around.
If a person wants to leave the hospital he cer
tainly has a right to go.
If the hospital doesn't give you the treatment
you want you're better off to try to get it
someplace else.
I'm too sick to go home and I can't get the kind
of treatment I need here. I don't know what to
do.
All I ask out of life is a chance to hold a good
job and take care of my family.
All I want out of life is to regain my health;
then I can go from there.
T F 130.
T F 131.
T F 132.
T F 133.
T F 134.
T F 135.
123
All I want out of life is for my wife to love me
again.
All I want out of life is happiness.
When I leave here I'm going to go home and I'm
sure I can make things work out better this
time.
It's all right with me if I leave the hospital.
I can get along o.k.
Although it is true that I really feel better in
the hospital I still intend to do the best I can
at home.
If I am released from here I will know how to
handle things better this time and can get along
fine.
124
DESCRIPTION OF CRITERION KEYS
Key
No. Label
1 Quiet
2 Quiet, Problem
with Alcohol
3 Marginal
4 Aggressive
5 Hostile
6 Unaccepting
7 Dependent-
Dissatisfied
8 Dependent-
Satisfied
Brief description of characteristics of
criterion group on which key was based.
9a Conforming
9b Non-Conforming
10a High Energy
10b Low Energy
Quiet, unobtrusive, somewhat social.
Alcohol the major problem in a quiet
otherwise untroublesome patient.
Quiet, isolated, schizoid.
Rebellious, active, difficult.
Overtly hostile.
Rejecting of hospitalization. Angry.
Dependent, dissatisfied, demanding,
complaining: a nuisance.
Dependent, anxious to get along, posi
tive attitude toward hospital.
Aware of appropriate social reactions.
Unaware or rejecting of social norms.
High activity level.
Low activity level.
It should be emphasized that the labels and des
criptions above are merely approximations of complex
behaviors and complete reliance on these descriptions is
not yet warranted.
Single keys were not used in predicting staff rat
ings of behavior. Combinations of keys were used. Some
single keys, did, however seem promising in predicting
extent of nurses notes as reported in the study.
On the following pages each criterion key is given
in complete detail. The items and direction for responses
are set forth together with the critical point which must
be reached before the key is scorable. When the key is
scorable values may be assigned to the behavior to be ex
pected on the ward in terms of the behavioral patterns
under study. These values are also given. Complete instruc
tions for the scoring process is given following presenta
tion of the criterion keys.
125
CRITERION KEYS
Criterion Key No. 1 (Quiet Patient)
Critical Point: 8 and above 20 Items Cri
F 6 F28 F78
T 8
F49 F97
F 9 T58 F101
Fll T65 F103
T15 F73 F114
F24 F77 T125
F81 F126
Values to be assigned when key is scorable:
Item: Value: Item: Value:
1 3 7 5
2 3 8 2
3 2 9 1
4 3 10 1
5 2 11 3
6 5 12 1
Criterion Key No. 2 (Quiet Patient, Problem with Alcohol')
Critical Point: 7 and above 12 Items
T28 T79
T29 T80
F32 T81
T36 T85
T44 F124
T62 F133
Values to be assigned when key is scorable:
Item: Value: Item: Value:
1 3 7 3
2 3 8 4
3 3 9 4
4 4 10 4
5 4 11 5
6 4 12 5
Criterion Key No. 3 (Marginal Patient)
126
14 Items Critical Point: 8 and above
F 8 F59
Fll T82
F13 F93
T26 F101
T44 F103
T56 F106
T57 F125
Values to be assigned when key is scorable:
Item: Value: Item: Value:
1 3 7 4
2 2 8 2
3 2 9 2
4 4 10 2
5 2 11 2
6 4 12 2
17 Items Critical
T33 T59 T9 6
T34 T67 T104
T40
F69 T106
T53 T76 T115
T55 177 T125
F56 T78
Values to be assigned when key is scorable:
1 6 Item: Value:
2 6 7 3
3 6 8 6
4 2 9 4
5 2 10 5
6 6 11 3
12 6
127
Criterion Key No. 5 (Hostile Patient)
8 Items Critical Point: 3 and above
T4 T117
F17 T118
T18 T120
T60 T127
Values to be assigned when key is scorable:
Item: Value: Item: Value:
1 6 7 4
2 7 8 4
3 4 9 7
4 4 10 7
5 1 11 4
6 6 12 4
Criterion Key No. 6 (Unaccepting Patient)
15 Items Criti
T6 T36 F58
F8 T49 F69
F27 T50 T73
