Upload
others
View
3
Download
0
Embed Size (px)
Citation preview
Boston University
OpenBU http://open.bu.edu
Theses & Dissertations Dissertations and Theses (pre-1964)
1956
A study of sociometric patterns and
behavior characteristics on a
mental hospital ward.
Langevin, Henry J., Jr.
Boston University
https://hdl.handle.net/2144/9171
Boston University
BOSTON UNIVERSITY
SCHOOL OF NURSING
A STUDY OF SOCIOMETRIC PATTEPJNS AND BEHAVIOR
CHAfu\CTERISTICS ON A MENTAL HOSPITAL WARD
Henry J. Langevin , Jr.
(B.S., Boston University Schoo l of Nurs ing , August 1955.)
A Field St udy Submitted in Partial
Fulfillment of the Requirements for
the Degree of Master of Science .
August-1956
First Reader: {j M ."-t-• Q fQ 1~J.v.... . Second Reader : L~ ..J-, X-"-':<""'""-.,.. • ~
ACKNOWLEDGMENT
The writer wishes to express
hi s appreciati on for the assistance
given in t he preparation of this study
by Winifred Gibson, Associate Professor
of Nursing ; Luci lle Sommermeyer, Asso
ciate Professor of Nur sing; and Dorrian
Apple, Visiting Assistant Prof essor of
Soci al Sciences.
· Appreciation i s also expressed
to t he graduate nurses, the nur sing
students, and the attendants who parti
cipated in making this study possible.
TABLE OF CONTENTS
LIST OF TABLES
LIST OF FIGURES . . . . . . . . . Chapter
I . INTRODUCTION . . . . . . . . . . . . . . Statement of the Problem Purpo se of the Study Scope of the Study . . . Limitations of the Study Review of the Literature Source of the Data Pattern of Presentation
II. PHILOSOPHY ..
Hypothesis
III. METHODOLOGY
. . . . . . .
Page
v
vi
1
1 2 3 4 5
10 10
12
14
15
Designing the Sociometric Instrument 15 Designing t he Behavior Instrument 15 Methods of Scoring . . . . . • . . 19 Collecting the Data . . . . . . . 20
IV. ANALYSIS OF THE DATA
To Test Hypothesis #1 To Test Hypothesis #2 To Test Hypothesis #3 To Test Hypothesis #4 Discussion of Findings
22
22 29 35 39 43
V. SUMMARY , CONCLUSIONS, AND RECOMMENDATIONS 48
APPENDIX
BIBLIOGRAPHY
Summary . . • Conclusions . Recommendations
iv
48 49 50
51
64
LIST OF TABLES
Table
1. Assignment of Patients to Status Groups
2. Relationship Between Sociometric .Status and Behavior Characteristics and Total Scores and Percentages • • . . . • . . .
3. Z Transformation Values and Probability Values for Correlation Coefficients of the Three Occupational Groups . . . • .
4. Disposition and Total Length of Hospitalization for the 16 Patients Included in this Study as of Three Mqnths Following the Collection of the Data . . . .
5. 2x2 Contingency Table of Sociometric Status and Hospital Disposition
6. Analysis of Age, Diagnosi s, and Length of Hospitalization of the 16 Patients
Page
30
32
39
41
43
Included in this Study .' . . . . . . . . . Appendix
7. Analysis of Age, Occupational Status, Sex, and Length of Time on Ward on the 14 Personnel Included in this Study . . . . Appendix
8. Ranked Sociometric and Behavior Rating Data for the 16 Patients as Determined by the lL~ Nursing Personnel . . . . . . . . Appendix
v
LIST OF FIGURES
Figure
I. Rank Order Correlations of Sociometric and Behavior Rating Data for Each of the 14 Nursing Personnel and for the
Page
Per sonnel Considered as a Group • • . . • 23
II. Rank Order Correlations of Sociometric and Behavior Rating Data for Nurses, Attendants, and Students as Members of Different Occupational Group s . . . . . 37
III. Rank Order Correlation of Sociometri c Rank and Length of Hospitalization . . . 42
IV. Linear Relationship Between Ranked Sociometri c and Behavior Rating Data for t he 16 Patients as Determined by the 14 Nursing Personnel . . . . . . . . Appendix
vi
CHAPTER I
INTRODUCTION
In recent years a number of studies1 have been under
taken for the purpose of examining group structure and i nte
gration i n mental hospitals with particular reference to
patient-patient, staff-staff, and staff-patient relationshi ps.
Thi s intensification of interest in the examination of the
social structure of mental hosp i tals and in patterns of
i nterpersonal relations on psychi atri c wards has led to t h e
realization that the attitudes directly or indirectly com-
municai:ed to a patient by staff members or other pati en ts
are of vital therapeutic signifi cance . Since nursing per-
sonnel are more intimately assoc i ated ~vith patients in t he
day by day experiences of hospital living than other members
of the hospital staff, their relationships with pati ents
should be the focus of more inten s i ve empirical research .
Statement of the Problem
A consideration of the h i storical background of
nurs i ng research in personnel-patient relationships on
psychiatric wards suggests areas which require continued
1David Rioch and P .. lfred H. Stanton, "Milieu Therapy, " Psychiatry, XVI (1953), 65-72.
-1-
-2-
attention. These are: (1) A need for new methodolog~cal
techniques which may be described, and repeated by other
i nvestigatorcS; (2) A need for studies which systematically
i nvestigate differen t a r ea s of interper sonal relation s on
men t a l hospital wards ; (3) A need for studi es which consi der
the total ward communi ty as a uni t of inquiry.
The present study is aimed at these needs. Speci
f i cally the problem is to undertake exploratory resear ch
into one aspect of group relations on a psychiatri c ward by
i nvest igating t he re l ationship between t h e feel i ng rea ctions1
of nurs ing personnel for pati ents and t h e i r perceptions of
the behavi or of t hese pati ents.
Purpose of the Study
TI1e main purpose of this study is to cons i der t h e
rela t i onship between behavior data ob tai ned from rat i ng
scales and sociometric data obtained from sociometri c tests
in order to answer the following questi ons.
1. Does a positive relatio~ship exist between h i gh sociometri c status and the more soci ally acceptable behavior characteri sti cs?
1The term " feeling reactions" as used in this study represents the feelings of nurs i ng personnel for patients accordi ng to the degree that they l i ke, feel indifference fo r , or dislike patients.
2The term " sociometric status" as used in thi s study represents a patient's ranking in relation to the other pati ents according to the feeling reactions of the nursing personnel . The best liked patient will obtain the h ighest soci ometri c status, whi le the least liked has the lowest status.
-3-
2. Are there behavior characteristics t hat differentiate patients of high, intermedi ate or low sociometri c status?
3 . When considered as members of different occupational groups, do nursing personnel -nurses, attendants, and students -- differ in their feelings for patients and in their perception s of the behavior of these patients?
4. Does a relationship exist between sociometric status and length of hospitalization ?
5. What are the implications of the results of this study for the direction of psychiatric nursing?
Scope of the Study
The data was collected in a sixteen-bed ward for
non-disturbed men which was part of a 225-bed private psy
chiatric hospital located in suburban Boston. The pati ents
ranged in age, from sixteen to seventy - five; in diagnosis,
from adolescent turmoil to senile psychosis; and in length
of hospitalization, from four week s to fifteen years. 1
This study is concerned with the way the nursing
staff (four nurses, four attendants, and six nurs ing stu
dents2) feel about and perceive the behavior of the sixteen
psychiatric patients who compri sed the total ward population
when t he data was collected.
1see Appendix A for a complete list of the sixteen patients included in this study.
2see Appendix B for a complete list of the fourteen nursing personnel included in this study.
-4-
This particular ward was selected because of the
variety of clinical cases available and the high personnel
patient ratio. A small hospital ward has the advantage of
giving the nursing staff ample opportunity to form impres
sions and feelings about all the patients on the ward. The
two methodological ins truments -- the sociometric test and
the behavior rating schedule -- were constructed for the
purpose of securing information about these feelings and
impressions.
Limitations of the Study
Since much of the emphasis of this study was
placed on the development of research techniques that could
be utilized for statistical purposes, the results are to be
considered exploratory in nature. Even though the items
selected for the behavior rating schedule make up but a
small proportion of possible behavi or variations, they were
consi dered of suffi c i ent importance to be included in the
study. Possibly other items could have been included; how
ever, practi cal considerations set the limits on the number
t hat could be used. The selection of numerical weights for
both the sociometric data and behavior data was done arbi
trarily since no standardized scales were available.
In any statistical analysis a small sample is always
a limitation especially in terms of the magnitude of the
results that can be considered significant. This study is
-5-
no exception. Even though the sixteen patients and the
fourteen nursing personnel comprised the total ward popula
tion rather than a ·sample from it, a comparatively high cor-
relation value was required in order to be significant.
In the discussion and interpretation of the results,
it is, of course, understood that any conclusions drawn are
not necessarily applicable to other mental hosp i tal popu
lations, but are strictly confined to generalizations about
the group studied here. However, some of the conclusion s
may be valid for similar groups; but no such extrapolation
is explicitly made, and the results can only be regarded as
tentative, pending further research.
Review of the Literature
The more important socio-psychiatric studi es re
lating to mental hospital structure and integration have
been more than adequately summarized by Rioch and Stanton, 1
and Stanton and Schwartz, 2 and will not be repeated here.
They suggest that the attitudes communicated to a patien t
by the hospital milieu is of vital importance in t he rapy .
A number of studies have been reported which deal
directly with interpersonal relations in the nursing
1Rioch and Stanton, op . cit.
2Alfred H. Stanton and Morri s S. Schwartz, The Mental Hospital, (New York: Basic Books, Inc., 195~
-6-
s ituation . Jenkins1 applied Hyde and York's sociogram tech-
nique to study interactions among nursing personnel, among
patients, and between the two on an acti ve mental hospital
ward. She found that the attitudes of nursing personnel
toward individual patients influenced the quality of their
nursing.
