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.