T28 T51 T106
T30 F57 T115
Values to be assigned when key is scorable:
Item: Value: Item: Value:
1 7 7 2
2 7 8 6
3 7 9 7
4 2 10 6
5 1 11 3
6 7 12 5
128
Criterion Key No. 7 (Dependent-Dissatisfied Patient)
14 Items Critical Point: 7 and above
T6 T33 F81
T8 T37 F82
T19 T39 F104
T23 F56 T126
T24 F76
assigned when key is scorable:
Item: Value:
7 3
8 4
9 6
10 6
11 4
12 4
Criterion Key No. 8 (Dependent-Satisfied Patient)
27 Items Critical Point: 13 and above
F7 F45 F82
T25 T46 F84
F30 F48 F100
T31 T58 T101
T32 F59 F104
F33 X 63 T105
F34 F76 T113
T42 F79 F125
F44 F80 T133
Values to be assigned when key is scorable:
Item: Value: Item: Value:
1 7 7 5
2 5 8 3
3 5 9 3
4 6 10 4
5 6 11 5
6 4 12 4
Values to be
Item: Value:
1 7
2 5
3 3
4 3
5 6
6 4
129
Criterion Key No. 9 (Conformist vs. Non-conformist)
This key is double scored.
25 Items - Score High if Critical Point
Score Low if Critical Point,.
F3 F54 T112
F4 F60 F117
T12 T68 F118
T13 F71 F120
T16 T75 F127
T17 F91 T129
F18 T102 T132
T22 F109 T134
T135
Key 9 a (Conformist) - High Score
Values to be assigned if key is scorable
Item: Value: Item: Value:
1 4 7 4
2 3 8 3
3 2 9 2
4 5 10 3
5 3 11 3
6 3 12 2
Key 9b (Non-conformist) - Low Score
Values to be assigned if key is scorable as low:
Item: Value: Item: Value:
130
Criterion Key No. n Key No.lO(High Energy vs. Low Energy)
inis key is double scored.
10 Items - Score High if Critical Point is above 5
Score Low if Critical Point is below 3
T23 F57
T28 T67
T33 T73
T35 T101
F56 T113
Key 10a (High Energy) - High Score
Values to be assigned if key is scorable as high:
Item: Value: Item: Value:
1 7 7 2
2 7 8 7
3 6 9 7
4 3 10 7
5 1 11 3
6 7 12 6
Key 10b (Low Energy) - Low Score
Values to be assigned if key is scorable as low:
Item: Value: Item: Value:
1 3 7 5
2 3 8 1
3 1 9 1
4 3 10 1
5 3 11 2
6 2 12 1
131
INSTRUCTIONS FOR SCORING
SELF-REPORT QUESTIONNAIRE
Materials Needed: Ten Criterion Keys
Prediction Work Sheet
Behavior Prediction Form
1. Score the questionnaire on each of the criterion keys.
Take each key in turn and simply count the number of
items which the patient has answered in the direction
on the key (if a patient has answered both True and
False or has omitted or changed an item, do not count
the item. Simply omit it.)
2. On each criterion key there is a critical point listed.
This is the number of items which the patient must an
swer in the given direction before the key is considered
scorable. If he doesn't have this many items, omit the
key as it is unscorable.
3. Each time you find a scorable key, turn to the Prediction
Work Sheet and enter the number of the key at the top of
a column marked "key.1 1 Now it is necessary to assign
values to this column for each of the 12 items listed
below the key number.
4. To find these 12 values you must return to the criterion
key. Hie bottom section of the key lists 12 items, each
with a value beside it. Simply copy these 12 values on
to the Prediction Work Sheet in the column beneath the
key number.
5. When all the scorable keys have been entered on the
Prediction Work Sheet together with the values for each
of the 12 items, it is necessary to compute a final
score. To do this you must now work across the rows
instead of up and down the columns as you have been
doing. First take Row 1, which contains all the Item 1
values for the keys you have scored. Add all the Item
1 values. Divide this total by the number of keys you
scored. This gives you an average score for Item 1.
Enter it across from Row 1 under "Average." This average
is the final score for the item. Do each of the 12 items
in turn.
6. Now copy each final score on to the Behavior Prediction
Form and scoring is complete. Circle the appropriate
number on the Behavior Prediction Form for each of the
final scores for the 12 items.
PREDICTION WORK SHEET*
Columns:
Key Key Key Key Key
Item 1
Item 2 ___ ___ ___ ___
Item 3 ___ ___ ___ ___
Item 4 ___ ___ ___ ___
Item 5 ___ ___ ___ ___
Item 6 ___ ___ ___ ___
Item 7 ___ ___ ___ ___