Kandler2 utilized a modification of the Moreno soci-
ornetric choice technique and the interview to investigate
the phases of rapport which may be determined by stated likes
and dislikes of patients and nursing personnel. The study
was made on three wards of a small psychiatric hospital
which does extensive research. She found that personnel
like patients who are young, friendly, attractive, have
something in common with them, who help on the unit, and who
present no problems. She also suggested that patients not
liked by the personnel are not liked by the other patients.
Using the Morenian choice process and nonparticipant
observation of personnel, Morimoto 3 attempted to learn more
1Elda Hoke Jenkins, "Interpersonal Relations in Psychiatric Nursing," (Unpublished Master's Thesis, School of Nursing, Boston University, 19L~9).
2Harriet M. Kandler, "A Study in the Elements of Rapport," (Unpublished Master's Thesis, School of Nursing, Boston University, 1950).
3Francoise R. Morimoto, "Favoritism in PersonnelPatient Interaction," Nursing Research, III (February 1955), 109-112.
-7-
about the differences in nurses' relationships with patients
they like or dislike. For the study, she utilized the per
sonnel assigned to a female active ward of a small psychi
atric hospital . Her findings indicate that personnel made
more contacts and spent more time interacting v1ith favored
than with nonfavored patients. Their approach to the
favored patient was a personalized one -- they did something
for her because she was a person. With the nonpreferred
patient their approach was more businesslike and primarily
concerned with fulfilling her physical needs -- they did
something for her because she was a patient .
Tudor1 utilized a sociopsychiatric approach of par
ticipant observati on to investigate a problem of mutual
withdrawal on a disturbed ward of a small mental hospital.
She concluded that mutual withdrawal between personnel and
patient not only served to reinforce and stabilize the
patient 's mental illness, but also entered into other prob
lems of patient-staff interaction on t he ward. That is, the
labeling of patients as "hopeless," "assaultive," "unrespon
sive," and "unable to tolerate closeness," served as a con-
venient rationalization for avoiding the patient, and
thereby perpetuating the process of mutual withdrawal.
1Gwen E. Tudor, "A Sociopsychiatric Nursing Approach to Intervention in a Problem of Mutual Withdrawal on a Mental Hospital Ward," Psychiatry, XV (1952), 193-217 .
-8-
The aforementioned studies conclude that nursing per
sonnel like patients who are helpful, young, friendly; spend
more time interacting with favored than with non-favored
patients; and withdraw from patients who are "hopeless,"
"assaultive," "unresponsive ," These results suggest that a
positive relationship exists between degree of liking, be-
havior characteri stics, and amount of interaction.
Although the present study is based on the study of
interpresonal relations as were the preceding four, it dif
fers in method and emphasis. Kandler's study bears some re
semblance to the present investigation in that she studied
the relationship between feeling reactions and behavior
characteristics. However, she did not attempt to study the
total patient population of any on~ ward, nor util i ze a
methodology rigorous enough for statistical analysis. There
are no nursing studies available that specifically attempt
to relate sociometric data with behavior rating data.
Sociometric studies of college students and school
children have indicated that positive relationships exist
between high social status and the more positive personality
characteri stics. In her study of the personality patterns 1 of school chi ldren, Northway concluded that recessive
~ary L. Northway, "Outsiders, A Study of the Personality Patterns Least Acceptable to Their Age Mates, " Sociometry, VII (February 1944), 10-25.
-9-
children, socially uninterested chi ldren, and soci ally in
effective chi ldren were least acceptable to thei r age mates.
Bott's1 study of nursery school children suggested that the
best adjusted children had many companions while the child
ren who presented distinct behavior problems had few com-
panions.
Bonney2 utilized sociome tric data and a personality
self rating scale to i nvestigate some factors related to
sociometric status in a men's domi tory. He found that a
high correlation ex isted between sociometr i c status and
positive personality characteri stics.
The most extensive research which has attempted to
combine sociometric and rating data has been that of Lemann
and Solomon. 3 The subjects were three small group s o f col
lege s tudents who lived together in " off-campus" dormitories.
Their findings not only agreed that sociometri c statu s and
personal i ty characteristics are positively related, but also
suggested that t he p i cture of a group obtained from rat ing
1H.M. Botts, "Observati ons of Play Activities in a Nursery School," Genetic Psychological Monographs, IV (July 1929), 44-88.
~.E. Bonney, R.E. Hobbit, and A. H. Dryer, " A Study of Some Factors Related to Sociometric Statu s in a Men's Dormitory, " Sociometry, V (November 1953), 287-293.
3Thomas B. Lemann and Richard L. Solomon, "Group Characteristics as Revealed by Sociometric Patterns and Personality Ratings, " Sociometry, XV (February 1952), 7-90 . This study contains an extensive bibliography of previous investigations in t hi s area of soc i ometric r esearch .
-10-
scale data may be connected with the independently obtained
picture of its sociometric structur e.
TI1ese interpersonal studies of community groups
have shown that soc i ometric status and behavior character
istics are directly related. Does such a relationship
exist in a mental hospital ward? The present study is an
attempt to answer this question.
Sources of Data
Each of the nursing personnel was requested to (1)
complete behavior profiles for all s ixteen patients, and
(2) list all patients sociometrically in rank order of
preference . Since fourteen nursing personnel were included
in the study, the raw data consisted of 224 behavior rating
schedules and fourteen sociometric forms. The scores were
then coded and placed in matric form to facilitate sta
t i stical analysis .
Three months after the materi al for the original
study was collected, the ward f i le was checked to determi ne
the location of t he sixteen patients included in the study.
This was done to determine if any relationship existed
between soci ometric status and length of hospitalization .
Pattern of Presentation
Chapter II will present the phi losophy underlying
this study. Following this, in Chapter III, will be a
-11-
detailed description of the methods used and t he ir develop
ment. Chapter IV will contain the analysis of the data, and
Chapter V will con s i st of the summary, conclusions, and
recommendations . suggested by this study. This final chap ter
will be followed by t he necessary appendixes and bibliogra
phy .
CHAPTER II
PHILOSOPHY
Usually a patient enters a psychiatric hospital
because he has experienced difficulty living in the outside
world. The t herapeutic function of this hospital is to
create an environment that will facilitate his return to
the outside world. Since the nursing personnel have the
longest and most intimate contact with the patient, much of
the task of gratifying his needs, facilitating his communi
cation with others, and enhancing his social participation
becomes their responsibility. However, the mental hospital
is not a social vacuum. Within the framework of continuous,
inescapable, and emotionally important relationships between
patients and nursing personnel, basic social needs exist
for the personnel as well as for the patients.
Of all the living beings, man is by far the most
dependent upon social relationships for his happiness and
growth. From his earliest years, the child develops by
entering into relationships with others and in communi
cating experiences with others . As he grows his behavior is
more and more affected by the relationship between himself
and others in the give and take of living.
-12-
-13-
In this course of development, the individual be
gins a differentiating process which is characteri zed by
hi s inherent preferences for others. He may approach those
who r espond to him or whom he wishes would respond to him;
or he may keep away from those whom he feels he cannot
interest or who repel him. Thus, a person gradually exerts
hi s preferences toward others in ways that hold greater s ig
nificance for him, and his choices are more or less con
sci ously made as he gains awareness and control over them.
Such a process represents the sustaining emotional r e in
forcement of the member of any group on one another.
Wi thin the group structure of a mental hospital
ward, as within any other social group, a hierachy of per
sonal relationships can be seen to exist. Attractions and
rejections exist among the patients, among the personnel,
and between members of both groups. Even though the nurs
ing personnel have a responsibility toward all patients,
and all patient s have a right to nursing care, as human
beings, the personnel are expected to have person~l . prefer
ences. Assuming that the nursing personnel prefer some
patients to the exclusion of others, are there behavior
characteristics that differentiate the preferred from the
non-preferred patients?
The sociometric studies reviewed in the· previous
chapter not only suggest but definitely conclude that
acceptance or rejection by one's contemporaries is closely
-14 -
related to the social adequacy of the individual's behavior
patterns. The point of view of thi s study is that this re
lationship exists within the structure of a mental hospital
ward as well as in the outside community.
Hypothesis
Formally stated, the hypothesis for this study is
as follows: Nursing personnel prefer patients who display
socially acceptable behavior characteristics. These pre
ferred patients return to the outside community qui cker than
the non-preferred patients.
From this hypothesis, four working hypoth eses have
been abstracted for testing in this study. They are as
follows:
1. A positive relationship exists between high sociometri c statu s (degree of preference) and socially acceptable behavior charac teristics.
2. There are behavior characteri stics that differentiate patients of high , intermediate,or low sociometric status. The higher a pat i ent's status, the more he will display socially acceptable behavior characteristics; the -lower hi s status, the more he will display less a cceptable characteristics.
3 . When consi dered as members of different occupati onal group s, nurses, attendants, and students do not differ significantly in their feelings for patients, nor in their perceptions of the behavior of these patients.
4. A po s itive relation ship exists between sociometric status and length of hospitali zation.
CHAPTER III
HETHODOLOGY
The first step in this study was to decide on the
methodological instruments to be used in collecting the
data. There were two distinct types of inforraation neces-
sary, that based on personnel preferences and that based on
perceptions of behavior . These required two separate areas
of planning , each ruled by different considerations.
Designing the Sociometric Instrument.
In order to ascertain the preferences of personnel
for patients , a modification of Horeno's1 soci ometri c
choice technique was used. However, before .deciding on the
specific sociometric questionnaire, a number of considera-
tions \vere involved. The f irst question to be decided in
making up the soci ometric instrument was wnat, if any, cri-
teria should be used . Sociometric criteria refer to t h e
different areas of choice (or rejection) that may be allm...,ed.
Among the school children t he " sit next to" and the "play
with" criteria are the ones most generally used; among col
lege students, the "choice of roorrrrnate" and " choice of
1 J.L. Moreno, .\~10 Shall Survive, (New York : Beacon House, Inc., 195B).
-15-
-16-
double dating companion" are frequently used. The most
common n~~ber of criteria in sociometric work is t hree.
Usually the sociometric test is given to a homo
geneous population like school children or college studen ts.
For this study, however, it was not feasible to use speci
fic criteria because of the age and sex differences of the
nursing personnel to be tested as well as differences i n .
age, phys i cal condi tion, and mental condi tion of t h e pati ent
group.
In order to overcome these difficulties in deter
mining the preferences of personnel for patients, and to
take i nto account as many di fferent components of soci o
metric status as possible, it was decided to ask t h e per
sonnel to list in rank order all the patients according to
the ease with which they were able to form fr i endly rela
t i onships.
To facilitate the choice process, a list of t he
six teen patients assigned to the ward was presented to
each of the nurs i ng personnel duri ng an informal intervi ew
s i tuati on. The instructions were to divide the patients
i n to three groups according to l i ke, feel indi fference for,
or dislike. Each group was to be subdi vided until a rank
order of all sixteen patients was obtai ned. Thus, each
patient was ranked sociometr ically from one to sixteen by
each of t h e nurs i ng personnel, tvith t he patient best liked
l
-17-
receiving the first choi ce and the patient least liked re
ceiving the last cho i ce.
To secure information about behavior characteri s-
tics that could be used in constructing the behavior sched-
ule, a group of forty affiliating nursing students were
asked to list the factors that t hey considered important in
liki n g or disliking patients. From these answers and from
items included in Cattell's1 " Source Traits Based on Self
Inventories" :the behavior schedule was designed. Specific
kinds of personal social behavi ors were assigned to each
item rather t han trait names or descriptive adjectives so
that as much as possible, the items would "mean" the same
t hing to all people. For example, the phrase "prefer to
r emain by himself, rather than with others" was used as one
of the items rather than the term "asocial."
Of t he original thirty-six items included on the
schedule, six were discarded as being too ambiguous.
The remaining thirty items were grouped into five
categorie s of six items each. (See Appendix C.) The five
groupings were :
A) Friendliness-Sociability
B) Gregariousness-Self-Sufficiency;
C) Dependence -Emotionality;
1Raymond Cattell, Descriptions and Measurement of Personality, (New York : World Book Co., 1946), 341 - 394' . .
-18-
D) Aggressiveness-Hostility;
E) Introversion - Social Inadequacy.
As suggested by the titles, the twelve items in
Groups A and B described socially desirable behavior char
acteristic s, while the eighteen items in Groups C, D, and
E were considered socially undesirable. The titles were
selected as a mean s of identifying the item groupings with
in the context of this study in order to facilitate data
analys i s. i\t the suggestion of three psychiatric nursing
speci alists, five i tems wer e reworded. They agreed in the
choice of items wi thin each grouping, and felt that the
thirty items were suitable for the purposes of this study.
Yhe final behavior schedule was designed by alter
nating items from each category in rank order. (See
Appendix D.) The selections were made as follows:
Group A items 4-10-15-18-24-30
Group B items 1- 7-12-21-27-29
Group c i tems 3-8 -13-19 -23-25
Group D items 2-6-11-16-20-26
Group E items 5-9 -14-17-22-28
The items were alternated in t h is manner in order to equate
the value of each category in the position of its items on
the form.
For the purpose of determining the intens ity of
re sponses t o each item, and for stati sti cal reasons, a five
-19-
point rating scale was used . Each item was to be answered
by circling the appropriate letter:
A represented always
U represented usually
SO represented sometimes
SE represented seldom
N represented never
Even though all r ating scales suffer fromtre differences in
the absolute rating tendencies of individual subjects, a
five point rating scale was sufficiently refined to meet
the requirements of this study.
Methods of Scoring .
The soci ometric choices of the individual subj ects
were transformed into sociometric scores according to t h e
method shown below.
Sociometric Choice 1 2 3 4 5 6 7 8 9 ~0 11 12 13 14 15
Sociometric scor e 16 15 14 13 12 11 10 9 8 7 6 5 4 3
High scor es signified preferred patients, and low scor es
non -preferred patients .
The behavior schedule i tems were scored according
to grouping . The twelve items in the two socially desir
able groups, A and B, were scored as follows: A-5, U-4,
2·
16
1
-20-
S0-3, SE -2, N-1. The eighteen items in the socially un
desirable groups, C, D, and E, were scored A-1, U-2, S0-3,
SE-4, N-5. The range for each group ran from 6-30; for
each schedule, 30-150. This was on the basis of 1-5 points
per item, six items per group, and five groups per schedule.
As seen by each subject, the higher the score, the more
desirable the patient's behavior is said to be. For ex
ample, a high score on groups A and B s ignifies that the
patient usually or always exhibits desirable behavior
characteristics, while a high score on group s C, D, and E
signifies that the patient seldom or never exhibits undesir
able behavior characteristics. The above analysis s imply
describes the process of scoring the items obtained from
the sociometric and behavior instruments . A more detailed
analysis will be presented in the following chapter .
Collecting the Data.
TI1e soci ometric test and the behavior schedule form
were ad~ini stered to the nursing staff on the ward under
study between March 13, and March 16. The personnel in
cluded four nurses, six nursing students, and four attend
ants. Three nurses and five students were f emale ; the
others vJere male.
Since ward experience of students was limited to
one month, the data was collected during their final three
days on the ward. In this way all subj ects had at l east
-21 -
one month 's experience i nteracting with the patients.
The fourteen subjects willingly cooperated in the
study . (The 11-7 night shift was not i ncluded . ) The data
was collected in t he following manner . An inforrnal con
ference was held wi th each subj ect . After the sociometri c
data had been collected, s ixteen behavior schedules were
fill ed by each sul~j ect, one for each patient on the ward .
The analysis of this data is presented in the following
chapter .
'
CHAPTER IV
ANALYSIS OF THE DATA
l:..nalysis of the Data to Test Hypothesis .ffol.
The rank correlation coefficient1
RHO •
was selected to test the hypothesis that a po s itive rela-
tionship exists bet\veen high sociometric status and soci-
ally acceptable behavior characteri stics. The correlation
coefficient is a measure of the degree of relationship
present between two variables, while the rank correlation
coeffi cient is the product moment coeff i cient applied to
ranked data .
In order to convert the behavior rating data into .
suitable form for rank order correlati ons, the scores r e-
ceived by t h e sixteen patients were transposed into ranks,
with the patient receivi ng the highest score being assigned
a rank of s ixteen, and the patient with the l owest score
receiving a rank of one. Following this procedure, both
1Allen L. Edwards , Statistical Methods for the Behavioral Sciences, (New York : Rinehard C! Co., 1954)-; 195.
-22-
FIGURE J
RRHI< ORDER CO!fR.ELil/1/JNS OFSDCJO METRIC RND BEH/JVId~ R FJTING 0 FJT/9 FOF{ E RCH OF IHE 14- /VUR SING- P E~S6NNEJ.. 1'9/'ID FdR TilE P£/?SONNEJ.. C.ONSI£>£J?ED 19.5 R GRoUP
l't---.::....!H-'-E.....;.R..:.....=O___:N_u__;R___;;_.5_£ __ 1 __ _.,. I. I
'Ill ~ ll. ~, ()
~I V>g ~ 8 ~ 7
' ~s ""~ ~3
2 1~--~----------
1 2 3 q.. 5 6 7 8 9 /IJ II 12 /3 14 IS /6 SOCIOMET~IC SCORES
RH&.I52.~
,, ___ =-..........;_,;......_ ____ ~----=----.·
IS
l ~ 12. ~~~ (J lb {f)!
<!:8 ~7 ~(,
§5 '44lfo Co3
2. 1~--------------~
I 2 3 <1- S 6 7 8 g /IJ II 12 13 14 15/6 So C/tJ ME 1/Uc. SCOI?£5
R. HO .38 2.
v,l ~ 1'2 ~~~ u /ll lllS ~8 ~1 ~6
~5 ~4-~3
2
I"
f!!/55 Y Hffl/) N f/ R S£ 2
/·------~~-------~ 1 2.. 3 + 5 6 1 8 s ltJ '' 11- 13 H 15 lo
5 OCIOM £/RIC. SCORf:..S r R H0 • .350
SIFJFF NUR.SE 'f-
"'' ~I ~II Y1 /()
~
~8 ~7 ~b
::t:s IJ.lq.. Q)3
2. , _____________ ~ I 2. 3 ~ 5 6 7 8 9 /6 II tZ 13 I~ IS 16
SoCJt:JMEil{/c scoRES /fHO .65.91-
.r 5/G-NIFICFJA/T J1T Tfl£ I o/t~ J..EVEL "~• St&-Nt~lt:IINI fiT "HIE. 5~ LE~EL
-~ 4--
F I G () ~ E I c ONTi 1'1"" li D
I
I 13
"' 12. ~~~ ~ /0 \)
CJ)9
eta ~1 ....... ~~
~5 ~4-1:03
2.
NUP..SIN~ STUPENT S
1~--~------------------~
13 ~I cc ,, ~I CI)Q)
~8 'Q 1 ~,
~s \'1,
~3 2
I 2. 3 ,_ 5 G 7 8 ~ 16 II 12-13 /4/S 16 So Uo M F-IP. tc S c.o ~ ES
RHO .?sa-t-
f?. S I /II G STUD /!'/VI 7
I 1 2. 3 q-5 b 7 e <J ltJ II 12..13 11--1s-16 50 C 10M E7RI C. S Ctl/(E. .S
!{If{) .]5() ..,..
13 ~I
~II a 10 U)g
't8 ~ 7 ~" ~6 ~4-~3
2. 1~--~------------------_j I 2 3 '1-5 " 7 8 9 /IJ II 13-..13 If-/~ /6
13
~' ~II \) /0 ell~
~g 'Q7 ~, ~.., ~,!) ~~ (t)3
2..
soc.toM~/f?..IC scoRES RJ.Jo.es9+-
1~----------------------~ I 2. 3 tf S ' 1 8 S to fl ~~ 13 I+ IS''
SO<:.I() ME7/fiC SC.b~E.S
RHO .4G.8
I
~I
~~~ ~ltJ ~~
8 ~7 ~6
~~ \t.t+ (0.3
2
-as-F IGU!iE. I c ONT//V(JED
NORSINCr STUDENI 3 NtJRS!N6 ST#IDENT 10 .
1--~----------~--------1 2 3 q- 5 6 7 8 a 10 II 12. • .1"3 /11-IS 16
~' ~ ., ~~~ v /0 W).9 ~8 ~ 7 ~(&.
~5 ~+
3 2.
SociOM£rR.IC ScoRES RHO ·838-Y
II
1~~--~~--------------1 2. 3 '1- 5 E, 1 e ~ 10 II ~~ 13 14- IS 16
5 0 C. -, 0 M E IR I c. S C::: b R E S R H o • soo-t+
~ /3 ~ ll
"'" {J'1 /fJ <t"g ~B ~
tr:1 ~b \&J5 ~+
3 2,
rrr1"Eh DfJN/ 12.
1------------------------~ I z 3 + s b 7 s 9 IO ,, 11!..13 14-15/h SOC/0 M£1"/f,-C SCaReS
RlfO .905 ~
I
-26-
/G-Ulf£ I COI/1/NVED
/f.,___.....:..:FJ:......:.T....:...T-=E:.:..N....:........:O::....:H:..;..N_..:........:.T_.:/_.::3:!:--.-~ ~----....:...fl~T~T~E_/_Y_D~n~N~T--~1+~~~
I I
~B ~12. , ,, v /IJ (1'\~
~8 ~ 7 <it'6 ~5 ~3
2.
J I
~I ~ ',, (J /~ "lg
~8 "S1 ~" ~5 ~4
3 a
~~~------------------~ I 2. 3 erSt. 7 9 0 /0 If IZ 13 11-/s-16 /I 2.. 3 t:f !J' 7 8 eJ ld 1112../J If-IS/~ soc.1o M£"1RI'- sco~E.s
RHO · '1 ttt- ~ SOc!.ltJMFIR/G SCORE'S
RHO • 67.S;c
f(, -t'01RL P~ SoN /VEL.
1:
.~:3 ~ <:)
Vlf V) ,,
"g ~
~· ~7 ~' "-4 ~4-
3 2 ~~~------------------~
I 2. .3 +-6' 1 6 e /IJ II 1"13 HIS /6 SocJO tiET.f<lt, scoJ?E s
RHO .876x
-27-
variables -- the sociometric data, and the behaviorr.ating_ ~ta
-- were in suitable form for analysis.
As can be seen in Figure I, individual rank order
correlations were done for the fourteen nursing personnel
as well as a composite correlati on for all personnel. The
data for t his latter correlation was obtained by s imply
adding all soci ometric and behavior rating scores and rank
ing the results from one to six teen.
In order to determine whether or not the data
contradicted the hypothesis to be tested, the null hypothe
s i s t hat the population correlation equals zero was set up
for testing . A hypo"tl1.esis that i s set up wid1. the pos
s ibility of its bei ng rejected at some defined probability
value i s called a 11null hypothesis. " In other words, if
the null hypothesis i s rejected at a specified level o f
confidence, ·the alternate hypothesis, t hat the population
correlation does not equal zero i s accepted. For t h e pur
pose s of this study, t he null hypothesis was rej ected and
the alternate hypothesis accepted at the five per cent
level of confidence. This means that a correlation value
equal to or larger than the one obta ined occur s by chance
less than five t~mes in one hundred.
In order to deteLliline the correlati on values that
can be accepted or rejected at specific significance values,
the t te~t is used .
-28-
r t =v,...--...,.z.--
1-R N-2
There i s, however, a much s impler method for f inding ou t
whether an ob served value of r is suffi c i ently large to
cause a rejection of the hypothesi s that the population cor
r elati on equals zero. Edwards ' 1 adaptation of Fisher's
table of "Va lues of the Correlation -Coefficient for Dif -
ferent Levels of Signi ficance" was used for this purpose .
From this table, a correla·tion value of . 623 is signifi cant
at t h e one per cen t level -of confidence1 and a value of
.497 i s s i gn i fic ant at the five per cent level. Th e null
hypothesis is rejected and the alternate hypothesis ac
cepted for a correlation value of .497 at t he f ive per cen t
level of con f idence .
In t hi s study the hypothesis that a positive rela-
tionship exi s ts bet\-veen soci ome tr i c status and posi tive
behavior characteri stics is held tenable at t he fiv e ue r ~
cent leve l of confidence for all correlation values above
. 497. -:..s shown in Figure I the correlation values of h e ad
nurse l; staff nurse 4; nursing students 5, 6, 7, 9, 10;
and attendants 12, 13, llj., and 15 are not only significant
at t he five per cent level but also highly s ignific ant a t
l Edwards, op. c i t . , 502 .
-~
-29 -
the one oer cent level . The correlation value of a ttendant • L
11 i s s ignificant at t he f i ve per cent l evel ; while t h e
values ob tained from a ssistant h ead nurse 2, and staff
nurse 3 , are no t considered s ignificant even t hough a
po s i tive r ela t i on ship i s suggested . The correlati on value
of . 876 for all t h e nursing personnel is a lso highl y s ig -
n i f icant at the one per cent level. ·
From the analysis of t hi s data three generalizati ons
c an be made about t he hypothesis t hat was tested . They
are a s follows :
1 . Al l the fourteen personnel preferre d patients who exh ibited socially acceptable behavior charac t eri sti cs.
2. TI1ese preferences were stati stically s ig n i ficant for twelve of t he per sonne l and highly s ignificant f or eleven of t he personnel.
3 . When considered a s a group, the nur s ing personnel exhibi ted a highly s ignificant preference for pat i ents whose behavi or patterns were socially acceptable .
- As dete~~ined by the methods and t h e data used in t hi s
study, the hypothes i s t hat a posi tive relation ship exi sts
between high soc i ometric status and soc i ally acceptable
behavi or characteri stics i s held as tenable.
Analysis of the Data to Test Hypothes i s #2.
The second hypothesis that wa s tested states as
follows: Th er e are behavior characteristic s that di ffer -
entiate patients of high , intermediate, or low soci ometric
-30 -
status. The higher a patient's status, the mo~ e he will
display soci ally acceptable behavior characterist i cs; t h e
lower his status, the more he will di splay less accep table
behavior characteri stics.
In order to test this hypothesis , further refine-
ment of the soc i ometric and behavior rati ng data was neces-
sary . As show in Table 1, the sixteen patients were classi-
fied ·· nto four status groups according to their sociometric
rank. These ranks were taken from the total group cor-
relation in Figure I .
TABLE 1·
ASSIG~~NT OF PATIENTS TO STATUS GROUPS
Patient p F I K N 0 E I G c D A J H M B L
Socia-metric 16 15 ~4 13 12 11 10 9 8 7 6 5 4 3 2 1 Rank
'
Socia- HIGH LOW INTER-metric Status HIGH 1 INTERMEDIATE MEDIATE LOW
A more involved method was used to refine the be-
havior rating data. As mentioned in the previous chapter
the thirty behavior characteristic items were classified
into five groups of six items each. The scores of each
I
-31-
group of items could range from s ix to thi rty. In order to
differentiate t he high from the low scores, a score of
e ighteen was selected as the upper limit for low scores,
and nineteen as the lower limit for the high scores for the
soci ally acceptable Groups ~and B. This process was
s imp ly r eversed f or the socially unacceptable Groups C,
D, and E. In other words, a high score on Groups A and B
indicated t hat a patient ranked high on these character
i sti cs, while a high score on Groups C, D, and E indicated
that a patient ranked low on these characteristi cs . Since
the populati on sample consisted of fourteen nursing pe r son
nel, and each status group included four patients, a maxi
mum score of fifty-six could be obtained for each of the
five behavior characteri s ti c groups .
As shown in Table 2, the high status patients,
t hat i s, the ones best liked by the nursing personnel
scor ed very high on Group A cha r acteristics, and moderately
high on Group B characteristi cs. These preferred patients
appreciated what other s di d for them, were considerate of
the f eelings of othe r s, enjoyed doing things to he lp
others, greeted people with a warm friendly smile, showed
interest in the people caring for them, and enjoyed the
company of others. These patients were also neat, clean,
and well groomed, talked sensibly and intelligently , en
j oyed and participated in soci al functions , wer e alert to
Group
~:~ o· A 0 <.
- :>-. :jo
B ttl ·r-1 G) Or-1 o..a cncu
I Cll c .uO o..:d (l)Cil 0 •r-1 0 l-1 tUG)
D s::-W ~~ • :>-. ~ r-I,.C: r-IU ttl E · r-1 G) Or-1 0 ,.0
Cl)('(j
TABLE 2
RELATIONSHIP BETWEEN SOCIOMETRIC STATUS P~~ BEHAVIOR CH.~~CTERISTICS IN TOTAL SCORES AND PERCENTAGES
'
High Status High Low Title Intermediate Intermediate
Scores % Scores % Scores %
F~iendli-n ss f£yiabil- 55 98 54 96 .4- 19 34-
9n~fi~s~-§rlf -suffi- 43 · ency
1 77 I 43 77 19 3L~
Dependenck I
¥~~tionall 2 3. 5 13 23.2 26 46 .4
ftes~ssiveL Hostility 0 0 0 0 13 23.2
1~bftover-t_~ oc·· al-
tnaaequac 7 12 . 3 15 26 . 8 30 53. 6 I --
Low Status
ScoreJ %
11 19.9
I I
17 30 . 4
37 65 -
13 23 . 2
47 83 . 9
I
I w N I
-33-
things going on around them, initiated and organized group
activities, and shared common interests with others. How-
ever, the differences in scores between Group A and Group B
characteristics is a reflection on the inability of patient
K to initiate and organize group activities.
The important differences between the scores of
the high status patients and those of the high intermediate
status group centered around Group C and Group E character
istics. Whereas the high status patients scored 3.5 per
cent on Group C characteristics and 12.3 per cent on Group
E characteristics, the high intermediate group scored 23.2
per cent, and 26.8 per cent respectively. These differ
ences suggest that even though the high intermediate group
received comparatively similar scores as the high status
patients on the socially acceptable characteristics, they
showed a tendency toward Dependence-Emotionality, as well
a s more Introversion-Social Inadequacy than the patients in
the high status group.
In examining the scores of the low intermediate
status group, a marked difference is noticed from the
scores of the two higher groups. These patients scored
over 60 per cent1 lower on Group A characteristics, and
33 per cent lower on Group B characteristics than either
1These figures represent differences between percentage scores and not percentages of total scores.
of t he two higher groups. A marked t r end is also noticed
in terms of Group C and Gr oup E characteristics. That i s,
progressively lower status i s equated wi th higher scor es
on socially unacceptab l e characteri s tics.
At t hi s point, Group D characteristi cs come into
focus. The patient s in the l ow intermediate status group
showed some s igns of refusing to follow rules and regula
tions, caus ing trouble and di sturbing other s, us ing people
to meet t heir own ends, making unnecessary requests and
demands, and consi dering t hemselves better than oth er s.
Even t hough a score of 23. 3 per cent is not high , it does
serve to further differentiate the pati ent s in t hi s l ovJ
intermedi ate statu s group from the patients in the highe r
groups .
The patients in the low status group r eceived t he
l owest scores on Group A and Gr oup B characteristics, and
t h e highest scor es on the socially unacceptable charac t er
istics of Group C and Group E. In compari son t o the high
status group , the patients in the l ow status group scored
78 . 9 pe 1.· cent lmver on Group A characteri stics, 4 6 . 6 per
cent lower on Group B characteristics, 61.5 pe r cent highe r
on Group C characteristics, 23 per cent higher on Group D
characteristics, and 71.6 per cent higher on Group E
characteristic s.
In comparison t o the high e r statu s groups, these
low status patients had more difficul ty in making up their
-35-
minds, worried mor e about themselves, ·were more grouchy
and i r ri table , f e lt sor rier for themselves, and got mor e
excited and ea sily up set. However, the maj or factor in
differentiating t h e l ow status group from t he oth e r pa
tients \vas in t e rms of the Group E cha r acteristic s. Ac
cording t o the nursing personnel, these least l ike d pati en ts
preferred t o r emai n by themselves, had difficulty in and
di s liked star t ing conversation, kept quiet and in t he back
ground, were suspic i ous of t he motives of other s, did not
mak e friend s easily, and disregarded t he fri endly approaches
of othe r s.
These finding s a r e in a greement with hypothesis
#2 . Soci ometric s ·tatus i s directly proporti onal t o soci
ally acceptable behavi or charac·teristics (Gr oup s A and B)
and inversely proporti onal t o soci ally unacceptable be
havior characteri stics (Gr oups C, D, and E) .
Analysis of the D~ ta to Test Hypothesis #3 .
Combined rank order correlati on s were done t o test
the hypothesi s t hat nurses, nur s ing s ·tuden ts , and attendants,
as members of different occupational groups do n o t di ffe r
s ignificantly in their f ee lings for patients, nor in their
p er cep t i ons of t h e behavior of these pati ents . Th e soci o
metric score s and behavior rating scores f or all member s of
each of the three occupat i onal group s were combined and
r an<:ed from one to s ixteen in a manner s imilar t o t hat u sed
for t he total group corre lat i on.
- 36 -
i s shovn! in Figure II, there was a general agree-
ment among the members of each occupational group concern-
ing t h e r e lation ship bet~veen soci ometric and behavior rat:-
ing data . The correlation values of . 635 f or nur ses, . 705
for a ttendan ts, and . 89L:- for student s are significant at
the one per cent level of confidence . There vJas greate r
intra-group agreement among t he students than among t h e
members of t he other two groups. Thi s was due to t h e fact
that t he indi vidual correlation scores of f i ve of the s ix
students were highly s ignificant at t h e on e per cen t l evel .
Possibly the studen ts, with limited p sychiatric nursing
experience, are more attracted to t h e patients whose
behavior pat t ern s are soci ally acceptable than are t h e
nurses and attendants.
Even though the correlation values f or all t hree
groups a r e significant at the one per cent level, t h e
present task ·was to test the r e lationship be tween t hese
t hree group s. Th.is was done by testing the di fferences
in correlation values between the nurses and attendants,
t h e nurses and s tudents, and t he students and at tendants.
I n order t o test t he s ignificance of the differ-
ence between t hese ·three pair s of correlati on values, t h e
f ollowi ng computations had to be completed . 1 The ob served
l "Edwards, op . cit. , 304-310 .
-.:37-
FIGIJI?/E. /1
R!INI<, OflDER CtHV{EJ./11/0NS OF SOCIOMEI!f/c_ F/NO
8£11/JViOFf Rf11/NG- /JFJTJCJ Ft>~ /VU~ScS, RTT£NDR NTS_, fiND 'STb-dE~TS 8.5 MEM/3~/f~ tJF OII=FE~£NT 0CCuPH7/aNf1L
&fleVP~.
NURSES /~--~~~~~~------------~
STU {)tENT .S
fj }1-
CI) 13 ll! j2_ ~ ,, <:! 10 lfl ~
~ 8 ~ 7 ~ ' ~r 'cl4 ~ 3
2.
V}/3 \u It \\:" ,, 0 v /b V\ ~
~ 8 ~ ;::.. 1 ~b \LI .f ~f
3 2.
1~--~~------------------~ ~~~L---------------------~ I 2.. 3 ~ 5 b 7 g 9 /0 If 12. 13 14- /$" /t I 2.. 3 +5' 7 8 .9 10 II 1213 /<f/S/-6
S 0 C. Itt 1\1 € T/f I C S C D ~ £ .S 5 OC /0 ,NJ ElkJC S (tJRF-5
/lHo .6J..)• /? /I{J • 2 9 4- r
I
13 Cl) /2..
~ II 'Q /4
~3 ~ B '() 7 ~b ~
~~ ~3
l.
/fEND ANI
',--------------------------~ 2 3 1-.) 4 7 8 9 /0 II IZ 13 ff I~ 16
S 0 C 10M E. 'T~ IC. 5 C 0/(E. .5 RHO. 70S X
-38-
correlation values were transformed into z 1 values by using
the necessary stati stica l tables . 1 The standard error of
the differ ence be tween the two z values were determined by
the f ol l owing formula .
SE --;-\ I 1 ,, 1 ; v N - 3 T N - 3
The observed difference between the two values wa s divided
by t he standard err or .
z = Zl I - Z2 I
SE
The significance values wer e determined by entering the
table of t h e normal curve. 2 The results are en tered in
Table 3 . Since hypothesis # 3 is in the form of a null hypo -
the s i s ; that is, that t he observed corre l ati on. values for
the three occupational group s do not differ s ignifican tly,
any probability value that does no t fall within the f i ve
per cent level of confidence will serve to hold the null
hypothesis as tenable. Referring to Table 3, a p value
of . 05 would be necessar y to conclude that the observed
correlati on values differ significantly at the five per
cent level . Since a value as low as . 05 does not appear
1Ibid.' 50 3. 2 Ibid., 490 -499 .
-39-
TABLE 3
Z TRANSFORMATION VALUES AND PROBABILITY VALUES FOR CORRELATION COEFFICIENTS OF THE THREE
OCCUPATIONAL GROUPS .
GROUPS z p
Nurse s - Students 1. 78 . 0750
Students - Attendants 1 .44 . 1498
Nurses - Attendants . 178 . 850
in Table 3 , the hypothesis that nur ses, attendants, and
nursing students as members of different occupational
group s do not differ significantly in their feelings for
patients nor in their perceptions of the behavior of these
patients is held as tenable .
Analysis of the Data to Test Hypothesis i/=L:- .
In order to test the hypothesis that sociometric
status and length of ho spitalization are significantly
related, the patients were ranked from one to sixteen for
-40-
both variables, and a rank order correlation computed .
Table 4 was used to rank the patients according to length of
hospitali zation .
The correlation value of .640 is significan t at the
one per cent level. (Figure III . ) Thus, the hypothesis
that sociometric status and length of hospitalizati on are
significantly related is accepted at the one per cent l evel.
The PHI correlati on coeff i cient1 was used to in-
vestigate the r elationship be tween sociometric status and
discharge rate .
Table 5 was constructed by dividing the patient
group into high and low status, and then clas sifying the
s ixteen pati ents according to wheth er or not t hey had been
discharged .
Using the formula
BC - AD
PHI -1( (A f C) (B f D) (A f B) (C f D)
with a, b , c, d, r eferring to the cell values in Table 5 ,
a PHI correlation coefficient of .500 was obtained. In
order to determine the significance of t he value obtained,
the x2 test was used .
x2 = N (PH1) 2
1Ibid . ' 18 7.
-41-
TABLE 4
DISPOSITION AND TOTAL LENGTH OF HOSPITALIZATION FOR THE 16 PATIENTS INCLUDED IN THIS STUDY AS OF THREE
HONTHS FOLLmVING THE COLLECTION OF THE DATA.
PATIENT DISPOSITION
A Transferred to state hospital
B Transferred to dis-turbed ward
c Discharged
D Discharged
E On ward
F Discharged
G Discharged
H On ward
I Discharged
J On ward
K Transferred to senile ward
L On ward
M On ward
N Discharged
0 Discharged
p Discharged
TOTAL LENGTH OF HOSPITALIZATION
9 months
3 years 3 months
2 years
4 month s
4 years 10 months
5 months
2 month s
6 months
6 week s
9 months
10 months
15 years 3 months
23 months
7 week s
6 weeks
7 weeks
I•
-+ 2. ...
RRNf\ ORPE'R CORR£J..fJ11tJAI oF S6CIOMt:. T{fiC. RRNJ( IJND "-EIVIff/ fJF H•S ,,TAI..Iti'Ji/•.V
l~r-------------------------------------~----~
~, -<l .... ~I . ~ N
~3 ' 'i:t' ~ 2::8 · V)
~1
u..' -Q
~s t-1.6 ~ ~3 -~
2 ,. I 2.. 3 ~ S 6 7 f3 ~ /D II 12. 13 llf- 15" 16
SOCitJ t1t:Tf(IG l?f.J/YI-( SCORE.S RHo • ~ 1-0ir
-43-
TABLE 5
2x2 CONTINGENCY TABLE OF SOCIOMETRIC STATUS AND HOSPITAL DISPOSITION .
HIGH STATUS LOW STATUS
Discharged
Not Discharged 2
2
6
The resulting value , x2 = 4 . is s ignificant at the five per
cent level of confidence . 1 s determined by the data used
in this study, length of hospitalization and discharge dis
position are s ignificantly related to sociometric status .
Discussion of Findings .
The findings of this study are in agreement with
the hypotheses tested . In general, the nursing personnel
preferred the patientswho se behavior patterns deviated as
little as possible from the behavior of the individuals in
the outside world . The se preferred patients r e turned to
the outside community after a comparatively short period
of hospitalization. In contrast , the non-preferred patients
1Ibid . ' 500 .
were characterized by deviant behavior pa tterns and by
l onger periods of hospitalization.
For example, patient P was best liked, and exhibi
ted the mo st acceptable behavior characteristi cs. He
. enter ed the hospital because of a mild depression and some
psychoneuroti c complaints. He i s a proud, successful
bus i ness man, a good organizer, and a man of many inter
es ·t s . He was di scharged after seven ·week s of hospi taliza
tion . Pati ent F and patient I were also he ld in high
esteem by t h e nur s ing personnel. They were a dmi tted t o the
h ospital because of acute psychotic reactions, but qui ckly
recovered and were convalescing at t h e time of this s tudy .
These t\vO patients a r e both young , handsome, consi derate,
and outgoing . Patient I remained in the hospi tal f or six
·week s, and patient F was di scharged after five months .
There was general agreement among the nur s ing per
sonnel concerning the non-preferr ed patients, especially
in relation to patient L . He rece i ved the l owes t status
scores, and the second lowes t behavior rating scor es.
Patient L was characterized as 11 totally unreceptive to the
friendly approaches of the pe rsonnel. 11 Such a s t atement is
a r eflection of fifteen year s of conditioning to hospital
life .
Patient B r anked lowest on the behavi or rating
scores, and next to lmvest on the soci ometric scores . He
-45-
was preferred by only one member of the nursing staff,
nurse 3. She was better acquainted with him than were the
other personnel . Since patient B seemed incapable of. form
ing 'Vvarm, friendly relationships with the personnel, all
but one nurse refrained from socializing with him .
The nursing personnel a greed that patient H felt
sorry for himself . They also agreed on hi s low soc i o
metric status and his undesirable behavior patterns. Even
though patient :r-1 sought support and warmth from others, hi s
low status suggests that he did not receive it .
Patient H was characterized by his complete inabil
ity to relate to people , socially and physically, and hi s
expressions of hopelessness for the future . His low socio
metric status was a reflection of h.i s inability to relate
to people .
These brief profiles suggest that the ability to
relate to others is a most important factor in achieving
high sociometric status on a mental hospital ward . Thi s
conclusion is reinforced by the differences in scores re
ceived by the high status and low status patien ts on Group
A and Group E characteristics . In general, the status
positions assigned to t h e patients in the psychiatric com
munity by the nursing personne l served to differentiate
them in t heir interpersonal affective behavior pattern s .
Gaining a high sociometric status position in the soc i al
-46 -
system of a psychiatric hospital ward means exhibiti ng be-
havior u seful toward living in the outside world.
Even though personal pre ferences and individual
fe e lings must be given proper consider ati on in nur se patient
r e lationships as well as in any other area of human r ela-
t ions, t h e fact r emains ·that all p sychiatric patients a r e
en t i tled to an op t imum amount of nursing care . A stated
goal in psychiatric nursing " is t o help the patient work
t hrough his problems by providing learning experiences
that give him new tools and a new perspective with which
to meet them . "1 The interpersonal environmen t of the
hospital war d should offer the patient the experience of
being accepted and r espected as a person in hi s own ri ght .
Since soci al development and ability i s so important a
part of clinical improvement and prognosi s, i t i s t h e f unc-
tion of t he nur s ing personnel to u t i lize all opportunities
t o aid t h e patients , especially t h e isolated ones , t o f orm
meaningful soc i al relation ships .
According to t he findings of this stu dy , an oppo-
site t r en d i s suggested. The patient s who needed t h e most
ass i stance in establishing relati on ships with others \·Jer e
t h e ones who were avoided , r ej ected, and ignor e d .
Ther e is no easy answer to t h e problem involved
in provi ding men t ally ill patients wi t h t he soci al
lnorothy E. Gregg , "The Psychia tric Nur ses Role, " American Journal of Nur s ing , LIV, (July 1954) , 848 .
re l ati onship s t hey r equi re . However , the real i zati on t hat
t h e h ospi tal comnunity i s structured by attractions and
r ej ect i ons growing out of the populati on's i ntimate asso
c i ati on wi t h one another , and t hat these attr acti on s and
r ej ecti ons are directly related to t h e social adequacy of
t he behavior patterns of the i ndi v i duals wi t hin t h e com
muni t y , i s t he ini t i al step i n t h e process of experimenting
\vi t h \vays of creating a t herapeuti c environmen t in wh i ch
all pa t i ents can move towar ds r ecover y.
CH/-.PTER V
SUH1Y1ARY , CONCLUSIONS , AI\TO RECOHMENDATIONS
Summary .
Sociome tric data obta ined from sociome tric t e sts
and behavior data ob t aine d from r ating sca les wer e u·til i zed
to investiga t e the relati on ship be tween the f ee ling r eac-
tions of nurs ing per sonnel f or pati ents, and t he ir pe rcep -
·tions of the behavior of the se pat i en ts . Th e data we r e
collected from the f ourteen nurs ing pe r sonne l a s signe d to
a sixteen bed , male n on -di s turbed p sychiatric ward . In
order t o de t ermine the existance of t hi s rela t i on ship , f our
hypotheses wer e tested . Th e y a r e as f o l lows:
1. ! positive r elationsh i p exi s t s bet'l.veen high soc i ometric status and socially accep table behavior cha r acteristics .
2 . There are beh avi or characteri sti cs that dif f er entiate patients of high , i ntermedia t e, or l ow soci ometric s tatus . TI1e highe r a patient 's sta t u s, the mor e he \vill di splay soci a lly ac ceptable behavi or characteri s tics ; t h e lowe r hi s s tatus, the mor e h e tvill di splay les s ac ceptable cha r acteri s tics.
3 . ~n1en con s ider ed a s membe r s of differ en t occu pationa l group s , nurses , a ttendants, and nur s ing s tudents do not di ffer s ignificantly in t he ir f ee lings f or pati ents, nor in t h e ir perceptions of t h e beh avior of t hese pa tien ts.
4 . A po sitive r e l a tion ship exi s t s bettveen sociometric s tatus and l engt h of hospital i zation .
-48-
-L:-9-
Conclusions .
These conclusions are not necessarily applicable to
oth er mental ho spital populations, but are strictly con-
fined to the group studi ed h er e . The conclusions may be
valid for similar groups ; but no such extrapolation i s ex-
plicitly made, and the results can only be regm:-ded as
tentative, pending further research . These conclusions are
as f ol lows:
1 . a) Each of the fourteen personnel preferred patients who exl1.ibi ted socially acceptable behavior characteristics .
b) These preferences were statistically significant f or twelve of the personnel, and highly s ignificant for eleven of the personnel , at the one per cent level.
c) 1/Jhen considered as a group, the nursing personnel exhibited a highly significant preference for the patients whose behavior patterns were socially accep·table .
2 . As determin ed by t h e five group s of items on the behavior schedule, sociometric statu s is directly proporti onal to socially acceptable behavior characteristics (Groups A and B) , and inversely proportional to soci ally unacceptable behavior characteristic s, (Groups C, D, and E) .
3. When considered as members of different occupat i onal group s, nurses, attendants, and nur s ing student s do not differ s ignificantly (five per cent level) in their feelings for patient s, n or in t he i r perceptions of the behavior of these patients.
L: . • a)A positive relationship exists betvveen sociometric statu s and length of hospitali zation . This relationship i s s ignificant at the one per cen t l evel .
-50 -
b) A positive relationship exi sts between sociometric statu s and di scharge di sposition . This relationship i s s igni f i can t at the five per cent level.
In general , the nursing pe r sonnel preferre d the
patients vJhose behavi or pa tte rns devi ated as lit t l e as pos-
sible from the behavior of the individuals in t h e out s ide
world . These prefe~ red patient s returned to the outside
conmmni ty after a comparati vely shor t period of hospi tal-
i zation . In contra s t , the non-preferred patients we r e
charac·ter ized by deviant behavior patterns and by l onge r
· eriods of hospital i zati on .
Re co1nmendati on s .
1. That t he methods u sed in this stu dy sh ould be applie d to othe r mental hospital wards in order to check on t h e val idity and reliability of the finding s of thi s study .
2 . That con siderati on shou ld be given to t h e use of s imilar social science techniques for the purpose of investiga ·t ing and evaluating t h e effectiveness of nurse-patien t · r elati onship s on mental hospi tal war ds.
3. That the applicati on of statistical methods to the study of inte r-persona l r elations in p sychiatric nursing should be given fur t her con s idera·tion .
-51-
APPENDIX A
TABLE 6
l'.NALYSIS OF AGE, DIAGNOSIS, AND LENGTH OF
HOSPI TALIZATION OF THE 16 PATIENTS INCLUDED
IN THIS STUDY
Patient Age Diagnosis Length of Hospitalization at Time of Study
A
B
c D
E
17
31
28
16
29
Transient Si tuati onal 6 month s Personality Disorder
Psychosis with Psycho- 3 years pathic Personality
Catatonic Schizophrenia 22 months
Adolescent Turmoil 7 weeks
Paranoid Schizophrenia L:. years, 7 month s ·--------------------F 26
G
H 30
I 26
J 75
K 75
Schizophrenic Reaction , 2 month s Paranoid T)~e with Depression
·----------------------~----~---------Manic Depressive Reac- 6 week s t i on , Depressed Type
Personality Disorder, 3 months Schizoid Personality
------~~--~----------Acute Shizophrenic 5 weeks Reaction
------·----------------~-------------Chronic Brain Syndrome 6 month s Associated with Cerebral Arteriosclerosis and Neurotic Reaction
Chronic Brain Syndrome 7 months Associated with Senile Brain Diseas e
---- ----·- - ·---------------------------54 Hebephrenic Schizophrenia 15 years L
-- -M 63 Manic Depres s ive Reac- 20 mon·ths
t ion , Depressed Type
(Concluded on next page)
Patient Age
N 20
0 55
p L:.
-52-
TABLE 6 (Concluded)
Diagno s i s Length of Hospitalization at Time of Study
Transient Situational Personality Disorder
Psychoneurotic Reaction, Depre ss ive Reacti on
Psychoneurotic Disorder, Depressive Reaction
4- week s
5 week s
-53-
APPENDIX B
TABLE 7
ANALYSIS OF AGE , OCCUPATIONAL STATUS , SEX, i\ND LENGTH OF TIHE ON WARD, OF THE lL:- NURSING PERSONNEL INCLUDED
IN THIS STUDY
Sta·tus Age Sex Length of Time on ~lard
1. Head Nurse 26 F 7 months
2 . Ass ' t . Head Nurse 30 M 2 . 5 months
3 . Staff Nur se 3Lj. F 15 mon t h s
4 . Staff Nu r se 38 ··· F 11 mon t h s
5 . rJur s ing Student 20 F 1 month
6 . Nur s ing Student 20 F 1 month
7 . Nur sing Student 20 F 1 month
8 . Nur s ing Student 20 F 1 month
9 . Nursing Student 20 F 1 month
10 . Nursing Student 2Lj. r-1 1 month
11. Attendant 21 M 7 months
12 . Attendant 24 H 6 month s
13 . :\ttendant 23 H 1 month
1L: . . Attendant SL~ H 5 months
-54-
APPENDIX C
CLASSIFICATION OF THE 30 BEHAVIOR CHARACTERISTIC ITEMS
INTO 5 GROUPS
Item Number Item
Group A -- FRIENDLINESS-SOCIABILITY.
4. 10. 15. 18. 24. 30.
Group B
1. 7 .
12 . 21. 27. 29.
Group c
3. 8.
13. 19.
23. 25.
Group D
2. 6 .
11.
Appreciates what others do for him. Is considerate of the feelings of others. Enj oys doing things to help others. Greets people with a warm friendly smile. Shows interest in t he people caring for him . Enj oys the company of others.
GREGARIOUSNESS-SELF-SUFFICIENCY.
Is neat, clean, and well groomed. Talks sensibly and intelligently. Enjoys and participates in social functions. Is alert to things going on around him. Initiates and organizes group activities. Has many interests which he shares with others.
DEPENDENCE -EMOTIONALITY.
Ha s difficulty in making up his mind. Worries about himself and possible misfortunes. Is grouchy and irritable. Is unwilling to help himself, wants others to do things for him. Gets excited and easily upset. Indulges in self pity, f eels sorry for himself.
AGGRES SIVENESS-HOSTILITY.
Refuses to follow rules and regulations. Goes out of his way to cause trouble, and disturbs others. Uses vile and profane language
-55-
Item Number Item
Group D -- AGRESSIVENESS-HOSTILITY (Continued)
16. Uses people to meet his own ends, plays one against the others.
20 . Considers himself better than others. 26 . Makes unnecessary requests and demands .
Group E INTROVERSION-SOCIAL INADEQUACY
5. Prefers to remain by himself rather than with other s.
9. Has difficulty in, and dislikes starting conversation.
14. Keeps quiet and in the background. 17 . Is suspicious of the motives of others. 22. Does not make friends easily. 28. Disregards the friendly approaches of others.
P,PPENDIX D
PERSONPL-SOCIAL CHARACTERISTIC BEHAVIOR SCHEDULE
The following statements describe some of the
'-'Jays that r.ve, as nursing personnel from our observations
of, and our interactions with patients perceive their be
havior . We are ilTte:Lested in the way you see the patient
as a r esult of your experiences on this particular ward .
You will see that each item is followed by the five fig
ures P, U, SO , SE , N. If you feel that the personal or
social behavior characteristics described in the item
always applies to the patient, c ircle the letter A; u sually ,
U; sometime s, SO; seldom, SE; never, N. All statements
a:r-e to apply only to t he patient whose name is on the
schedule . You vJill be given a separate schedule for each
of t' 1e other pati en ts . Remember we are interested only
in how you a s a person see t hese patients .
A Always
u Usually
so Sometimes
SE Seldom
N Never
-56-
-57-
Fame of Pati ent . Code :fj: . -----
1. Is neat, clean, and well groomed A u so SE N
2 . Refus es t o f ollow rules and r egula - A u so SE N t i ons.
3 . Has difficulty in making up hi s mi nd . A u so SE N
4 . Appreciated what othe r s do foJ..~ h i m. A u so SE N
5 . Pre f ers t o r ema in by himself t han A u so SE N wi·th othe :. s .
s . Goes ou t of hi s r..vay to cause t r oubl e A u so SE N and di s turb others.
7 . Ta l k s sensibly an d i ntelligently . A u so SE N
0 tvorries about h imself and possible A u so SE N o . mi sfor t une.
g . Has di f ficul t y in , and di sliJ:es A u so SE N starting conversation::::
1 · . I s con sider ate of t he feel ing of A u so SE N oth.er s .
11. Uses vile and profane langua ge . A u so SE N
12 . En . oys and p a r t i c ipates i n soc i al A u so SE N functi on s.
13 . I s grouchy an d irr i table . A u so S'"' N
14-. Keep s quiet and i n t he background . A u so SE r
15 . .:.'..nj oys do ing t hings to he l-y others . A u so SE N
16 . Uses peop le t o meet h is own ends, A u so SE N plays one a gainst ·the o·ther
17 . Is su soici ou s of t h e motives of A u so SE N o thei.·s.
18 . G_eets peop le ~·Jith a ·war m, f riendly A u so SE N smile .
- 58-
19 . I s un~1illing to helD himse lf , wants A u so s N other s to do things for him.
20 . Considers himself better than oth ers . u so SE r1
21. Is alert to things going on aroun d him. A u so SE N
22 . Does not make friends easily . A u so SE N
23 . Gets excite d and easi ly upset. A u so SE N
2L: .. Shows interes t in t he people caring A u so SE N f or him .
25. Indulges in self pity , feels sorry A u so SE N for himself .
26 . Mal'es unnecessary demands and re- A u so SE N quests .
27 . Initiates and organizes group A u so SE N ac tivities.
28 . Disregards t h e friendly approache s A u so SE N of others.
29 . Ha s many interests which h e shares A u so SE N with others .
30 . Enjoys t he company of o thers . A u so SE N
- ------
-.09-
RP,t!VD)X E lfJBL..E 8
1<. FJ N K £ D S 0 c I o M E T R t' c... fl N D B £ H fJ V I tJR R R 7 /N lr D 19 T A ftJf< THE 16 P fJ71ENrs 115 D£7 elf 1'11/fFP PY THE ,,_NvR:>ovG Pt=/fs6/I/I'(EL
I PnTJENTS
A B C. D ~ f. F G H I :r ~ L M N o P
J... ~c;END: p~t~ o/'INEJ,.. RRE JDeHiiFJ~D Rct-oRPIN& 76 i9Pf>~NJJJJt B PAllflfT~ flf?E IJJEN1JFIEI) IJccoRDI#&ItJ ~PPFNbi.X IJ
s 0CIOME.1RIC.. ANI> BEHIJVit>l( tA711'1(r 5~6RES 1}~£!?. 1/li'IKED .1/11 fit 19/3/fl ;oPG- f/(611 HlfrH i o ,J..Dip . 1m. uPP~R HRLI!- ()F ~l'fcH, C ELJ.. R£ P~E SE.NTS SOC!IOMf!.l~l'e. SCIJ~~~ u,HH£ THE }.$UJE~ HA~F RIPI(f. SEN1S 8£Hf:JVIt>~ friiiiNC,. S c •R~S. Fo~ ~X A Mfltf) 1Hl TOTAh S C&IES R' C.£ lu r=o 8 y PA71EN T P /IVIJIC II~~ 'IHJIJT HE. u,19~ THE. 13 ~S -r LJK~D PRIIEHT, fiND AJ.SO E1!f18JT~D 1"HIE f1tJS-r 5 OCtAI...J..y )9C,C~PTJ48J...E. BE HI9VI61~ CHJfJRI#Jt:.- ' I IICS.
•60-
P PEN D!X E:. /fsURf. IV
l..JNE191{ RE/,.fJTI~/'I.SHIP Bf"/k!E£1'-/ flf3NJ(ED S6CIOM£11/C FINO BEHIJVItJ~ RIJT//Y6- DIHFI FoRTH£ 16 PFJi IeNTS Fl~ DeTeRMIIYeP B)' Til£ If- NIJ/f':!J//1/6- P~--I?~OI'I.NEL
PFJTJfiiT A p IFNT B
I ,, /0
:'1~ ~& ~7 vb V)
I I I
t ' \ I \ ' I '
~ \1 ~' 2 ~ ~
I I 2 3 1-5 6 7 B 0 /IJ 1/ 12..13 ,.
3 -2.
NtJ~SII(fr PE'RSO/'fNt:l. RHo . I 84
I I z 3 + s l> 7 8 9 II II ll.. 13 fl- .
NU~ S/Nc;. P£:1( S()N/'1£1.. RHO ,.2.b0
GRrtPHS REP~ESEN/ Dr:JTA oF RPPEND/X E. IRJ3J.E 8
I l'l.
'I lb
V)g ~ s Q:: 0 7 o, "'5' ,..
3 2 .@,
I I 2.··3 -,_ S '-1- B-! ltJ--II 12. ·,3 H
N VR S//Y G- P~lf S.OIY /VEL ~ HO -.380
ll.t--__.P:.._.;,.:;II...:..T..:..;I E~-H....;.....;._T--=D---~ I /(I-
I 14 ' II 'o -0
V) 16 \4~
~8 U1 {'Jb
5 4 3 2 I I
I 2 "3 <J- 5 6 7 S 0 /IJ II 12.13 I~ /Vuf< S)/'16-- PE~ ~DN/f€ i-.
RHO .710 I...~C.END:
• 50C:./OME71?1' SCoRES 81 fillY I Ott< s c: 0 RES
- f> I -
F I G u R E I V C. t:. t/~) It/ v f P
PfJ11EN1 E ~r---~----------------~ 15 ~ 13 I
" V) /0 , ~ 9 ' yJ ~ a 'e--e" ~7 Col')6
5 4 3 2.
\ I \ I ()
,.Gl ,. \
(/1 \ I ~
I
I I I I I I I
\ I 'I
~ I \ I \
I ' I \ \ \
~ .. "
I I 2- 3 q. S ~:, 7 8 9 10 II 12. 13 l<f-N lJ~ SIIYG- PE~ SD/V NE l
Rf-lo • 62.0
/b·~--~~--------------~
15 I I I
" fJ} /0 ~.9 <c::g ~
t~1
"6 5 1-3 ;. 1~--------------------~
I 2. :3 4f- .5 6 7 S 9 JtJ II ll. 13 14-IV UR 5/NG- PE ~ S6HNtL.
Rti0.980
12. II
'-'? )() UJ-' ~8 ~7
"'" 5 t, 3 2
PfJI/ENi F
1~--------------------~ I 2. 3 4- 5 6 7 8 9 It> 1/ l'l- 13 14
NOR 311\16- PER .S ol'll'l~l. f?HO.B60
PI/TIENI H ,,~--~~----------------15 I 13 12..
'r C/J /IJ '..l..l ~
~8 t.l7 "'~>
5 / q.' .3 2.~.-
1~------------------~~ 1 2 3 tJ- 5 6 7 a s /() '' 12 13 ''~
Nu ~ S./N G PEitStJNNt 'RfoiO .700
. .
FtGuf?E tV CoNTti\IUPt;;
~~~=P..~~=-,_..:...,.K"'""ijF--......r-----?.t
15 14 13 12. I I
,_, /0 \41-9 ~8 ~ 7 V)b
.5 4 3 2 J~· ---------------4
I 2. ~ 4 6 6 7 B ~ 10 I/ 12 13 14-N tJR St#& P£,.,.StJNNEL
f{HO ·''-1
I I 2 3 4- 5 b 7 8 9 /tJ II 12. 13 l'f IV vR SI;V6- Pt~SINN~L.
~ Ho .5,0
Pfl71~NI ~ /~.~-~----------------
I ,, V) /0 ~9 ~Q 8 v7 "'c.
5 4 I
3 4c 1L-11L-----------~
I 2.. 3 <1- ~- 6 1 8 9 /~ 11 IZ. 13 ltf-
I 1'2. ,
~ /6 u.i~ ~8 ~1 V)f,
s <+
I
3 / 2.
N VII.. SIN & F£RS6NN£1-R H() . 600
P8TIEIVT L
•
I I 2. 3 4 6 b 1 6 9 /6 It I 2. 13 I,_.. . /Vt.J~ S/Nf:T PFflStJIVNEL
RHO , 82.4
- 63-
F I G U R £ I V C ol'l TIN u E 1J
PRT ~NT M
POT/EN/ 0 '~~~-=~~~~------~
" t) I () /\ I} (.'ti ~ ,~ : \
. ,, / \ VJ ,. \ij9
~8 t:..J1 '?~
5 + 3 2.
· 1~----~----~------~ I 'Z... 3 tf S l> 1 S 9 /fJ II 12..13 It-
If vi( S //Y 6- P £1( S(j"f I'( fL.
R IJO .1- 8 8
I , /0
VJg
~" '07 v, <.1)5
4-3 2
'o, ~
P'JTIENT N
• I
'~ I \ I
' I \ I
I I
~
' -I 2. 3 ~ 5 6 7 8 S /~ II 12. 13 14-
,, Cl] /0
~3 ag tj7 VJ'
s 4 3 2.
h uP. SIN& PE/156~/VEJ.. RHo.s9s-
1~------------------~ ' 2.. 3 +- .s-" 7 e s lfJ ,, 12..1"> ,,_
IV vi( SINo P~lfSoA/N£1.. f(HO .7&7.
BIBLIOGRAPHY
Ba les, Robert F., Interaction Process Analysis, Cambridge: Addison-Wesley Press, Inc., 1951.
Barker , Roger G., 1'The Social Interrelations of Strangers and Acquaintances," Sociometry, V (February 1942),
' 169-179.
Black, Kathleen , "Agpraising the Psychiatric Patient's Nursing Needs,' American Journal of Nursing, LII (June 1952), 718-721.
Blake , Robert, "The Interaction-Feeling Hypothesis App lied to Psychotherapy Group," Sociometry, XVI (August 1953) , 253-265.
Blake, R. R., and Ramsey , G.V., Perception, An Approach to Personality, New York : Ronald Press, Inc., 1951.
Bonney, M. E., Hoblit , A. H., and Dryer, A:H., "A Study of Some Factors Related to Sociometric Status in a Men's Dormitory," Sociometry, XVI (November 1953), 287 -293 .
Botts , H.M., "Observations of Play Activities in a Nursery School, " Genetic Psychological Monographs, IV (July 1928), 44-88.
Cartwright, Derwin, and Zander, Alvin, Grout D~amics, Evanston, Ill.: Row, Peterson & Co., 95.
Cattell , Raymond, Descriptions and Heasurement of Personality , New York: World Book Co., 1946.
Caudill, W., Redlick , R.C., Gilmore, H.R., and Brody, E.B., "Social Structure and Interaction Processes on a Psychiatric Ward," American Journal of Orthopsychiatrx, XXII (1952), 314-334.
Caudill, ~if . , and Stainbrook, E. , "Some Covert Effects of Connnunication Difficulties in a Psychiatric Hospital, " Psychiatrx, XVII (1954), 22-43.
Criswell, J.H., "Sociometric Methods of Measuring Group Preferences," Sociometrx, VI (1943), 398-408.
- 64 -
-65-
Devereux, George, "The Social Structure of a Schizophrenic Ward and Its Therapeutic Fitness," Journal of Clinical Psychopathology , VI (1944), 231-265.
Edwards, Allen L., Statistical Methods f or the Behavioral Sciences, New York: Rinehard & Company, Inc., 1954 .
Friedman , M.B., and Leary, T.F., "The Interpersonal Dimension of Personality , " Journal of Personality , XXIII (1954), 129-154.
GAP Report No. 33 , "Therapeutic Use of the Self , 11 Group for the Advancement of Psychiatry, 3617 W Sixth Ave., Topeka, Kansas, June, 1955.
Gregg, Dorothy E. , "The Psychiatric Nurse 's Role , " American Journal of Nursing , LIV (July 1954), 848-850.
From Custodial to New York:
Hare, A. P. , Borgatta , E.F., and Bales, R.F., Small Groups, New York: Alfred Knopf Co., 1955.
Romans, George, The Human Group, New York: Harcourt Brace & Co ., 1950.
Jenkins, Elda H., "Interpersonal Relat ionship in Psychiatric Nursing," Unpublished Master.' s Thesis, School of Nursing , Boston University, 1949 .
Jennings, Helen H. , Leadershi9 and Isolation, New York : Longmans, Green & Co., 1 50 .
Kandler, Harriet M. , "A Study in the Elements of Rapport, Unpubli shed Master 's Thesis, School of Nursing, Boston University, 1950.
Kandler, H. , Behymer , A. F., Kegeles , S. S., and Boyd, R. W., "A Study of Nurse-Patient Interaction in a Mental Hospital , " American Journal of Nursing , LII (1952), 1100-1103 . . .
Kraus , Stefan , "Considerations and Problems of Ward Care for Schizophrenic Patients," Psychiatry , XVII (1954), 283-292.
-66 -
Lemann , Thomas B., and Solomon , Richard L., "Group Characteristics as Revealed by Sociometric Patterns and Personality Ratings, " Sociometry, XV (February 1952) ' 7-90.
Mode , Elmer B., Elements of Statistics, New York : PrenticeHall, Inc., 1951 .
Morimot o, Francoise R., "Favoriti sm in Personnel-Pati ent Interaction, " Nursing Research , III (February 1955) , 109-112.
Morino , J.L., Who Shall Survive, New York : Beacon House Press, Inc., 1953.
Northway, Mary L., "Outsiders, A Study of the Personality Patterns Least Acceptable to Their Age Mates," Sociometry, VII (February 19.!}4), 10-25.
Peplau , Hi ldegards, Inte~ersonal Relationships in Nursing , New York: Putnam & o., 1952 .
Riod1 , David, and Stanton, Alfred H., "Milieu Therapy, " Psychiatry, XVI (February 1953), 65 -72.
Rowland , Howard, "Friendship Patterns in the State Mental Hospital, " Psychiatry, II (August 1939), 363-373.
Sears, R. R., "Experimental Studies in Projection," Journal of Social Psychology, XVI (1936), 151-163.
Stanton, Alfred H., and Schwartz, MorrisS., The Mental Hospital, New York: Basic Books, Inc., 1954.
Szurk , S. A., " Dynamics of Staff Interaction in Hospital Psychiatric Treatment of Children," American Journal of Orthopsychiatry, XVII (1944), 652-664.
Tudor, Gwen, "A Sociopsychiatric Nursing Approach to Intervention in a Problem of Mutual Wi thdrawal on a Mental Ho spital Ward, " Psychiatry, XV (1952), 193-217.