Item 8 ___ ___ ___ ___
Item 9 ___ ___ ___ ___
Item 10 ___ ___ ___ ___
Item 11 ___ ___ ___ ___
Item 12 ___ ___ ___ ___
*The average patient will have about four scorable
keys. Enter additional columns above if they are
needed, however.
FINAL
SCORE
(Average)
BEHAVIOR PREDICTION FORM
A. Implementing Behavior
Item
1 Needs a lot of attention and reassurance............... 1
2 Complains a lot; needs to be noticed...................1
3 Demanding, insistent, resistive to routine; wants
things his way......................................1
4 Anxious to please; goes out of his way to be helpful....1
B. Dependency Orientation
5 Admits he is very dependent upon others................ 1
6 Insists very strongly that he is independent and can
get along on his own. .............................1
7 Does whatever others tell him; seldom making suggestions
of his own......... 1
8 Makes demands on others but somehow their attempts to
help him meet with failure (rate hospital behavior
primarily)..........................................1
C. Emotional Control
9 Has such a high energy level that it is a real problem
to hold himself in check............................1
10 Lets his impulses go freely (doesn't hold them in) 1
11 Holds his impulses in but has to watch himself rigidly
in order to do it............. 1
12 Holds his impulses in for awhile and then cuts loose....1
Patient's Name
Ward__
Rater
Date
2 3 4 5
2 3 4 5
2 3 4 5
2 3 4 5
2 3 4 5
2 3 4 5
2 3 4 5
2 3 4 5
2 3 4 5
2 3 4 5
2 3 4 5
2 3 4 5
A rating of 7 indicates the patient will be predicted to show this
behavior in unusual degree. A rating of 1 indicates the patient
will show very little of the behavior. A 4 rating is about average.
STAFF RATING OF BEHAVIOR FORM
(Item Content Identical to that
in Prediction of Behavior Form)
Place a check at the point you feel most accurately
describes the patient as you have seen him. For
example, if Patient Jones needs a great deal more
attention and reassurance than most of the patients
you see, mark him "7". (If he needs a great deal less,
mark him "I". If he's about average, mark him "4",
and so on). NOTE: RESERVE THE 4x RATING FOR A
PARTICULAR PURPOSE. Use the 4x rating only for those
few patients whose behavior is so changeable that it
is simply impossible otherwise to rate him on the
sentence in question.
Not
True
1 2 3
About
Average
4 5 6
Very
True
7
I
A.
Implementing Behavior.
Needs a lot of: attention and reassurance. (Item 1)
1 2 3
4x
4 5 6 7 B. Complains a lot; needs to be noticed. (Item 2)
1 2 3
4x
k 5
6 7 C. Demanding, insistent, resistive to routine;
wants things his way. (Item3)
1 2 3 4 5 6 7 D. Anxious to please; goes out of his way to be
4x
helpful. (Item 4)
*This form was contained on one sheet in the study done. Notations
in parentheses did not appear on the original form and are included
here for explanatory purposes.
Patient's Name__
Ward__
Rater
Date '
STAFF SATING 0F BEHAVIOR FOIH (continued)
Not About H* dependency
True Average Triife
1 2 3 4 5 6 7______A. Admits he is very dependent upon others.
5x
1 2 3 A 5 6 7 B. Insists very strongly that he is independent
and can get' along on his own.
1 2 3 4 5 6 7 C. Does whatever others tell him; seldom making
55 suggestions of his own.
1 2 3 4 5 6 7 D. Makes demands .on others but somehow their
5x attempts £o help him meet with failure.
(ratd hospital behavior primarily).
III. Emotional Control.
1 2 3 4 5 6 7______A. Has such a high energy level that it is a
4* real problem to hold himself in check.
1 2 3 4 5 6 7 B. Lets his impulses go freely (doesn’t hold
55 them in).
1 2 3 4 5 6 7_______C. Holds his impulses in but has to watch him-
55 " self rigidly in order to do it.
1 2 3 4 5 6 7______ D. Holds his impulses in for awhile and then
5x cuts loose.
(How confident $re you in this rating? Circle one: Very confident,
moderately confident, mildly confident, barely confident, just guessing)
Item 5)
Item 6)
Item 7)
Item 8)
Item 9)
Item 10)
Item 11)
Item 12)
t-*
w
Ln

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

On The Relationship Between Anxiety And Aggression In Nine-Year-Old Boys

PDF

Conflict And Generalized Conflict In Verbal And Motor Responses

PDF

Methodological Problems In The Identification Of Suicidal Behavior By Means Of Two Personality Inventories

PDF

Behavioral (Social) Intelligence: A Factor Analysis

PDF

Consensus Of Role Perceptions In A Welfare Planning Council

PDF

Transformational Processing Of Sentences Containing Adjectival Modifiers

PDF

A Study Of The Relationship Of Temperament Variables To The Ability To Make Certain Judgments Of Emotional Behavior

PDF

A Factor Analytic Investigation Of The Effects Of Specific Visual Contentof Motion Pictures Used In Psychological Measurement

PDF

Intellect After Lobotomy In Schizophrenia: A Factor-Analytic Study

PDF

A Study Of The Relationships Between Factors Found In Cattell'S Sixteen Personality Factor Questionnaire And Factors Found In The Guilford Personality Inventories

PDF

Economic Differentiation And Social Organization Of Standard Metropolitanareas In The United States: 1950

PDF

Creativity In Children: A Study Of The Relationship Between Temperament Factors And Aptitude Factors Involved In The Creative Ability Of Seventh Grade Children With Suggestions For A Theory Of C...

PDF

Differential Fertility Behavior And Values In Rural And Semi-Urban Costa Rica

PDF

A Factor Analysis Of The Semantic-Evaluation Abilities

PDF

An Analysis Of The Selection Criteria For Assignment Of Students To Advanced Placement Classes In The Los Angeles Unified School District

PDF

A Factor-Analytic Study Of Military Leadership

PDF

Determinants Of Intercounty Migration: California, 1970-1973

PDF

A Study Of The Effects Of Generalized Expectancies Upon Accuracy Of Interpersonal Perception

PDF

The Effects Of Prior Part-Experiences On Visual Form Perception In The Albino Rat

PDF

Status Consistency Among The Clergy

Asset Metadata

Creator Leonard, Calista Verne (author)

Core Title Prediction Of Overt Behaviors In Hospitalized Psychiatric Patients

Degree Doctor of Philosophy

Degree Program Psychology

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag OAI-PMH Harvest,Psychology, clinical

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Winder, Clarence L. (committee chair), Guilford, Joy P. (committee member), Shneidman, Edwin S. (committee member), Van Arsdol, Maurice D., Jr. (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c18-345132

Unique identifier UC11359093

Identifier 6411256.pdf (filename),usctheses-c18-345132 (legacy record id)

Legacy Identifier 6411256.pdf

Dmrecord 345132

Document Type Dissertation

Rights Leonard, Calista Verne

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA