"MY
W1

This is to certify that the
dissertation entitled
The Relationship Between Cognitive
attributions of Loneliness and Coping
Responses to Loneliness in University
Students
presented by

Shelley B. Smithson

has been accepted towards fulfillment
of the requirements for

Ph .D . degree in Counseling Psychology

4571/ / 1/

a M ajor professor

Date (/3/87

“HP“"LL‘é‘. . n- ”L .i. inn 042771

 

4;...“ fl WP-

 

 

 

M86 leURNING MATERIALS:
Place in book drop to

LIBRARIES remove this checkout from
.agszzggle. your record. FINES will

 

___, be charged if book is
returned after the date
stamped below.

 

 

 

 

‘a ..‘ [I / ’O'
I» y""»fir¢«~é -- ‘3 3 9

’h

 

 

 

 

 

 

 

 

THE RELATIONSHIP BETWEEN CDGNITIVE ATTRIBUTIONS OF
LONELINESS AND OOPING RESPONSES TO LONELINESS
IN UNIVERSITY STUDENTS

By

Shelley B. Smithson

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

[DCTOR OF PHILOSOPHY

Department of Health Education. Chunseling Psychology,
and Human Performance

1986

PBSTRACT

THE RELATIONSHIP BETWEEN COGNITIVE ATTRIBUTIONS or
LONELINESS mo COPING RESPONSES TO LONELINESS
IN UNIVERSITY STUDENTS
By

Shelley B. Smithson

This dissertation looked at the relationships between attribu-
tions of loneliness and manners of coping with loneliness in univer-
sity undergraduate students. Loneliness was entered into the research
design as a third variable of interest. with attributions and coping
behavior investigated in terms of degree of loneliness being experi-
enced. The sample consisted of students who were both lonely and not
lonely. Criteria were established to investigate hypotheses both
among the total sample of students (83 males. l16 females) and among a
Lonely subsample (l9 males. 30 females). In general. the total stu-
dent sample demonstrated a range of loneliness typical of student
populations seen on other large university campuses in the United
States. Attri butions of causality and stability of loneliness were
measured using the theoretical model of Weiner.1 Coping behavior was
investigated according to the paradign of Lazarus and Folkman.2 in

which coping is seen as either emotion-focused or problem-focused in

orientation.

Shelley B. Smithson

Hypotheses predicting relationships between one form of coping or
another depending on the attributions of stability and causality of
loneliness were generally not supported. Coping behavior was not
found to be predicted in the directions hypothesized. nor was coping
behavior strongly related to attributions of loneliness. Correla-
tional data and analysis of variance suggested that attributions that
were unstable and external were found to have a moderate and signifi-
cant relationship with emotion-focused coping. suggesting that indi-
viduals are mcwe prone to self-comforting coping strategies when they
are feeling that the problem of loneliness is changeable and tempo-
rary. This was seen to a further degree with cluster analyses. These
analyses suggested that coping was high for moderately lonely students
with external and unstable attributions for their loneliness. whereas
students who were very lonely or slightly lonely and showed greater
tendencies toward stable and internal attributions for their loneli-
ness engaged in significantly less coping behavior. This study sug-
gests that perhaps those students who are most lonely are more apt to
see their situation in negativistic terms and feel less motivated to

seek change.

 

1B. Weiner. Achiexement motixation and attfltution thenrx
(Morristown. NJ: General Learning Press. l974).

2R. S. Lazarus & s. Folkman. fitness. appraisal. and coping (New
York: Springer. 1984).

This dissertation is dedicated to my husband. Jim Heavenrich.
whose unending support. encouragement. and humor were most sustaining

in my completion of this project.

ACMGVLEEXSLENTS

I wish toiextend my appreciation and gratitude to the following
individuals for their respective contributions to this prodect.

To my doctoral dissertation committee members: Chairperson
Dr. Richard.Johnson. who provided me'with guidance. encouragement. and
thoughtful academic feedback from the proposal stage through the final
draft of this dissertation; Dr. Rochelle Habeck. who spent consider-
able time helping me proceed with my conceptual izations and organiza-
tion of the data; Dr. Steve Raudenbush. for the numerous statistical
and design consultations he provided for me: and Dr. Dozier Thornton.
who engaged me in most thoughtful reflection over the concepts of the
study and the instrumentation to be used. These four individuals each
expressed in his or her own way sincere interest in my work and
provided invaluable instruction throughout this process.

In addition. acknowledgments are extended to the following
people: Mary Haas. Acting Director. Michigan State University
Residential Housing Programs: Lynn Snellman. M.A.. for her fine
computer and statistical instruction and continual availability:

Dr. Gregory Holmes. for serving as a valuable role model and inspiring
student who preceded me in this same doctoral program: and Mary Bowen.
for her many hours of time and typing skills in the preparation of

preliminary drafts of this dissertation.

Finally, I wish to acknowledge my appreciation of my parents,
Barbara and Paul Smithson, whose values on the pursuit of knowledge

have been a most important source of information for me.

vi

TABLE OF CONTENTS

L IST 0F TBLES O O O O I O O O O O O O O O O O 0

Chapter
I.

II.

III.

STATEMENT OF NEED AND INTENT OF STUDY .

The Dissertation Study . . . . . . . .
Hypotheses O O O O O C O O O O O O O 0

OVERVIEW OF LITERATURE . . . . . . . . .

Definitions of Loneliness . . . . . .
Incidence and Prevalence of Loneliness
Theoretical Orientation to Loneliness

Correlates of Loneliness . . . . . . .

Loneliness and Social Skills . . . . . . .

Loneliness in Adolescence and Early Adulthood

Review of Attribution Theory . . . . . . .
Loneliness and Attribution Theory . . . .

Review of Theories of Coping Response

EmommY O O O O I O O O O O O O O O O 0

Description of the Sample . . . . . . . .
Procedures . . . . . . . . . . . . . . . .

Rationale for Selection Criteria for Lonely

SUbJeCtS 0 O O O O I O O O O O O O O I O O
Instrunentation..............
Student Information Sheet . . . . . . . .

Revised UCLA Loneliness Scale .

Causal Attribution and Stability Attribution

Scales . . . . . . . . . . . . . . . . . .
Attributional Style Questionnaire . . . . .
CopingInventory..............
Depression Adjective Checklist . . . . . . .

Coopersmith Self-Esteem Inventory--Adult Form

“SignOfthGStUdyeeeooeeeeeoeeo

vii

Page

TO
13

l7

l7
T9
21

33
37
In
51
54

65

67
68

72
74
75
75

77
79
82

85

Analyses . . . . . . . .
Post Hoc Analyses . .
Summary . . . . . . . .

IV. RESULTS . . . . . . . . .

Results on Main Variables
Locus of Stability . .
Locus of Causality . .
Loneliness . . . . . .
Coping Behavior . . .

Hypotheses Tested . . .
Hypothesis 1 . '
Hypothesis
Hypothesis
Hypothesis
Hypothesis
Hypothesis
Hypothesis
Hypothesis
Hypothesis

Pest Hoc Analyses

OGNO‘UT->WN

Differences Between Lonely and Nonlon

Tendencies in Attribution Sty

6

I g o o o o O o o o o .40 o o o o

Subsanples............

Post Hoc Analyses on Ceping Behavior

Cluster Analyses . . . . . . . . . .

Discussion . . . . . . .
Summary . . . . . . . .

V. SUMMARY AND CONCLUSIONS .

Summary of the Study . .
Conclusions . . . . . .

Differences in Ceping According to
Styles . . . . . . . . . . . . . . . . . . . .
Differences Between Emotion-Focused and Problem-
Focused Coping According to Attributional
Group . . . . . . . . . . . . . . . . . . . .
Attribution and Loneliness . . . . . . . . . . .

Further Notions From Cluster Analyses and Other

At

1

o o O O o 0 O o o o o o 0 o e o o o O o o o o
o o o o o‘< o g o g o . o o o o o o O o o e o

tributional

POSt HOC Ana‘yses e 0'. e e o e e o e o e o 0

Discussion . . . . . . . . . .

Implications for Future Research . . . . . . . . .

viii

Page

87
87
88

9O

93

93
100
102
105
107
109
109
113
119
120
121

125
126
129
130

135
137
143
149
152

154

154
157

157
158
159
161

162
176

Pa ge
APPEN DI CES

A. INSTRUCTIONAL LETTER MD (DNSENT INFORMATION . . . . . 186
B. INSTRJMENTS AND QUESTICNNAIRES USED IN STUDY . . . . . 189
C. (ROSS-TABULATION INFORMATION (N LONELINESS TEASURES . 207

EFERENCES 0 O O O O O O O O O O O O O O O O O O O O O O O O O 211

Table

4.1

4.3
4.4

4.5

4.6
4.7
4.8

4.9

4.14

4.15

LIST OF TABLES

Reliabilities and Frequency Statistics: Total Sample .

Reliabilities and Frequency Statistics: Lonely
SUbsam p] e O O O O O O O O O O O O O O O O O O O O I 0

Pearson Correlations: Total Sample . . . . . . . . . .
Pearson Correlations: Lonely Subsample . . . . . . . .

Pearson Correlations: Attributions With Depression and
se] f-Estem O O O I O O O O O O O O C O C O O O O O 0

Pearson Correlations: Attributions With Loneliness . .
Mean Coping Scores . . . . . . . . . . . . . . . . . .
Proportion of Samples in Attributional Groups . . . . .

Model of Sunmary Data Tables for Analysis of Variance:
Mean Scores for Coping Response . . . . . . . . . . .

Summary of Results From Analyses of Variance (Total
smp16) O O O O O O O O O O O O O O O O O O O O O O 0

Summary of Results From Analyses of Variance (Lonely
SUbsmp1e) O O O O O O O O O O O O O O O O 0 O O O 0

Summary Table of Mean Scores for Emotion-Focused Coping
Behavior by Attribution (Total Sample) . . . . . . .

Summary Table of Mean Scores for Problem-Focused Coping
Behavior by Attribution (Total Sample) . . . . . . .

Difference Between Emotion-Focused and Problem-Focused
Coping (Total Sample) . . . . . . . . . . . . . . . .

Summary Table of Mean Scores for Emotion-Focused Coping
by Attribution (Lonely Subsample) . . . . . . . . . .

Page
94

95
96

97

98
99
105
108

110

112

114

115

115

116

117

4.16

4.20

4.21
4.22

4.23

4.24

4.25
4.26
4.27

4.28
4.29
4.30

Sunmary Table of Mean Scores for Probl em-Focused Coping
by Attribution (Lonely Subsample)

Difference Between Emotion- and Probl em-Focused Coping
(Lone1 y SUDsamp16) I O O O O O I O O O O O O O I O O O

Loneliness (UQA Scale) by Attributional Style: Total
smp1e O O O O O O O O O O O O O I O I O O O O I O O O

Loneliness (UQA Scale) by Attributional Style: Lonely
Subsample

Loneliness (UCLA Scale) by Attributional Styl e--Total
Sample (With Locus of Causality Measured by Causality
Question on Student Loneliness Questionnaire)

Regression of Variables on Loneliness: Total Sample . .
Regression of Variables on Loneliness: Lonely Subsample

Attributional Style by Loneliness (Revised UCLA Scale):
Tota‘ smp1e O O O O O O O O O O O O O O O O O O O 0 O

Attributional Style by Loneliness (Revised UCLA Scale):
LonelySubsample...................

Comparisons of Lonely and Nonlonely Subjects . . . . . .
Coping by Attributional Style

Emotion-Focused Coping by Attri butional Style: Total
smp1e O O O O O O O O O O O I O O O O O O O O O O O O

I-Tests: Comparison of Lonely and Nonl onely Groups . . .
Cluster Analysis of Variables

Cluster Analysis on Observations: Cluster Means . . . .

xi

Page

118

123

127

128

129
131
133

134

135
137
138

139
141
145
MB

G-iAPTER I

STATEMENT OF NEED AND INTENT OF STUDY

Loneliness is a widespread phenomenon in our society. Loneliness
affects people of all ages and encompasses a range of emotions and
situational components. Although there are different causes and
courses to the experience of loneliness. the common denominator to all
people who describe themselves as lonely is that they suffer from a
longing for satisfying social contact that is missing. Although
loneliness is frequently associated with social isolation. al oneness.
and solitude. it differs from these states in that loneliness refers
to the specific state of longing for either more social contact with
others or social contact of a more rewarding nature. Whereas social
isolation objectively describes an individual's social life. and
aloneness and solitude both describe conditions of social isolation or
separateness. loneliness pertains to a painful feeling of desiring
what currently is not being experienced.

Most people experience some form of loneliness at one time or
another in their lives. Loneliness is a culturally defined phe-
nomenon. given that each culture establishes standards for the desired
level of social contact and involvement. Certain cultures foster a

greater degree of loneliness in their members than do others. The

American culture has been cited by numerous theorists as a lonelier
culture due to the emphasis on autonomy. independence. materialism.
individualism. and competition (Moustakas. 1961: Sadler. l979; Weiss.
l978). Thus. forces in American culture contribute to difficulty
achieving social ease and comfort. In all cultures. primary attach-
ment needs and interaction needs deriving from mutual dependence and
desire for companionship result in specific ways in which people form
relationships and cohesive groups. Loneliness becomes a concern when
an individual subjectively feels a persistent feeling of nonattachment
or lack of inclusion where such involvement is desired. Thus. lonel i-
ness is not an objective state. but rather a subjective one that can
only be accurately reported by the subject. versus accurately detected
by an observer.

There are many different causes for loneliness. as well as
different courses of intensity and duration of loneliness. The word
lonely typically carries a negative connotation. conj uring up images of
unhappiness. dejection. sadness. and desperation. Most people can
think of individuals they have known who always appear lonely. They
are suffering from what appears to be a chronic condition. There are
also individuals who report going through a particularly lonely period.
or who identify certain circumstances that can result in a temporary
bout of loneliness. Particular roles in life seem to carry higher
risks for loneliness. such as the full-time mother whose children have

all grown and left home. or the small-town minister whom all confide

in. and yet has no one s/he can turn to for solace and comfort. always
feeling under the scrutinizing eyes of all.

Loneliness also changes in intensity and affect. depending on the
duration of loneliness and the etiology of the loneliness. One person
may feel acutely lonely for a short period of time. whereas another
may experience some sensations of loneliness. only to become more
chronically lonely and unhappy as no easy solutions are found to
remediate the difficulty. One person may describe him/herself as
always having been lonely. suggesting either a characterological
predisposition to feeling apart from others or from life. or
encounters with persistently difficult circumstances.

There are sudden life changes and gradual transitions that can
foster loneliness. given the right set of ingredients. Widowhood.
imprisonment. illness. and occupational demise can cause potent
feelings of loneliness in those who are described a such.

The subjective experience of loneliness cannot be isolated from
the cultural context in which it occurs. In a society such as ours.
which is highly oriented toward individual achievement and autonomy.
there are strong undercurrents encouraging people to forego comforts
of social closeness or dependency for the cultural values of success
and performance. In addition. there are prescribed conditions for
normal social adjustment. thus establishing certain social norms that
are identified as desirable to achieve. In a society that typically

does not encourage people to be alone and even uses isolation as a

common form of punishment in the rearing of children. Americans have
the tendency to be frightened of being alone.

Loneliness connotes a sense of deprivation and desperation. Thus.
loneliness is an internal experience that signals to an individual
that his/her object world is deficient in a manner that creates
emotional distress (Gordon. l976). Remedying the experiences of
loneliness can involve making changes in whom one rel ates to. how one
seeks friends. or in the internal expectations one has for social
intimacy and involvement. Peplau and Perl man (l98l. p. 3l) further
defined loneliness as the internal perception and effect of believing
one's social contact is deficient. They believed that cognitions play
an important role in the self-description of loneliness. Cognitions
can affect how people make contact with others and how people appraise
their relationships. thus affecting the level of satisfaction or
dissatisfaction one may feel with his/her relationships.

There are a number of theoretical approaches to individual and
social development that address loneliness. Developmental theory of
the self (ego psychology) addresses the inevitable experience of
loneliness in young children during the process of ego formation. when
separation from symbiosis results in a "psychological bi rth" as a
separate person (Mahler. l975). This theory recognizes the possibil-
ity of current life events reawakening these archaic feelings of
loneliness. Eriksonian theory (Erikson. 1950) recognizes ongoing
developmental tasks within life-cycle stages that lead to the forma-

tion of a cohesive identity. implying that failure to successfully

accomplish stage-specific tasks can result in lonely feelings. Social
psychologists look at the interaction between individual traits and
social milieu as resulting in lonely periods for people. and cognitive
theorists believe in the importance of understanding cultural norms of
cognition and intraindividual differences in cognition as defining
what constitutes loneliness for an individual.

The range of approaches to loneliness suggests that there is a
continuum of normality to pathology in which one's loneliness can be
viewed. Most individual/social developmental approaches recognize the
inevitable experience of loneliness in the growth process. and some
theorists have addressed the importance of loneliness as part of the
process of maturation (Guntri p. 1968; Mijuskovic. 1980; Moustakas.
1961: Sullivan. 1953; Winnocott. 1958). Moustakas (1961) believed
that loneliness accompanies life transitions. decision making. and
loss. and saw loneliness as a normal experience. providing the
opportunity for growth and creativity. He stated that loneliness only
becomes pathological or harmful if one's resources are such that
coping with loneliness or keeping it in perspective are seriously
diminished.

Considering the universality of the experience of loneliness.
there has been surprisingly little attention to loneliness in psycho-
logical research. Most people are familiar with lonely figures in
fictional literature. such as Robinson Crusoe (MW Defoe.
1720/1960). Holden Caulfield in Salinger's W (1961).
or Harry Haller. the "Steppenwolf" in Hesse's W (1929).

Loneliness has been a major themeein literature and continues to have
a place among the themes in contemporary works. However. the recogni-
tion of loneliness as a worthy area for psychological investigation
has been relatively recent. Fromm-Reichmann (1959) attributed the
dearth of research in this area as due to a "collusion" among
researchers to avoid a subject that almost invariably is a reminder of
unpl easantness to most people. Thus. there is a need to further the
research efforts that have begun over the past 10 or 15 years.

This researcher looked at the problem of loneliness among the
college-aged population. Studies have indicated the prevalence of
loneliness to be higher among adolescents and young adults than any
other age group (Brennan & Auslander. 1979; Ostrov & Offer. 1980).

This study looked at the relationship between the attributions
college students make of loneliness and the coping behaviors they
report using during periods of loneliness. Due to the fact that
college students are entering a phase of their lives in which their
cognitive capacities are encouraged and drawn upon to provide
explanations for many new phenomena they face. it is*worthwhile to
learn more about the nature of their cognitive styles in relation to
reactions to the experience of loneliness. Research on the coping
behavior of college students who are lonely is very limited. but
preliminary research has demonstrated that cognitions students have
about their loneliness seem to influence their course of loneliness

(Cutrona. 1982: Paloutzian 8. Ellison. 1980).

In learning more about the interrelationships among attributions
regarding loneliness and responses to loneliness in undergraduates.
more information can be gained as to whether cognitive attributions
play an important role in experiences of loneliness and in determining
coping behavior. as well as looking at the types of coping behaviors
used. Further value from this type of study is learning about
improved methods of intervention with lonely students. as well as
gaining insight into preventive resources for students to help
ameliorate negative aspects of loneliness. The goal of this type of
research is to help researchers and clinicians understand what types
of coping styles are effective for people. and whether the way one
appraises a lonely experience influences the use of adaptive coping
responses.

This study served as a first step in this inquiry by investi-
gating the usefulness of understanding coping behavior as influenced
by cognitive attribution. A previous study (Revenson. 1981) looking
at the relationship between attributions about loneliness and coping
responses to loneliness served as a foundation from which the present
study was developed. The study itself is referred to later in this
paper as it represents the effort on the part of its researcher to
relate the constructs of causal attribution and coping according to
recent directions in loneliness research.

Attribution theory has promoted research in looking at the

relationship of attributions to emotions and behavioral responses.

In the following section. literature regarding some important studies
relevant to this area of research is discussed.

Attribution studies of achievement motivation have provided
meaningful data indicating that explanations of locus of causality and
stability are associated with certain emotions and expectations of the
future. With regard to negative events. internal/unstable attribu-
tions are associated with feelings of control and positive expecta-
tions for future outcomes. Internal/stable attributions are asso-
ci ated with depressed affect and low motivation. External lstable
attributions are associated with lack of control and depressed affect.
Finally. internal/stable attributions are believed to be related to
depressed affect and low expectancy of positive future outcomes
(Peplau. Russell. & Heim. 1979; Selignan. 1979; Weiner. Russell. 8.
Lerman. 1979). In the application of attribution theory to lonel i-
ness. it can be hypothesized that. dependi ng on the attributions made
regarding the causes and permanence of the lonely situation. coping
will either be more focused on palliative measures or on taking action
to do something about the problem. A person may not go about con-
sciously identifying causes and ascriptions of changeability to these
causes. but at a more automatic level of thinking. these assumptions
people make are all the more powerful.

Cutrona (1982). Peplau and Perlman (1982). and Revenson (1981)
all looked specifically at attributions of loneliness in college
students and the relationship of causal attributions to coping

responses. The initial works of Cutrona and of Pepl au and Perl man led

Revenson to design a study to look at the relationship between
Weiner's attributional styles and the reported coping responses with
loneliness. The coping model that lends itself to this research is
Lazarus and Folkman's (1984) emotion- versus problem-focused coping.
which is based on the premise that coping is effortful behavior (ver-
sus reactive) in the sense that it is behavior called upon in the
service of mobilizing against psychological stress or distress.
Although people may not actually be consciously aware that they have
engaged purposively in particular responses to contend with stress. it
was Lazarus and Folkman's contention that a person can identify coping
behaviors when asked to. and that people have a sense of how responses
(whether thoughts or behavior. per se) are serving to protect them or
help them ward off the stressor.

Revenson (1981) studied the relationship between attributions of
causality and stability of loneliness and coping responses (emotion-
versus probl em-oriented). She hypothesized that people with internal/
unstable attributions would select more problem-oriented coping
responses than emotion-oriented responses. She also hypothesized that
people with external/stable attributions for loneliness would select
more emotion-oriented coping responses than probl em-oriented responses
because of the lack of personal control or sense of change implied in
that cogiitive style. Revenson used the Ways of Coping Inventory
(Lazarus 8. Cohen. 1980) to measure coping behavior. having adapted
some of the items to read more appropriately for loneliness. Revenson

found results that were in the predicted directions. but only found

10

one significant interaction effect of stability and causal attribution
with type of coping responses endorsed. She also found that respond-
ents with internal attributions. regardless of stability attributions.
endorsed more of both types of coping responses.

Revenson stated in personal communication that her study served
as a good starting point for further studies that can be conducted to
look at the same relationships. Revenson did not use a sample of
people who were all lonely. She recruited volunteers (.N = 137)
between the ages of 18 and 49 and administered the New York University
Loneliness Scale to them. She then asked the subjects to think of
times when they had been lonely and to indicate their coping responses
on the Coping Inventory. Thus. the data collected were from a sample
ranging from nonlonely to very lonely people. and the data were

retrospective in many cases.

W
The present study looked specifically at the relationships among

attributions of loneliness. coping responses to loneliness. and degree
of loneliness within a sample of undergraduate students. Locus of
causality and locus of stability were the attributions of interest
with respect to attributional style. and coping response was regarded
within the problem- and emotion-focused coping paradign of Lazarus and
F01 kman (1984). The sample of subjects included students who were
lonely as well as students who were not lonely at the time the
research was conducted. Students provided 561 f-report data with
respect to their current state of loneliness. their attributions about

11

causality and stability of loneliness. and their manner of coping with
loneliness.

The study furthered the questions examined by Revenson's (1981)
work by expanding on her research in a number of ways. A more select
sample was chosen (instead of studying a general adult population».
Subsamples were studied of lonely and nonlonely students within the
whole sample. and hypotheses were expanded or in somercases altered.
A coping instrument designed specifically for coping with loneliness
was used to measure coping responses. a departure from the previous
research cited. The state of loneliness and attributions of loneli-
ness were also measured with different instruments. which this
researcher identified as having more sound psychometric integrity and
greater applicability to the research. Depression and self-esteem
were two additional variables included in the research design to allow
for the possibility of further explanations or clarifications of
results obtained and to have a fuller description of the sample of
subjects. Methods of analyses were extended to include more correla-
tional analyses and cluster analyses for descriptive purposes.

A critical contribution to this research was the performance of
analyses on a subgroup of lonely students. This provided for a
greater understanding of the experience of loneliness and the nature
ofiattributions of loneliness and coping with loneliness among people
currently experiencing loneliness. The formation of subsampl es of
lonely and nonlonely subjects allowed for comparisons along a number

of interesting variables. In addition. it was seen as crucial to

12

identify a group of students who were providing current information
about themselves. in order to add strength to a study that did have
some students providing retrospective data in the whole sample. Thus.
for purposes of linear regressions and correlations. there werelnore
powerful analyses that could be run with the use of the large N of the
total sample of subjects. whereas to further explore individuals. the
subsample provided data that were reported from current experience
rather than reported in part from memory.

Thus. the underlying conceptualization of the study was that a
particular attributional style about causality and stability of
loneliness would lead to a more probl em-oriented or more emotion-
oriented approach to coping1with loneliness. Attributional style was
conceptualized within Wei ner's framework of locus of causality and
locus of stability. in which locus of causality is seen as either
internal or external. and locus of stability is seen as either stable
or unstable. This framework thus yields four attributional styles:
internal/stable. internal/unstable. external/stable. and external/
unstable. Thus. it was conceptualized that attributional styles would
relate toia certain level of each type of coping response in compari-
son to the other attributional styles and that. within each attribu-
tional style. differences between amount of problem-focused and
emotion-focused coping would be seen.

The hypotheses concerned the strength of relationships between
the variables of interest and the comparison in coping behavior

between subjects of differing attributional stances toward their

13

loneliness. Additional hypotheses concerned the relationship
between the degree of loneliness and attributions of loneliness and
coping with loneliness. The following sections discuss the rationale
further for the hypotheses and then state the hypotheses for the

study.

Hxnotheses
The value in looking at the relationship between attribution and

coping with loneliness was in learning about the effect of attribu-
tion with regard to an emotionally upsetting and often ambiguous
condition. More insights into the relationship between attributions
and coping patterns can lead to further knowledge regarding ways of
intervening with attributions to enhance coping or can lead to knowl-
edge regarding the relative effectiveness of different coping behav-
iors. The hypotheses that follow are concerned with the relative use
of emotion- or problem-focused coping. depending on the explanations
and cognitive considerations students had regarding their degree of
loneliness. The hypotheses thus address the degree to which students
made use of problem-solving or palliative-oriented strategies during
lonesome periods.

It is important to learn how to predict the ways that individuals
are likely to cope with loneliness because a particular kind of coping
behavior can lead to successful or unsuccessful resolutions of a
problem. Both emotion- and probl em-focused coping are seen as vital

to adequate coping. and loneliness is viewed as a particularly

l4

troublesome condition for coping by virtue of the painful ness. anxi-
ety. and social stigna associated with it. Problem-solving strategies
are necessary in order for an individual to feel a sense of mastery
and self-efficacy in dealing with loneliness. leading to less antici-
patory anxiety of future states of loneliness and allowing the indi--
vidual to take on certain opportunities or risks that an individual
might otherwise shy away from. Not making adequate use of emotion-
focused strategies can also result in problematic adjustment. in that
the individual who copes predominantly by "taking action" on the
problem does not increase the sense of tolerating uncomfortable emo-
tions without needing to "do something" about them rigit away. An
individual might decide that coping with loneliness is best achieved
by signing up for a social-skills class. reading a self-help book. and
looking at all of the positive ways to view loneliness. but if s/he
cannot sit down and have a good cry. the use of the problem-solving
strategies may be at the expense of not integrating the discomfort
s/he feels. This discussion is important to bear in mind as the
"backdrop" for the research hypotheses that follow.

Wham: Students having stable attributions of their

loneliness (as measured by the Grant Attributions Scale) will

show more use of emotion-focused coping than students who have

unstable attributions about their loneliness (measured by the

Coping Inventory).

This hypothesis was based on the assumption that stable attribu-
tions are associated with low expectations. lack of hopeful ness. and

lack of belief in change. The same measures apply to the following

hypotheses.

15

W: Students having stable attributions of their
loneliness will use less problem-focused coping than students
having unstable attributions about their loneliness.
W: Students who have both stable and internal
attributions of their loneliness will use more emotion-focused
coping than students who engage in any of the other three com-
bined attributional styles toward loneliness (stable/internal.
sta bl e/e xternal. unstabl e/external ).

This hypothesis was based on the assumption that students who
have stable/internal attributions about their loneliness should feel
the greatest degree of emotional distress. including feelings of self-
bl ame. hopelessness. and low motivation to make change.

W: Students who have unstable/internal attributions

of their loneliness will engage in more problem-focused coping

than students representing the other three attributional styles.

This hypothesis was based on the assumption that these students
should feel a sense of control over a phenomenon they see as change-
able.

The following two hypotheses concern differences in coping

within attributional groups.

W: Students who have stable attributions of their
loneliness will use more anotion- than probl em-focused coping.

This hypothesis was based on the assumption that stability is
associated with a lack of expectation that things can change. thus
discouraging the use of problem coping and encouraging palliative.

comforting measures of coping.

W: Students who have unstabl e/internal attributions
of their loneliness will show more use of problem-focused than
emotion-focused coping.

16

This hypothesis was based on the assumption that people with
unstable/internal attributions will feel a sense of motivation and
personal control over problems.

The last group of hypotheses concerns the relationship of
loneliness to attributions about loneliness and coping with loneli-

T1955.

11131911125151: Students who have stable attributions of lone-
liness will experience greater degrees of loneliness than will
students who engage in unstable attributions about their loneli-
ness (loneliness as measured by the UCLA Loneliness Scale or the
Likert scale on loneliness on the Loneliness Questionnaire).

W: Students who have stable as well as internal
attributions of their loneliness will experience the greatest
degree of loneliness compared to students who engage in other
attri butional approaches to their loneliness.

W: As the degree of loneliness increases. the use of
emotion-focused coping will increase. based on the earlier pre-
diction that students who have stable and stable/internal attribu-
tions about their loneliness will have a greater tendency to use
emotion-focused coping compared to other students.

O-IAPTER II

OVERVIEW OF LITERATURE

QeflnitjanoLLmeJJness

Loneliness is a common experience to many people. and yet is
found to be a neglected area in research for being such a pervasive
problem. In addition to the already mentioned aspect of loneliness
being a rather untouched field of inquiry due to the difficult feel-
ings it can arouse in people. there are also conceptual issues that
result in loneliness being a less adequately defined topic for inves-
tigation. Therefore. in approaching the study of lonely students. it
is necessary to adhere to the preference of the more prominent
researchers in the field. who have chosen to look at loneliness as an
identifiable andlneasurable phenomenon. which islnultidimensional in
nature. Although there are different causes. courses. and etiological
viewpoints on loneliness. the attention to the ever-present complaint
of lonely individuals that a social sense of belonging or having inti-
macy is miSsing provides for the following as a workable definition of
loneliness.

Peplau and Perlman (1982) defined loneliness as "an unpleasant
experience that occurs when a person's network of social relations is
deficient in some important way. either qualitatively or quantitat-

ively" q» 31). Although these researchers respected a diversity of

17

18

emotions that people label when feeling lonely. the aspect of special
deprivation is always the key element. Weiss (1973) called this the
relational deficit in lonely individuals. and he identified two
subtypes of loneliness-~emotional loneliness and social isolation.

He stated that emotional loneliness is the profound sense of
abandonment one feels when lacking an intimate relationship. He
described social isolation (social loneliness) as the feeling of
isolation that exists for an individual who may have a few special
ties with people. but does not fit into a social network for contact
and affirmation.

Sullivan's (1953) definition of loneliness emphasized the need
for intimacy. He stated that "loneliness . . . is the exceedingly
unpleasant and driving experience connected with inadequate discharge
of the need for human intimacy. for interpersonal intimacy" (p. 290).
Sullivan's theoretical thinking was embedded in a biological/psychi-
atric foundation. and thus his definition emphasized the instinctual
need that he saw human beings driven to satisfy. Fromm-Reichmann
(1959) viewed loneliness in a similar light. She believed that there
is a universal need for intimacy that stays with people from infancy
onward.

Definitions of loneliness may vary in the degree to which they
emphasize objective versus subjective aspects of being lonely. and the
degree to which they identify developmental versus social/cognitive
factors in explaining vulnerability to loneliness. What they all have

in common is (a) the acknowledgnent that people have an intrinsic need

19

to have close. satisfying contact with other human beings; (b) lone-
liness alerts to a longing for what currently is not being experienced;
and (c) loneliness is accompanied by distressing feelings.

Loneliness is differentiated from depression. a state in which a
person feels hopelessness instead of the longing associated with
loneliness. It is also distinguished from bereavement. which is the
sense of sadness over someone (or something) that has already been
lost. Loneliness is rather a reaction to the absence of what is

desired or cherished.

IncidenmndflmalencLoLLoneJJness

Survey research studies have indicated that loneliness truly is a
major problem for many people in our society. Sermat (1980) reported
that among a number of studies he conducted with college student and
adult populations. he found that no more than 1 or 2% of respondents
11m: experienced loneliness. while between 10 and 30% stated that
they had pervasive feelings of loneliness. with up and down fl uctua-
tions. during significant portions of their lives. The most prevalent
reason given for being lonely was a serious breakdown in intimate.
personal communication with other people.

Weiss (1973) reported that in a survey study he did on the gen-
eral American adult population. 26% of the respondents in his sample
answered in the affirmative to the question of whether they had felt
lonely during the previous 2 weeks. In a telephone survey conducted
by Wei 55. he found that 14% of all women and 9% of all men asked if

they were lonely in the past week responded in the affirmative.

20

Marital status was shown to be an important variable in the studies
conducted by Weiss. Of the subjects he interviewed. 27% of unmarried
females reported feelings of loneliness. and 23% of unmarried men
reported loneliness as well.

Lynch (1975) explored the psychosomatic aspects of loneliness and
concluded that loneliness that becomes chronic actually influences an
individual's risk status. The most compelling research cited in this
book is that of the incidence of divorce or bereavement in the history
of cardiac patients. Lynch demonstrated that individuals who suffer
from chronic loneliness. defined by Lynch as a gnawing desire for
companionship and/or grieving for the absence of a particular person.
had five times the probability of developing disabling cardiac disease
as other people. hence the title of his book. W. Lynch
believed that the autonomic responses of the sympathetic nervous
system are overly worked throughout the psychological state of loneli-
ness. resulting in a vulnerability to vascular dysfunction.

More typical studies on loneliness have focuéed on the loneliness
of particular subpopulations. Lopata (1969. 1973) did some extensive
fiel d-research studies of widows and found that 48% of one sample of
widows reported loneliness as the leading problem following the death
of the spouse. and another 22% of the respondents reported that lone-
liness was an important issue for them.

Peplau. Bikson. Rock. and Goodchilds (in Peplau 8. Perlman. 1982)
have studied loneliness in old age. Many older adults. especially

those who are widowed. live alone. Interestingly enough. only 15% of

21

older men and women living alone report being lonely. Although 15% is
still a sizable proportion of an age group. nevertheless. this is
contrary to the impression most people have that the vast majority of
people living alone are lonely. There researchers emphasized the fact
that al oneness cannot be equated with loneliness. that older people
often have accrued social relationships over time that have special
and enduring meanings for them. and that people in later life have a
reservoir of memories and life experiences that provide for a particu-
lar satisfaction or fulfillment that can make a1 oneness a quieting.
tranquil solitude versus gnawing isolation.

A large-scale study by Rubenstein and Shaver (1979) of adults
between the ages of 18 and 87 found that there was a significant
inverse relationship between loneliness and age. with young respond-
ents therefore most at risk for being lonely. Other researchers have

found similar results (Blau. 1973; Dyer. 1974; Lowenthal. Thruner. &

Chiriboga, 1976).

IbeocethaLQnientationJLLoneJJness

Loneliness is not a one-dimensional experience. Rather. a
variety of sentiments and emotions accompanies the experience of
loneliness. Rubenstein and Shaver (1982) surveyed American adults who
were lonely. trying to ascertain the range of feelings and strategies
for dealing with loneliness. Their study was based on information
from 3.500 respondents. The most common feelings to be mentioned by

lonely individuals were sadness. depression. boredom. sel f-pity. and

22

longing to be with one special person. A factor analysis of these
feelings resulted in four predominant feeling categories: desperation.
depression. impatient boredom. and sel f-deprecation. These authors
also factor analyzed the reasons for being lonely that respondents
provided. and found five factors: being unattached. alienation. being
alone. forced isolation. and dislocation.

Weiss (1973) described the symptoms of loneliness. as based on
whether an individual is suffering from emotional or social isolation.
The individual plagued by social isolation is apt to feel out of place
and may feel lonely for friends to share common interests with. This
kind of lonely individual is inaccessible to affirmation of others and
may report feeling lonely even when supposedly surrounded by friends.
This person usually is emotionally engaged in a search for activities
and/or a search for a network or group in which to be accepted as a
menber.

The symptoms Weiss described for emotional isolation have a more
piercing. driving quality. He stated that the symptoms are reminis-
cent of the child who has been abandoned.

We might reasonably suspect that the loneliness states of adults

are developmentally of the earlier childhood states. They may

have been modified by the new strengths and understandings of
maturation. but still they seem like the childhood syndrome in

fundamental ways. (p. 20)

In emotional isolation. Weiss identified a sense of pervasive appre-
hension. inability to concentrate. over-vigilant behavior. and diffi-

culty in rel axing. This kind of loneliness can create a sense of

hrpersensitivity to minimal cues in the environment. which can

23

otherwise feel barren and desolate to a person who feels devoid of
intimate ties with anyone.

In reading Weiss's rather poignant descriptions of lonely people.
one is struck by similarity of the symptomatology he described with
the nature of loneliness as explained by ethol ogist/object-rel ations
theorist John Bowl by (1973) and ego psychologist Margaret Mahler
(1975) in their developmental perspectives on loneliness. Bowlby
looked at loneliness as an emotional experience that has survival
functions. In the development of emotional cues relevant to the
survival of the species. loneliness would have served to signal sepa-
ration from the group. leading to proximity-seeking behaviors and
hel p-eliciting behaviors. The over-autonomous person would have been
at risk for protection and safety. and would have represented a loss in
support and contribution to the group's needs. Much of the vigilant
and desperate tone to loneliness symptoms connotes a sense of hyper-
arousal and fear. feelings that ultimately motivate a person to move
out of the situation bringing on such feelings.

Mahler's (1975) seminal work. WWI.
paid great heed to the stages of ego development. which hinge on the
successful mastery of a hierarchy of continuous stages evolving from
an autistic and then symbiotic relationship to the caretaker to a more
individuated and separate tolerance of existence. This successful
individuation and separation process leads to a more secure identity
and a reliance on an internal feeling of safety. warmth. and related-

ness to carry one through during times of minimal quality contact with

24

others. This realization of object constancy. althougi alluded to in
her writings as a stage-specific task. is actually a lifelong process
of growth as people endure repeated separations. losses. and transi-
tions througiout life.

Moustakas (1961) contended that loneliness is an integral aspect
of human existence. in part due to our developmental origins. and that
coping with loneliness in a successful manner involves drawing inward
to further know one's loneliness and sense of pain. He promoted a
mastery of loneliness via this inward. contemplative process. in much
the same way as other existential thinkers. He believed that cultural
constraints against being alone and isolated foster an anxiety about
loneliness that impedes the positive growth that can be realized by
succumbing fully to the experience of loneliness.

Social psychologists have emphasized the role that society and
cultural norms play in the creation of lonely feelings (Lamont. 1979;
Moustakas. 1961; Polansky. 1979; Zimbardo. 1979). Although these
theorists recognized the underpinnings of loneliness within a context
of ego-developmental structure. they nonetheless believed that the
most potent forces for allowing lonely feelings to regerminate or
reappear are due to social factors that. in essence. interact with
one's own emotional or social vulnerability. Such factors that are
identified were (a) social indicators. such as marital status. income.
education. and occupational status; (b) communal or kinship systems:

(c) socio/political aspects of society. such as economic pressures and

25

standards: and (d) cultural norms regarding privacy. social networking.
sel f-disclosure. and processes of social adjustment.

Social cognitive theorists look toward individual variation in
cognitive process (attribution. appraisal. discrepancy models) to
identify how broader social forces affect people differently (Beck.
1976: Jones. Hansen. 8 Smith. 1980; Ostrov 8. Offer. 1980; Selignan.
1975).

Two questions regarding individual differences in experiences
with loneliness are: To what extent do intrapersonal factors play a
part in causing someone to become lonely? And. why is it that some
people may have bouts with loneliness. whereas other people seem to
suffer from a chronic loneliness? Looking at studies that have inves-

ti gated personality correlates to loneliness sheds light on these

questions.

ConnelateLoLLoneJJness

One of the most consistent findings in research on loneliness is
that lonely people often feel worthless. incompetent socially. and
unl ovable. Studies have explored the relationship between loneliness
and sel f-appraisals. loneliness and attitudes toward others. and social
skills in varying social settings.

A number of studies have demonstrated significant correlations
between loneliness and sel f-esteem. Low sel f-esteem may be associated
With loneliness either as a causative factor and/or as a result of the
experience of loneliness (Wood. 1972). From either perspective. sel f-

GSteem affects one's feelings of acceptability. one's expectations of

26

others. and one's level of comfort in social settings. Much of the
research on loneliness and sel f-esteem has been conducted with student
samples. although some studies have involved specific subpopulations
of other types.

Loucks (1980) in his work on construct validity on the Bradly
Loneliness Scale found loneliness related to variables such as low
self-esteem. self-criticism. and uncertainty of self-view. Brennan
and Auslander (1979) found loneliness related to sel f-pity in a group
of adolescents. as well as to feeling rejected by one's parents and
feeling unpopular among peers. Barret and Becker (1978) identified
feelings of sel f-derogati on and negative emotions among lonely widows.
Rubenstein and Shaver (1980) found that sel f-depreciation. including
feelings of being unattractive. feeling stupid. and feeling ashamed
were correlates of loneliness in a large-scale study with an adult
sample. Pal outzian and Ellison (1979) found a correlation of -.57
between sel f-esteem (high) and loneliness as measured by the abbre-
viated loneliness scale.

Rogers's (1961) work on sel f-congruency provides some insights as
to the potential for self-esteem problems to lead to loneliness. His
theory of the self discusses loneliness as an estrangement between a
person and hi s/her true inner feelings. Rogers contended that people
develop facades for feeling acceptable to others. and yet feeling
untrue to themselves. the definition of low sel f-esteem for Rogers.
Thus. the feeling of discrepancy between true and outer self can lead

to an alienated. or lonely position with oneself. Lowenthal et a1.

27

(1976) reported greater loneliness among older men who believed they
saw themselves as different from the ways others viewed them. lending
some support to this notion of sel f-discrepancy leading to loneliness.
Belcher (1973) reported a significant relationship between loneliness
and a Q-sort measure of sel f-ideal discrepancy (I = .75). Moore
(1972) also reported that lonely students have higher discrepancy
scores between self and reflected sel f-concepts than do nonlonely
students.

Russell. Pepl au. and Cutrona (1980) found a correlation of -.49
between scores on the Revised UCLA Loneliness Scale and the Texas
Social Behavior Inventory. a measure of social sel f-esteem. This study
was conducted on undergraduate college students. Other studies have
shown a correlation between low sel f-esteem and loneliness among under-
graduate students (Jones. Freemon. 8. Goswick. 1981; Jones. Hanson. 8
Smith. 1980; Moore 8. Sermat. 1974; Paloutzian 8. Ellinson. 1980;
Russell. Peplau. 8. Ferguson. 1978). Cutrona. Russell. and Peplau
(1979) found that self-esteem was an important factor in whether new
college students experienced only initial loneliness at the outset of
the school year or whether freshmen reported loneliness to have endured
over a 7-month period in which a survey on loneliness was conducted.
Students with low sel f-esteem were more likely to report persistent
feelings of loneliness after the year was in progress. whereas students
with higher sel f-esteem appeared to make friends and make an overall

more positive adjustment to school.

28

It is believed that social competence is related to self-esteem.
given that low sel f-esteem can hinder one's efforts to relate success-
fully to others. Zimbardo (1977) contended that a person with low
sel f-esteem is more likely to be passive. overly sensitive to criti-
cism. and more socially anxious. Loneliness can result from ambiguous
social situations. and people with low sel f-esteem may tend to inter-
pret ambiguous social exchanges more negatively than others (Jacobs.
Berschei d. 8. Wal star. 1971). People who devalue themselves may also
have the tendency to assume that others appraise them in a similar
manner. thus leading to potentially rejecting or hostile mannerisms
toward others. affecting the potential for new relationships. Ber-
scheid and Wal star (1978) and Solano and Batten (1979) found that
lonely college students had self-disclosure problems in mixed-sex
dyads. Goswick and Jones (1981) found that lonely students were more
sel f-focused in their attention in social situations. suggesting that
perhaps students with low sel f-esteem have less empathy or responsive-
ness to others. This study can also be interpreted in the reverse--
that loneliness already existing in someone can lead to sel f-absorption
and consequently hinder further relating to others.

Loneliness can also be viewed as a cause of low self-esteem.
especially as loneliness becomes more chronic. Parmel ee and Werner
(1978) described the stereotype of a lonely person as socially
awkward. cold. unattractive. and unable to make friends easily. Gordon
(1976) and Mil ner (1975) both identified the state of being friendless

as equated with being a social failure in our society. Thus. the

29

inadequacy that stereotypically is associated with feeling lonely can
create or exacerbate low sel f-esteem. James (1908) presented one of
the first theories of low sel f-esteem. stating that sel f-esteem was
based on one's actual successes in relation to one's idealized
perception of oneself. Thus. the sel f-appraisal of being a social
failure would increase the discrepancy between one's idealized image of
oneself and the realistic confrontation with one's social state. Low
sel f-esteem can result from the lack of correspondence between one's
pretensions and the social deficits one experiences. Given that one's
self appraisals can be highly subject to one's style of attribution. it
is possible to speculate that social difficulties attributed to per-
sonal inadequacies would have greater effect on sel f-esteem than situa-
tional constraints. Miceli. Morasch. and Peplau (1979) asked college
students to evaluate people who were lonely as to reasons for their
loneliness. When the perceived reason was attributed to internal
causes. the individual was seen as less likable. less resourceful. and
was assumed to have less self-esteem.

Depression is also an emotion that has been correlated with lone-
liness. One can see how depression can be viewed as a causative factor
with loneliness. given that depressive individuals often withdraw from
others as part of their depressive state. In persistent loneliness.
especially loneliness that one perceives is not under control. loneli-
ness could be viewed as a causative factor for depression. in that a
person begins to feel that his/her attempts to override the loneliness

are futile.

30

Empirical investigations have frequently found correlations
between loneliness and depression (Bradburn. 1969; Bradley. 1969;
Bragg. 1979; Ellison 8. Paloutzian. 1979; Perlman. Gerson. 8 Spinner.
1978; Russell. Peplau. a Cutrona. 1980). For example. Bradley (1960)
found a significant relationship (I = .37) between loneliness and the
0 scale of the MMPI. a scale measuring depressed affect. Shaver and
Rubenstein (1980) found that 60% of a lonely sample of adults from a
mid-sized city in Massachusetts listed depression as an emotion they
experienced in their loneliness. The factor analysis of emotions
associated with loneliness that Shaver and Rubenstein performed on
their data yielded a host of depressive labels. including melancholy.
sorry for self. empty. and sad. These authors conceptualized the
depression people report as a part of loneliness as a reactive depres-
sion to the state of loneliness. Perl man et a1. (1978) reported that
lonely people are more unhappy. less satisfied. more pessimistic. and
more depressed than nonlonely subjects.

Further examination into the conceptual relationships between
depression and loneliness have led some researchers to consider
loneliness as a subset of depression. Horowitz. French. and Anderson
(1982) asked 40 subjects to describe with adjective labels someone
they knew to be depressed and someone they knew to be lonely. These
subjects were students in an introductory psychology class at Stanford
University. A hierarchical clustering procedure was used to analyze
the adjectives applied to each prototype. The "depressed" person was

demonstrated by this procedure to be considered in a broader. more

31

varied manner than a lonely person. It appears that loneliness may
include features of depression. whereas depression does not neces-
sarily encompass loneliness. These researchers concluded that the
lonely prototype is nested within the depressed prototype. Depression
can result from nonsoci a1 events such as the loss of one's job. ill-
ness. poor academic performance. or failing at some avocational pur-
suit. Loneliness. however. is viewed as being more exclusively
triggered by a loss of a significant relationship or by deficits in
one's social need satisfaction.

One of the more serious relationships seen between depression
and loneliness is the association of suicide with social isolation
(Becker. 1974; Col son. 1973). Indirectly. the evidence for signifi-
cant suicide rates among adolescents (Jacobs. 1971) and the evidence
for loneliness in college students being related to campus suicides
(Lamont. 1979) raises concerns about understanding the potential for
interaction of loneliness and depression.

Bragg (1970) proposed the idea that there are depressed lonely
people and nondepressed lonely people. He compared groups of lonely
students who were first matched for their degree of loneliness. but
who scored differently on the Beck Depression Inventory for degree of
depression. Results supported the idea that depression is reflective
of a more negative picture of one's world and a more global sense of
dissatisfaction. whereas nondepressed lonely subjects are dissatisfied
primarily with social concerns. but not necessarily with other areas of

their lives. Bragg also found that people who scored higher on the

32

Beck Depression Inventory who were also lonely were people who attrib-
uted the causes for their loneliness to social versus environmental
reasons.

Young (1982) contended that the relationship between depression
and loneliness can be explained by the fact that lonely individuals who
attribute their reasons for loneliness to internal traits or deficien-
cies in social ability will also tend to become depressed as a result.
He saw depression. as did Shaver and Rubenstein (1980). as a reactive
state to loneliness.

Russell. Pepl au. and Cutrona (1980) addressed the concern of
overlap between loneliness and depression from a psychometric view—
point. These researchers found a correlation of .55 between loneliness
(as measured by the revised UCLA Loneliness Scale) and depression (as
measured by the Beck Depression Inventory). Although this correlation
provided strength for the revised UCLA scale for concurrent validity
for the instrument. in performing linear regression to measure the
predictability of affiliative tendency. depression. social desirabil-
ity. and social risk taking as determinants of loneliness. only 14% of
the variance in the loneliness scores could be accounted for by the
Beck Depression Inventory scores.

Weeks. Michel a. Pepl au. and Bragg (1980) performed structural
equation analysis on data collected from 330 psychology students at
UCLA on degree of loneliness and affective states. Their analyses
yielded results supporting the impression that loneliness and
depression are correlated but clearly different constructs. Neither

33

appeared to be a direct cause of the other. These researchers con-
tended that they shared common origins. Both states appeared stable
over a 5-week period.

Thus. research on the correlations found between loneliness and
depression has provided some contradictory viewpoints. On one hand.
there are theorists who have conceptualized loneliness as a separate
construct from depression. although seeing both loneliness and depres-
sion as sharing in similar affective states. There are researchers
who have identified depression as a component of loneliness. while
seeing loneliness as a reaction to social dissatisfaction that goes
beyond depression. In reverse. there are researchers who have cited
evidence for loneliness being understood as a subset of depression. or
a "type" of depression that results from more causes that are all
concerned with lack of intimacy or loss of intimacy. Little research
has looked into the question of depressed persons being more vulnerable
to loneliness. This is an interesting question that the correl ational
data have raised. Research up to this time has investigated the sta-
tistical differentiation of loneliness and depression and the descrip-
tive differences between the two. with a tendency to see depression as
either a result of loneliness or a derivative of the same conditions
that may also lead to loneliness. Samples of subjects for this
research have most typically been college students.

LonelinesundjociaLSkflls
Studies that have explored the social -skill differences between

lonely and nonlonely subjects have found some interesting differences

34

between these two groups. Most studies have been based on sel f-report
data. Poor communication and social skills have been implicated in
the loneliness that adolescents experience. These difficulties are
seen particularly with respect to initiating contact with others
(Brennan 8. Auslander. 1979; Zimbardo. 1977). Social-skill problems
are also implicated in loneliness in the elderly (Perlman et al..
1978) and in divorced subjects (Jones et al.. 1980). NeViano and
Gross (1976) reported that lonely alcoholics were higier on social
inhibition than nonl onely alcoholics. Horowitz and French (1979)
found that lonely students reported problems of inhibited sociabil-
ity. such as initiating communication in groups. enjoying social
gatherings. giving up control in relationships. and introducing one-
self. Among college students. loneliness was associated with lower
social risk taking. lower affiliative motivation. less expression of
affection for others. and less sel f-discl osure. Pal outzian and
Ellison (1979) reported a negative correlation of -.55 between posi-
tive social skills and loneliness in a study identifying correlates
with loneliness measured by their abbreviated loneliness scale.

The appraisal one has of his/her social skills has implications
for how one copes if lonely. Paloutzian and Ellison (1979) found that
lonely students who saw themselves as having positive social skills
were more apt to become involved in social activities with friends as
a coping device. whereas students lacking confidence in their social
skills tended to use more "escapist" coping devices like drinking.

overeating. or sleeping. Given that students evaluate the social

35

provisions received from friendships asunore important than the social
needs met by family relationships or even by love relationships. the
effect of social-skill difficulties can have a large effect on coping
and the persistence of loneliness.

Some studies have provided evidence that lonely college students
indeed have less social contact than do nonlonely students (Jones et
al.. 1980; Russell et al.. 1980; McCormack 8 Kahn. 1980). Lonely
students date less frequently. spend more time alone. and engage in
fewer social activities.

Very little research has been conducted to actually observe
social-skill behavior in lonely versus nonlonely subjects. The few
studies that have been done have used college student samples and
provided some interesting results. Solano and Batten (1979) found
that college students whoiwere lonely had particular difficulty with
self-disclosure in mixed-sex dyads. but interestingly enough. higher
self-disclosure in same-sex dyads. Perhaps lonely students are
inhibited about heterosexual relations or lack confidence in this
area. but have a tendency to share inappropriately in same-sex
relationships. Platt and Spivack (1975) tested subjects for social
competence in problem solving. Subjects were asked to provide
solutions to 11 interpersonal situations that needed to be assigned
explanations for the outcomes provided. Results were significant and
demonstrated that lonely people (as measured by the UCLA Loneliness

Scale) were less able to think of ways to solve problems posed by

36

interpersonal situations. They were comparable to nonlonely subjects
on a control situation that was not interpersonal in content.

Jones. Hobbs. and Hockenberry (1980) asked lonely college
students to engage in videotaped discussions with opposite-sexed
strangers who were also college students. Lonely subjects differed
from nonlonely subjects in that they asked fewer questions of their
social partner. responded more slowly to partners' statements. changed
the topic of conversation more frequently. and made fewer personal-
attention statements about their conversational partner. These
studies indicated the degree to which social competency can influence
the comfort level people have in interpersonal situations. suggesting
that social skills may have more to do with loneliness than social
isolation or physical isolation.

Other personality correlates with loneliness are low social risk
taking and hostility. Sermat (1980) conjectured that repressed hos-
tility may be associated with loneliness because people may feel an
accumulated sense of dissatisfaction if there are developmental defi-
cits in terms.of parental security and warmth. and consequent fears of
entering into interpersonal interaction. Passivity and anxiety also
show repeated correlations with loneliness in individuals and also seem
likely to be characteristics associated with a history of problematic
social learning (Moore 8 Sermat. 1974). These authors contended that
lonely peeple tend to become overly concerned with external pressures

and approval and are less internally motivated.

37

LonelinessJLAdoJescetmndjarJLAdthbmd
Peplau and Perl man (1979) and Cutrona (1980) all represent the

recent trend in loneliness research that looks at more normative
populations and focuses on the social and cognitive factors involved
in loneliness. One reason this kind of research is becoming more
prevalent in loneliness inquiries is that identifiable risk factors
and reported behaviors can be associated with one another. providing
for an observable loneliness process. One can infer why a college
student is lonely. due to separation problems that are reactivated by
the upheaval in leaving home and undertaking a less certain and more
demanding environment. but if one asks a student how s/he feels when
lonely. what their thoughts are about loneliness. and what types of
behavior they engage in when lonely. a more dynamic picture results.
These theorists believe that loneliness is both an emotional and cogni-
tive process. in that one's evaluations and expectations influence
one's assessment of relational deficits.

With regard to late adolescence and early adulthood. the study of
loneliness is most pertinent because there is a psychological "shift-
ing" occurring which can provide the necessary prerequisites for
loneliness. In Wei ss's framework. a student who has always relied on
a best friend in my: school for the identity of being the popular.
successful over-achiever in a moderately competitive school may find
totally different circumstances and challenges awaiting him/her at a
large university. A student can be vulnerable to both emotional and

social isolation. according to Wei ss's terminology.

38

Adolescence has been termed by 8105 (1961) as the second
"separation-individuation" phase of human development. recapitul ating
early separation-individuation phase that occurs within the first
3 years of life. College students represent the older range of
adolescence. when certain tasks of leaving hone. asserting their
individuality in a new way. and confronting new challenges and
conflicts undoubtedly promote this second separation-individuation.
The ambivalence of becoming more separate can be particularly painful
at this time of a person's life. when it is realized that one is
becoming more responsible for one's decisions and commitments.

Erikson (1950) identified the particular life-cycle stages of
adolescence and young adulthood as the identity versus role confusion
and intimacy versus social isolation stages. both addressing specific
tasks that are necessary to achieve in order to psychologically have a
sense of a meaningful self and to create intimate relationships with
people. In addition to the pressures many college students experience
in adjusting to college life. our culture places great emphasis on
one's social popularity. competition. and achievement as hallmarks of
success. imposing difficult and often contradictory demands on young 1
adults (Gordon. 1976). Thus. the combination of changes in a college
student's social milieu. the new demands being encountered academi-
cally. and the developmental transactions occurring at this point in
the life cycle make the study of loneliness in undergraduate students

a rich and worthwhile endeavor.

39

Several studies have indicated that there is severe loneliness
reported in college students. Phillips and Pederson (1972) tried to
assess the "general mood" of students and found that boredom. lonel i-
ness. and a sense of futility were the most prevalent problems. Chris-
taans (1965) identified loneliness as debilitating to concentration and
related to the lack of motivation. Ostrov and Offer (1978) found that
14% of a male sample aged 16 to 20 years felt very lonely. as did 12%
of a female sample of the same age span. Brennan and Auslander (1979)
found that 10 to 15% of adol escents sampled (a 9.000 sample group)
reported feeling severely lonely. Lamont (1979) wrote about the diffi-
cul ties of students on large college campuses. Lamont believed that
suicide rates and alcoholism rates on college campus are related to
loneliness experienced by students. thus providing clinical reasons of
great importance for learning more about loneliness in college stu-
dents. Mij uskovic (1979) spoke of the development of a reflective
sel f-consciousness during adolescence (as did Piaget. 1952; Selman.
1971). in which the adolescent and young adult senses his/her own
separateness in a new. existential manner. inviting introspection and
increased differentiation from the environment.

Cutrona has been one of the few researchers to investigate
loneliness and social adjustment among undergraduate college students.
In her New Student Study (UCLA. 1979). she found that approximately
75% of a sample (N greater than 300) of new students in a large
university setting experienced at least occasional loneliness since

they arrived on campus. Of that 75%. over 40% reported that their

40

loneliness had been severe. At the end of the first academic year.
9 months after the study was initiated. 25% of the original 75%
were still experiencing loneliness (Peplau 8 Perlman. 1982).

Cutrona was particularly interested in investigating the course
of loneliness in freshmen during their first year at college. One
striking difference between those students who overcame loneliness and
those who did not was in the expectations that students had of them-
selves at the beginning periods of their lonely states. These stu-
dents who had expectations that were positive regarding their capacity
to overcome their loneliness were able to make new friends. Such
cognitions that these students reported were: (a) satisfaction with
personality. (b) belief that changeable aspects of their personalities
may have been partly responsible for their loneliness. and (c) belief
in the need to lower standards for relationships. Students who blamed
their loneliness on what they saw as unalterable aspects of their
personalities were more apt to still be suffering from loneliness at
the end of their first year. Another finding was that students who
overcame their loneliness were more likely to attribute causes of
their problem to both situational conditions in their lives as well as
to personal characteristics. instead of primarily labeling personal
qualities as the causes of their loneliness. Cutrona found that
students who made unfavorable comparisons of themselves with other
students. believing that others had "better" friendships or more
friendships. were also likelier to persist in feeling lonely. Cutrona

concluded among a number of summary comments that it would be

41

worthwhile to investigate the actual behavioral responses that students
V use to cope with loneliness to further differentiate the coping pat-
terns that result from the different expectations and cognitive attri-
butions made about their lonely situations.

Two other researchers in the field of loneliness. Peplau (Pepl au
8 Perlman. 1982) and Revenson (1981). have drawn upon Weiner's (1974)
theory of attribution to provide more insigits into how lonely people
appraise their situations. Peplau's work has focused on theoretically
linking Wei ner's research to loneliness. and Revenson's work has
explored the relationship between cognitive attribution of loneliness

and coping behavior.

W
In understanding the applicability of attribution theory to the

study of loneliness. some major concepts and contributions of attri-
bution theory are cited. One assumption of attribution theory is that
people attempt to construct a meaningful explanation or account for
situations they encounter. in part to feel more control over situa-
tions that can otherwise feel vague and confusing (Schank 8 Abel son.
1977).

Attribution theory deals with an individual's knowledge of cause
and effect. Different from cognitive appraisal. which is concerned
with how people make evaluations and judgnents. attribution theory
focuses on the explanations people have for the phenomena they

appraise as occurring in their phenomenological world.

42

Heider (1958). considered the founder of attribution theory.
identified three assumptions of an individual's manner of explaining
what happens between the self and the environment. (a) A person's
behavior is contingent upon the way s/he perceives the social world.
(b) People desire to control and predict what is going on around them.
(c) People apply perceptual properties of their objective world to
their social. interpersonal world (i.e.. people apply cause-and-effect
reasoning to interpersonal transaction). Hei der also postulated four
perceived causes of success and failure in situations that people
could identify as achievement related: ability (or power). effort.
task difficulty. and luck. Heider believed that these four aspects of
control (intrapersonal and extrapersonal) were generalizable to many
situations and were the basis for causality constructs.

Much research on attribution has focused attention on these
aspects of cognitive construction of events. Attribution theorists
generally are interested in looking at the ways people form attribu-
tions about phenomena as well as investigating how attributions affect
attitudes. emotions. and consequent behavior. Causal attribution
refers to the process of identifying causes for events. Determini ng a
particular cause for a situation or event is most complex because most
sociopsychol ogical situations or experiences that people have (incl ud-
ing their own affective states) are not uni dimensional. and causative
factors are chronologically and conceptually difficult to distinguish
or differentiate. Events are associated with other events. both

spatially and temporally. Heider spoke of a 'bause" as an event that

43

precedes the event of interest and can be shown to have an invariant
relationship to the particular event of interest.

The degree to which an individual makes causal attributions can
influence the degree to which that individual feels control over
him/herself and his/her environment. This is particular true as indi-
viduals identify causes that are modifiable (Kelley. 1971). Weiner
(1974) and his associates have done extensive research into the effect
of causal attribution on the degree of control people feel they can
exercise over themselves and their environment. Weiner also included
in his attributional research the concept of stability attributions--
attributions regarding the degree to which an individual sees an out!
come as amenable to change in the future.

Wei ner's theory incorporated the concepts of internal control
(Heider's effort and ability) as internal causality phenomena. He
incorporated Hei der's concepts of luck and task difficulty (or 'envi-
ronmental" difficulty) as external locus of causality phenomena. He
then "crossed" the concept of stability with locus of causality.
believing that some'identified causes of problems or events are
changeable (or stable) to the perceiver. His theory of attribution
has. therefore. identified two cognitive dimensions that he believed
people employ to define their success or failure at life's challenges
in a way that provides meaning for them. He studied the level of
motivation and belief in future success in subjects who had failed in
certain achievement-related situations. He identified that whether

individuals look to internal or external factors in determining the

44

cause of their failure influences their emotional reaction (i.e..
depression. helplessness. surprise. unconcern) to their failure situa-
tion. He a1 so found that whether people view their dilemma as change-
able or not changeable influences the degree to which they feel they
can be successful in future attempts with the difficult situation.
The general predictions of his theory are that internal attributions
lead to a greater sense of personal responsibility and. in some cir-
cumstances. more sel f-blame. His theory predicts that stable attribu-
tions lead to more debilitation in terms of believing in hopeful ness
of a more positive outcome in the future. For example. if an indi-
vidual believes that s/he failed a math test because of being innately
unintelligent. the individual will more likely than not believe (or
expect) to be unsuccessful on the next test. and the consequent behav-
ior may be reflective of a hopeless. defeated attitude.

Other studies using Wei ner's attribution theory and its predic-
tions about consequent effect and behavior have consistently found
results demonstrating that internal/stable attributions are associated
with depression. decreased motivation. and lowered sel f-esteem
(Harvey. 1981; Hugdahl. 1980; Zemore 8 Johansen. 1980). Additional
studies have supported the idea that depressed people attribute posi-
tive outcomes to external/unstable causes (implying less controlla-
bility and reflection of self) and negative outcomes to internal/
stable causes. whereas nondepressed subjects do the reverse (Forsyth.
Donelson. 8 McMillan. 1981: Selignan et al.. 1979). Althougi Weiner's

attribution model has been heavily associated with depression. and

45

loneliness has been demonstrated to be a distinct construct separate
from depression (Bragg. 1979; Peplau. Russell. 8 Cutrona. 1979: Weiss.
1973). the application of the theory to mood disorders does have
relevancy for the study of attribution and loneliness. Loneliness.
like depression. may or may not have clear precipitants.

Wei ner's research focused on the consequences of attribution for
performance and motivation. Other researchers have been interested in
other aspects or consequences of attribution--primarily as attribution
relates to attention to one's environment. effect of attributions on
attitudes. and the ways in which differences in individual attri bu-
tion styles relate to different causal identifications. Still other
researchers have looked at the degree to which the emotional effect
of arousal depends on the source to which the emotional arousal is
attri buted. thus resulting in emotions being label ed according to the
explanations provided for arousing events. Dweck (1975) explored the
effects of altering the attributed cause of failure on children's
responses to these failures. Dweck worked with 12 children who had
difficulties with mathematics and told some children. upon completion
of some experimental tasks. that their failure was due to poor abil-
ity. whereas other subjects were informed that they would increase
their successes if they made more effort. Those given the low-ability
explanation for their failures still evidenced debilitating responses
after subsequent attempts at the tasks and were more upset than chil-
dren who made repeated attempts at the tasks. based on increasing their

effort. These results were similar to those of the studies done on

46

learned helplessness (reviewed by Wortman 8 Brehn. 1975) and psycho-
logical response. Studies dcne by Kukla (1972) and Riemer (1975) also
suggested a similar relationship between learning and motivation (as
discussed earlier in rel ati on to Weiner's work). suggesting that
individuals who are high in achievement motivation are more likely to
take personal responsibility for both success and failure. whereas
people low in achievement motivation are more likely to attribute
success to external factors and failure to internal and unchangeable
factors.

Further research on consequences of attributions has looked at
issues pertaining to the effect of attributions on attentional pro-
cesses (information searching). labeling of arousal states. and atti-
tude change. Girodo (1972) exposed students to a gruesome film and
told one-half the group of subjects that an irritating gas would be
released in the air during the presentation of the film. The other
half of the subjects were told that there would be harmless compressed
air (the accurate explanation). Those students who did not attribute
their anxiety during the film to the alleged noxious air were found to
have much greater recall of details about the film. thus demonstrating
that. in trying to understand why one feels as s/he does during expo-
sure to a stimulus. searching behavior is engaged in. causing greater
attention. Attributions are apparently a way that people can organize
stimuli. and actually limit the amount of concentration required to
attend to what is happening around them. If one has an expectati on

beforehand as to what one might experience. the expectation (causal

47

attribution) then might reduce one's alertness or concentration.
Schacter and Singer (1962). in a classic experiment. found that sub-
jects labeled drug-induced states of arousal as being that of
euphoria. anger. sadness. and so on. depending on the external cues
they had been exposed to. The researchers concluded that a person's
response to arousal is affected by the source to which the arousal is
attributed. Pittman (1975) found that the source of attribution can
also affect attitude. Students were exposed to two sources of
arousal: presenting a speech on a topic they disagreed with and
experiencing an electric shock. When students were told that their
arousal was most likely a result of advocating a position in their
talk that was contrary to their beliefs. they changed their attitudes
toward the topic material. whereas students who attributed the arousal
to the anticipation of the shock did not experience any attitude
change.

Storms and Nisbett (1970) demonstrated an interesting consequence
of attribution on behavior in their treatment of inscmniacs. They asked
on group. who were led to believe that they were participating in dream
research. to wake up at a certain time nightly to take an “arousal"
pill as part of creating interesting dream material (really a placebo).
The other subjects. also inscmniacs. were given placebos that were
presented as tranquilizing pills. Those subjects who attributed the
waking up during the night as due to the "arousal" pill they had taken
experienced falling asleep as gradually more easy than those who had no

positive explanation as to their arousal during nighttime.

48

Strong (1973) suggested the applicability of such attribution
studies to the field of social psychology and psychotherapy. stating
that such studies have provided an empirical understanding of one's
motivation to make important sel f-changes. Thus. in terms of lone-
liness and coping behavior. the ability to choose coping behaviors may
be contingent on the attributions one has made about his/her current
state of loneliness. These studies have indicated that the effect of
causal attributions is relevant to emotional. cognitive. and behavioral
responses by people to events going on around them.

Responsibility for one's actions is an area of causal attribution
that was also of interest to Heider (1958). He proposed five main
ways of approaching personal responsibility for actions: (a) global
association. whereby any association with an event involves personal
responsi bility: (b) extended commission--whereby persons are respon-
si ble for actions that appear to be causally linked to subsequent
events: (c) careless commission. whereby persons are held responsible
for all actions that could have been foreseen; (d) purposive commis-
sion. whereby a person must intend a particular effect in order to be
held responsible for it: and (e) justified commission. in which people
are held responsible for their actions only if there are not any
convincing reasons for their actions from their environment. Much
further research. particularly done by Holmes and Strickland (1970)
and Collins and Hoyt (1972). has been concerned with the causal attri-
butions one makes dependent on the degree of choice or intention one

feels in a given situation. It has been found that most people

49

attribute causes to internal factors in situations where no environ-
mental determinants can be logically identified to describe an indi-
vidual's behavior or responses. If the environment is not perceived
as contributing to a given situation (i.e.. a person feels lonely at a
university. and yet the student perceives that the university is not a
lonely place. and that there is no explanation externally for the
loneliness). then an individual is more likely to identify internal
causes for his/her dilemma. Justified commission. whereby someone
places responsibility on external 5. diminishes as the actions or
feelings go against the expectations created by the environment.

Kelley (1971) proposed that people determine the cause of events
by means of attending to three variable situations: (a) evaluation of
an event over time (consistency). (b) stimuli (distinctiveness). and
(c) degree of agreement or disagreement as to responsibility for
situation (consensus). Kelley addressed both attribution of causality
and stability.

Kelley demonstrated that the more severe the consequences of a
situation. the more concerned an individual becomes with finding an
attribution to explain the phenomenon. This was further studied by
Harvey. Harris. and Barnes (1975). This relates to Kelley's notion of
distinctiveness in attri bution theory. The more severe an action or
event. the attribution will tend to be either applied to the actor or
not. depending on the unusualness (distinctiveness) of the event for
the individual. Thus. the more discrepant one's situation feels to

oneself. the implication is a desire to find an external cause.

50

Another important implication is the usefulness of helping an indi-
vidual recognize the normalcy. consistency. or pattern of events that
s/he has overlooked in responding to crisis situations as a way of
forming more accurate attributions of causality and stability.

Jones and Nisbett (1972) focused their attribution research on the
tendency of observers to attempt to provide either intrapersonal-factor
or situational-factor explanations for actors' behaviors. The attribu-
tions of people were determined by these researchers to be affected by
the valence (importance) of the situation and the observer (hedonic
relevance). They made a significant contribution to attribution
inqui ries by conducting experiments that demonstrated the discrepancies
between actor and observer attributions of the same event (i.e.. the
performance of the actor [subject]). This was meaningful in attempting
to understand how people could conceptualize events in such different
ways. depending on whether an individual was personally involved in the
event or affected by it. or was merely an onl coker.

Causal chaining. by which an individual can identify a series of
causes. one leading to another. is another important aspect of attri-
bution. in that enabling someone to broaden his/her thinking on caus-
al ity allows for more variables to be taken into consideration.
Brickman. Ryan. and Wortman (1975) have done important research on
causal chaining. indicating that as people are able to identify more
underlying causes. these influence consequent behavior more than imme-
diate causes peOple identify for events. The important implication
from this research is that behavior is affected by analyzing one's

51

accepted causal inferences. and permitting oneself to question exactly
what are the crucial determinants in a given situation or event. This
research team also found that internal causal attributions led to more
of a sense of control and responsibility on the part of subjects than

external ca usal attributions.

LoneJJnesunLAttLibutianbecu
Peplau. Russell. and Heim (1979) believed that attribution theory

lends itself well to loneliness. An analysis done on the reasons
people gave for being lonely resulted (by way of factor analysis) in a
two-dimensi on typology very congruent with the dimensions of locus of
causality and stability proposed by Weiner (Michela. Peplau. 8 Weeks.
1981). Pepl au stated that loneliness is also like a failure expe-
rience and that one's cognitive attributions regarding his/her loneli-
ness will determine the expectancies and affects of lonely people.
thus affecting their expectations for successfully overcoming lonel i-
ness.

What may initially ”set off" depression or loneliness may not be
accountable for the continuance of the depressive state or loneliness
state (difference between preci pitant and maintaining causes). A stu-
dent may arrive on a large university campus. feel alienated from
his/her fellow students in the awkwardness of trying to make new
friends. and then continue to feel lonely when he/she is not accepted
into the honors classes. The student may label the cause of consequent
loneliness as the belief that the campus is filled with unfriendly

people. whereas the cause may be more attributable to the student's

52

being unable to direct his/her energies toward membership in a
different academic group (hurt over the rejection) where s/he may find
acceptance and a beginning reference group. In much the same way that
Beck (1976) and Ellis (1958) had depressed clients look at the cogni-
tive assumptions they used to explain their depression. so as to under-
stand how they perpetuated their depression. so the attributions in
loneliness may serve to perpetuate loneliness instead of allowing for

a redefinition of the situation.

The authors made some predictions about how people are likely to
cope with loneliness. depending on the attributions and consequent
reactions they have. She and her colleagues predicted that depressed
affect is the result of loneliness when internal and stable attri bu-
tions (jointly) are made. Weiner et a1. (1978) found these results in
the achievement domain. Peplau and her colleagues also believed that
feel ings of pessimism and hopelessness about the future will be
characteristic only for individuals who ascribe their loneliness to
stable causes. The link between stability and expectancy has been
demonstrated by extensive evidence in achievement behavior (Weiner.
Nierenberg. 8 Col dstein. 1976). Pepl au further predicted that feelings
of shame. guilt. and embarrassment will be most often associated with
internal attributions that an individual feels s/he has control over.
such as effort. and that these serve as motivating feelings for
overcoming the problem. She stated that "whether people respond to

loneliness with depression or hostility. with passive withdrawal or

53

active striving.lnay depend on their personal explanations for lone-
liness" (in Pepl au 8 Perl man. 1982. p. 142).

Studies done with lonely college students have lent support to
the applicability of Weiner's attribution theory and Pepl au's predic-
tions. For example. Bragg (1979) reported that among lonely college
students. severe depression was associated with causal attributions to
internal. stable characteristics such as one's physical appearance.
certain personality traits. and fears of rejection. Anderson (1980)
provided evidence regarding the relationship between attribution and
coping response. He found that less-effective interpersonal behaviors
were found in lonely students who believed that their social problems
were due to what they saw as enduring personality traits that pre-
vented them from having adequate social skills. thus coping with
loneliness by keeping at a distance fron people. Students who
believed their loneliness was due to a lack of effort or poor selec-
tion of interpersonal strategies had more effective interpersonal
behavior during their periods of loneliness. Lazarus and Fol kman
(1984) reported evidence regarding coping behavior of students with
regard to other types of pressures (other than loneliness) but found
strong relationships between students' cognitions (for example.
regarding test performance) and the ways the students coped with their
pressures (i.eu. test anxiety). Thus. there is evidence that the
relationship between cognitive appraisal and coping is an important

area for more study and lends itself well to college populations.

54

Pepl au predicted that people will choose an active mode of coping
when they make internal/unstable attributions because of the implica-
tions for control and changeability over one's situation. She found
partial support for this from Cutrona's New Student Study (Peplau.
Russell. 8 Heim. 1979). Weiner et a1. (1979) hypothesized that coping
efforts of people who make external/stable attributions will be char-
acterized by more emotionally oriented responses because external and
stable attributions have been linked to lower expectations and dimin-

ished sense of personal control.

W

The categorizing of coping responses into emotionally oriented
and problem oriented is the result of research in the area of coping
behavior. largely pioneered by the research teams of Pearlin and
Schooler (1978) and Lazarus and Folkman (1984). Pearlin and Schooler
defined coping simply as "the things that people do to avoid being
harmed by life strains. At the very heart of this concept is the
fundamental assumption that people are actively responsive to forces
that impinge upon them" (p. 1). They defined coping responses as the
specific behaviors. cognitions. and perceptions in which people engage
when actually contending with their life problems. These two theo-
rists believed that coping behavior is directed toward (a) changing
the situation in which the stressful experience origi nates: (b) trying
to redefine the meaning of the stressful or harmful situation: and (c)
controlling. in the sense of managing the stress that arises from the

situation.

55

Coping behavior has been explored from a number of different
theoretical biases. These will be reviewed. and then a rationale will
be presented as to why Lazarus and Folkman's paradign is viewed as
applicable for a study of loneliness in college students.

Earlier theoretical formulations on copi ng behavior emerged from
two very different areas of study. One was the field of animal
experimentation. in which coping was viewed within a drive-reduction/
behavioral approach. The other area of study to investigate coping
was the field of ego psychology. based on neo-analytic concepts of how
the ego functions. both consciously and unconsciously. as an autono-
mous part of the psyche. heavily oriented toward maintaining equilib-
rium of drives within the person's emotional self.

Christ (1981) extended the precedents established on adaptive
behavior in animals by looking at the psycho-physiology of coping.
concentrating on the cardiovascular responses of people undergoing
stressful situations. He derived an active versus passive model of
copi ng. in which he identified different cardiovascular responses to
each of these modes of coping with stress. Hypertension was a
particular area of research for Obrist. and he used hypertension as
a model for examining the psychobiology of behavioral responses to
psychotherapeutic situations. Collaborative research with Cohen (Cohen
8 Obri st. 1975) resulted in a stage-response theory to stress. whereby
skel etal/motor and cardiovascular responses were seen as the result of
central nervous system operation. Cognitive attributions of situations

and behavioral responses were then seen to have an effect on the

56

physiological coping responses. Thus. conclusions from this research
were that cognitions play a significant role in determining physio-
logical stress responses. as in the elevation or reduction of blood
pressure.

This is reminiscent of the tremendous effect Selye's work has had
in the area of adaptational "cost" to people. which he proposed in his
eminent work on the General Adaptation Syndrome. in which he presented
a three-stage process by which people respond to stressful experiences:
the initial alarm reaction. the stage of resistance. and exhaustion
(Pelletier. 1977). Selye's research served to distinguish between
initial physi cl ogical responses that prepare people for either a
"fight" or "flight" reaction to trauma (which can be conceptualized as
the body's initial autonomic coping response) and a more extended
exposure to stress. which can become deleterious to someone unless new
coping mechanisms are established. Selye's research helped pave the
way for more refined ways of looking at coping. whereby coping was not
viewed merely as an automatic motion to difficult or unpredicted
trauna. but rather involved intentional learned behavior.

The ego-psychol ogy literature on coping has stressed the use of
the cognitive capacities of the ego to result in adaptive behavior.
Coping is defined in this literature as realistic thoughts and behavior
that result in problem solving and stress reduction. Most of the ego-
psychol ogy-minded coping theorists have formulated hierarchies of
coping behavior. evaluating coping strategies from least to most

effective and from least organized to most organized. Menni nger

57

(1963) identified five orders of "regulatory devices" that are ranked
according to the degree of internal psychological organization indi-
cated. with higiest-order devices being sel f-control. crying. use of
honor. thinking through. and talking about a problem. In Menninger's
formulation. in which coping behavior is clearly viewed as behavior
including cognitions. coping has rational as well as regulating func-
tions for the ego.

Valliant (1977) also approached coping from the viewpoint of ego
processes. His four levels of defenses proceeded from psychotic
mechanisms as the most primitive defensive response to trauma to the
most mature responses. such as humor. anticipation. suppression. and
altruism.

Blom (1984) defined coping as "specific behaviors having the
qualitative characteristics of active problem solving and attempts to
master situations that are stressful and challenging" (p. 9). His view
differed from the paradign offered by Lazarus (1977) in that emotion-
oriented coping was seen by Blom as management of emotions that is a
form of adaptation to stress but not coping in the sense of being
oriented toward mastery of a given situation or problem. Thus Blom
saw emotion-oriented responses as defensive adaptational responses.
and probl em-oriented responses as coping behavior that has a purpose
in achieving "active mastery and life satisfactions" (81 cm. Ek. 8
Kul karni. 1984).

Haan (1977) viewed coping as an active effort one makes to

effectively respond and master a crisis or personal dilemma. She

58

stated that important aspects of this intentional behavior are manage-
ment of troubling emotions. information seeking. interaction with
others.lnaintaining realistic perceptions of the problem. and having
flexibility in behavior and cognition with which to problem solve and
maintain emotional stability. Haan saw coping as the healthiest of
three response patterns of life's problems: coping. defending. and
fragnenting. Defending was seen as an active process in which an
individual attempts to deny important emotions and keep out of con-
sciousness the conflicts or anxieties raised by certain situations.
Thus. defending is a process of withdrawing from a situation and
closing off sources of awareness limiting one's capacity to learn more
or increase one's awareness of the problematic situation. Fragnenti ng
is seen as a process in which an individual misconstrues what is
happening and therefore disengages with reality as a way to tolerate
the pain or conflicts associated with personal problems or crises.
Fragnentaticn is therefore seen as a dysfunctional process that leads
to protecting the individual. at the expense of effortful behavior to
change the course of the problematic dilemma at hand.

Murphy (1976). influenced by her studies of children. made the
distinction between defending and coping as based on whether a behavior
is oriented toward mastery (coping) or protection (defending). She
stated that coping implies creative efforts at restructuring the
reality one is confronted with. nesulting in goal-oriented behavior

toward desired outcomes.

59

Garmezy (1981) studied the coping behaviors of children who were
living in sustained stressful situations. such as being mothered by
depressed or schizophrenic mothers. or being raised in tougi. inner-
city neigiborhoods. He identified problem-solving skills. hopeful-
ness. regulation of emotional reactivity. optimism. impulse control.
and future orientation as coping traits that the more resilient chil-
dren displayed in adjusting to their chronic situations. Garmezy
viewed coping as a process or chain of behaviors that involves
researching behavior early in the process. followed by developing
momentum to respond. with direct action on the problem as a conse-
quence. and confining one's action to the parameters of the situation
involved. Garmezy stated that there is a range in effectiveness in
coping. with some individuals able to lend much creativity and flexi-
bility to their coping style. whereas others are more constricted in
their responsiveness.

White (1974) saw caping as one of the forms of adaptation
available to people. in a manner similar to Haan (1977). Coping is
distinguished from mastery and defense. in that coping involves new
learning in order to adjust to a novel situation. Coping is called
forth when a dilemma is such in part because the individual is ini-
tially at a loss as to how to respond to the dilemmas or else has no
immediate behavior in his/her repertoire with which to respond. White
saw coping as the process of contending with the discomfort of dilem-
mas in ways that allow for forward movement. whereas mastery is the

process of engaging in behavior already identified as resolving the

60

problem. Thus. coping is characterized by behaving in a confident and
resilient manner. before the outcomes of behavior or the known suc-
cessful strategies can be employed. This view of coping therefore
implies that coping involves strength and persistence. Defense
behavior is marked by behaviors oriented toward managing anxiety.
Defensive reactions occur when anxiety has arisen in response to a
situation that overwhelms an individual.

The psychology of disability has provided for coping definitions
that are focused on assessing what attitudes and behaviors result in
positive adjustments to chronic disability. Wright (1960) believed
that people adjust to disability in either a positive manner. in which
they are mobilized to accept what has happened and emphasize positive
aspects of their life. or by succumbing. in which they identify with a
sick role and maintain negative attitudes. Coping is seen as a
process involving acceptance. mourning losses. and realistically iden-
tifying expectations of success for the future. Succumbing is charac-
terized as emotionally refusing to accept what has happened and
therefore being caught in a perpetual struggle of feeling inadequate
and identified with unrealistic standards of normalcy. Thus. this
approach to coping addresses the need for positive sel f-definiticn in
accepting dilemmas and progressing. This view also emphasizes the
need for an individual to separate him/herself from the societal norms
one has been enculturated into for purposes of redefining standards of
success. Wright saw coping as constructive and active. and succumbing

as destructive and passive. Vash (1981). in a similar vein. discussed

61

the need for flexibility with regard to internal (autopl astic) aspects
of trauma and the environmental (alloplastic) options available for
adjustment. Coping involves focusing inward to understand oneself.
develop sel f-acceptance in the face of trauma. and identify internal
resources for change. including value change and self-actualizing
potential of problems. Coping also involves assessing how the envi-
ronment affects one's feelings about a personal trauma. and identi-
fying ways to interact more positively with one's environment
(including dealing with social biases. constrictions. and assumed
beliefs).

Lazarus and Folkman's (1984) model of coping was based on the
assumption that coping actually provides both an ego-regulating
function and a mastery function. Lazarus is a cognitive theorist who
sees coping as contingent on the cognitive appraisals one makes
regarding harmful or threatening circumstances. In an extensive study
that was done by these authors in 1979 on 100 adults whose coping
responses to stressful situations were recorded over a 12-month period.
a factor analysis of the different responses yielded a two-factor
inventory of probl en- versus emotion-focused items. The Ways of Coping
Inventory was derived from that study (Lazarus 8 Cohen. 1980). as well
as a Revised Ways of Coping Inventory (Lazarus 8 Cohen. 1983). They
defined emotionally oriented coping responses as those responses that
are directed at making the individual more comfortable from the stress
or distress being experienced from the perceived threat or problem.

Probl em-oriented responses were defined as responses that attempt to

62

take action to modify or change the situation. in an active. "taking
action" manner. This can be accomplished by focusing on internal
personal factors. or on external factors. or both. Both types of
coping can involve cognitive as well as behavioral strategies.

Lazarus and Folkman adopted the person-in-environment model of
coping. in which psychological stress or distress resides neither in
the situation nor in the person. but in the transaction between the
two. This model lends itself well to loneliness. in that most theories
of loneliness look at the interaction between an individual and his/her
social milieu for explanations regarding the cause and continuance of
loneliness. These theorists agreed with Cutrona. Peplau. Weiner. and
others that cognitive attribution can play a key role in determining
consequent behaviors in reaction to upsetting situations and psycho-
logical states.

The importance of learning about coping with loneliness is because
of the effect coping can have on integration and learning. Lazarus and
Felhman stressed the fact that coping per so does not imply effective
or ineffective responding. but is the responsiveness of an individual
doing what s/he is capable of in responding to stress. The importance
of learninglnore about how coping can be effective versus ineffective
seems important. however. in determining the individual's prognosis
for acquiring a changed internal capacity to understand and benefit
from the current demanding situation or else to have the maladaptive
situation be perpetuated. Coping can result in new insights. psycho-

logical changes. improved use of resources. and with regard to

63

loneliness. an appreciation of what loneliness has to offer instead of
a dreadful feeling of malaise. If loneliness can help people become
more acquainted with their inner selves. as Moustakas upheld. then
perhaps coping strategies. such as spending time alone to think. are
strategies that lead to greater sel f-trust.

Paloutzian and Ellison (1982) asked 260 college students what the
most effective means of dealing with loneliness were for them. The
most useful strategies cited were: (a) getting alone to think. (b)
listening to music. (c) spending time with close friends. (d) reading.
and (e) eating. It is interesting that among the top five choices.
three involved being alone. In two other surveys done regarding
coping responses to loneliness. results were similar to those of
Pal outzian and Ellison (1979). Studies done by Cutrona. Russell. and
Peplau (1979) and Shaver and Rubenstein (1979) indicated a range of
coping responses involving time alone. time with friends. and
pleasure-oriented responses. The study done by Cutrona et a1.
involved college freshmen. and the survey done by Shaver and Ruben-
stein encompassed responses from a range of adult age groups. Palout-
zian and Ellison categorized the responses they found into the
categories of reflective solitude. contact with friends. sensual 1y
oriented responses. searching responses. nonsocial diversion. passiv-
ity. and prayer. The other studies indicated responses that fit into
these categories as well. In comparing Lazarus and Folkman's paradign
of emotion-oriented and probl em-oriented coping responses. the cata-

gories presented by survey data specifically on loneliness coping seem

64

to cover both palliative and taking-action types of measures. What
these survey studies did not provide is information on the range of
responses used by individuals. or the effectiveness of responses for

individuals who were inclined toward one coping style versus another.

O-iAFTER III

LETHODGJBY

This chapter includes (a) a description of the sample. (b) dis-
cussion of recruitment of the sample. (c) discussion of the instruments
used. (d) presentation and discussion of the research design. (e) a
summary of the study and hypotheses. and (f) a summary of the analyses
performed.

This research study looked at the relationship between attribu-
tions lonely students make about their state of loneliness. and the
manner in which they cope with their loneliness. The attributions of
interest were stability and locus of causality attributions. Attribu-
tions of stability are concerned with the degree to which an individual
sees a situation as changeable or unchangeable. Attri butions of caus-
al ity are concerned with the degree to which an individual attributes
the cause of a particular situation to internal or external sources.
This attributional model is based on Weiner's (1974) theory of attribu-
tion. The coping paradigm that was used in relating attributional
styles of subjects to coping response was that of Lazarus and Folkman
(1984). This paradign specifies two types of coping. emotion-focused
and probl em-focused coping. Loneliness was assessed in subjects through

the use of (a) a scale designed to measure loneliness. (b) a question

65

66

"Are you lonely?" and (c) a Likert scale on loneliness. These are
described under Instrumentation in this chapter.

Theoretical assumptions of this study were the following: (a)
that loneliness causes people to experience distress that results in
coping effort: (b) that people have a need to make sense out of
ambiguous and stressful situations. such as loneliness. and that
attributions fulfill this purpose of prediction and control; and (c)
that people engage in purposeful cognitions and behaviors to cope with
difficult circumstances. This coping can be oriented toward amelior-
ating distress (emotion-focused coping) or toward taking action on the
problem at hand (problem-focused coping) (Lazarus 8 Fol kman. 1984).

For purposes of this study. loneliness was defined as a painful
emotional experience that results from the deprivation of desired
relationships with others. Loneliness is characterized as a longing
for human contact or intimacy. and can result from desiring an emo-
tionally close relationship (such as a love relationship) or from
desiring social contact with people with whom one identifies (such as
a person suffering from cancer who longs to talk with other cancer
patients even though s/he may have close relationships with family or
friends). Loneliness can result from missing a particular person one
longs for. or from wishing a relationship one has the expectation of
having (Gordon. 1976: Perl man 8 Peplau. 1981; Sullivan. 1953: Weiss.
1973).

67

WW
The sample consisted of 199 undergraduate students from Michigan

State University. Eighty-three of the students were male. and 116
were female. The ages of the sample ranged from 17 to 24 years. with
95% of the sample falling between the ages of 18 and 21. The break-
down by year in school was as follows: 69 freshmen. 47 sophomores. 54
juniors. and 28 seniors. Of the entire sample. there were 27 students
who identified themselves as members of a sorority or fraternity.

The majority of the student sample lived in university housing.
residence halls on the Michigan State University campus. One hundred
sixty-three students lived in the residence halls. including eight who
were members of a Greek house off campus. Nineteen students lived in
a sorority or fraternity house. and nine students lived in their own
apartments off campus. Thus. the student sample was primarily a group
of students living within the university environment. Ninety-four of
the residence hall students lived in what are considered to be smaller
residence hall complexes. and 76 of the students lived in the larger
residence hall complexes. Therefore. 86% of the sample resided on
campus in residence halls. 10% lived in sororities or fraternities.
and 4% lived in their own apartments. 0f the 86% of the sample living
on campus. 6% lived in small residence halls (fewer than 800 students)
and 44% lived in larger residence halls.

In terms of background of students. 38% of the sample grew up in
an urban area or city (50.000 inhabitants or more). 25% of the sample

grew up in an average-sized town or small city (20.000-50.000

68

inhabitants). and 10% of the sample reported growing up in a rural
area. Students in the sample came from a range of family sizes. with
only 5% of the sample reporting having no siblings. and the majority
of the sample (83%) reporting having between two and five siblings.
One hundred ninety-one of the subjects reported growing up in a
household in which the mother was present. and 176 of the subjects
reported the presence of their father. thus leaving 13% of the sample
growing up in fatherl ess homes or homes where someone other than the

biological father was present.

Emcedunes

Initially. contact was made with the Housing Director at Michigan
State University to discuss the interest in conducting a survey study
on loneliness. attributions. and coping behaviors. Strategies were
discussed for recruiting as random a sample as possible of under-
graduate students at the University. With the approval of the Acting
Director of Housing. permission was granted to contact residence hall
directors to set up plans for collecting data on campus.

The sample of students was recruited from residence halls. from
undergraduate education courses. and from two off-campus 'Greek"
houses (a sorority and one fraternity). A description of the recruit-
ing method follows.

Students in the residence halls were recruited on a voluntary
basis by resident advisers who had been informed about the study.

Meetings were conducted with residence advisory staffs in five

69

residence halls on campus that had been selected by the Michigan State
University Acting Director of Housing as representing the best sex.
age. and academic-interest variation on campus. These halls were
also selected to provide for a diversity in "type" of student. such
that students in this study would represent variation in academic
performance. extracurricular activities. and maturity levels.

The nature of the study was explained to resident advisers. with
emphasis placed on the privacy and confidentiality students would be
guaranteed if they participated. Resident advisers were asked to
identify volunteers on their balls or. if they desired. pre-select
students they felt were lonely. It was left to the discretion of
the residence hall advisers to whether they approached particular
students or asked for volunteers on their respective halls. This was
done in accordance with specifications from the Michigan State Univer-
sity Human Subjects Committee to provide for the maximum comfort of
students asked to participate in the study.

In the meetings held in each residence hall with the respective
group of resident advisers. student volunteers were given the opportu-
nity to ask questions about the study. how the results were to be
used. and why there was interest in exploring the question of coping
with loneliness and attri butional styles of students experiencing
loneliness.

Resident advisers were then given packets of the instruments.
which were all numbered in a like manner per packet so as to identify

one subject with each set of the forms. A letter of information about

70

the study. students' rights in participating in research. a consent
form included with the letter. and a code number matching the numbered
questionnaires and data forms were provided to each student. Resident
advisers were asked to collect materials within a week and submit them
to their residence hall director. Data were then obtained from the
resident directors.

Students were also recruited from Education 200 cl asses at the
University. These classes have general undergraduate enrollment
primarily from students within the College of Education. Students
were told the nature of the study and what their participation would
involve. Data packets were handed out to all students present in the
classes. with the instructions that they were to return the packets.
filled in or left empty. to their instructor within a week. The
instructors then left the packets for collection in Erickson Hall at
their offices.

Recruitment was conducted in one sorority house and one fraternity
house. for purposes of providing some comparison data between resi-
dence halls and off-campus group living. The study was presented to
the students living in these houses at their house meetings. and
packets of the questionnaires were left for students who were inter-
ested in participating. Students in each house decided that they
felt comfortable with handing their packets in to their house
president. and the packets were collected fron these individuals.

All students were informed that they could request feedback on

their own data by contacting me and providing me with their code

71

number. appearing on their instructional letter. Students were also
told that they could get in contact with me if they were interested in
having a synopsis of the study results sent to them.

Because the questionnaires required a minimum of a half hour to
complete. it was necessary to have students complete the forms on
their own time. individually. They were encouraged to go through the
forms in one sitting. and were asked not to discuss their form with
anyone before filling it out. An advantage to having students
complete their questionnaires privately is that there was a reduced
opportunity for social anxiety to inhibit responses. as may have
occurred in group settings. A disadvantage is the concern that
concentration may have been compromised by some degree if students
filled out the forms over a nunber of sittings.

It was assumed that in striving for a total of 200 respondents.
there would be a sizable sample of lonely students within this total
and that there would be a range in terms of loneliness. Although it
was desirable at the outset to specifically recruit lonely subjects.
the recruitment technique was to identify a need for students expe-
riencing loneliness to participate. but that students who did not feel
they were lonely at the present time could participate as well. This
was done because previous research has indicated that there can be a
stigna associated with loneliness. and that indeed there may have been
students who were assessed as lonely by the loneliness scale used in
the study who did not necessarily say they were lonely or want to be
identified to themselves as participating in research because they

72

were lonely. Students were told that they would hopefully find the
questionnaires and scales thought provoking and helpful in gaining
self-awareness. Students were also made aware of resources for coun-
seling and a telephone number to the researcher if there were any
negative repercussions felt by completing the instruments in the
study.

BationaleioLSelectioLQciteLta
ior—Lonaluublasts

A decision was made to select a group of lonely subjects from the
entire sample to look at relationships between variables among lonely
subjects and to make comparisons between lonely and nonl onely sub-
jects. Additionally. there was interest in looking at which items on
the dependent measures might successfully discriminate lonely from
nonlonely subjects.

There were three measures in the study that identified lonely
subjects. The first measure was the Revised UCLA Loneliness Scale.
The second was a direct question. "Are you lonely?" The third was the
Loneliness Questionnaire. asking students who were lonely to circle a
number from 1 to 10 indicating the degree of loneliness they were
experiencing. It was decided that a particular score and above on
either the UQA Loneliness Scale or on the Likert scale would allow
for subjects who either measured objectively as lonely based on a
standard instrument or self-reported as lonely. Sol ely depending on
a minimum score and above on the UCLA Loneliness Scale to determine the

sample did not seem sufficient. given that there were students who

73

received scores at or below the mean on the UCLA Loneliness Scale who
nonetheless indicated that they were lonely on one of the other
measures. It was therefore decided that any students scoring at one
standard deviation above the mean or higier on the UQA Loneliness
Scale or at one standard deviation above the mean or higher on the
Likert scale would be included in the Lonely sample. Students who
endorsed the answer "Yes" to the question "Are you lonely?" but did
not score at one standard deviation above the mean on either instru-
ment were excluded from the Lonely subsample since their only indi-
cator of loneliness was not corroborated by one of the other measures
as indicating "strong" loneliness. For purpose of identifying a
sample of lonely subjects on which to do further study and analysis.
it seemed more sensible to make the inclusion criteria somewhat strin-
gent. This sample yielded a Lonely sample of 49 subjects. 30 females
and 19 males.

For purposes of comparing a Lonely sample with a Nonlonely sample.
criteria needed to be established for choosing the Nonlonely sample as
well. It was decided that students scoring at one-half standard
deviation below the mean on the UCLA Loneliness Scale and responding
negatively to the question "Are you lonely?" would be included in the
Nonlonely subsample.

The UQ.A Scale tends to measure loneliness by means of assessing
degree and quality of interpersonal relationships. Therefore. it was
considered possible for a student to experience an inner sense of

loneliness despite ostensibly satisfying relationships. and thus to

74

report loneliness although receiving a low score on the UQ.A measure.
The maximum score a student could have on the UCLA Loneliness Scale is
35. This score was considered "not lonely" in previous research
studies that were conducted with college-aged students (Peplau 8

Perl man. 1982). Likewise. there were students who scored at 35 or
below who did endorse low numbers on the Likert scale. but these
students were permitted in the Nonl onely sample. given that the Likert
scale may have been misinterpreted by students who thought that they
were supposed to circle a number whether lonely or not. There were
students who endorsed "no" on the direct question. "Are you lonely?"
and who also had low UQA scores. but did circle a low nunber on the
Likert scale. It was presumed that these students may have thought
that they had to circle a number. Therefore. the Likert sel f-rating
measure was considered less reliable. Again. there would have been a
sacrifice of subjects had there been a more stringent requirement of
including only subjects scoring 35 and below on the UCLA Scale and
responding "no" to the loneliness question who did not endorse the
Likert scale at all. These inclusion criteria for the Nonlonely

sample yielded a sample of 60 students (24 males. 36 females).

Insttumentation
The instruments used in the study were as follows (see Appendix B
for copies):
1. The Student Information Sheet
2. Loneliness Questionnaire

3. The Revised UCLA Loneliness Scale

75

4. Grant Attribution Scales

5. The Attributional Style Questionnaire--Negative Situations
Subscale

6. Coping Inventory

7. Lubin Depression Adjective Checklist--Fonm A

8. Coopersmith Sel f-Esteen Inventory-~Adult Form

W

The Student Information Sheet asked students to respond to some
simple demographic items and then asked questions about the students'
present status regarding loneliness. This sheet provided identifying
information. such as age. sex. year in school. and living situation.
In addition. the questions regarding loneliness provided some concur-
rent validity data for the Revised UCLA Loneliness Scale and some data

for the correlational analyses.

W
The Revised U0.A Loneliness Scale was used to measure loneliness.
The scale is a 20-item list of statements that asks respondents to
indicate on a 4-poi nt Likert scale how frequently the statements are
true for the respondent. Statements reflect interpersonal involve-
ment. feeling of attachment to others. degree of feeling understood by
people. and satisfaction with relationships. The scale has no mention
of the words "lonely" or "loneliness" in any of the statements. and
items are phrased in both positive and negative ways. to reduce

response bi as.

76

The scale was developed by Russell. Pepl au. and Cutrona (1980)
and is a revision of an earlier UQA Loneliness Scale by Russell.
Pepl au. and Ferguson (1978). The items for the scale were generated
from interviews with college students and adults in both nonclinical
and clinical settings. The scale was revised in order to change
items from all negatively phrased items to a combination of positively
and negatively worded statements. Social desirability was a second
concern with the original UCLA Loneliness Scale. due to the desire for
people to appear less lonely than they actually are (Gordon. 1976).

The revised scale was administered to a sample of 237 college
undergraduate students at two different universities. A coefficient
alpha of .94 was obtained for internal consistency of the items.
Loneliness scores were also significantly correlated with responses to
questions regarding social activities and social relationships of the
respondents. supporting concurrent validity. Discriminant validity
was supported by the comparison of a correlation of .71 between the
scale and a sel f-label ing loneliness index and a lower correlation
between the scale and measures of mood and personality characteri s-
tics. The resul ts have indicated that the scale does measure a dis-
tinct construct from that of depression or social anxiety and has been
used in a number of studies with college students. The reliability

estimate of this instrument for the large sample was .89.

77

CausaLAttributiannLStabflJtv
AttLibuthnJcales

One set of attribution scales used was written by Peter Grant
(1984) at the University of Saskatchewan. He developed scales to
measure loneliness attributions on a sample of middl e-aged adults.

The overall scale is a 20-item instrument consisting of statements
about loneliness that one indicates agreement or disagreement with on
a Likert scale response format. The reliability figures for the
scales reported by Grant were .71 (causality) and .74 (stability).
There was a moderate correlation between the two scales of .52. These
scales. which were an extension of Revenson's (1982) scales from her
study on causal attribution and coping with loneliness. were the most
applicable scales available that measure attributions of stability and
locus of causality regarding loneliness.

The Grant Attribution Scale. subscale of Locus of Causality. was
altered for purposes of the analyses. based on the reliability coeffi-
cients (alpha) that were obtained on the entire sample. The original
scale consisted of four items that are concerned with both statements
attributing loneliness to situational determinants and statements that
attribute loneliness to personality traits. Grant (1984) reported a
reliability coefficient on a group of middle-aged. lonely people as
.71 for this four-item scale. The reliability for the sample in this
study was .45. indicating that college-aged students respond differ-
ently to the instrument. The five items were divided into two sub-
scales. one consisting of three statements addressing the extent to

which the respondent believes his/her loneliness is attributable to

78

the situation one finds oneself in. and the other containing the two
items attributing loneliness to internal personality factors. These
two new subscales yielded alpha coefficients of .69 and .80. respec-
tively. These two subscal as are thus termed an External subscale and
an Internal subscale.

A reliability estimate (coefficient alpha) of .76 was obtained on
the five-item Locus of Stability subscale of the Grant Attribution
Scale. This compares with a reliability coefficient of.74 reported
by the author of the instrument (Grant. 1984) for his Locus of Stabil-
ity subscale.

One additional change that was made in the use of the Grant
Attributional Scale for this research study was the extension of the
Likert scale from a five-number response set for each item to an
eight-number response set for each item. This was done in order to
provide for increased variance in the response set to the instrument.
and for the avoidance of encouraging responses falling toward the
middle number of the original odd-numbered set.

An alternative way of assessing students' attributions about
loneliness was to ask them directly on the Loneliness Questionnaire
whether they saw their state of loneliness as attributable to factors
internal to them or external to them. and whether they saw their state
of loneliness as modifiable or resistant to change. Each of these
questions was a forced-choice question allowing for a 'yes" or "no"
response. Obtaining these data allowed for the possibility of group-

ing students if necessary into attributional-style cells as an

79

alternative to using correlational methods of analysis if reliability
or variance difficulties with the independent measures were extreme
enough to warrant some analysis of variance tests to be run. The data
from these direct questions also provided for the comparison of
responses to the different attribution instruments. as a way of fur-

ther understanding the task of measuring attributions.

WWW

Given the scarcity of instrumentation available concerned with
measuring attributions about loneliness. inclusion of a more widely
used attributional-style questionnaire in the data collection was
seen as necessary for purposes of comparing the overall attribution
instrument with the instrument designed specifically to measure
attributions about loneliness (the Grant Attribution Scale). This was
seen as an opportunity to compare attributional style regarding lone-
liness with attribution style in general. as the selected instrument.
the Attributional Style Questionnaire. measures attributional style
across a range of positive and negative situations. Given that there
have been questions posed regarding the belief that attribution style
may be more state specific versus uniform across situations. the
inclusion of two different instruments could both address concerns
pertaining to concurrent validity for the Grant Attribution Scale and
further the exploration as to the nature of attributional style in

general.

80

The Attributional Style Questionnaire (Selignan. Peterson.
Semmel. von Baeyer. Abramson. 8 Metalsky. 1980) assesses the extent
to which individuals show characteristic attributional tendencies.
The test measures the three dimensions of locus of causality. stabil-
ity. and globality for positive and negative situations. There are 12
situations that subjects are asked to respond to. and questions
regarding each event are posed. with answers indicated on a 7-point
Likert scale. A strength of the scale is that stability and locus of
causality are treated as orthogonal concepts. which is a departure
from Weiner's theory but in accordance with how loneliness theorists
have applied Weiner's general framework to loneliness studies. Sub-
jects are asked to generate a cause for each of the events presented
and then respond to a set of standardized questions regarding each
event.

This questionnaire yields a number of different subscales. The
authors of the questionnaire advised using composite scores that
provide summation scores of the three different attributional dimen-
sions measured: locus of stability. locus of causality. and globality.
The authors further advised using composite scores for positive situa-
tions posed in the questionnaire. and for negative situations posed
in the questionnaire. thus yielding the different composite attribu-
tional scales. The reliability estimates (alpha coefficients) pro-
vided for these composite scales are .72 for attributions of positive
events and .75 for attributions of negative events. The reliability

estimates for this study sample for the composite attributional styles

81

were very similar: .74 for the combined attributions for positive
events (.N = 173) and .70 for the combined attributions for negative
events (N = 173) (.66 and .58 for the Lonely subsample. respectively).
Reliability estimates on this study sample were lower for scores that
were summed either by attributional dimension alone (of interest to
this author were the causality and stability scores) or summed by a
further scale division into attributional dimensions by positive or
negative events (i.e.. positive causal scale. positive stability
scale). Those reliability estimates are provided below.

The composite scale for negative situations was chosen as a
measure of the predictor variable for both the total sample and the
Lonely subsample. It was decided to use only the subscale since it
seemed most relevant to the issue of measuring attribution regarding
loneliness. considered a negative situation. The use of the composite
scale raised concerns in two areas. One is the inclusion in the
summary score of the attributional dimension of global ity. which is
not a variable of particular interest in this study. However. this
additional aspect of the measure. while adding some complexity to the
interpretation of the measure. nonetheless was regarded as a potential
source of information that might offer interesting additional
insights. A second concern was with the interpretation of a composite
score in general. in that it is difficult to know to what extent a
particular attributional dimension score is accounting for the entire
summed score. Again. compromises in measurement accuracy were made in

including this instrument. with the option of having strong

82

interpretive possibilities should there be very strong correlations
between this scale and other variables. The following is an example
of a negative situation posed on the questionnaire and the types of
questions that accompany each item: "You have been looking for a job
unsuccessfully for some time." Following each statement. the respond-
ent provides his/her own reason for the situation described. and then
indicates to what degree the situation results from internal or exter-
nal factors (causality) and to what degree these factors will be
present again in the future (stability). (There are also two other
questions that are concerned with global ity and importance of the

situation to the respondent.)

W

A combined inventory of items from two extensive loneliness
studies was used in this research. Cutrona's New Student Study
(1979) and Pal outzian and Ellison's Loneliness and Quality of Life
Study (1979) both yielded similar coping lists that were empirically
derived from interviews with approximately 360 undergraduate college
students. There are overlapping items. and scrne items exclusive to
one or the other. The items have him face validity and content
validity for both relevance to coping with loneliness and categoriza-
bility into probl em- or emotion-focused strategies. A pilot study was
conducted using four expert raters who were asked to assign items to
either a probl em-focused category or an emotion-focused category. and

there was 100% agreement on 23 of 27 items presented to the raters.

83

After investigating responses from the Emotion-Focused and
Problem-Focused Coping subscales. the correlation of L = .56 (p, <
.001) indicated a need to look into the possibility of refining these
subscal as A factor analysis of the Coping Inventory yielded two
factors that rationally looked like an emotion-oriented subscale and a
problem-oriented subscale. There were items. however. that did not
correspond conceptually well enough to remain in the particular
scales. After further examination of correlations between scale items
and overall scale scores. items were deleted. yielding two scales
using 19 of the original 35 Coping Inventory items. The two subscal as
have a correlation of 1: = .47 (p < .001). The Emotion-Focused Coping
subscale contains 12 items and has an alpha coefficient of .77. The
Probl em-Focused Coping subscale contains six items and has an alpha
coefficient of .72.

Examples of items from the Emotion-Focused Coping subscale are
(in response to being asked how often respondent uses the following
ways of coping with loneliness). "Think about possible benefits of
being lonely. such as thinking you are learning to be more sel f-
reliant or growing from the experience" and "Take your mind off
feeling lonely by using drugs or alcohol." As can be seen by these
two items. the Emotion-Focused Coping subscale incl udes items that are
considered both passive or escape items. and items that are cognitive
ways of feeling better. At the finest level of analysis. these items
yielded a more sound factor by being grouped together in one scale.

whereas attempting to form an emotional subscale of passive items and

84

a subscale of constructive or cognitively positive items did not
provide for sufficient reliability estimates to make them worthy of
use in this study.

Examples of Probl em-Focused Coping subscale items are "Try harder
to be friendly to other people (like making a greater effort to talk
with students in classes)" and "Do something to improve your social
skills (learning to dance. learning to be more assertive. improving
conversational skills.. etc.)." The reliability estimate for various
subscales of Problem items proved inadequate for use in the study.
and therefore a scale was formed of as large a number of items as
possible with the highest reliability and the lowest correlation with

the Emoti on-Focused Coping subscal e.

W
The Depression Adjective Checklist--Form A (DAQ) (Lubin. 1981)

is a 32-item checklist consisting of adjectives characteristic of
depression and contentment. Examples from the checklist of depressive
adjectives are Nilted" and "miserable." Examples from the checklist
of positive adjectives are "safe" and "enthusiastic." Scores are
determined by adding up the number of depressive adjectives that are
endorsed plus the number of positive adjectives that are not endorsed.
There are 22 depressive adjectives and 10 positive adjectives on the
checklist. Form A of this checklist was used in this study. The use
of this checklist provided a measure of an emotional state that has

been higily correlated with loneliness. and therefore was considered

8S

important as a way of further describing the sample of students in
this study.

Internal consistency of this checklist has been reported to be
.81 for male subjects and .86 for females. based on a sample of
adults. These items on the checklist were originally drawn from a
pool of 171 items that successfully discriminated at a significance
level of .01 between well-adj usted adult males and males rated as
markedly or severely depressed after undergoing psychiatric inter-
views. Normative data on the DAG. have been derived from an adult
sample with a mean age of 44.5 years. students of my: school age.
college and graduate students. and adolescent psychiatric popul ations.
A correlation for college students between this measure and a train
measure of depression. the College Inventory of Depression (Lubin.
Nathan. 8 Nathan. 1982). was found to be .68 (p < .01). Correlations
between scores on the DACL (Form H) for females and males and the
MAAQ Depression Scale (Zuckerman 8 Lubink. 1965) were found to be .80

and .79. respectively (p < .01).

AdulLEocm
The Coopersmith Sel f-Esteem Inventory--Adult Form contains 25

statements that respondents endorse as being "like me" or "unlike me."
Examples of such items are "I have a low opinion of myself" and "Most
people are better liked than me." The definition of sel f-esteem used

as a basis for this inventory is that sel f-esteem

86

refers to the evaluation a person makes and customarily maintains
with regard to him- or herself. 'Sel f-esteem' expresses an atti-
tude of approval or disapproval and indicates the extent to which
a person believes him— or herself capable. significant. success-
ful. and worthy. (Coopersmith. 1984. p. 5)
The adult form was adapted from Coopersmith's School Short Form. a
25-item inventory derived from the original SO-item Sel f-Esteem
InventorynSchool Form. The Adult Form is basically a rewording of
the 25-item School Short Form that poses statements in more mature
language. Reliability coefficients are not provided on the Adult Form
of this inventory by the author of the inventory. A study conducted
on 103 college students by Bedeian. Geagud. and Zmud (1977) reported

KR 20's of .74 for males and .71 for fenales.

DengLoLtbLStudx
This research study is a descriptive study using data derived

from a voluntary group of student subjects. The main variables of
interest were investigated by means of a fixed-effects model of
analysis of variance. with planned comparisons. Paired-sample
.t-tests were used for one main hypothesis. Post-hoc analyses
involved use of linear regression and cluster analysis. Analyses were
conducted on the total sample as well as on the Lonely subsample.
Attempts were made to obtain subjects from as random a population as
possible. Thus. efforts were made to recruit students from a variety
of university settings. providing for a range of personality. demo-
graphic. and academic characteristics.

The data collection was conducted according to principles of

naturalistic field survey research. in which anonymous survey data

87

were collected in a manner so as to preserve the integrity of unbiased

results.

Amuse:

The main analyses used to measure the strength of the relation-
ship and direction of relationship between attribution style. loneli-
ness. and coping response was analysis of variance with a priori
contrasts. This method of analysis requi res the following assump-
tions:

1. Variables are normally distributed and are continuous.

2. There are linear relationships between variables.

3. Observations on variables are independent of one another.

The variables in this study satisfied these assumptions. Fre-
quency distributions on all of the instruments used to measure the
independent and dependent variables. and scattergrams to indicate the
different pairwise relationships among variables. ensured confidence
in adhering to these assumptions. One variable. depression. was found
to have a positively skewed distribution. The log transformation of
observations on this variable was used in assuming the strength of

association with other variables.

W
In addition to the prediction equations run to measure the

multiple correlations of different variables with coping behavior.
correlations were obtained between all of the instruments used in the

study and between the instruments measuring loneliness and attribution

88

and particular coping items of interest. These correlations provided
further information on the sample of subjects. as well as clarifying
the relationship between different pairs of independent variables.
These correlations were derived for both the overall sample and for
the subsample of lonely subjects. Multiple linear regression equa-
tions were also run on loneliness. with attribution. sel f-esteem.
depression. and demographic variables as predictors.

I-tests were conducted to compare the overall sample of students
and the Lonely subsample on a number of key variables. Thus. lonely
and nonl onely subjects were compared on demographics. affective meas-
ures. attribution measures. and the dependent variables of emotion-
and probl em-focused coping. Cluster analyses were run to obtain
further information on variables and observations. The cluster
analysis done on the variables in the study was performed by oblique
rotation methods. The Fastcl us cluster procedure was used to identify

groups of observations among the total sample.

Sumner):
This chapter has provided the "plan" for the research study.

Discussed were the sample of subjects. instruments used in the study.
design of the study. testable hypotheses. and analyses employed to

test these hypotheses. The research study consists of both analyses
run on the original hypotheses. as well as post hoc analyses done on
subsamples of the original sample of student subjects. The post hoc

analyses provided further understanding of the sample of students and

89

addressed the issue of describing in greater detail the phenomenon of
loneliness in the lonely students. The section on instrumentation
provided explanations for the derivation of scales and the rationale
for using particular scales. Given the relative paucity of measurement
tools in the area of loneliness. attribution. and coping behavior.
instrumentation became a major focus of the endeavor. Factor analysis
of the coping instrument and item analysis on the attributional scales
resulted in a more greatly refined instrument. Thus. instruments were
not used in final form until there was satisfaction that. at the
finest level of analysis. the highest reliability possible was
obtained. along with rationally appearing items. The statistically
stated hypotheses are included in Chapter IV. as the results of each

hypothesis are presented and then discussed.

OiAF’TER IV

RESULTS

This chapter provides results and interpretation of results
regarding performance on measures of*variab1es. results of main
hypotheses. and post hoc analyses. Relevant discussion pertaining to
the findings follows the presentation of data.

The first section of the chapter presents descriptive and
statistical data pertaining to the major instruments used in the
study. This is organized according to the main variables of interest
in the study. with tables containing corresponding reliability.
frequency information. and Pearson correlations. Observations
pertaining to patterns of responses on the attribution and coping
instruments are provided. as well as descriptive information on
reasons for being lonely that were endorsed by the subjects in the
research.

The second section restates the hypotheses and presents the
results of the hypothesis testing. along with explanations of the
analyses used. The third section then presents the post hoc analyses.
followed by discussion and interpretation of these results. integrat-

ing these with results of the original hypotheses of the study. A

90

91

fourth section provides a summary of all of the results of the data
analysis.

In exploring the hypotheses for this study. decisions had to be
made regarding what samples to test the hypotheses on. given that
there was a general sample of students recruited. not all of whom
were lonely. In that the hypotheses are concerned with the relation-
ships between attribution and coping in reference to loneliness. it
seemed most relevant to discuss these in terms of a lonely sample of
subjects. The general sample recruited for data collection numbered
199 students. It was decided to use all of the data from the sample
in looki ng at the hypotheses. as well as to test the hypotheses on a
select group of "lonely" students. as discussed in Chapter III.
Therefore. the layout of this chapter entails presentations of data
with respect to a general sample of 199 subjects. as well as to a
subsample of lonely subjects (Lonely subsampl e). The hypotheses
tested on the whole sample need to be regarded in view of the fact
that this whole sample included subjects who responded to measures
"retrospectively." as well as subjects who were lonely at the time of
data collection and were responding to the questionnaires with current
information about themselves. One concern was the combining of retro-
spective and current reporting by subjects. and this is looked at
further in the discussion section of this chapter. As the hypotheses
are presented. reference is made to the results both to the general
sample (total sample) and to the subsample of lonely subjects. which

is called the Lonely subsample.

92

The main variables of interest in this study were (a) locus of
stability (unstable/stable). (b) locus of causality (internal/
external). (c) loneliness. and (d) coping response (emotion-focused
and probl em-focused). The main independent variables were locus of
stability and locus of causality. The outcome variables were both
emotion-focused coping and problem-focused coping. The reader is
reminded that the two types of coping responses were considered inde-
pendent outcome variables. Thus. the focus of the research was on the
investigation of the relationship between attribution style and coping
response. for each of the coping styles of interest. Loneliness was
entered into the research design in significant ways. One was with
respect to using subjects' measures on the various loneliness scales
as a way to form a subsample of lonely subjects for the purpose of
further understanding correlates of loneliness and drawing comparisons
between lonely and nonl onely subjects on particular variables. Lone-
liness was also looked at with respect to three main hypotheses and
some post hoc analyses that considered the combined effect of loneli-
ness and attribution on coping response. Loneliness was a1 so investi-
gated in its own rigit as an outcome variable in multiple linear
regression post hoc analyses.

The first section that follows contains tables of reliabilities
and frequencies for all of the instruments in the study. as well as
tables of Pearson correlations for all of the variables of interest.

Data are tabled for both the total sanple and the Lonely subsanple.

93

Locus_o.f_5tab_fl_i.tx
Locus of stability was measured by the Grant Locus of Stability

subscale. Of a possible range of scores from 10 to 40. with 40 repre-
senting the most unstable attributions. the mean response for the
whole sample was 29. with a .512 of 6.4. This scale was slightly
skewed negatively. This subscale mean for the Lonely subsample was
24.6. with a SD of 6.6. The distribution on this scale for the Lonely
subsample is indicative of a direction slightly more oriented toward
stable attributions. (See Tables 4.1 and 4.2.)

The correlations between locus of stability and locus of causal-
ity (measured by the Grant Locus of Causality subscale) were .1: = .53.
p < .01 for the total sample and 1: = .44. ,p < .01 for the Lonely
subsample. The locus of stability correlated moderately both among
the total sample and the Lonely subsample with the UCLA measure of
loneliness. the Likert sel f-rating scale on loneliness. and with low
sel f-esteem and depression. (See Tables 4.3 through 4.5 for specific
breakdowns of correlations in reference to loneliness measures.)

Correlations between locus of stability and the previously
mentioned correlates were slightly lower for the Lonely subsampl e.
which is expected due to the decreased variance within the more homo-
geneous subsample.

One final set of correlations worth mentioning here is the cor-
relations between the locus of stability and the Selignan Attribu-

tional Style Questionnaire Subscale for Negative Events. These

94

Table 4.1: Reliabilities and Frequency Statistics: Total Sample

 

 

Scale/Measure Alpha Mean ‘50 Range
Likert scale .00 3.7 2.3 1-9
Revised UCLA scale .90 39.9 9.7 22-70
Grant Locus of
Stability subscale .76 29.0 6.4 10-40
Grant Locus of
Causality subscale .58 27.7 5.6 12-40
Emotion-Focused
Coping scale .77 34.0 5.8 15-46
Problem-Focused
Doping scale .72 15.3 3.5 4-23
Lubin Depression
Adjective Checklist .66 8.9 6.6 0-31
(Fonm A)

Coopersmith Self-
Esteem Inventory .35 17.1 5.5 1-25

 

95

Table 4.2: Reliabilities and Frequency Statistics: Lonely Subsample

 

 

Scale/Measure Alpha Mean ‘50 Range
Likert scale .00 4.1 2.2 1-9
Revised UCLA scale .91 51.7 8.3 30-70
Grant Locus of
Stability subscale .80 24.6 6.6 10-39
Grant Locus of
Causality subscale .64 24.8 5.98 12-40
Emotion-Focused
Coping scale .67 32.7 5.3 18-45
Problem-Focused
Coping scale .59 15.1 3.3 6-22
Lubin Depression
Adjective Checklist .71 15.0 7.7 2-31
(Form A)

Coopersmith Sel f-
Esteem Inventory .53 12.1 5.1 1-24

 

9MB

>ococcou

.Eooumo-u_om 0;» Loumocm one .0coum ecu Logo—z 0:»

._o. v.N««

.mo. v.N«

.mco.u:n_cuum .ono.o-o_noum-_mccouc_ ecmzou

use coumoca ecu .ocoom ozu Lozm_c oz» .muco>w o>.umooz ecu o_auma:m oc_mcco_umu:o o_>um .oco_u:a_cau< cosm__om ecu cu mcouoc .ooziam<o

.>Louco>c. eunummuu_om no.5ncooo0u ecu cu ncouoc .mm

a

.commmocaoo co _o>o_ ecu cognate ecu .ocoom ecu cu;m_: och .um_.xoozu o>muuoqe< co_mmocnon :_nao any cu mcouoc au<o

 

 

 

acme0o
-- «can. so. «a_. No. vom3u0a
-se.ao.a
acmooo
-«mm. -- .oN. .m_. mo.- unmade.
1:0.uoEu
. . . . >o___aoum
so «ON .. seam cam~ - so mecca
. . . . >o....=.u
«a- «m. aemm .. «emw - so mecca
No. wo.: ce-.- «am~.u u- .moz:om<
no. «4.. «eon. we. «emm.- .. «awm.- ««_o. ««~N.- ««~:.- __. «m..- oc.n _um
c..- «m..- «eua.n «a:~.n «nan. «mom.. in «nwm. «ame. gene. «ksm.u nc.u mo.- <1-au<
ao.- _..- «4.3.- a:.. «awn. ««_o.- seam. .. a«:m. «4mm. ««m~.r _o.: No.- <49:
. . . . . . . . . . >_oco_
-- 4o «e~m - «em. . «.N «a- . «am: ««:m a- «esm «cam . n- me one. 30:
m a
am—.u «-.- «em:.- m..- «MN. ««N:.- «awe. «can. «asm. n: «o~.: 0—. n..- unw“_“
> oco
.. _c. m_. No. «~_.- —_. «enm.- ««m~.- «a:m.u «wN. -1 mo.- co. :0” 0L“
«emN. «eaw. mc.r No. mo.- «m_.| no. .o.: I: 0—. so. in NO.r xow
ma. «m—. wc. in 00.: 00. mo.: No.i mo. n—.1 :0. NO.: I: 00<
ac_o0u mc_aoU .
> > .m > O >
as... as... dam...” gum...” . a... a... .3. is shun... an...“ 33...... a. .2
-6039... -5295 a I. - 2 .:
o.oEmm .muOh "moo—un_oLL6u camcmom ”m.a o_amh

€97

>ucoucou ozu coumoLm

.Eooumo-..om ecu coumoLe oz. .oLOOm oz. Logm_: as»

.>couco>c. sodomy-..om cu.5mcooooU ozu ou uco.oc .mm

0—0.

V “:3.

.mc. v.N«

.mco.u:a.cuua .ono.m-o.noum-.mccouc. ecmzo.

ecu .oLOum on. Locm_: och .mu:o>m o>.umuoz to. 0—momaam oc.meco.umoso o.>um .oco.u:n.cuu< cmEm..om ecu cu undue; .oozudm<u

a

.co.muocooo .o .o>o. use coumocm ecu .oLoom us. Locm.: 0:» .un..xoo:u o>.uuoae< co.mmoceoo :.n:. ecu cu nco.oL 4u<m

 

 

 

mc.o0u
-- ««_:. 3.. w..- mo. vameoo.
nee—note
ac_ooo
««.a. -- mN. .- .- vomauo.
-eo.uOEu
. . . . >u...nmum
a. MN -- «as: «an: - .o «soc.
>u nmeou
~_.- - 5...... - 3.- ......33
mo. - .22.- Sr - .8252
m.. .~. «cam. «awe. «m~.- - «m~.- «um.- m..- .o.- no. mo.- no. .um
-- m..- «on.- .0. «mm. «m~.- - o.. «soc. «ame. «m~.- ~o.- mo.- <u-4u<
mo. ~o.- «on.- «ma.- «mN. «um.- m.. - mo. 3... o..- s..- o..- <.u:
> oco
m..- ... in: 2.- .3. S: to... 8. - ...... 2.- 8.- ..o. 96" .6...
. . . . . . . . . . . . . 9.22
.w: - o~ - .N a. .o .o - «em: a. - ¢.a -. o~ - me no ucox..
> oco
- me. o.. co.- m..- mo. m~.- o..- m~.- o~.- -- oo.- mo. so» at“
-- o.. ... «mw. mo.- mo.- so.- u... oc.- me. oc.- .. m..- xom
3.- .o. S. 8.- 8. no. 8.- a..- .6. mo. no. - -- 2?
oc.oou m:.e09 -
> . > >
......m... .. ...... ”...“... ..
£8395 .EoCQfi . a . . -m< a a x x: > <
o.oEmmn:m >.oco. ”aco.uo.occ09 concede “4.: 0.9m»

98

Table 4.5: Pearson Correlations: Attributions With Depression and
Self-Esteem

 

Attributions Depressiona Self-Esteemb

 

lotaLsampJe (N = 199)

Locus of causality .25 -.45
External itens -.03 .11
Internal items .28** -.59**

Locus of stability -.42 .45**

LmeJLsubsamble (n= 49)

Locus of causality .01 -.42
External items -- .11
Internal items -.13 -.40*

Locus of stability -.36* .54**

 

Note: Locus of causality is scored in direction of increasing values
signifying increasing externality of attributions. Locus of
stability is scored in direction of increasing values signifying
increasing instability of attributions.

aThe higher the Depression Adjective Checklist score. the greater
the level of depression.

bThe higher the Self-Esteem Inventory score. the higher the self—
esteem.

!p < .05.
"p, < .001.

99

Table 4.6: Pearson Correlations: Attributions With Loneliness

 

 

Attributions UCLA Likert Rating
IntaLsnmple (N = 199)

Locus of causalitya .44 -.15
External itens -.08 -.01
Internal items .58** .23"

Locus of stabilityb -.41** -.43**

LnneJLsubsampJe (n = 49)

Locus of causality .45** .14
External items -.17 .04
Internal items -.48** .04

Locus of stability -.30*‘ -.21

 

aThe higher the score. the more external the

score.

bThe higher the score. the more unstable the score.

in < .05.
**n < .001.

100

correlations for the total and Lonely subsamples. respectively. were
-.27. ,p < .01 and -.42. .p < .01. These correlations suggest that
locus of stability as it relates to the problem of loneliness is
different from one's attributions toward negative events in general.
This may be due in part to the fact that loneliness involves a strong
emotional component. whereas the negative events presented on the
Selignan Questionnaire tend to be more event specific. involving social
or achievement situations that have an immediate effect. and therefore
different from conditions that arouse an on-going feeling of lonel i-

0655.

Locus_o.f_CausaJ_1tx

Locus of causal ity was measured by the Grant Locus of Causality
subscale. Of a possible range of scores from 10 to 40. with a higher
score representing greater externality. the mean score for the total
sample was 27.7 with a SD of 5.6. The mean score for the Lonely
subsample was 24.8 with a SD of 5.98. Thus. this subscale was
normally distributed for both samples. The distribution on this
subscale for the Lonely subsample is indicative of a direction slimtly
more oriented toward internal attributions. Reliability coefficients
of each sample were moderate. The alpha coefficient for the total
sample was .58 and for the Lonely subsampl e. .64. (See Tables 4.1 and
4.2.)

Locus of causality correlated with locus of stability in the

direction of internal causal attributions associated with unstable

101

attributions (or external causal attributions associated with unstable
attributions of stability) (1: = .53. p < .01 [total sample] and .r =
.44. p < .01 [Lonely subsampl e1). Among both samples. locus of caus-
ality. in the direction of internal attributions. was also found to
correlate with low sel f-esteem and loneliness as measured by the UCLA
scale. (See Table 4.6 for summary of correlations with loneliness.)
Although internality is often associated with depression. these corre-
lations were small between locus of causality and depression for both
samples. whereas depression was found to correlate significantly with
locus of stability among both samples. Internality appeared to be
related more highly with low self-esteem than with an emotional expe-
rience of depression. (See Tables 4.3 and 4.4.) Table 4.5 specifies
correlations between both the internal items and external items of the
Locus of Causality subscale with thelneasures of self-esteem and
depression.

As was found with the locus of stability. locus of causality
correlated only moderately with the Seligman Attributional Style»Ques-
tionnaire for Negative Events. This is important in that the mod-
erate correlations (I = -.23. p < .01 for the total sample and L =
-.23. p < .05) for the Lonely subsampl e) indicates that attributions
of causality for loneliness occurred differently from attributions for
specific negative social or achievement events. As was speculated in
reference to the locus of stability. perhaps loneliness. due to the
nature of its experience as a personal emotional problem. alters the

way causality is perceived.

102

Loneliness

The main measure of loneliness in the study was the UCLA
Loneliness Scale. This measure had high alpha coefficients for both
samples. .90 for the total sample and .91 for the Lonely subsample.
The mean score on loneliness among all of the subjects was 39.9 on a
scale that ranged from 22 to 70. Thus. the scale was skewed somewhat
positively for the total sample. The mean loneliness score for the
Lonely subsample was 51.7 among a range of scores from 30 to 70. This
distribution approximates a normal distribution for this sample. (See
Tables 4.1 and 4.2.)

There were correlations identified between loneliness and
depression. as well as between loneliness and low self-esteem.
Pearson correlations between loneliness and depression for the total
sample and the Lonely subsample were .52 (p < .05) and .16 (p > .05).
respectively. Correlations between loneliness and low sel f-esteem for
the total sample and the Lonely subsample were .61 (p < .05) and .32
(p < .05). respectively. (See Tables 4.3 and 4.4.)

The UCLA measure of loneliness correlated positively with the
sel f-rating of loneliness by students among the total sample (1; = .39.
p < .01). The correlation between these two measures among the Lonely
subsample was a negative correlation and was not significant (1': =
-.l4. 9, > .05). These correlations indicate that the objective meas-
ure represented by the UCLA Scale and a subjective Likert measure were
perhaps allowing for different types of loneliness to be endorsed.

Most of the items on the UCLA measure are related to loneliness that

103

is the result of emotional or social deprivation of close or support-
ive relationships. Perhaps the Likert rating scale identifies
endorsements of loneliness that are of a different nature. such as
loneliness associated with alienation. boredom. or anxiety. In look-
ing at the compari son of correlations of each of the measures of
loneliness to depression and low sel f-esteem. it was observed that the
self-rating of loneliness correlated more highly with depression than
low sel f-esteem (r, = .62. p < .01 and r, = .42. p < .01. respectively).
whereas the UCLA measure correlated more highly with low sel f-esteem
than depression (L = .51. p < .01 and 1: = .52. p < .01. respectively).
(These are correlations among the total sample.) Thus. perhaps the
UCLA measure was identifying a type of loneliness characterized by the
low-sel f-image problems of low sel f-esteem. whereas the sel f-rating
measure identified a form of loneliness characterized more by hope-
lessness or negative feelings of the future. It is also possible that
people who were not wanting to identify themselves as lonely on the
sol f-rating measure were measured as such when it came to the more
objective questions asked on the UQA Scale. Thus. the small correl a-
tions between the two loneliness measures raised questions about the
social stigna involved in admitting to loneliness. (See Tables 4.3
and 4.4.)

Loneliness (UCLA) correl ated significantly with locus of
causality. and in investigating this correlation. it was seen that the
internal items of that scale accounted for the correlation (I: = .58. p

< .001 for the total sam ple: j: = .48. p < .001 for the Lonely

104

subsample). (See Table 4.6.) Loneliness (UCLA) also correlated at a
significantly more moderate strength with locus of stability (1 =
-.41. p < .001 ) for the total sample in the direction of stable
attribution. This correlation was -.30. p < .05 for the Lonely
subsample. (See Table 4.6 for complete data.)

W. The most selected reason for being
lonely was having no friends or feeling a lack of good friendships.
Twenty-five percent of the students who responded to this question
indicated that their reason for loneliness was having no friends or
feeling a lack of good friendships. Another 20% indicated that they
were homesick. Approximately 15% indicated that they had no boyfriend
or girlfriend. or had recently broken up with a romantic friend.

Eight percent indicated that they did not know why they were lonely.

A cluster analysis on variables provided evidence that different
reasons for feeling lonely were related to other aspects of the
subjects. For example. feeling homesick or missing important people
who were away was more symptomatic of younger students. as would be
expected.

Attribution style was associatedrnore with having no boyfriend or
girlfriend. Having no friends or lacking good friendships was most
associated with loneliness. depression. low self-esteem. and a tendency
to view situations as stable and internally caused. Reasons seldom
stated were (a) feeling a lack of sex. (b) not knowing the meaning of
life. (c) having personal problems. and (d) having roommate diffi-

cul ti es.

105

WE
There were 35 coping strategies. both cognitive and behavioral.

that students rated (from "never used" to "often used") to indicate
their coping style. Twelve of these items made up the Emotion-Focused
scale and six of these items made up the Problem-Focused scale.

(Other items were not included due to reliability difficulties with
including them.) (See Appendix B for exanples of scales.)

The average coping scores for the total sample and the Lonely
subsample are presented in Table 4.7. These scores are the mean
values for the designated groups. Coping responses varied from
approximately 55% to 60% endorsement of the possible higiest scores on
each coping subscale. Thus. it appeared that students were engaging
in moderate levels of coping and that the differences in coping.

depending upon loneliness or attribution. were minimal.

Table 4.7: Mean Coping Scores

 

 

Emoti on-Focused Probl em-Focused

(Possible Score (Possible Score

Range: 12-48) Range: 10-24)
Total sanple 34.15 15.38

Lonely subsample 32.77 15.18

 

106

On the whole. correlations between coping behavior and other
variables were very modest. (See Tables 4.3 and 4.4 for all corre-
lational data.) Among the total sample. the degree of relationships
between these variables was essentially the same. The highest corre-
lati on found for emotion-focused coping with other variables was again
with problem-focused coping (.39. p = .001). The correlation of
emotion-focused coping with locus of stability was .20 (p = .002).
This relationship was extremely moderate.

Among the Lonely subsampl e. emotion-focused coping had some mod-
erate relationships with a few other variables. It correlated most
higily with problem-focused coping (.41. p = .001). It correlated
very moderately with the Grant Locus of Stability subscale (.23. p =
.05). Emotion-focused coping also correlated again very moderately
with the causality question on the Loneliness questionnaire in the
direction of externality (I: = .25. p < .05). This correlation was in
the direction of more emotion-focused coping used as an individual
perceives his/her state of loneliness as unchanging or unchangeable.
This was contrary to what was expected in this relationship. (See
Tables 4.3 and 4.4.)

Probl em-focused coping was not found to be related strongly to
most other variables as well. The only correlation worthy of special
comment is the correlation between problem-focused coping and the
Likert sel f-rating measure of loneliness among the Lonely subsample

(1; = .42. p = .002). The more lonely a student claimed to be. the

107

less likely the student was to use problem-focused coping. Problem-
focused coping was not found to correlate to any significant degree
with the UCLA measure of loneliness in either sanpl e.

Correlations between coping behavior and affective states of
depression and sel f-esteem were basically inconsequential. (See
Tables 4.3 and 4.4.) The very small correlations all suggested.
however (among the range of nonlonely to lonely subjects). that both
enotion- and probl em-focused coping were not associated positively
with affective distress. but rather increased with sel f-esteem and

absence of depression.

IondencjesJILAttflbuticLStxle

One interesting result was the finding that more subjects were
grouped into (a) unstable and external attributional style or (b)
stable and internal attributional style. This finding was not inde-
pendent of loneliness. in that nonl onely students did not demonstrate
a. preference for a particular attributional style regarding loneli-
ness.

Table 4.8 identifies the four attribution "cells" and the numbers
of people who were assigned to each cell. The col umns in the table
are associated with three different selections of students: the total
sample. the Lonely subsampl e. and the Nonlonely subsample. The per-
centages that each attribution cell number represents for the number

of subjects in the given cell are also provided.

108

Table 4.8: Proportion of Sanples in Attributional Groups.

 

 

Total Lonely Nonlonely
Samplea Subsampleb Subsample
Attribution (N = 199) (n = 49) (n = 58)
n 5 n % n %
Unstable/
internal 29 14.6 6 12.2 13 22.0
Unstable/
external 71 35.7 16 32.6 16 27.0
Stable/
external 24 12.0 9 18.4 15 25.0
Stable/
internal 75 37.7 18 36.8 14 24.0

 

Note: Grant Attribution subscal es were used as determinants of attri-
bution style (cells formed by median split method).

aFor the total sample: x2 (3. N = 199) = 43.87. p < .05.

t'For the Lonely subsanple: X2 (3. n = 49) = 7.87. p < .05.

Therefore. there was representation across all four attribution
styles. but there was preference for the unstable/external and stable/
internal orientations among the total sample and the Lonely subsample.
However. among a group of nonlonely subjects. their reported attribu-
ti ons about loneliness (answered from the perspective of imagining
that they were lonely) took on a little different pattern. Among this
group. the numbers per attribution cell appeared to be more similar.
and indeed there were no significant differences between cells (see

Table 4.8). Thus. attributions about loneliness appeared more varied

109

among the sample of students who were not lonely at the time of the

data collection.

flxoothesoLIostsd

Analyses of variance. along with some a priori contrasts. corre-
lational analyses. and paired sample .t-tests. were all forms of analy-
sis used to investigate the main hypotheses. The particular type of
analysis used for each respective hypothesis is indicated with the
presentation of results for each of the hypotheses addressed.

The first five hypotheses were all tested by means of an analysis
of variance. The model of the analysis is provided in Table 4.9. The
reader is referred to specific tables of data that are indicated for

each particular hypothesis.

W

Students having stable attributions of their loneliness will
show more use of emotion-focused coping than students who have
unstable attributions of their loneliness.

Null hypothesis stated in terms of the model of analysis:

Locus of stability will have no positive significant effect on
emotion-focused coping such that the mean value of emotion-
focused coping for people with stable attributions of their
loneliness will be the same as or less than the mean value of
such coping among subjects with unstable attributions.

Stated symbolically: HO: ”Al .5 ”AZ 5 ”E

”1‘ ”Al ’ UA2

Legend: Where U and ”AZ are mean scores for emotion-focused
coping for subje s with unstable and stable attributions.
respectively. and ”E is the overall mean score for emotion-
focused coping.

110

Table 4.9: Model of Summary Data Tables for Analysis of Variance:
Mean Scores for Coping Response

 

 

 

 

Factor A Factor B
Locus of Stability Locus of Causality
Internal External
Unstable U u u
Al 81 A1 2
(U1) ( 2 A1
Stable U 1 U U
(‘63? (‘64? .2
”31 ”32 U“

 

aIn hypotheses that follow.

”5 refers to U for emotion-focused coping

Up refers to U for problem-focused coping

“0 refers to difference between emotion- and problem-focused
coping
Linear model for two-way fixed effects analysis of variance:
X11=U+ a1+ 31+ 811
Where x1] is single observation
U is contribution of dependent variable
31 is contribution of Factor A

D1 is contribution of Factor B

e11 is error term for observation

 

111

W. The sample mean for emotion-focused coping
among subjects with stable attributions of their loneliness was 33.06.

and the sample mean for emotion-focused coping among subjects with
unstable attributions was 35.24. These two means did not prove to
be significantly different from one another. as there was no signifi-
cant main effect for locus of stability on emotion-focused coping:
E (1.195) = 2.45. p > .05. (See Tables 4.10 and 4.12.) Therefore.
the null hypothesis was not rejected. The analysis of variance per-
formed on the total sample did not suggest that locus of stability
affected emotion-focused coping to a significant degree. In addition.
the emotion-focused coping mean score for subjects with stable attri-
butions for their loneliness was lower than the mean score for stu-
dents with unstable attributions (rather than higier). indicating
further absence of support for the belief that stable attributions are
associated with increased amounts of emotion-focused coping. The
correlation of locus of stability with emotion-focused coping among
the total sample was 1: = -.20. .p < .01. There was a slight linear
relationship seen between unstable attributions for loneliness and
emotion-focused coping with loneliness. Thus. this correlation was
also indicative of a relationship between these two variables in the
opposite direction than was hypothesized in the alternate hypothesis.
.EQLthLLnnaJLsubaample. There was no significant effect found
for locus of stability on emotion-focused coping among the Lonely
subsample as well: E (1.45) = 2.87. p > .05. The group mean scores

of emotion-focused coping among students with unstable and stable

112

.o.. v.ma«

 

 

.mo. v In...
. . . . mc.>um mc.ooo
n. mm . om :N . cooZuon cococomw.o
. . . . mc.ooo
mm mm mum m.o com3o0maso.noLo
. . . . mc.ooo
.1 u.»O— m «...nwm N m.~ N UUmSUOmlcoquEm
... .n. u u
co_oumtouc_ ......u c..: >a__..=.u >o___n..m
eoc.nsoU to meoOA mo maoo. 0.3m.cm> oeoouso

co.um.cm> wo ooL30m

 

.c.oemm .oBOFV oocm.cm> mo mom>.mc< ecu. au.:mom to >Loessm no..: o.nmh

113

attributions. respectively. were 34.09 and 31.70. The main effect
approached significance (p = .097) but again in the reverse direction
to that hypothesized. (See Tables 4.11 and 4.15.) The direction of
difference supported the possibility that students with stable attribu-
tions for their loneliness had a tendency to engage in less. not more.
emotion-focused coping. The null hypothesis was not rejected. One
further note is that within this subsample of students. the correl a-

ti on between locus of stability and emoti on-focused coping was -.23 (p
= .05). This significant correlation demonstrated the existence of a
slight association between unstable attributions and increased use of

emotion-focused coping.

another-1.22

Students having stable attributions of their loneliness will use
less probl em-focused coping than students having unstable attri-
butions of their loneliness. based on the assumption that stu-
dents with stable attributions will feel less belief in the abil-
ity to make an impact on their loneliness.

Null hypothesis stated in terms of the model of analysis:

Locus of stability will have no significant effect on problem-
focused coping. such that the population mean for probl em-focused
coping associated with stable attributions will be the same as
(or more than) the population mean for probl em-focused coping
associated with unstable attributions.

Stated symbolically: HO: UAl _>_ ”AZ 2, Up

”v Um < UAZ

Legend: Where ”Al and ”AZ are mean scores for probl em-focused
coping for subjects with stable and unstable attributions. and Up
is the overall mean score for probl era-focused coping.

11h

.oocmo.m.cm.m mo m.o>o. ecmecmum use eocomoLoom o:.m>LM_Oz "ouoz

 

 

mm.. mm.. mm.~ mm. mo.>um mc.ooo
coozuon oococomw.n
. . . . mc.ooo
mmo m.m N. . am. comaoownso.noLm
. . . . mc.ooo
mm: m: . o: nm N vom360muco_uoEm
M ...-_ > u > M
......w c_mz s....=.u u___nmom
co.uomcouc_ eoc.neou to maooo .o m3604 o.nm.cm> decouso

co.um.cm> mo ooL30m

 

.o.oEmmnam >.oco.v oocm.om> mo mom>.mc< Econ mo.:mom mo >Lmesem “...: o.nmh

115

Table 4.12: Summary Table of Mean Scores for Emotion-Focused Coping
Behavior by Attribution (Total Sample)

Locus of Causality
Locus of Stability

 

Internal External
Unstable 34.34 35.60 35.24
(n=29) (n=7l) (n=100)
Stable 32.58a 34.54 33.06
(n=75) (ns24) (ns99)
Total 33.46 35.09 34.15
(33104) (3:95) (ne199)

 

aThis group was found to have significantly less emotion-focused
coping than the other three groups by means of an a priori contrast:
.1 (195) = -2.56. t prob. < .05.

Table 4.13: Summary Table of Mean Scores for Problem-Focused Coping
Behavior by Attribution (Total Sample)

Locus of Causality
Locus of Stability

 

Internal External
Unstable 15.44 15.59 15.55
(n=29) (ns7l) (nelOO)
Stable 14.98 15.95 15.22
(n=75) (ns24) (ns99)
Total 15.21 15.77 15.38
(n=104) (119-95) (n=l99)

 

Note: A priori contrast on testing comparison of unstable/internal
group with all other groups proved insignificant for the
unstable/internal group mean being higher for problem-focused
coping: t (195) = -.O8. 1 prob. > .10.

116

mm. There was no main effect for locus of
stability on problem-focused coping: therefore. the null hypothesis was

not rej ected; E (1.195) = .019. p > .05. The mean score for problem-
focused coping for students with stable attributions was 15.22. and for
students with unstable attributions. this mean score was 15.55. The
overall mean score for probl em-focused coping for the entire sample was
15.38. (See Tables 4.10 and 4.13.) Looking at the table of corral a-
tions (Table 4.3). one sees that there was little evidence of an asso-
ci ati on between locus of stability and problem-focused coping as indi-

cated by this analysis as well (I: = .07. n > .05).

Table 4.14: Difference Between Emotion-Focused and Probl em-Focused
Coping (Total Sanple)

 

Locus of Causality
Locus of Stability

 

Internal External
Unstable .29 .37 .34
(n=29) (n=7l ) (n=100)
Stable .22 .22 .21
(11:75) (n=24) (n=99)
Total .26 .30 .28
(n=104) (n=95) (n=l99)

 

Note: Difference scores were derived by subtracting a weighted
problen-focused score fron a weighted emotion-focused score for
each subject. This was done to account for unequal nunbers of
items in each of the coping subscales. The difference score
means are therefore the means of the weighted difference scores.
The weigited difference scores can have values fron -3 to 3.
with -3 representing total endorsenent of all problem-focused
items to the exclusion of emotion-focused items: +3 represents
the opposite endorsenent pattern.

117

W There was no significant effect seen
for locus of stability on probl em-focused coping for this subsample as
well: E (1.45) = .134. p > .05. The mean score of problem-focused
coping for students with unstable attributions was 15.63. and the mean
score for students with stable attributions. 14.81. (See Tables 4.11
and 4.16.) The null hypothesis was not rejected. Lonely students did

not appear to engage in a higier level of problem-focused coping if
they had unstable attributions of their loneliness.

Table 4.15: Sunmary Table of Mean Scores for Emotion-Focused Coping
by Attribution (Lonely Subsample)

 

Locus of Causality
Locus of Stability

 

Internal External
Unstable 35.83 33.43 34.09
(n=6) (n=l6) (Jr-=22)
Stable 31.72 31.66 31.70
(n=1 8) (n=9) (n=27)
Total 33.77 32.54 32.77
(n=24) (n=25) (n=4 9)

 

Note: An a priori contrast done to see whether the group mean for
unstable/internal subjects had a significantly higier use of

emotion-focused coping was not significant; 1‘. (45) = -1.l9.
.t prob. = .24.

118

Table 4.16: Sunnary Table of Mean Scores for Probl em-Focused Coping
by Attribution (Lonely Subsample)

 

Locus of Causality
Locus of Stability

 

Internal External
Unstable 15.00 15.87 15.63
(11:6) (n=l6) (n=22)
Stable 14.38 15.66 14.81
(n=18) (n=9) (n=27)
Total 14.69 15.76 115.18
(,n=24) (n=25) (n=4 9)

 

Note: An a priori contrast done to see whether subjects with
unstabl e/internal attributions for loneliness engaged in a
significantly higher level of problem-focused coping proved
insignificant: t (45) = -.21. _t prob. = .83.

The correlation of problem-focused coping with locus of stability
for this subsample was not significant as well (n = .14. ,p > .05).
adding further support to the belief that unstable attributions are not
associated with problem-focused coping to a greater degree than stable
attributions. In fact. the mean scores of problem-focused coping for
all four attributional-style groups (see Table 4.16) were all within
'L5 points of one another. demonstrating that the range seen in
problem-focused coping (see Table 4.2) was not attributable to varia-

tion imposed by the locus of stability.

119

Hummus-.1

Students with stable and internal attributions of their loneli-
ness will use more emotion-focused coping than students who engage
in one of the other three attributional styles (unstable/inter-
nal. stable/external. or stable/internal). This is based on the
expectation that these students will feel sel f-blame for their
loneliness. and will see the situation as hopeless.

Null hypothesis stated in terms of the model of analysis:

The means for emotion-focused coping for the four stated
attribution groups will be estimates of the same population
mean. (All means are fron one population.)

Stated symbolically: Ho: U1. U2. U3. U4 the same or less than
population means

”1: U1) U2: U3: U4

Legend: Where Ux is the population mean for emotion-focused
coping for each attribution group. and U1 15 the population mean
for emotion-focused coping of the stable and internal attribu-
tion population.

M. An a priori contrast of the data yielded a

significant difference in emotion-focused coping between stable/
internal attribution group and the other three attribution groups: :1:
(195) = 2.56. p < .05. This difference was in the opposite direction
from that hypothesized. Emotion-focused coping was found to be sig-
nificantly lower for the internal/stable group than for the other
three groups. rather than higher. with a mean emoti on-focused coping
score of 32.5. (See Table 4.12 for summary of data.) Of further
interest is that the unstabl e/external group had the higiest use of
emotion-focused coping. in terms of the raw data. Therefore. the null

hypothesis was not rejected. There appeared to be slightly more

120

emotion-focused coping on the part of people who had unstable and
external attributions of their loneliness. Thus. people who felt that
their loneliness was caused by external circumstances and was change-

able tended to use more palliative measures of coping.

Wu“. An a priori contrast of the four
attributional groups with respect to emotion-focused coping on the

Lonely subjects yielded no significant differences in emotion-focused
coping dependent on attributional style: :1; (45) = -1.19. ,p > .05.
(See Table 4.15 for summary of data.) The null hypothesis was not
rejected.

Months-3154

Students who have unstable and internal attributions of their
loneliness will engage in more problem-focused coping than
students in the other three attributional groups. based on the
assumption that these students should feel a sense of control and
motivation over a phenomenon they see as changeable.

Null hypothesis stated in terms of the model of analysis:

The mean of the population of students having unstable and
internal attributions of their loneliness will be the same as or
less than the means of students from the other three populations
representing the other three attributional groups.

Stated symbolically: Ho: U1 5. U2. U3. U4

H]: U] > U29 U3: U4
Mammals. An a priori contrast of the data yielded
no significant differences among attributional group means with

respect to probl em-focused coping: t (195) = -.0879. p > .05. (See

121

Table 4.13 for summary of data.) This result was consistent with the
finding that there was no significant combined effect of locus of
stability and locus of causality on problem-focused coping. (See Table
4.10.) The null hypothesis was not rejected. Probl em-focused coping
among students did not appear to increase significantly with an
unstabl e/internal approach to one's loneliness. According to this
analysis. all four attributional groups appeared to engage in a simi-
lar level of problem-focused coping. Therefore. the group means were
considered estimates of the same population parameter.
W- No significant difference was seen
between the unstable/internal attributional group for probl em-focused
coping contrasted with the other three groups: 1 (45) = -.20. p > .05.
(See Table 4.16 for summary of data.) The means among the four
attributional groups of the Lonely subsample were also very close in
value for problem-focused coping. Thus. the null hypothesis was not

rejected for this subsample as well.

Wheels:

Students who have stable attributions of their loneliness will use
more emotion- than problem-focused coping than students with
unstable attributions. This is based on the assumption that
stability is associated with a lack of expectation that things
can change. thus discouraging a probl em-focused orientation and
encouraging sel f-ccmforting measures of coping.

Null hypothesis stated in terms of the model of analysis:

Locus of stability will have no significant effect on difference
between emotion- and probl em-focused coping. in the direction of
stable attributions being associated with greater difference
between coping styles.

122

Stated symbolically: Ho: ”Al 5 ”AZ 5 UD

”1= Um UA2
Legend: Where ”Al and ”AZ are the population means for the
difference scores between emotion- and probl em-focused coping for
those with stable and unstable attributions of their loneliness.
00 is the population mean for difference between coping style
for all students.

W. Analysis of variance indicated no signifi-
cant effect for locus of stability on difference in use of each coping

style: E (3.195) = 1.23. p > .05. The outcome variable of difference
in use of coping style was measured by difference scores. These
difference scores were derived by subtracting a weighted problem score
from a weighted emotion score. and the scale that resulted ranged from
-3 to 3. It was necessary to calculate weighted difference scores
since the subscales for emotion- and problem-focused coping had
unequal numbers of items. The mean difference scores for the stable
and unstable groups of subjects were .21 and .34. respectively. show-
ing a slightly greater use of emotion-focused coping. (See Table
4.14.) The null hypothesis was not rejected. Students appeared to
engage in similar levels of emotion- and problem-focused coping.
irrespective of their locus of stability orientation.
WEI-(Mills. Analysis of variance indicated no
significant effect for locus of stability on difference in use of each
coping style for this subsample as well: E (1.45) = .98. p > .05. The
mean difference scores for the stable and unstable groups of students

among the Lonely subsample were .17 and .23. respectively. (See Table

123

4.17 for summary of data.) The null hypothesis was not rejected on

the basis of this analysis also.

Table 4.17: Difference Between Emotion- and Problem-Focused Coping

(Lonely Subsample)

 

Locus of Causality

Locus of Stability

 

 

Internal External
Unstable .48 .14 .23
(n=6) (n=1 6) (n=22)
Stable .24 .02 .17
(n=18) (n=9) (n=27)
Total .36 .08 .20
(n=24) (n=25) (n=49)
Note: See note to Table 4.14 for explanation of how these mean scores
were derived.
onotbesisj

Null

Students who have unstable and internal attributions of their
loneliness will show more use of problem-focused than emotion-
focused coping. based on the assumption that people with unstable
and internal attributions will feel both motivation and personal
control over their problen of loneliness.

hypothesis stated in terms of the model of analysis:

This hypothesis was investigated by making an inference about the
means between emotion- and probl em-focused coping for people

in the given attributional group. This was done by means of a
difference score between the two types of coping for this
attributional group. Thus. the null hypothesis was stated as
follows:

124

There will be no difference between the population means of
emotion- and probl em-focused coping for people with unstable and
internal attributions of loneliness. as the mean of differences
scores will be equal to zero.

Stated symbolically: Ho: 11 = 0

H]: D < 0 (stated as "1 ess than zero" due to
the direction in which the difference
scores were computed for the sample data)

Legend: D is the population parameter of difference between

motion- and probl em-focused coping (a negative number signifies
more problem- than auction-focused coping).

W A paired samples .t-test run on the mean
difference scores for students with stable and internal attributions
yielded nonsignificant results. These difference scores were derived
by subtracting a weigited probl em-focused coping score from a weigited
emotion-focused coping score. and the scale that was derived ranged
from -3 to 3. The mean difference score for students with unstable
and internal attributions of their loneliness was .28; _t (74) = .03.
p > .05. There was very sligit favoring of emotion- over problem-
focused coping that was not meaningful in degree. This was not in the
direction expected. and therefore the null hypothesis was rejected.

EoLJheJoneJLsubsampJe. The paired samples t-test run on the
difference scores of students within the Lonely subsample who had
stable and internal attributions of their loneliness yielded
nonsignificant results for differences between emotion- and problem-
focused coping. The mean difference score was .48: j; (17) = .61. p >
.05. Again. a very sligit favoring of emotion-focused coping over
problem-focused coping was seen. but at a nonsignificant level. The

null hypothesis was not rejected.

125

W

Students who have stable attributions of their loneliness
(measured by the Grant Attribution Scale) will experience greater
degrees of loneliness (as measured by the UCLA Loneliness Scale
or the Likert scale) than students who have unstable attributions
of their loneliness.

Null hypothesis stated in terms of the model of analysis:
Locus of stability will not bear a positive linear relationship
with loneliness. such that the tendency to see loneliness as a

stable phenomenon will not correlate with an increase in degree
of loneliness experienced.

Stated symbolically: Ho: ny < .20

(.20 selected as the lowest correlation acceptable for purpose of
significance. given that lower correlations. irrespective of
significance. indicate extrenely sligit relationships)

W A correlation between loneliness (as
measured by the UCLA Loneliness Scale) and locus of stability was .41:
p = .001. The correlation among the total sample for the degree of
loneliness as measured by the subjective indication. the Likert scale
rating. with locus of stability (also measured by the Grant Attri bu-
ti on Scale). was .43: p = .001.

W. The correlation between scores on the
UCLA Loneliness Scale and the Grant Locus of Stability subscale was
.30: p < .05. The correlation within this subsample for the Likert
scale measure of loneliness and the same locus of stability measure was
.21; p > .05.

The correlations were all in the direction of indicating a

positive relationship between stable attributions of loneliness and

126

increased degree of loneliness. It appeared that stable attributions
of loneliness. attributions that saw loneliness as difficult to affect
or change. were associated with an increased degree of reported

loneliness.

uxootbesisi

Students who have stable and internal attributions of their
loneliness will experience fire greatest degree of loneliness
compared to students who have either stabl e/external attribu-
tions. unstable/internal. or unstable/external attributions.
This is based on an expectation that stable and internal attribu-
tions lead to a sense of hopelessness or passivity for dealing
with problems.

Null hypothesis stated in terms of the model of analysis:

This hypothesis was investigated by use of an analysis of
variance a priori contrast. The null hypothesis was that there
will be no difference in the degree of loneliness between stu-
dents who have stable and internal attributions of their loneli-
rress and students who have stable/external. unstable/internal.
or unstable/external attributions of their loneliness.

Stated symbolically: Ho: U1 = U2. U3. U4

H]: U1> U2: U3: U4
Legend: U] is the mean of loneliness for the population of
students having stable and internal attributions. 02. U3. and U4

are means of loneliness for fire populations of other
attributional groups.

WW. There were sigrificant differences found
among the means for loneliness (UCLA measure) for the four attribu-
ti onal groups; 1 (194) = 4.92. p < .001. An a priori contrast done of
their mean loneliness scores indicated that fire group with stable and
internal attributions of loneliness had a significantly higrer mean
loneliness score (44.6) than the other three groups. Thus. fire null

hypothesis was rejected based on this analysis. (See Table 4.18.)

127

Table 4.18: Loneliness (UCLA Scale) by Attributional Style: Total

 

 

Sample
Group Count Mean SD
Unstable/internal 29 40.75 9.62
Unstable/external 71 35.43 6.80
Stable/external 24 37.54 9.60
Stable/internal 75 44.67 10.17
Total 199 39.92

 

Note: An a priori contrast analysis conparing fire stable/internal
group with the other three groups reached significance:
I (194) = 4092' a < .0010

This hypothesis was also investigated with the use of another
rneasure of locus of causality. the direct question on the student
information sheet asking students whether they thoth that their
loneliness was caused by internal or external factors. Students
placed into attributional groups by means of their responses.to this
question and their scores on the Grant Attributional Scale for locus of
stability yielded four attributional groups as well (n = 130 for this
sample). Students in the stable and internal attribution group by
this group formation also had a mean score of loneliness on the UQA
measure that was significantly higrer (44.60); 1; (126) = 2.85. p, < .05
than mean scores for the other three groups. (See Table 4.20.)

The null hypothesis was rejected. given firat the mean score on
loneliness for the sample of students with stable and internal

attributions suggested that they did indeed experience a greater

128

degree of loneliness and hence came from a different population than
did the other students.

.Egn_the:Lgnely_subsample. According to an a priori contrast
analysis. the group representing students with internal and stable
attributions for fireir loneliness had a significantly higrer mean score
for loneliness (55.66) compared to fire ofirer three attributional
groups: 1 (45) = 2.36. p < .05. Thus. students who had stable and
internal attributions for their loneliness were found to have a higher
degree of loneliness than students in the other three attributional
groups. (See Table 4.19 for summary of data.) The null hypothesis

was rejected on the basis of this analysis as well.

Table 4.19: Loneliness (UCLA Scale) by Attributional Style: Lonely

 

 

Subsample
Group Count Mean 75D
Unstable/internal 6 54.50 4.03
Unstable/external 16 49.25 8.66
Stable/external 9 46.44 11.58
Stable/internal 18 55.66 4.97
Total 49 51.73

 

Note: An a priori contrast conparing fire stable/i nternal group with
the other three groups reached significance: 1'. (45) = 2.36. p <
.05.

129

Table 4.20: Loneliness (UCLA Scale) by Attributional Style--Tota1
Sample (With Locus of Causality Measured by Causality
Question on Student Loneliness Questionnaire)

 

 

Group Count Mean .SD
Unstabl e/i nternal 19 38.10 8.37
Unstabl e/external 41 37.68 8.79
Stable/external 41 44.02 10.38
Stabl e/internal 29 46.00 11.54

Total 130 41.60

 

Note: An a priori contrast conparing fire stable/internal group with
the other firree groups reached significance; 1; (126) = 2.85. p
< .01.

methesisi

As the degree of loneliness increases. the use of emotion-focused
coping will increase. based on fire belief that students who are
lonelier will need to engage in more palliative measures of sel f-

comfort.
Null hypothesis stated in terms of the model of analysis:

There will be a positive linear relationship between emotion-
focused coping and loneliness. with loneliness measured by the
UCLA Loneliness Scale.

Stated symbolically: Ho; P < .20

H]: ny > '20

(.20 chosen as least acceptable degree of correlation considered
meaningful)

Legend: P is the population correlation coefficient for x.
emotion coifing. and y. loneliness. two variables assumed to have
a bivariate normal distribution.

W. The correlation for the total sample was
-.11 at the .05 level of significance. indicating that an extremely

130

slight relationship was found in the direction of emotion-focused
coping relating inversely with degree of loneliness.

EoLtbLLonoJLsuhsample- The correlation found for this
subsample was -.02 at a nonsignificant level.

This hypothesis was also investigated with fire use of the Likert
scale of loneliness as another means of testing this hypothesis. The
correlation between loneliness. as measured by the Likert scale. and
emotion-focused coping for fire total sample was -.22. p < .01. For
the Lonely subsample. this correlation was -.20. p > .05. On the
basis of these analyses. the null hypothesis was not rejected. as
there was no evidence that firese two variables in question correlated

linearly in a positive direction.

EostJoanalxsos

Multiple linear regressions were run on loneliness (as measured
by the UCLA Scale) to gain further insigrts into the relationships
between loneliness. attribution. and coping behavior. For the total
sample. a hierarchical regression equation was run. including the
variables of problem- and emotion-focused coping; stability. causal-
ity. and their interaction; sel f-esteem; and depression. These varia-
bles togefiier accounted for 46% of the variance on the UCLA Scale. and
the regression equation was significant: E (7.190) = 23.48. p < .001.
(See Table 4.21.) The variable accounting for the most variance in
loneliness was locus of stability (stable attributions). which con-
tributed 17% of the variance to the total amount. The partial regres-

sion coefficient did not reach significance. Low sel f-esteem

131

accounted for 16% of the variance on loneliness. This partial regres-
sion coefficient did reach a level of significance: E (7.190) = 24.65.
p, < .001. Variance accounted for by fire ofirer variables totaled 13%.
and no one variable was responsible for more than 7% variance on
loneliness. The partial regression coefficient for locus of causality
(more internal attributions associated with increasing loneliness)

also reached significance. with E (7.190) = 7.52. p < .001.

Table 4.21: Regression of Variables on Loneliness: Total Sample

 

 

Variable R2 R2 Change aa
Locus of stability .17 .17 -.Ol
Locus of causality .24 .07 -.18
Interaction of
stability/causality .25 .01 .07*
$6] f-BStean 042 016 - 035
Depression .46 .04 .25
Emotion-focused
coping .46 .OO -.02
Problem-focused
coping .46 .00 .O6*

 

Note: Hierarchical multiple regression procedure. Loneliness meas-
ured by the Revised UCLA Loneliness Scale.

a Standardized regression coefficients.

Q. < .05.

132

As can be seen from Table 4.21. locus of stability was entered
first in the equation. with the expectation firat stable attributions

may predispose one to being more vulnerable to loneliness. Although

from the tabled variance changes (R2) it appears that locus of
stability accounted for fire largest proportion of variance in loneli-
ness. computing the partial correlation coefficient for fire ofirer
variables revealed that low sel f-esteem accounted for 22% of the
unexplained variance on loneliness after locus of stability had been
entered into the equation. The reader is rerrrinded that locus of
stability and low self-esteem had a correlation of .46. p < .001 for
the total sample. These two variables shared some common variance and
contributed significantly to the variance of loneliness as measured by
the UCLA Scale (for the total sample).

The same variables were entered in a hierarchical regression
equation for the prediction of loneliness as measured by the UCLA
Scale on the Lonely subsample. These variables accounted for 26% of
the variance on the dependent variable. (See Table 4.22.) Locus of
causality accounted for 12.6% of the variance on the dependent vari-
able. and locus of stability was responsible for 9.5% of the variance.
This regression equation was not significant: E (1.47) = 4.27. p >
.01. The relationships between locus of stability and locus of caus-
ality with loneliness were in fire direction of stable and internal
attributions being associated with increased degrees of loneliness.
The partial regression coefficient for locus of causality was signifi-

cant; E (1047) g 406], n < 0050

133

Table 4.22: Regression of Variables on Loneliness: Lonely Subsample

 

 

Variable R2 R2 Change Ba
Locus of stability .09 .09 -.O3*
Locus of causality .22 .12 -.41
Interaction of
stability/causality .22 .00 .OO
Self-esteem .23 .01 -.10*
Depression .24 .01 .12"
Emotion-focused .24 .00 -.03*
coping
Problem-focused .26 .02 .13
coping

 

Note: Hierarchical multiple regression procedure. Loneliness meas-
ured by the Revised UCLA Loneliness Scale.

aStandardized regression coefficients.

in < .05.

Hierarchical regressions on length of time lonely failed to
prove significant for eifirer sample. The regression equation for the
Lonely subsample approached significance; E (44) 8 1.97. ,p = .095. Of
the variables that were entered into the equation. the only one that
contributed a significant degree of variance was the degree of loneli-
ness (14% of the variance on length of time lonely). The prediction
was in the direction of length of time lonely being predictive of
a gradually decneasing degree of loneliness. thus suggesting that

134

chronic loneliness may be of a less intense degree than shorter epi-
sodes of loneliness. which may be acute and dealt with more effectively
or quickly as a result.

Post hoc analyses on loneliness by attribution were done to
further investigate possible differences between groups (separate from
Hypothesis 8 pertaining to only one group). These indicated that the
unstable/external group of students within both samples had signifi-
cantly lower UQ.A loneliness scores than students in the other three
groups. (See Tables 4.23 and 4.24.)

Table 4.23: Attributional Style by Loneliness (Revised UCLA Scale):
Total Sample

 

 

Group Count Mean .SD ‘ErRatio .ErProb.
Unstable/
internal 29 40.75 9.62 13.577 .000
Unstable/
external 71 35.43 6.80
Stable/
external 24 37.54 9.60
Stable/
internal 75 44.67 10.17

Total 199 39.92

 

Note: The Scheffe method of post hoc comparisons identified three
subsets of homogeneous groups:

Subset l: Unstable/external. stable/external
Subset 2: Unstable/internal. stable/external
Subset 3: Unstable/internal. stable/internal

135

Table 4.24: Attributional Style by Loneliness (Revised UCLA Scale):
Lonely Subsample

 

 

Group Count Mean 75D .ErRatio ErProb.
Unstable/
internal 6 54.50 4.03 3.745 .0174
Unstable!
external 16 49.25 8.66
Stable/
external 9 46.44 11.58
Stable/
internal 18 55.66 4.97
Total 49 51.73

 

Note: The Scheffe method of post hoc comparisons identified two
subsets of homogeneous groups:

Subset l: unstable/internal. unstable/external. stable/external
Subset 2: unstable/internal. unstable/external. stable/internal

DittenonooLBotxesILLoneJx
WW

Investigations of the differences between lonely and nonlonely
groups were conducted to derive more information about loneliness. It
was hypothesized that there would be significant differences between
lonely and nonlonely students with respect to depression. self-esteem.
and age (with younger students experiencing more loneliness). Other

variables of interest for exploration purposes were attributional

136

style. living environment (i.e.. off campus. residence hall). number
of siblings in family of origin. and sex.

I-tests for independent samples were run. comparing fire lonely
group of students (the Lonely subsample) with the nonlonely group of
students (Nonlonely subsampfle). Significant.trtests are described
below.

There were significant differences found with respect to attribu-
tional style. depression. and sel f-esteem. Table 4.25 provides the
statistical information. As can be seen by investigating the mean
values and I-probabilities. lonely students were higher for locus of
causality. in the direction of internal ity. higher on locus of stabil-
ity for the direction of stable attributions. lower in selfeesteem.
and more depressed.

There were no differences found between the two groups with
respect to age. sex. year in school. living arrangements. or nunber of
siblings present in the household. Nor were there any significant
differences found in the level of emotion- or problem-focused coping
reported between the two groups. taking into consideration the fact
that the nonlonely group was reporting "retrospectively" with respect

to describing their coping behaviors when lonely.

137

Table 4.25: Conparisons of Lonely and Nonlonely Subjects

 

 

Two-Tailed
Variable Mean Value .IEValue .p-Valuesa
Locus of causality
Nonlonely group 29.72 -5.O4a <.OOO
Lonely group 24.83
Locus of stability
Nonlonely group 30.99 5.65a <.000
Lonely group 24.65
Selfiesteem
-Non1onely group 20.28 -9.50a <.OOO
Lonely group 12.14
Depression
Nonlonely group 5.65 7.36a <.ooo
Lonely group 15.02

 

a_I-val ues for pooled variance estimates were used when E-ratios
for variances were not significant at the .05 level or less. I-val ues
for fire separate variance estimates were used when the E-ratio for
variance proved significant at the alpha level (Norusis. 1982. p. 47).

WW
There was a significant difference found between attributional

groups with respect to emotion-focused coping among the total sample
by means of a post hoc comparison analysis. Students having internal
and stable attributions for fireir loneliness actually used less
emotion-focused coping than other students. and students with unstable
and external attributions used mom emotion-focused coping: E (3.195)
= 3.45. p < .05. (See Tables 4.26 and 4.27.) Again. this difference

138

was not large. and yet it was interesting because it was in the
opposite direction to that anticipated.

It is also interesting that probl em-focused coping did not differ
among any of the four attributional groups. althougr there was one
group that appeared less lonely in both fire total sample and in the

Lonely subsample (the unstabl e/internal group). (See Table 4.26.)

Table 4.26: Coping by Attributional Style

 

 

Emoti on-Focused Probl em-Focused
Coping Coping
A. Man:
Total sample 34.15 15.38
(possible score (possible score
range 12-48) range 6-24)
Lonely subsanple 32.77 15.18
3- Wimp
Total sanple
Unstabl e/ i nternal 34.34 15 .44
Unstable/external 35.60* 15.59
Sta b1 e/external 34.54 15 .95
Sta b1 e/i nternal 32.5 8" 14 .98
Lonely subsample
Unstabl e/internal 35 .83 15 .OO
Unsta bl e/external 33 .43 15 . 87
Sta bl e/external 31. 66 15 .66
Stabl e/internal 31.74 14.38

 

aOn Scheffe post hoc comparisons. results of nonsi gnificance were
E (195) = .610. p > .10 for the total sample and E (45) = .949. p >
.10 for the Lonely subsanple.

'Group was significantly higrer for anation-focused coping.

'"Group was significantly lower for emotion-focused coping.

139

Table 4.27: Emotion-Focused Coping by Attributional Style: Total
Sample (d1 = 3.195)

 

 

Group Count Mean .SD ErRatio ‘ErProb.
Unstable/
internal 29 34.34 5.26 3.450 .0176
Unstable/
external 71 35.60 6.05
Stable/
external 24 34.54 5.40
Stable/
internal 75 32.58 5.65
Total 199 34.15

 

Note: The Scheffe method of post hoc comparisons identified two sub-
sets of homogeneous groups:

Subset l: unstable/internal. stable/external. stable/internal
Subset 2: unstable/internal. unstable/external. stable/external

The coping strategies used with the most frequency by the sub-
sample of lonely students were:

*1. Tell yourself that your loneliness would not last forever.
*2. Remind yourself that you actually do have good relation-
ships with other people.
3. Take naps or sleep longer at night to get away from the
feelings of loneliness.
4. Go for a walk by yourself.
*5. Try to figure out why you are lonely.
6. Spend time with a close friend just to be together.
7. Distract yourself from feeling lonely with a quiet activity
such as reading a book. listening to music. or watching TV.
*8. Think about things you can do to overcone your loneliness.
**9. Work extra hard to succeed at some activity (such as study-
ing extra hard for an exam. putting extra effort into prac-
ticing an instrument. etc.).

i"Items appearing on Emotion-Focused Scale.
**Item appearing on Problem-Focused Scale.

140

These coping items represent both thinking and action types of
strategies. The thinking strategies. while considered emotion-focused
strategies in this study. nonetheless are oriented to comforting an
individual in a constructive manner. The behavioral strategies
endorsed above are both active-constructive and passive in nature
(such as sleeping or watching TV).

There was a relatively low amount of endorsement of what may be
considered passive-destructive kinds of coping. such as drinking.
going to bars. or becoming sexually involved for companionship. There
was also an absence of certain very constructive strategies. such as
seeking help from a therapist or doing something to improve social
skills. The least endorsed item was seeking help fron a counsel or.

I-tests performed on gender to explore differences between males
and females with respect to coping behavior indicated that females
engaged in significantly more emotion-focused coping than did males.
although this difference was very small. The mean for females was
37.9. and the mean for males was 35.1 (1; [197] = -3.21. p < .01).
Females also had a higrer mean for probl em-focused coping (3.81)
compared to males (29.1): 1(197) = -3.76. p < .001. Again. this
difference. althougr significant. was very slight. (See Table 4.28.)

Lonel ier students appeared more lethargic and withdrawn in their
coping choices than did students in general. For example. lonely
students resorted to more sleeping and overeating compared to the
general student sample. Forty-three percent of fire lonelier sample

endorsed the item of frequent sleeping on an "often" basis compared to

141

25% of fire total sample. Only 51% of fire lonelier students stated

that they engaged in some type of sports activity "sometimes" or

"often" as a way to cope with their loneliness. compared to 65% of fire

total sample. Thirteen percent of the lonelier students endorsed the

"sports activity" item on an "often" basis compared to 26% of the

total sample.

Table 4.28: I-Tests: Comparison of Lonely and Nonl onely Groups (311 =

 

 

105)
Lonely Not Lonely Two-Tailed

Variable (n=47) (n=10) I-Val ue
Age 19.10 19.30 -.69
Year in school 2.00 2.10 -.45
No. of siblings in fanily 3.60 3.50 .25
UCLA 52.55 30.94 18.40*
Emotion-focused coping 32.90 34.20 -l.71
Problem-focused coping 15.10 15.80 .83
Locus of stability 24.40 30.90 -5.81*
Locus of causality 24.50 29.70 -4.94*
I'lSQ-Negativea 4 .54 3 .91 -4.94*
Depression 15.06 5.65 7.36*
Sel f--Esteemc 12 .00 20 .33 -9.50*

 

aAttributional Style Questionnaire-Negative Situation items. The
higher the score. the greater the tendency toward internal. stable.

and global attributions.

bDepression measured by Lubin Adjective Checklist. The greater

the score. the higrer the level of depression.

CSeTf-esteem measured by the Coopersmith Sel f-Esteem Inventory.
The higrer the score. the greater one's sel f-esteem.

in < .001.

142

Lonelier students used certain positive strategies less often
than did students as a whole. For example. in response to the coping
strategy 'Remind yourself that you actually have good relationships
with other people." 36% of fire lonelier sample indicated use of this
strategy "often" compared to 49% of the total sample. With respect to
the coping strategy of telling oneself that one's loneliness will not
last forever. 40% of the Lonely subsample endorsed this item as a
frequently used strategy compared to 54% of fire total sample. A
third positively oriented item. "Thought about good qualities that you
possess." all lonely students (in the Lonely subsample) endorsed this
item. However. 57% of this subsample endorsed the item as being used
"sonetimes" or "often" conpared to 84% of the total 55in a.

Hierarchical regressions that were run on both emotion- and
probl em-focused coping were indicative of the "imperviousness" coping
seemed to demonstrate in this study to other variables. as these
variables were defined and measured. The multiple linear regression
conducted on emotion-focused coping for both samples resulted in less
than 12% variance of emotion-focused coping being explained by the
prediction equation (E [189] = 1.68. ,p > .05 for the total sample: E
[44] = 1.43. p > .05 for the Lonely subsample). The multiple linear
regressions run on probl em-focused coping resulted in less than 6% of
the variance of probl em-focused coping explained by fire prediction
equations (E [189] = 1.83. ,p > .05 for the total sampl e: E [44] = .58.
p > .05 for the Lonely subsample). Neither procedure seemed to

improve on the minimal relationships found with the Pearson

143

correlations. Thus. loneliness. attribution. depression. and sel f-
esteem were not identified as probable predictors of coping behavior.
as measured in terms of sel f-reported problem- and emotion-focused

coping.

ClustoLAnast-es

Two different cluster analyses were performed on the data to
obtain more descriptive information on the sample. One such analysis
was done on the variables included in the study. There were eigrt
variables identified by the method of cluster analysis. Interesting
groupings of variables were the following:

1. Loneliness (measured by the UCLA Scale). low sel f-esteem.
depression. internal and stable attributions for negative situations
(the Selignan Attribution Style Questionnaire). and loneliness caused
by few friends constituted one cluster.

2. A second cluster comprised emotion-focused coping. probl em-
focused copi ng. and being fenal e.

3. The Grant Attribution measures of locus of stability and
locus of causality clustered with reason for loneliness attributed to
the lack of 91 rllboyfriend. Given that this is a commonly endorsed
reason for being lonely. students' attributions about their lonely
situation were associated with their perceived reason for being
lonely.

4. A fourth "logical" cluster was the grouping of age and year
in school with homesickness and missing certain people who are far

away .

144

All of the eigrt variables combined accounted for 12% of the
variance among variable distance for the 25 variables entered into the
cluster analysis. These clusters all had eigenvalues between .91 and
.97.

A second. more revealing cluster analysis was performed on all of
the observations in the sample. This analysis yielded five clusters
that bear interesting relationships among 11 variables that were
entered into the cluster analysis. The reader is referred to Table
4.29 for a summary table of the statistics for this cluster analysis.
The analysis was done by the centroid method of clustering by distance
among scores on all variables per observation.

All five groups identified in the cluster analysis are worthy of
more discussion because of fire distinguishing characteristics of their
results. One such cluster comprised 12 students who had the highest
mean score for loneliness. a score that was over 1 standard deviation
above the mean for fire total sample (UCLA measure) and had fire higrest
Likert rating score as well. This cluster group was characterized as
having (a) the most stable and internal scores on the attribution
measures. (b) more stable attributions for negative situations. (c)
the lowest sel f-esteem. and (d) the highest depression scores as well.
These students had coping scores that were below the mean for fire
total sample. including the lowest coping score for probl em-focused
coping. There was little difference between the levels of emotion-

and problem-focused coping used by firese subjects.

1115

 

meo.aoLo .chmLoo >cme mc.>m:

 

mm. .m. No.. o.~ >u.caoomc. we moc..ooo .m

. . . . o... .0 mc_c.oe soc; o.=o¢
mm mm so . o N :36. eo~.mrommco>m :. on ozone .s

eco.om.c.u\oco.cu>on mc.>mg uoz

mm. mm. o... o." .=o_aaa_caoa to >.___a.om
mco.u:n_cuum mo >u..mccouc. .o

uomucOU .mexom mo xom.

um. Na. w~.. o.m czou ..mEm c. o: ctoLo
>._Em. c. mac..a.m .0 Loose: .m

nee—note ecum.ocr.oo;om mc.>oc Lo .oo£um mc.x.. .02

«m. me. on._ o.m >o_u c_ a: axons
moon .mcac c. a: ozone .:

. . . . mc.ooo eomaoo.-Eo.noLa
am m: m: . o m mc.ooo eomaoo.-co.uosm .m

>m3m cm. o.oooo mc.mm.t

. . . . xo.moeoz m:..oou

R 3 mm _ o s .2
.oo;Om c. Lmo> .N

um..xuocu o>.uuono< co.mmocooo

. . . . >L0uco>c. Eooumm-._om

_m as - N: N o m .moz-am<
<49: ..

o:.m>com.m eoc.m.oxu eoc.m.oxu co.um.cm> Loam: u
ecooom co.ucoooco co.um.cm> Loom:.u .

 

mo.nm.cm> mo m.m>.mc< Leum:.u "au.: o.nmb

146

A second cluster (n = 30) was the next loneliest group. with a
mean UCLA group score also beyond 1 standard deviation above fire
sample mean. but a Likert rating score just slightly above the sample
mean. These subjects had a sel f-esteem mean score for their group
that was below the mean score for the total sample. and therefore did
not appear to be suffering particularly from low selfeesteem. The
cluster mean scores on attributions for their loneliness*were in the
stable and internal direction. like fire first cluster. The mean scores
for both emotion- and probl em-focused coping were sligrtly above the
mean values for the total sample.

A third cluster (n = 52) of subjects was characterized by (a)
having a group mean score on the UCLA measure for loneliness that was
slightly hig‘rer firan the total sample mean. (0) moderate sel f-esteem.
(c) coping levels below the sample means. and up locus of stability
and causality means that were in the direction of stable and internal
attributions of loneliness. Like the first cluster discussed. this
group hadlnoderate levels of coping behavior. along with an internal
and stable attribution approach to loneliness.

A fourfir cluster group (n = 69) was characterized by a mean UCLA
Loneliness score well below the sample mean. This group had a high
sel f-esteem score. well above the sample mean. and cluster means for
attributions in the direction of instability and externality. This
cluster had the highest emotion- and probl em-focused coping scores of

the five clusters.

147

The fifth cluster (n = 36) was not a lonely cluster either. The
mean UCLA score was 29.8. This group had the highest sel f-esteem
score and the lowest depression score. but low coping scores. below
the total sampfle mean scores. This group also had lows of stability
and causality mean scores in the direction of unstable and external
attributions.

The cluster analyses indicated that indeed there were different
groups of subjects among the entire sample with respect to coping.
loneliness. attribution. and affective states. (See Table 4.30.)
Loneliness-was seen to relate to attribution in that all of the lonely
clusters.(three) had mean attribution scores in the direction of
stable and internal attributions. Mean scores on the Attributional
Style Questionnaire also corresponded accordingly. Loneliness was
higrest among the cluster group characterized by the most internal and
stable attributions for loneliness. The locus of stability dimension
of attribution seemed to play a critical role in relation to loneli-
ness. given the range of locus of stability mean scores across the
clusters from the most lonely to the nonlonely clusters. These results
regarding the relationships seen between attribution and loneliness
corresponded to the results presented earlier in this chapter regard-
ing the analyses of variance run on the total sample and the Lonely
subsample. crossing loneliness with attributional cells. Lonely
subjects were seen to have from low to moderate levels of self-esteem
and depression. Lower degrees of or absence of loneliness seemed to

be related to external and unstable attributions of loneliness.

.Eooumor..om oz» Loco.z ecu ..mwro .o omcm. o.n.mmoov ocoom on. Locm.s one .>Louco>:_ Eooummr..om su.Em.oo00ue

.mco.u:n.cuum .mno.m ecm ..mccouc. .o.amom ocmzou
>ocoocou on. Loumocm 0:. ..mr.. cocoa ecu .o:m.: och .o.momn=m mco.um:u.m o>.ummoz--o..mcco_umoad o.>um .mco.u:a.cuu<u

.mmoc..oco. .o mco.u:n_cuum oz. .m:.0uxo once on. ..waro .o omcm. o.n.mmoo. cocoa on. cozm.: echo

.mmoc..oco. .o mco.uen..uum ecu o.nmumc: duos oz. ..maro .0 «meme 0.3.mmoo. 0.00m oz. Locm.: echo

 

1148

 

e.s~ m.» o... ..o. o.m a... ~..~ ... m.em om. e.w. N. ..m .m
m... .... a... N... m... m... ...: E :m 8. ..m. a .m ...
..s m..~ m... o.om m.. o... m.~m m. m.m~ ... m.o. «.m. ..m .m
.... ...: ...... ..Nm N... ...: n3 ... he. a. m... a .u .N
a.m ..m. m.e. o.mm o.: ..om m..m m.. :.sm we. m.m. me u.m ..
-.mmmwa ewwwwM. -mmwmmmu -mmemm“ uhwmu a....na=mu .....anum au.... <58: o..n¢ om< Looma.o

 

.mm. I.m. memo: Loam:.u .mco.um>comao co m_m>.mc< Loam:.u "cm.: 0.5m»

149

Emotion-focused coping behavior varied widely among the cluster
groups. (See Table 4.30.) Among lonely subjects (three clusters). it
was seen that coping behavior for both emotion- and probl em-focused
coping increased wifir unstable and internal attributions. and decreased
with stable and internal attributions. Among the two nonl onely cl us-
ters. one group was above the mean for coping. particularly with
respect to emotion-focused copi ng. and the other group was below the
mean. again particularly with respect to emotion-focused coping. One
nonl onely cluster also had a problem-focused coping mean score that
was above the mean scores for probl em-focused coping for both the
total sample and the Lonely subsample. These clusters did not differ
particularly in terms of degree of loneliness. attribution style
(unstable/external). or levels of sel f-esteem and depression. Impli-

cations of this last result are discussed in Chapter V.

Discussion

Relationshi ps were seen as predicted between stable attributions
and degree of loneliness. as well as between internal attributions and
degree of loneliness. Attribution style was not seen to bear a strong
or even moderate relationship with coping behavior. The relationships
that were seen were found to be in the reverse directions to what had
been predicted. Thus. coping behavior was seen to rel ate only slightly
to both attribution and degree of loneliness. posing furfirer contem-
pl ation on the nature and prediction of coping behavior in response to

loneliness. Lonely students were seen to be depressed. to have low

150

self-esteem. and to engage in less positively oriented coping behaviors
than less lonely students. Correlations between affective states and
loneliness indicated that there were most likely other affect states or
components to the lonely experience for students as well. Absence of
close friendship. homesickness. and romantic relationship problems
were common reasons for loneliness among students.

In general. students were found to cope at moderate levels. and
there were no strong differences between the amounts of emotion- and
probl em-focused coping. Coping behavior in response to loneliness was
not seen to increase among lonely students as degree of loneliness
increased. Nor was coping behavior seen to increase by type of cop-
ing. emotion- or probl em-focused. depending on fire presence of
unstable/internal attributions or stable/internal attributions for
loneliness. respectively. Rather. coping behavior appeared to
increase slightly with both decreased loneliness and unstable/
external attributions for loneliness.

Cluster analyses indicated that indeed there were different
groups of subjects among the total sample with respect to coping.
loneliness. attribution. and affective states. External and unstable
attributions seemed related to lower degrees of or absence of loneli-
ness. Coping behavior was seen to vary more widely than was evident
with analyses of variance. especially emotion-focused coping. It was
seen that. among lonely subjects. increased coping appeared related to
unstable and internal attributions. whereas low coping appeared

related to stable and internal attributions. Loneliness was seen to

151

relate to attribution in that the lonely clusters (three) had subjects
Inaintaining internal and stable attributions for loneliness. Lonely
subjects were seen to have from low to moderate levels of sel f-esteem
and depression. Loneliness was worst among fire cluster group charac-
terized by the lowest mean scores for internality and stability. This
cluster was also characterized by fire lowest degree of sel f-esteem

and the highest level of depression. The internal/external dimension
of attribution (locus of causality) appeared to play a critical role
in relation to loneliness. given that the cluster group with an exter-
nally oriented mean score on this attributional dimension was charac-
terized by a loneliness mean score almost 1 standard deviation below
the total sample mean (on the UCLA measure). These results corres-
ponded to the findings of the analysis of variance performed on the
total sample and the Lonely subsample with respect to loneliness
across attributional cells.

Irtests and other descriptive analyses supported the correlations
and cluster analyses. indicating that lonely students'were likely to
suffer from low selfeesteem and depression and to maintain stable as
well as internal attributions. Younger students were slightly more
lonely than older. and many students indicated lack of friendships or
absence of a romantic relationship as the precipitant. Coping levels
were less for more lonely than less lonely students. Women engaged in
slightly more coping that men. and emotion-focused coping tended to be

favored over problen-focused coping.

152

An interesting finding was that there were very few lonely people
characterized as having unstable and internal attributions for their
loneliness as compared to other lonely students. It may be that.
given an unpleasant situation such as loneliness. students felt that
internal causes for loneliness were more likely to be resistant to
change. and therefore the propensity for internal and stable versus
internal and unstable combinations for attributions. This seems to be
important. given that the attribution literature has suggested that
the unstabl e/internal attributional style is more associated with a
feeling of motivation. personal power and control. and attribution to

the self for positive changes.

Sumau
This chapter presented the major hypotheses and their results.

along with post hoc analyses run on the data. Of the main hypotheses
presented. those supported were that loneliness increased along with
stable attributions and increased most significantly with attributions
that were bofir stable and internal in nature. The post hoc analyses
provided for some interesting findings that lent further support to
the impression that coping behavior tended to decrease as loneliness
increased. particularly if the attributions for loneliness were stable
and internal. Furthermore. post hoc compari sons and descriptive
characteristics of the clusters identified in the cluster analyses
revealed the tendency for emotion-focused coping to increase with
exteer and stable attributions of loneliness. There was not much

variation seen in degree of problem-focused coping with respect to

153

degree of loneliness or type of attributional style. Rafirer. the
range seen in probl em-focused coping occurred across attributional
groups and degrees of loneliness.

Lonely students were found to engage in a moderate level of coping
and reported using a variety of cognitive and behavioral strategies.
More solitary thinking or activity strategies were endorsed at higrer
levels by lonely students firan strategies involving social interaction
or involvement. On the whole. students across both the total sample
and the Lonely subsample reported from sligrt to moderate levels of
passive measures of coping. such as using drugs. sexual involvement.
or drinking as a way to escape the pain of loneliness. Rather. more
benign or positive forms of coping were endorsed. with fire exception of
(a) going to a counselor for help or (b) seeking social skills or
assertiveness cl asses. two strategies one may have thougrt would be
appealing.

I-tests comparing lonely and nonl onely students and regression
analyses on loneliness both indicated that low sel f-esteem and depres-
si on were predictors of loneliness. Oopi ng behavior. interestingly
enough. was not found to be strongly or even moderately predicted by

any of the variables in the study.

O-IAPTER V

SUMMARY AND CWOLUS ICNS

SumarLoLtbLStudx
This dissertation looked at the relationships between attribu-

tions of loneliness and manners of coping with loneliness in univer-
sity undergraduate students. Loneliness was entered into the research
design as a third variable of interest. with attributions and coping
behavior investigated in terms of degree of loneliness being experi-
enced. The sample consisted of students who were both lonely and not
lonely. Criteria were established to investigate hypotheses both among
the total sample of students (N = 199) and among a Lonely subsample (n
= 49). In general. the total student sample demonstrated a range of
loneliness typical of student popul ations seen on other large univer-
sity campuses in the United States. Scores of 40 or greater on the
UCLA Revised Loneliness Scale are considered to be indicators of
loneliness. with 40 representing moderate loneliness all the way to 80
(the highest score possible) representing severe loneliness. The
sanple mean score for the 199 subjects in this study on the UCLA
measure was 39.9. As would be expected in a normal distribution.
close to half of the students had scores of 40 or greater (89 students
scored above the sample mean). and approximately 19% of the students

(38) scored beyond 1 standard deviation above the mean. Thus.

15h

155

approximately half of fire sample was considered lonely with respect to
the UCLA measure. with 19% of the sample regarded as very lonely.
Coping behavior was investigated according to the paradigm of
Lazarus and Folkman (1983). in which coping was seen as either
emotion-focused or problem-focused in orientation. Hypotheses
predicting relationships between one form of coping or another
depending on the attributions of stability and causality of loneliness
were generally not supported. Coping behavior was not seen to be
predicted in the directions hypothesized. nor was coping behavior
strongly related to attributions of loneliness. Correlational data
and analysis of variance suggested that instability of attributions
and externality of attributions were found to have a moderate and
significant relationship with emotion-focused coping. suggesting that
individuals are more prone to sel f-comforting coping strategies when
they are feeling that the problem of loneliness is changeable and
temporary. This was seen to a further degree with a cluster analysis
that identified three lonely groups of students. One lonely cluster
had significantly higher emotion-focused and problem-focused coping
scores than an even lonelier group. An important distinguishing
feature of this higher-than-average coping cluster was fire greater
tendency toward unstable and external attributions for loneliness than
seen in the lonelier cluster. A third slightly lonely cluster with a
tendency toward stable and internal attributions had significantly
lower coping scores for both emotion-focused and probl em-focused cOping

than the moderately lonely cluster. This seems to suggest that coping

156

was high for moderately lonely students with external and unstable
attributions for their loneliness. whereas students who were very
lonely or slightly lonely and showed greater tendencies toward stable
and internal attributions for their loneliness engaged in signifi-
cantly less coping behavior.

Lonely students were seen to suffer from depression and low self-
esteem. However. other effects not identified in this study were also
assumed to contribute to loneliness. Lonely students were more likely
to be female than male. Reasons stated for being lonely were the
absence of an intimate/romantic partner. homesickness or missing
people who were far away. and feeling a lack of good friendships.

This study raised the question that perhaps those students who are
most lonely are more apt to see their situation in negativistic terms
and feel less motivated to seek change. Given that attempts to change
a situation are one important aspect of coping behavior. and yet the
affective components of loneliness that accompany negative cognitions
work against sel f-promoting behavior. redefining the problem as a way
to motivate coping is seen to be extremely important in a student
population.

The findings of this study can only be considered within the
scope of the parameters of the population studied. Thus. external
validity is limited to university students studying within large
university settings. There were not significant differences found in
attribution. coping. or loneliness for students depending on year in

school or living environment within the university setting. Nor were

157

there significant differences found pertaining to previous history of
family background or environment. Moderate alpha coefficients for
some instruments used in fire study may have placed some limitations on
the strength of the analyses. thus raising some concern regarding
internal validity. However. fire findings as such are of interest and
suggest areas to pursue in further research on university populations

with respect to coping with loneliness.

Conclusions

WW
tthtflbutionaLStxles

Students with stable attributions were hypothesized to use more
enotion- than probl em-focused coping based on the assunption that they
would feel more pessimistic about changing their situation.

1. Students with stable attributions were not found to have more
emotion-focused coping than students with unstable attributions.
Rather. the reverse relationship was seen. Students with unstable
attributions had higher emotion-focused coping scores.

2. Students with stable attributions were also hypothesized to
use less problem-focused coping than students with unstable attribu-
tions. This was not seen. according to analyses of variance and
correlation. However. one cluster of students. characterized as stu-
dents with both stable and internal attributions. was found to have a
slight decrease in probl em-focused coping compared to the other cl us-
ter. The difference. however. was very slight and not considered

worthy of examination for significance.

158

3. Students with stable and internal attributions were
hypothesized to have the highest level of emotion-focused coping.
Rafirer. the tendency was for this group to have lower coping scores
in general than other attribution groups. Thus. students who were
assumed to have the most negative view of their loneliness were those
who appeared to engage in the lowest degree of coping and did not
engage in more enotion-focused coping than other students.

4. On the other hand. students who had unstable and internal
attributions for their loneliness were hypothesized to have the great-
est degree of probl em-focused coping. since it was assumed that these
students would feel most control over fireir situation and most opti-
mism for change. However. as mentioned with reference to Hypothesis
2. students did not differ very much with regard to problem-focused
coping in relation to attribution. Rather. probl em-focused coping was
endorsed at a moderate mean level for students representing all of the
attribution groups.

WW

WW
tLAtttibutionalJinoup

5. Students with stable attributions for their loneliness were
hypothesized to use more emotion-focused than probl em-focused coping
than students with unstable attributions. based on the assumption that
these students would be in more need of palliative measures of coping
and would feel less motivated to pursue change. Analysis of variance
on difference scores failed to find significant differences across

groups. In fact. there was really no difference found between

159

emotion-focused and probl em-focused coping between students of stable
and unstable attributional styles. Even within attributional groups.
it was seen that emotion- and problem-focused coping seemed to go hand
in hand rather than differing in degree within attributional groups.
6. Students who had unstable and internal attributions for
loneliness were hypothesized to use more problem- than emoti on-focused
coping. based on the belief that the majority of coping effort would
go into problem-solving and taking-action activities. However. as
stated previously. probl em-focused coping did not exceed emotion-
focused coping for any attributional group as measured by analysis of
variance or cluster analysis. Cluster analysis showed that all five
clusters used more emotion-focused than problem-focused coping. with
only one group showing much of a sizable difference. A cluster
characterized by unstable and external attributions had a 5.6 point
differential favoring emotion-focused coping. Again. this suggests
that the least at-risk group for loneliness was engaging in the most
coping. This also points to the slightly greater use of emotion-

focused over probl em-focused coping.

AttnihutimndJonaJJnoss

7. Students who had stable attributions regarding their loneli-
ness were hypothesized to have greater degrees of loneliness than
students with unstable attributions. This was based on the notion
that a low belief in ability to affect one's loneliness would lead to

increases in the feeling of loneliness. A moderate rel ationship was

160

found between loneliness and stable attributions. in the direction of
increased loneliness for students with stable attributions. Multiple
linear regression analysis suggested that locus of stability in the

"stable" direction is a predictor of loneliness to a moderate degree.

8. It was further hypothesized that students who had both stable
attributions of loneliness and attributions of internality for cause
of loneliness would have the higrest degree of loneliness. This was
based on the assumption that loneliness would be worse for those
feeling that they have personal and relatively enduring characteris-
tics that lead to loneliness. There were significant differences
found among attribution groups for degree of loneliness. with this
group having a significantly higher loneliness mean score (on the UCLA
measureh. This was found by way of analysis of variance and cluster
analysis. However. the cluster analysis also identified a lonely
group (although not the loneliest) that had internal and unstable
attributions for loneliness. Thus. the cluster analysis raises the
possibility that it isrnore likely that it is only among the loneliest
of people that attributions will often be both stable and internal for
loneliness.

9. It was hypothesized that as students became lonelier. they
would use more emotion-focused than probl em-focused coping strategies
to respond to their level of distress. What was seen instead was
little relationship between loneliness and emotion-focused coping.
both via Pearson correlations and cluster analysis. The cluster

analysis identified a very lonely group that used a bel ow-average

161

level of emotion-focused coping and a moderately lonely group that
used a level of emotion-focused coping equal to the total sample mean.
Emotion-focused coping was also seen to play a small role as a

predictor in loneliness. according to multiple linear regression.

EurtbeLNotionsinonfiustoLJnastos
anthboLEosLfloanalxsss

Cluster analyses indicated that indeed there were different
groups of subjects among the total sample with respect to coping.
loneliness. attribution. and affective states. External and unstable
attributions seemed related to lower degrees of or absence of lonel i-
ness. Coping behavior was seen to vary more widely than was evident
with analyses of variance. especially emotion-focused coping. It was
seen that among lonely subj ects. increased coping appeared related to
unstable and internal attributions. whereas low coping appeared
related to stable and internal attributions. Loneliness was seen to
relate to attribution in that fire lonely clusters (three) had subjects
maintaining internal attributions for loneliness. Lonely subjects
were seen to have from low to moderate levels of sel f-esteem and
depression. Loneliness was worst among the cluster group character-
ized by fire lowest mean scores for internality and stability. This
cluster was also characterized by fire lowest degree of sel f-esteem
and the higrest level of depression. The internal/external dimension
of attribution (locus of causality) appears to play a critical role in
relation to loneliness. given that the cluster group with an exter-

nally oriented mean score on this attributional dimension was

162

characterized by a loneliness mean score almost 1 standard deviation
below fire total sample mean (on the UCLA measure). These results
correspond to the findings of the analysis of variance performed on
the total sample and the Lonely subsample with respect to loneliness
across attri buti onal cells.

I-tests and other descriptive analyses identified lonely students
as more likely to suffer from low sel f-esteem and depression and to
maintain stable as well as internal attributions. Younger students
were slightly more lonely than older. and many students indicated
lack of friendships or absence of a romantic relationship as the
precipitant. Coping levels were less for more lonely versus less
lonely students. Women engaged in slightly more coping than men. and
emotion-focused coping tended to be favored over probl em-focused

coping.

Discussion

The hypotheses as to how attribution of loneliness would predict
coping behavior were not supported by the results of this study.
Actually. there was not any strong relationship seen between attribu-
tion and coping. Those relationships seen were in directions that
were opposite from what was expected. This occurred in two instances.
One was the finding that locus of stability in the direction of unstable
attributions for loneliness was related to more emotion-focused cop-
ing. People who see their situation of loneliness as changeable and
see the causes of their loneliness as amenable to intervention may

feel hopeful about their situation and cope by doing more palliative

163

things for themselves. Contrary to fire expectation that. under
duress. coping would resort to a more emotionally oriented form of
trying to make oneself feel good. emotion-focused coping decreased
slightly. The reader should be reminded of the fact that small dif-
ferences are being discussed here. Nonetheless. that coping behavior
which is emotion-focused was not found to be related to stable
firoughts about loneliness is an intriguing finding. Considering the
fact that people who see problems as less open to change usually feel
worse and might logically be expected to cope with more palliative
measures than others. this finding raises questions about influences
on coping behavior. It is possible that this type of individual
actually feels more depressed about his/her situation and is therefore
less apt to find ways of obtaining pleasure or attending to his/her
needs. Depression was found to be correlated with stable attributions.
and the effect of depression on behavior may explain in part the lack
of a positive relationship between emoti on-focused coping and stabil-
ity. Sel f-esteem was also found to be correlated significantly with
stability. Low sel f-esteem may cause lonely people to feel unworthy
of attempting to feel better when they are feeling dejected or alone.
Perhaps people feel firat their loneliness is their fault and feel they
must endure fire discomfort in a sel f-punitive manner.

An interesting finding was that there were very few lonely people
characterized as having unstable and internal attributions for their
loneliness as compared to other lonely students. It may be that given

an unpleasant situation such as loneliness. students feel that

164

internal causes for loneliness are more likely to be resistant to
change. and therefore the propensity for internal and stable versus
internal and unstable combinations for attributions. This seems to be
important. given that the attribution literature has suggested that
the unstable/internal attributional style is more associated with a
feeling of motivation. personal power and control. and attribution to
the self for positive changes. Students may feel frightened or unduly
concerned when they see in themselves qualities that they believe are
implicated in their loneliness condition. and may be constricted in
their firinking as to how they migrt work with their observation or
sel f-awareness to improve upon their situation. A student. for
example. who feels ostracized by fellow students due to inferior
academic performance may isolate him/herself and become less secure
both socially and academically. If this particular student could cope
with fire situation by asking a student for help (i.e.. with material
from a course) or seeking academic counseling. the student could
counteract feelings of loneliness via firese contacts and could develop
increased confidence. One serious implication of diminished coping in
the very students needing to engage in higrer coping levels is that
students may forego feedback or clarification of themselves that they
would otherwise obtain through talking with friends. meeting with a
counsel or. or taking initiative with a new activity.

In comparison. students with unstable and external attributions
for fireir loneliness coped slightly more than other students with

respect to both emotion- and problem-focused coping. This. too. was

165

contrary to the expectation that students with internal and unstable
attributions would be the 'lriglrest" copers with respect to problem-
oriented coping strategies. In fact. students who view their state of
loneliness as caused by circumstances external to their immediate
control and see these circumstances as temporary are possibly less
threatened by fireir condition of loneliness and apt to feel more
motivated to do something about their concerns. They also appear more
motivated to engage in sel f-comforti ng behaviors. In looking at the
most widely used emotion-focused coping strategies. it was seen that
students seem to use ego-enhancing methods more than escapist methods
of sel f-comfort. Thus. students who have this more optimistic view of
loneliness may see loneliness as manageable and therefore are more
capable of both finding positive ways to feel secure as well as
engaging in "taking-action" strategies. Students who may be regarded
as the least at-risk for their condition of loneliness have a slight
tendency to be the better copers. Students who experience the least
loneliness (among lonely students) are also the better copers. Attri-
bution does seem to play a role in this finding.

Wei ner's (1974) fireory on attributions in achievement situations
can be helpful in interpreting these results. His theory postulates
that people with internal and stable attributions for perceived
failures will experience dejection and problems with motivation.
Therefore. in considering the results of reduced emotion-focused
coping among people with stable and internal attributions for their

loneliness. it appears plausible that the absence of emotion-focused

166

coping can be regarded as due to a lack of motivation. Coping behav-
ior. rather than being considered a sel f-comforti ng or actual izing
behavior that is reactive in nature. may be contingent on the indi-
vidual's belief that there is some hope for change before self-
comforting or probl em-solving measures can be undertaken.

Weiner's theory can also apply to the lack of results seen with
respect to the expectation of finding more probl em-focused coping
among those students who had unstable and internal attributions for
fireir loneliness. The theory postulates that attributing a problem to
external causes results in assigning a problem to either a task diffi-
culty or unfortunate luck. If a student is able to perceive the
causes of loneliness as either resulting from an unlucky situation
(such as being placed with an unfriendly roommate) or sees fire task to
be mastered (meeting people) as a concrete task that in time will be
mastered. this student may feel more motivated to take care of him/
herself and engage in coping behavior. The student who can find an
external attribution may be acting in a sel f-preservative manner.
S/he may be more likely to perceive external circumstances as tempo-
rary or malleable. given the multitude of explanations one can often
identify and the emotional distance one can create for oneself in the
avoidance of internalizing the causes of loneliness.

In research conducted by Weiner. Russell. and Lerman (1979) on
affective responses to failure situations in which subjects assigned
external causes for explaining the failures. it was found that these

causal explanations were accompanied by particular affect states. If

167

luck was the external cause identified. subjects experienced surprise
as an affect. If other people were blamed as the external cause for
failure. then aggression was the affect experienced by subjects.
Thus. interpreting the cause of one's loneliness to external sources
may increase the likelihood of a situation being perceived as
unstable. in part because the emotions elicited are more sel f-

actual izing emotions. allowing for an emotion-cognition process

that is constructive versus passive.

It is reasonable to raise the speculation (based on the above
discussion) that perhaps the relationships seen among attribution.
loneliness. and coping can all be explained simply by relationships
between loneliness and coping. One might wonder if loneliness is
associated with less coping behavior as loneliness increases. since an
individual may become more unhappy and less motivated in general to
respond to problems. This is an alternative way of approaching the
relationship among these variables.

The cluster analysis on observations indicated that coping
behavior did not necessarily increase or decrease with degree of
loneliness. Rather. it was seen that a cluster of subjects who had a
mean score on the UQA scale that was below the total sample mean
(regarded as a "not lonely" mean score) had considerably higher
emotion-focused coping mean scores than the ofirer clusters or the
total sample.

This indicates that there were nonlonely subjects in this cluster

group who reported a high level of emotion-focused coping when they

168

had previously been lonely. (These subjects were most likely
reporting retrospectively. given that they were not experiencing
levels of loneliness at the time of data collection.) It is of
interest that the cluster group characterized by the lowest mean
score on the Revised UQA Scale also reported the hi grest emotion-
focused coping level when lonely in the past. Is there some associa-
tion between their current level of low degree of loneliness and their
propensity to cope in a higrer manner than others when experiencing
the distress of loneliness? This leads to fire speculation that per-
haps people who are less lonely and indicate a higr level of emotion-
focused coping when lonely might actually be people who engage in a
certain level of coping behavior on an ongoing basis. Their retro-
spective reports of coping behavior may actually reflect a certain
capacity to cope that is an integrated part of their personality.

One might believe that these subjects inflated their sel f-reports
of coping behavior. given that they had been lonely and therefore may
not have been reporting accurately. However. fire reader is reminded
that there was another cluster group that had an even lower mean
score for degree of loneliness (on the UCLA measure) that reported a
very low mean score for emotion-focused coping. suggesting that indeed
these groups differed somehow with respect to coping behavior. The
only distinguishing feature between these two clusters that seems
worthy of comment is that fire high emotion-focused coping cluster was
66% female compared to a 47% proportion of females in the low emotion-

focused coping cluster. In terms of depression and sel f-esteem. both

169

clusters measured high on sel f-esteem and low on depression. Given
that the correlation between emotion-focused coping to gender was
higier for females than males. the presence of females in the afore-
mentioned cluster may explain in part the tendency toward sel f-
comforting behaviors employed to cope with loneliness.

Thus. coping does seem related to attribution and to loneliness.
with attri bution playing an important role. It may be that file
"internal" factor is the more important with regard to relationships
filth coping. in that correlations of attribution were higier for
internality with loneliness.

It may be that lonely students who suffer from depression and/or
low self-esteem are those who have the more negative attributional
stance and the lower coping level. Students who are lonely who have
less stability or internality combined as an attributional approach
may also suffer less from depression and/or self-esteem (as suggested
by correlations in Table 4.5) and therefore are more apt to have energy
for effective coping.

In conceptualizing the three main variables of interest in this
study. attribution. loneliness. and coping. there were several ways to
see their interrelationships. One was to conceptualize attribution
as affecting degree of loneliness. such that coping ability or level
is associated to attribution via loneliness. Perhaps a student has a
negative outlook on many situation in life. not‘Just loneliness. and
therefore is more prone to feel lonely. lhis may then lead to coping

behavior that is correlated with both fiie degree of loneliness being

170

experienced aS‘well as with the particular attributional style that
the student characteristically applies in interpersonal situations.
The scores on the UCLAiLoneliness Scale correlated moderately among
both the total sample and the Lonely subsample with the Seligman
Attributional Style Questionnaire subscale for negative events in =
199. 1: = .38. p = .OOl and n = 39. .1: = .29. ,p < .01). These correl a-
tions suggested that indeed there may be a tendency for a particular
response set to difficulties to make one more vulnerable to emotional
distress. This level of distress may fiien serve to affect one's
motivation to cope. The correlations of the negative subscale of the
Attributional Style Questionnaire were minimal with emotion-focused
and probl em-focused coping for both the total sample and the Lonely
subsampl e. fiius suggesting as well that coping migit be a result of
the interaction between loneliness and attribution. Lonel ier students
tended to cope slightly less. with attribution playing an important
role in this relationship. given the role attribution seems to play in
the experience of being lonely.

Another way to view these variables is that loneliness affected
both attribution and coping. such that it was initially one's state of
loneliness (or trait) that affected one's outlook and one's energy for
coping. Thus. there might be a spurious relationship between attribu-
tion and coping. A third possibility was that coping actually pre-
ceded both. such that one's coping ability determined to what degree
a student became lonely. which in turn affected his/her outl ook on

loneliness.

171

Considering the first twc» the conceptualization of attribution
influencing one's vulnerability to loneliness and influencing coping
seems most supported by the data. After all. attribution was a
predictor of loneliness and was associated moderately with emotion-
focused coping. General attributional approaches to negative situa-
tions seemed to be more stable and internal for subjects who also
demonstrated a greater tendency for loneliness. also supporting the
notion that a particular attributional outlook on life can lead to
vulnerability to loneliness.

Therefore. it seems useful to raise the possibility of looking at
loneliness as an outcome variable in this study. Considering the way
in which coping behavior seemed to bear only a minor relationship with
loneliness in terms.of straightforward correlations. and that attri-
butional style was seen to relate to both coping response and loneli-
ness. the three variables can be conceptualized in the following
manner. Perhaps attributional style is an important predisposing
factor to loneliness wiinn(a) loneliness in general associated with
increasingly stable attributions and (b) unstable/internal attribu-
tions associated with fiie greatest degree of loneliness. If one's
general framework toward the explanation of emotional and social
experiences is to interpret them as difficult to affect. especially as
attributions become more internal in a causal direction. then coping
may play an additional role in contributing to loneliness. Attribu-
tion may influence one's motivation to engage in coping. It is pos-

sible that the correlations seen between loneliness and coping (see

172

Tables 4.3 and 4.4) were low because attribution played an intervening
role. and therefore the correlations were indicative of the low to
moderate levels of coping among both lonely and nonl onely people that
would have taken on different patterns with attribution regarded as a
covariate. When attribution is taken into account. it appears that
individuals with external and unstable attributions for their loneli-
ness had the higier levels of coping among the range seen (in refer
ence to emotion-focused coping) compared to ofiier groups of people (as
evidenced by the cluster analyses and the post hoc comparisons dis-
cussed in the Post Hoc Analyses section in Chapter IV). People with
the stable and internal attributions were found to have the lower
levels of emotion-focused coping. (The reader is reminded of fine fact
that there were virtually no differences identified for problem-
focused coping according to attributional styled

Thus. attribution along with one's premorbid coping ability (with
the consideration that attribution may affect one's motivation to
c0pe) may work in combination as prerequisites for experiencing
loneliness with respect to types of loneliness and intensity of
loneliness. There are more questions raised by the data that are
answered by the conceptualization of loneliness as fire result of
particular attributional patterns and coping ability. One might ask
how it is that an attributional approach can create or lead to suscep-
tibility to loneliness. It is plausible that if an individual has the
propensity to interpret problematic social interactions or separate-

ness from others as the result of internal traits or characteristics.

173

this may engender a feeling of despondency. hopelessness. or even
denial of dissatisfaction socially or emotionally. The result may be
(a) a sense of despair fiwat internal characteristics are too difficult
to change. (b) a belief that change may create unpleasant feelings of
sel f-disavowal. or (c) a fear that change may leave fine individual
open to the expectation that one's emotional life should improve.

Thus. individuals who are prone to internalization may have
resultant concomitant attributions of stability to protect them from
the risk of change. the unpleasantness of deliberately trying to
change personal traits. or the expectations that they should feel more
connected and involved with others. If coping skills are limited. or
motivation is limited due to one's outlook or attitude. fiien this will
only compound the problem. leading the individual to feel in less
conflict by resigning oneself to his/her loneliness.

What is difficult to assess from this study is the degree to which
coping behavior can operate independently of the situation coping
occurs in or the attributional stance one has toward one's situation.
In other words. how can coping behavior be enhanced so that it becomes
a behavior less rooted in affective or cognitive conditions of one's
situation and more determined by one's acquired learning of heal thy
coping mechanisms? One interesting observation about coping with
loneliness is that the more direct coping mechanisms that had to do
with actually discussing one's loneliness as a means of dealing with
it were relatively unpopular mechanisms. One is reminded of Gordon's

(1976) notion that loneliness is seen as failure and that coming to

174

terms.with loneliness is difficult to do by directly acknowledging
loneliness. It makes sense that coping may in large part be learned
by simply repeating what we see others do around us. With loneliness.
this has serious implications since people often do not talkIabout
feeling lonely and therefore may not impart knowledge to others about
effective ways of understanding and dealing with the experience. Few
coping mechanisms speak to an acceptance of one's loneliness. but the
one that addresses this on the Ceping Inventory was widely endorsed by
students. "Think about possible benefits of being lonely. such as
thinking that you are learning to be more self-reliant by growing from
the experience." Working with people to more directly delve into
their loneliness and learn from it seems to offer potential that is
missed by people wholengage in minimization or denial strategies. It
is possible that there was not as great a range seen in coping behav-
ior as expected because certain ways of coping are not easily identi-
fiable even to the subjects themselves and that those strategies
provided speak to the more common range of strategies employed by most
people when they experience loneliness. These commonalities are mini-
mization. denial. distraction. involvement in studying. and looking at
ways to overcomerloneliness. Perhaps the more differentiating strate-
gies are difficult to identify without having subjects actually study
themselves during episodes of loneliness. so that the greater indi-
vidual variations in coping can be more effectively observed and

measured.

175

One concern regarding coping with loneliness that this study
could not adequately address is the difference. if any. between one's
coping behavior in general and one's coping with a specific situation
such as loneliness. Lazarus and Folkman (l984) raised the possibility
that coping may be primarily situation-specific. Therefore. it is of
interest to«examine the possibility that perhaps people do not regard
loneliness as a situation that offers many alternatives for coping.
In the lonely cluster group having the highest coping levels for the
different lonely clusters identified. fiiere were still approximately
one-third of finese 39 subjects who were coping at about half the
coping capacity potential tapped by the Coping Inventory. Given
Pearlin and Schooler's (1978) contention fiiat caping involves (a)
managing stress. (b) making changes in the trauma-inflicting situa-
tion. and (c) redefining the identified crisis or problem. the unhap-
piness or resignation one may have toward loneliness may unify the
very adaptive and goal-oriented purposes of coping behavior.

It cannot be assmned that people know how to cope with a given
problem just because they are in distress. nor can it be assumed that
coping will always result in constructive change or adaptation in
people. Particularly regarding the topic of loneliness. this is a
subject generally not widely discussed or presented as a life
difficulty that people should be prepared to encounter. Perhaps
university campuses would profit by presenting loneliness in a more
normative manner as a life issue that occurs with frequency and

prevalence. and as Moustakas (l96l) contended. offers great

176

opportunity for growth and enrichment. Coping may best be approached
as a form of behavior (cognitive and behavioral) that is learned via
teaching. modeling. practicing. and refinement versus being seen as
naturally involving ego strength or skill.

There is an implication that attribution can strongly affect or
cause emotions and behavioral responses to such emotions. One
important‘question raised by this study is how coping behavior can be
taught or enhanced in people such that it is more independent of
precursors such as prenorbid personality traits or cognitive styles.

Coping behavior appeared to occur in fairly equal levels between
emotion- and problem-focused coping. Rather than finding significant
differences among individuals for their preference of probl em- or
emotion-focused coping. apparently students had fiie tendency to have a
general level of coping ability or coping performance. The more
coping mechanisms a person is aware of or familiar with. the more s/he
is likely to use all of those at his/her disposal. According to
Lazarus and Folkman (l984). use of both types of coping is optimal.
and predictions in this study that people within different attribution
“cells" would have a predominant coping style did not seem evident.
Thus. the more an individual learns about coping mechanisms that are
problem solving or palliative in nature. the more enhanced his/her

coping style is apt to become.

WW
It is important to look at ways in which the conceptual izations

of coping. attribution. and loneliness in this study may be expanded

177

for further research that can succeed in enlightening people as to the
critical aspects of coping behavior.

It is possible that a more fruitful way of looking at coping is
to view coping as a process that changes in reaction to a difficult
problem. such that measuring one's coping strategies at a static point
in time or asking for a summary of one's coping resources does not
provide an accurate picture of how an individual is dealing with a
life problem. For example. in this study it is possible fiiat it was
difficult to identify differences between emotion- and probl em-focused
coping because these were measured at the same time. and therefore any
differences that may have occurred in amount of or preference of
coping depended on what point in the history of a problem the particu-
lar coping occurred. It is also possible that choices in both types
of coping vary as reaction time to a problem increases. or according
to the success or failure of initial coping strategies. Perhaps some
people are initially emotion-focused copers. regardless of their
appraisal of a problem. and others are problem-focused copers. It
could be speculated that the appraisals they maintain about their
problem affect their effectiveness of carrying out coping strategies.
and the degree of success or failure leads to further coping choices.
Perhaps an individual who maintains stable attributions of loneliness
and is an initial problem-focused coper may become frustrated if the
difficulties with loneliness do not become ameliorated quickly. hence
resorting to the use of emotion-focused coping responses to deal wifii

the continued problem. An individual may have a definite pattern to

178

his/her coping or may engage initially in problem-focused coping. but
not long enougi to see results. and then engage in more emoti on-
focused coping as s/he becomes more desperate to feel comfortable.
Coping may be viewed differently. as a result of the examination
of the cluster analysis on observations and the absence of variability
in coping levels seen among lonely individuals. Coping appears from
file results of this study to possibly comprise a set of cognitions and
behaviors that are practiced in more subtle. ongoing ways for indi-
viduals that protect such individuals from experiencing more serious
crises (such as himer degrees of loneliness). Thus. among the lonely
subsets within this study. coping behavior was engaged in at moderate
levels and was possibly affected by feelings of low sel f-esteem and
depression that would deplete an individual and reduce effective self-
enhancing behavior. Contrary to expectation. more severe loneliness
did not motivate coping. One group of individuals who had a fairly
low mean score for loneliness on both the UCLA scale and the Likert
sel f-rating scale was characterized by a high mean score for emotion-
focused coping. (This was not the case for probl em-focused coping.)
Perhaps these individuals never experience a higw degree of loneliness
because they have learned a set of strategies for dealing with inter-
personal problems on a daily basis that protects them from extreme
loneliness. Or if this group is subject to more pronounced degrees of
loneliness at given times. they may have a familiarity with a greater
range of coping strategies that protect them and reduce the length of

time they are lonely or the degree of discomfort they feel. One might

179

hypothesize for furfiier research that coping is an ongoing process
rather fiian just a reactive form of behavior that rises and ebbs in the
face of different issues or problems. such that an individual's "pre-
morbid" coping level takes on a preventive role in helping prepare for
and withstand duress. For those with low coping abilities. when
problems are of the magnitude experienced because of previously low
ongoing levels of adequate responsiveness on a continual day-to-day
basis. coping may not increase.

Research fiiat would have people actually identifying their pro-
cess of coping. such as the study done by Lazarus and Fol kman (T985).
may provide more specific information regarding coping with the prob-
lem of loneliness. It is possible fiiat fiie amount of coping an indi-
vidual engages in is difficult to measure without recording changes
that occur in what these authors have termed fiie "coping process."
Thus. not only may an individual have a certain limited reservoir of
coping strategies. but may have many at first in confronting a problem
and then taper off to staying with the few strategies that seem to be
most effective. It is also possible that coping is reflective more of
a combination of personal traits and coping behavior as well as of fine
particular problem for fiie individual. thus making a relationship
between attribution and coping more difficult to identify. For example.
greater degrees of loneliness for one person may be both uncomfort-
able. but also challenging. and thus we might see probl em-oriented
coping mechanisms on the part of fiiat individual. such as looking for

new friends or taking a class in assertiveness. Another person migit

180

experience a strong degree of loneliness as uncomfortable and very
unsettling or contrary to hi s/her sel f-image. and become anxious.
needing to drink or sleep frequently to escape from fiie dilemma. In
both cases. these individuals may have stable attributions of loneli-
ness. the first individual saying to him/hersel f. "This loneliness is
here to stay unless I do something about it!" and the second indi-
vidual may say. "This problem is getting file best of me. I'm not sure
how it came about or what to do. or how it will go away. I just want
to be free of feeling it for awhile." Coping may proceed from a more
automatic response out of habit to behavior that is more intentional
once a problem has provoked more conscious thought on the difficul-
ties.

It would be of value to learn more about the nature of emotional
threat fiiat students attribute to loneliness. According to Fol kman
and Lazarus (1983). every stressful encounter or situation can be
perceived as beneficial. benign. or threatening. Appraisals that are
of a fiireatening nature are then associated with different coping
mechanisms fiian appraisals of a situation that are benign or positive.
Thus. if a student sees loneliness as an opportunity to master a
situation and reacts with feelings of challenge to fiie situation. then
that individual may demonstrate a preference for probl em-focused
coping. A student who sees loneliness as threatening may be more
likely to respond by using more emotion-focused coping mechanisms.
Thus. a further refinement of the study could involve fine separation

into subsampl es of students who have "challenge" responses. "benign"

181

responses. or "threat" responses to loneliness. to see whether initial
appraisals are more predictive of coping in expected directions and
whether particular coping strategies appear more dominant than others.
Fol kman and Lazarus's (l983) research on students coping througiout
the process of a midterm examination provided evidence that attri bu-
tion. coping. and emotion are intricately involved with one another
and that differences do emerge in copi ng and emotional response.
depending on the initial appraisals made about fiie stressful situa-

ti on.

One other aspect of attribution that would be of benefit to
explore in further research is the degree to which internal attribu-
tions of causality regardi ng loneliness are associated with skill
deficit (ability) versus lack of effort. Weiner and his associates
(1979) viewed skill and effort as the two important aspects of
internal attributions. with skill deficits being perceived as more
immutable to people and effort deficits feeling more malleable to
people. It is conceivable that skill attributions of causality of
loneliness (such as not feeling like a good conversationalist) leave
students feeling more distressed than lack-of-effort attributions that
may leave students believing that they simply had not tried hard
enough (i.e.. to be social or outgoing). It would be of interest to
explore the degree to which different internal attributions create
different emotional responses that in turn affect coping behavior or

degree of loneliness.

182

Improvements in the design over this past study would include the
following:

l. Having people record their own account of their lonely
experience.

2. Having a measure of observation of people who are experiencing
loneliness to provide for other than subjective data (i.e.. in-depth
interviews or peer ratings on loneliness).

3. Identifying a baseline coping level for subjects that could
then be compared with an ongoing recording of coping measures used.

4. Grouping subjects according to type of loneliness such that
relationships among attribution. loneliness. and coping could be
studied within the context of a particular loneliness paradigm.

5. Attempting to differentiate people experiencing more of a
"state" loneliness versus "trait" loneliness. for example. by obtain-
ing affective measures that identify chronically depressed from
reactive depressed people as a way to separate different forms of
loneliness.

6. Maintaining regular personal interview contact with subjects
in the study in order to capture in a more descriptive manner the on-
going qualitative aspects of their experience of loneliness.

Specific methodological changes:

7. Improving reliability on instruments whose reliability
coefficients run in the moderate range--either by selecting some
different instruments or conducting item analyses in order to refine

several measures.

183

8. Obtaining some subjects from a clinically lonely sample in
order to acquire a subsample experiencing intense degrees of loneli-
ness.

9. Improving upon attribution measures used in order to increase
- both the number of items in the scales and including more choices of
items in terms of content (so as to improve content val idity in rel a-
tion to loneliness).

The further study of attribution of loneliness. degree of loneli-
ness. and coping with loneliness is considered valuable because of the
commonality of loneliness among people and because of the importance
in understanding adaptive behavior. Lynch's (T976) seminal work on
the effect of loneliness on people in later life is indicative of the
social importance loneliness plays in society. in that people are
truly at risk physically and emotionally for loneliness that becomes
chronic and isolating. Polanski's (1979) notion of situational versus
sel f-imposed loneliness also speaks to the role people play in their
loneliness and the importance of gaining more insigit into the dynam-
ics that create emotional distress from loneliness.

Results of this study compared in some interesting ways to results
reported by Revenson's (l98l) study discussed in Chapter II. The
reader is reminded of Revenson's interest in looking at attribution
of loneliness and coping behavior in an adult population (also
students) irrespective of current state of loneliness. As found in
Revenson's study. type of coping seemed less indicative of someone's

coping style than amount of coping. However. in ofiier areas of

18h

findings. this study differed from Revenson's work. This study added
three important features to make this study more comprehensive. A
coping instrument was used that was specifically designed for coping
with loneliness. as compared to use of a general coping measure.
Second. the investigation of attribution. loneliness. and coping
behavior was explored among all attribution groups instead of just
among two attribution cells. as done in the Revenson study. Third.
the investigation of the variables of interest was done both among the
total sample. as well as among a Lonely subsampl e. so fiiat relation-
ships within a lonely group of people could be further understood or
assessed. With these added dimensions to the work. Revenson's finding
that internal attributions for the causality of loneliness prompt more
coping behavior was reversed. In fact. internality of attribution was
associated with both increased degrees of loneliness and decreased
coping behavior. The following findings were also departures from the
previous study. Locus of stability was also seen to rel ate to loneli-
ness. and in the direction of less-stable attributions prompting a
slimt increase in coping behavior. Low self-esteem and depression
were seen to be major affective components of the lonely experience
reported for some of the lonely students. Attributional style of both
stable and internal attributions was associated with the highest
degree of loneliness and the lowest level of coping.

Thus. in helping students understand the etiology of loneliness.
choices in reacting to and managing loneliness. and learn from lonely
experiences. it appears necessary to help them see personal responsi-

bility in a less-threatening and more growth-promoting manner.

185

One final question in research on coping with loneliness is the
extent to which coping affects overall adjustment. It can be assumed
that there are reciprocal influences in the relationships between the
variables looked at in this study. such fiiat students who cope more
effectively with loneliness may then either experience less degree of
loneliness subsequently or may find that they navigate better through
subsequent episodes of loneliness. One may wonder to what extent
caping behavior may have already influenced fiie degree of loneliness
that some students were reporting. given that there was some variation
in the length of time students reported experiencing loneliness.

It is possible that initial coping ability may actually influence
one's cognitive perceptions of his/her loneliness. thus eifiier
increasing or decreasing one's experienced degree of loneliness.

Thus. the extent to which a student has the ability to creatively
respond initially to loneliness may make the loneliness experience
feel eifiier controllable or not controllable. temporary or immutable.
or due to internal versus external circumstances. A student whose
initial coping efforts are not very successful may begin to believe
that the causes of the loneliness are due to personal ineptness. which
could be a faulty but convincing cognition. This student may fiien
resort to less-effective coping and have a continued experience of
loneliness. Thus. the interrelationships between attribution. loneli-
ness. and coping are complex and need to be pursued further in

research.

APPENDICES

APPENDIX A

INSTRUCTIONAL LETTER AND CONSENT INFORMATION

186

Loneliness Study

 

Code Number
Dear Student:

This introduction letter and information sheet contain
important information that is necessary to read before you
continue to look through the forms in this folder or choose
to fill them out. I went to tell you about the study and
about your rights as a research participant.

I am interested in learning about the experience of loneli-
ness in college students, from the standpoint of how
students eXplain loneliness to themselves, and how they cope
with loneliness. I want to see if there is a relationship
between the thoughts that students have about loneliness and
how they choose to deal with their loneliness. This can
lead to important ways of knowing how to better help
students when they feel lonely, and can indicate if there
are certain ways of thinking about loneliness that lead to
more effective coping styles.

To participate in this study, the only requirement is that
you are an undergraduate college student. It does not
matter whether or not you feel lonely at this time. As you
fill out the questionnaires, you can try to think back to a
time when you have felt lonely or you can imagine what you
think you would do if you were to feel lonely (for the
questionnaires that ask about attitudes towards loneliness
and coping).

Included in this folder is a general information sheet
asking about background and living information, some
questionnaires about thinking style and attitudes toward
loneliness, a list of coping responses, a self-esteem
inventory, a depression inventory, a loneliness scale, and a
form with some specific questions about loneliness. In
filling out these forms, please remember that there are no
right or wrong answers. Also, it is not necessary to spend
too much time on any one item or question. A time period of
forty-five minutes will be needed to complete all of the
items. The forms can be filled in during one sitting or in
several sittings. Some students find that choosing a time
of day that is relaxing for them enables them to concentrate
on questionnaires and move through them with greater ease.

Before proceeding with looking at the materials, the

I87

following information provides you with a list of your
rights regarding participating in the research. As I stated
in the information meeting, you are under no obligation to
participate in this study. Please read the following items
carefully.

(1) You have the right to additional information about this
study after your participation has been completed.

(2) You are free to discontinue your participation in the
study at any time. Should you decide after you have
sent materials to me that you would like to have your
materials withdrawn, you can call me and give me the
number indicated in the upper right hand corner of this
letter, as this is the number that appears on all your
forms. There is no need to identify yourself by name
should you need to contact me to have your forms
removed from the study and discarded.

(3) It is possible that some of the items or statements in
the questionnaires about moods and feelings could be
upsetting for you. You may find that in responding to
certain items, you experience unpleasant feelings.
Please take this into consideration in thinking about
participation in the study.

(4) Should participating in the study lead to some
troubling feelings, you have the right to contact me
for suggestions for assistance. Included with this
packet is an information sheet on resources for dealing
with loneliness, and with other types of feelings or
problems. There is a description provided for each of
the resources listed.

(5) Results of the study will be treated in a confidential
manner and you will remain anonymous. Results of the
study will be available to you upon request. The
anticipated time of completion is December of 1985.

(6) All forms and questionnaires received from students
will be discarded once data has been coded and
tabulated.

(7) In choosing to complete the enclosed forms you give
your consent to participate in this study, acknow-
ledging that you have read the above information.
Please retain this letter for your personal records, as
it contains important information and your code number.

Any requests about this study, in terms of assistance,
further explanations, or requests for results, can be

188

directed to am» If you should want to contact a faculty
member regarding any aspect of this study, my faculty
research adviser is Dr. Richard Johnson, Department of
Education at M.S.U. He can be reached at 353-1824.

For those of you participating in the research, I have
written a paper on loneliness and coping with loneliness
that I will be handing out to students once the data is all
collected from my study. I will bring copies of this paper
to R.A.s in the residence halls, or the off-campus houses or
classes that have participated in the research, so that all
participants are sure to get a copy. If a student would
like feedback on his/her forms, please contact me at the
number provided below. If you give me your code number
(make sure to keep this letter with the code number on it!)
I 'will be glad to locate your submitted materials and
discuss the results with you.

Thank you very much for participating in the research. I
hope that you learn more about yourself and have new ideas
to think about as a result of completing this packet of
forms.

Shelley Smithson
Ph.D. Candidate in
Counseling Psychology
332-0231

APPENDIX 8

INSTRUMENTS AND QUESTIONNAIRES USED IN STUDY

I89

PERSONAL INFORMATION SHEET

Sex

 

Age

 

Year in college

 

Live in dormitory

 

fraternity/sorority house

 

in fraternity/sorority, but living in
apartment

 

What type of environment did you grow up in?
rural area
small town (less than 20,000 people)
average sized town (20,000-50,000 people)
city (50.000-200,000 people)

large city/metropolitan area (over 200,000 people)
(including suburbs)

How many siblings were in your household?

Of these siblings, how many were _____ brothers
sisters
other children

(i.e cousins living
with you)

How many adults were living in your household?

 

Please indicate what adults these were: mother

father

step parent

relative (i.e.
grandparent,
aunt, uncle)

I90

QUESTIONS ON LONELINESS

1.

Do you feel lonely at the present time? yes

no

 

don't know

If you do feel lonely, please indicate below how lonely
you feel. the number 10 indicates severe loneliness,
and the number 1 indicates only slight feelings of
loneliness. Please indicate along the scale of numbers
where you see yourself.

1 2 3 4 5 6 7 8 9 10

How long would you say that you have been feeling
lonely?

less than a week
about two weeks

about a month

a couple of months
six months

more than a half year

more than a year

 

I believe that the main reason I am lonely is:

 

Do you believe that the reason you have indicated above
has more to do with you as a person or with the
situation you find yourself in?

me asaperson the situation
surrounding me

Do you believe that there are ways for the reason (you
have indicated for your loneliness) to be changed or
modified so that you won't continue to feel lonely?

yes no

 

PLEASE NOTE:

COpyrighted materials in this document
have not been filmed at the request of
the author. They are available for
consultation, however, in the author's
university library.

These consist of pages:

Appendix B, page 191 (Revised UCLA Loneliness Scale)

page 192-198 (Seligman Attributional Style
Questionnaire)

page 199 (Check Ljst, DACL Form A.)

page 200 (Questionnaire)
page 201-203 (Coping Inventory)

 

 

 

 

page 204-206 (Grant Attribution Scale-1984)

University
Mic rofilms

International
300 N Zeeb Rd., Ann Arbor, MI 48106 (313) 761-4700

I9]

Revised UCLA Loneliness Scale
Directions: Indicate runv often you feel the way described in

each of the following statements. Circle one
number for each.

Never Rarely Sometimes Cften

 

1. I feel in tune with the people 1 2 3 4
2. I lack companionship l 2 3 4
3. There is no one I can turn to 1 2 3 4
4. I do not feel alone 1 2 3 4
5. I feel part of a group of friends. 1 2 3 4
6. I have a lot in common with the l 2 3 4

people around me
7. I am no longer close to anyone 1 2 3 4

8. My interests and ideas are not 1 2 3 4
shared by those around me

9. I am an outgoing person 1 2 3 4
10. There are people I feel close to 1 2 3 4
11. I feel left out 1 2 3 4
12. My social relationships are l 2 3 4

superficial
13. No one really knows me well 1 2 3 4
14. I feel isolated from others 1 2 3 4
15. I can find companionship when l 2 3 4

I want to

16. There are people who really 1 2 3 4
understand me

17. I am unhappy being so withdrawn 1 2 3 4

18. People are around me but not 1 2 3 4
with me

19. There are people I can talk to 1 2 3 4

20. There are people I can turn to l 2 3 4

I92

Seligman Attributional Style Questionnaire

Name

 

Date

 

Phone“

 

DIRECTIONS

Phase try to vividly imaging yourself in the situations that iollow. Ii such a situation happened to you.
what would you ieei woold have caused it? Whlie events may have many causes. we want you to pick only
Ciie— the major cause it this event happened to yo... Please write this cause in the blank provided aiter
can event. Next we want you to answer some questions about the cause and a final question about the
situation. To summarize. we want you to:

1)
2')
3)
4i
5)

6)

Read each situation and vividly imagine it happening to you.

Decide what you feel would be the major cause oi the situation it it happened to you.
Write one cause in the blank provided.

Answer three questions about the cause.

Answer one question about the situation.

Go on to the next situation.

YOU

1)

(.3

4)

5)

YOU

6)
7)

8)

9)

10)

193

MEET A FRIEND WHO COMPLIMENTS YOU ON YOUR APPEARANCE.

Write down the one mayor cause

 

is the cause oi your friend's compliment due to something about you or something about the other
person or circumstances? (Circle one number) '

Totally due
to the other
person or Totally due
circumstances 1 2 3 4 5 6 7 to me

In the future when you are with your lriends. will this cause again be present? (Circle one number)

Will never
again be Will always
present 1 2 3 4 5 6 7 be present

Is the cause something that just affects interacting with friends or does It also influence other areas
of your lite? (Circle one number)

influences

just this influences
particular all Situations
situation i 2 3 4 5 6 7 in my life

How important would this situation be it it happened to you? (Circle one number)

Not at all Extremely
important 1 2 3 4 5 6 7 important

HAVE BEEN LOOKING FOR A JOB UNSUCCESSFULLY FOR SOME TIME.

Write down one major cause

 

Is the cause of your unsuccessful job search due to something about you or something about other

people or circumstances? (Circle one number)

Totally due to

other people Totally due
or circumstances 1 2 3 4 5 6 7 to me

In the future when looking for a job. will this cause again be present? (Circle one number)

Will never
again be ' Will always
present 1 2 3 4 5 6 7 be present

Is the cause something that lust influences lcolzing for a (db or does it also iniluence other areas of
your lite? (Circle one number)

influences

just this Influences
particular all situations
situation 1 2 3 4 5 6 7 in my life

HOW important would this situation be if it happened to you? (Circle one number)

Not at all Extremely
important 1 2 3 4 5 6 7 important

I914

YOU BECOME VERY RICH.

11) Write down the one maior cause

 

12) is the cause of your becoming rich due to something about you or something about other people or
Circumstances?

Totally due
to other people Totally due
or circumstances 1 2 3 4 5 6 7 to me

13) In your financial future. will this cause again be present?

Will never
again be Will always
present 1 2 3 4 5 6 7 be present
.14) Is the cause something that just affects obtaining money or does It also influence other areas of
your life?
Influences
just this Influences all
particular situations in
situation 1 2 3 4 5 6 7 my life

15) How important would this situation be If It happened to you?

Not at all Extremely
Important 1 2 3 4 5 6 7 Impdrtant

A FRIEND COMES TO YOU WITH A PROBLEM AND YOU DON'T TRY TO HELP THEM.

16) _ Write down the one malor cause

 

17) Is the cause of your not helping your friend due to something about you or something about other
people or circumstances? (Circle one number)

Totally due
to other people Totally due
or circumstances 1 2 3 4 5 6 7 to me

18) In the future when a friend comes to you with a proclem. will this cause again be present? (Circle
one number) '

Will never
again be Will always
present 1 2 3 4 5 6 7 be present

19) Is the cause something that lust affects what happens when a friend comes to you with a problem or
does it also influence other areas of your life? (Circle one number)

Influences
lust this influences all
particular situations in
situation 1 2 3 4 5 6 7 my life

20) How important would this situation be if it happened to you? (Circle one number)

Not at all Extremely
important 1 2 3 4 5 6 7 important

I95

YOU GIVE AN IMPORTANT TALK IN FRONT OF A GROUP AND THE AUDIENCE REACTS NEGATIVELY.

211 Write down the one maior cause

 

22) Is the cause of the audience reacting negatively due to something about you or something about
other people or circumstances? (Circle one number)

Totally due
to other people Totally due
or circumstances 1 2 3 4 5 6 7 to me

23) In the future when giving talks. will this cause again be present? (Circle one number)

Will never
again be ‘ Will always
present 1 2 3 4 5 6 7 be present

24) Is this cause something that just influences giving talks or does It also Influence other areas of your
life? (Circle one number)

Influences
lust this Influences all
particular situations in
situation 1 2 3 4 5 6 7 my life
25) How Important would this situation be If It happened to you? (Circle one number)
Not at all Extremely
Important 1 2 3 4 5 6 7 important

YOU 00 A PROJECT WHICH IS HIGHLY PRAISED.

26) . Write down the one major cause

 

27) Is the cause of being praised due to something about you or something about the other people or
circumstances?

Totally due
to other people Totally due
or circumstances 1 2 3 4 5 6 7 to me
28) In the future when doing a project. will this cause again be present?
Will never
again be WIII always
present 1 2 3 4 5 6 7 be present
29) Is this cause something that just affects doing protects or does It also Influence other areas of your
life.
Influences
lust this Influences all
particular situations in
situation i 2 3 4 5 6 7 my life
30) How important would this situation be if it happened to you?
Not at all . Extremely
Important 1 2 3 4 5 6 7 Important

196

YOU MEET A FRIEND WHO ACTS HOSTILELY TOWARD YOU

31) Write down the one major cause

 

32) Is the cause of your friend acting hostile due to something about you or something about other
peeple or circumstances? (Circle one number)

Totally due
to other people Totally due
or circumstances 1 2 3 4 5 6 7 to me

33) in the future when interacting with friends. will this cause again be present? (Circle one number)

Will never
again be Will always
present 1 2 3 4 5 6 7 be present

34) is the cause something that Just Influences Interacting with friends or does it also Influence other
areas of your life? (Circle one number)

Influences
Iust this Influences all
particular situations In
situation 1 2 3 4 5 6 7 my life

35) How important would this situation be if it happened to you? (Circle one number)

Not at all Extremely
Important 1 2 3 4 5 6 7 Important

YOU CAN'T GET ALL THE WORK DONE THAT OTHERS EXPECT OF YOU.

36) Write down the one malor cause

 

37) is the cause of your not getting the work done due to something about you or something about the
other people or circumstances? (Circle one number)

Totally due -
to other people Totally due
or circumstances 1 2 3 4 6 6 7 to me

38) In the future when doing the work that others expect. will this cause be present? (Circle one number)

Will never
again be Will always
present 1 2 3 4 5 6 7 be present

39) is the cause something that Just affects doing work that others expect of you or does It also
Influence other areas of your life? (Circle one number) .

Influences
lust this Influences all
particular . situations in
situation 1 2 3 4 5 6 7 my life

40) How important would this situation be if it happened to you? (Circle one number)

Not at all Extremely
Important 1 2 3 4 5 6 7 Important

I97

YOUR SPOUSE (BOYFRIEND/GIRLFRIEND) HAS BEEN TREATING YOU MORE LOVINGLY.

41) Write dOwn the one major cause

 

42) Is the cause of your spouse (boyfriend/girlfriend) treating you more lovingly due to something abdut
y0u or something about other people or Circumstances?

Totally due
to other people Totally due
or circumstances 1 2 3 4 5 6 7 to me

43) In future interactions with your spouse (boyfriend/girlfriend) will this cause again be present?

Will never
again be Will always
present 1 2 3 4 5 6 7 be present

44) is this cause something that )ust affects how your spouse (boyfriend/girlfriend) treats you or does it

also influence other areas of your life?

Influences
lust this Influences all
particular , situations in
situation 1 2 3 4 5 6 7 my life
45) How important would this situation be If it happened to you?
Not at all Extremely
Important 1 2 3 4 S 6 7 important

YOU APPLY FOR A POSITION THAT YOU WANT VERY BADLY (0.9.. IMPORTANT JOB. GRADUATE
SCHOOL ADMISSION. etc.) AND YOU GET IT. .

46) ' Write down one major cause

 

47) Is the cause of your getting the position due to something about you or something about other
people or circumstances? (Circle one number)

Totally due -
to other people Totally due
or circumstances 1 2 3 4 5 6 7 to me

48) In the future when applying for a position, will this cause again be present? (Circle one number)

Will never
again be Will always
present 1 2 3 4 5 6 7 be present

49) Is the cause something that lust Influences applying for a position or does it also Influence other
areas of your life? (Circle one number)

influences

just this . influences all
particular Situations in
Situation 1 2 3 4 5 6 7 my life

50) HOW Important would this situation be if it happened to yOu‘.’ (Circle one number)

Not at all Extremely
Important 1 2 3 4 5 6 7 important

198

YOU GO OUT ON A DATE AND IT GOES BADLY.

51) Write OOWn the one maior cause

 

52) is the cause of the date going badly due to something abdut you or something about other people or
Circumstances? (Circle one number)

Totally due
to other pebble Totally due
or circumstances 1 2 3 4 5 6 7 to me

53) In the future when dating, will this cause again be present? (Circle one number)

Will never
again be Will always
present 1 2 3 4 5 6 , 7 be present

54) is the cause something that just influences dating or does it also influence other areas of your life?
(Circle one number)

Influences

(ust this influences all
particular situations in
situation 1 2 3 4 5 6 7 my life

55) How important would this situation be if it happened to you? (Circle one number)

Not at all Extremely
Important 1 2 3 4 5 6 7 important

YOU GET A RAISE.

56) Write down the one maior cause

 

57) is the cause of your getting a raise due to something about you or something about other people or
circumstances?

Totally due
to other people Totally due
or circumstances 1 2 3 4 5 6 7 to me

$8) In the future on your job. will this cause again be present?

Will never
again be Will always
present 1 2 3 4 5 6 7 be present
5'3) Is this cause something that j'.-:‘. attests getting a raise or does it also influence other areas of your
life?
Influences
just this Influences all
particular ‘ situations in
situation 1 2 3 4 5 6 7 my life

60) How important would this situation be if it happened to you?

Not at all . Extremely
important 1 2 3 4 5 6 ' 7 important

I99

CHECK LIST

 

DACL FORM A
By Bernard lubin
Name l/o With/é Age Sex
Date Highest grade completed in school

 

DIRECTIONS: Below you will find words which describe different kinds of moods
and feelings. Check the words which describe How You Feel New - - Today. Some
of the words may sound alilm, but we want you to check all the words that describe
your feelings. Work rapidly and check 11} of the words which describe how you

 

 

W
1. [:1 Wilted ~17. C) Strong
2. [3 Safe 18. D Tortured
3. E] Miserable 19. El Listless
4. D Gloomy 20. D Sunny
5. Cl Dull 21. El Destroyed
6. D Gay 22. D Wretched
7. E] Low-spirited 23. C] Broken
8.0 Sad ‘ 24. E) Light-hearted
9. U Unwanted . 25. D Criticized
10. E] Fine 26. CI Grieved
11. E] Broken-hearted ‘27. D Dreamy
12. D Down-cast 26. D Hopeless
13. U Enthusiastic 29. D Oppressed
14. C] Failure 30. C) Joyous
15.D Afflicted 31. C] Weary
16. D Active 32. D Droopy

w MC ”1 metwummemmmunumni- mammaevmmmvm

like Unlike

I
re
3
is

DUDDDDDDD

DDUDDDDDDDDDD
eeesasaaesasa

E] 23.
D 24.
[j 25.

DDUDDDDDDUDDDUUDDUUDDUDDU

PesewewN-i

200

. Things usually don’t bother me.
. I find it very hard to talk in front of a group.

. There are lots of things about myself I'd change if I could.

I can make up my mind without too much trouble.
I'm a lot of fun to be with.

I get upset easily at home.

It takes me a long time to get used to anything new.
I'm popular with persons my own age.

My family usually considers my feelings.

. I give in very easily.

. My family expects too much of me.

. It’s pretty tough to be me.

. Things are all mixed up in my life.

. People usually follow my ideas.

. I have a low Opinion of myself.

. There are many times when would like to leave home.
. I often feel upset with my work.

. I’m not as nice looking as most peeple.

. If I have something to say, I usually say it.
. My family understands me.

. Most people are better liked than I am.

. I usually feel as if my family is pushing me.

I often get discouraged with what I am doing.
I often wish I were someone else.

I can’t be depended on.

© 1975 by Stanley Coopersmith. Published in 1981 by Consulting Psychologists
Press. All rights reserved. It is unlawful to reproduce or adapt this form without
written permission of the Publisher.

20l

COP I NG INVENTORY

 

Instructions: Please read the following list of c0ping behaviors. Indicate to
the right of each item the frequency with which you use each behavior to c0pe
with loneliness. If you are not currently feeling lonely, then please think
about a time when you have been lonely, and indicate the frequency that you used
these particular coping behaviors.

Strategy Never Seldom Sometimes Often
I. Go for a drive

 

 

 

 

2. Talk with a counselor or therapist
about ways to overcome loneliness

3. *Try harder to be friendly to other
people (like making a greater
effort to talk with students in
classes)

4.**Told yourself that your loneliness
would not last forever, that things
would get better

5.**Tell yourself that it's not so
important to be popular or to have
a boyfriend/girlfriend

6. Talk to a friend or relative about
ways to overcome your loneliness

7. Take your mind off feeling lonely
by using drugs or alcohol

8. Go shepping to buy things or spend
money.

9.**Remind yourself that you actually
do have good relationships with
other people.

10. Pray to God by yourself

11. *Try to find new ways to meet people
(such as joining a club moving into
a dorm, going to dances, etc.)

12. Take your mind off feeling lonely
through some sports activity

13.**rbink about possible benefits of
being lonely (like thinking that you
are learning to be more self reliant
or growing from the experience)

202

Strategy

14.

15.

16.

17.

18.‘

19.*

20.

21.

22.

23.**Tbink about things you do extremely
well (such as schoolwork, athletics,

24.*

25.

26.’

27.‘

Take naps or sleep longer at night
to get away from the feelings of
loneliness

Go to a play or movie
Go for a walk by yourself
Eat more than usual (such as

snacking frequently or eating
more at mealtime

*Try to figure out why you are

lonely

*Read the Bible or some other

religious/spiritual literature
Go out to the bars

Spend time with a close friend
just to be together

Do something helpful for someone
else (such as helping a classmate

with homework, doing volunteer work,

etc.)

artwork, cooking, etc.)

*Thought about good qualities that

you possess (intelligence, sensi-
tivity, self-sufficiency, warmth,
etc.)

Distract yourself from feeling
lonely with a quiet activity such
as reading a book, listening to
music, or watching TV

1'Think about things you can do to

overcome your loneliness

*Tell yourself that you are over-

reacting, that you shouldn't be
so upset

Never

Seldom

Sometimes

Often

203

Strategy

Never

Seldom

Sometimes

Often

 

28.

29.

30.

31.

32.

33.‘

34.‘

35.

'Do something to improve your social

skills (learning to dance, learning
to be more assertive, improving con-
versational skills, etc.)

Do something to make yourself more
physically attractive

Become sexually involved with
someone

*Get a job or get more involved

in a current job

i'Work extra hard to succeed at

some activity (such as studying
extra hard for an exam, putting
extra effort into practicing an
instrument

*Tell yourself that most other

people are lonely at one time
or another

*Take your mind off feeling lonely

by deliberately thinking about
other things (anything other than
your loneliness)

'Get more involved in your studies

 

*

fifi

Problem-focused Subscale items

Emotion-focused Subscale items

201-1

GRANT ATTRIBUTION SCALES - 1984

Please tell me how much you agree or disagree with the following statements about
your loneliness. Please circle one number under each statement between “1" and
"8" - with "1" representing strong disagreement and "8" representing strong
agreement.

1. I feel I can deal successfully with my lonely feelings.

Strongly Strongly
Disagree Agree
1 2 3 4 S 6 7 8

2. When I feel lonely, it is because I am in the wrong place at the
wrong time.

Strongly Strongly
Disagree Agree
1 2 3 4 5 6 7 8

3.**Loneliness cannot be cured, you have to learn to live with it.

Strongly Strongly
Disagree Agree
1 2 3 4 S 6 7 8

4. I feel lonely because of things that are outside of my control.

Strongly , Strongly
Disagree Agree
1 2 3 4 5 6 7 8

5.**Loneliness always goes away.

Strongly Strongly
Disagree Agree
1 2 3 4 S 6 7 8

6. I feel confident about my ability to handle my lonely feelings.

Strongly Strongly
Disagree Agree
1 2 3 4 5 6 7 8

7. If a person is lonely, it's because he or she hasn't made enough of
an effort to fight it.

Strongly Strongly
Disagree Agree
1 2 3 4 5 6 7 8

8.**Once lonely, always lonely.

Strongly Strongly
Disagree Agree
1 2 3 4 5 6 7 8

9. Sometimes I feel unable to cepe with my lonely feelings.
Strongly Strongly
Disagree Agree

1 2 3 4 5 6 7 6

205

10.**I believe my loneliness will pass.

Strongly Strongly
Disagree Agree
1 2 3 4 5 6 7 8
11. I am a lonely kind of person.
Strongly Strongly
Disagree Agree
1 2 3 4 S 6 7 8
12.**I often feel that loneliness is here to stay.
Strongly Strongly
Disagree Agree
1 2 3 4 5 6 7 8
13. *If I try hard enough, I can always avoid feeling lonely.
Strongly Strongly
Disagree Agree
1 2 3 4 5 6 7 8
I4. I only feel lonely now and again.
Strongly Strongly
Disagree Agree

15. ’Loneliness is usually caused by unfavorable circumstances.
Strongly Strongly
Disagree Agree

16. I am able to control by lonely feelings.

Strongly Strongly
Disagree Agree
1 2 3 4 5 6 7 8
17.**Even when I feel lonely, I know that sooner or later this will change.
Strongly Strongly
Disagree Agree
1 2 3 4 5 6 7 8
18. *If I feel lonely, it is usually because of the situation I am in.
Strongly Strongly
Disagree Agree
1 2 3 4 5 6 7 8
19.**No matter what I do, I'll always feel lonely.
Strongly ' Strongly
Disagree Agree

1 2 3 4 5 6 7 8

206

20. *I am more likely to feel lonely than other people because of my

personality.

Strongly Strongly

Disagree Agree
1 2 3 4 5 6 7 8

 

**

Items on Locus of Causality Subscale

Items on Locus of Stability Subscale

APPENDIX C

CROSS-TABULATION INFORMATION ON LONELINESS MEASURES

207

In response to the question, "Are you lonely?” 78 students checked
”yes," and 14 students checked ”don't know.” In terms of the Likert scale
for indicating degree of loneliness experienced (between and 10),
students endorsed numbers ranging from one to 10. The mean response of the
110 students who checked a Likert scale number was 3.7, and the standard
deviation was 2.3. On the Revised UCLA Loneliness Scale, the mean score was
39.9, and the standard deviation was 9.7. This mean score is very similar
to mean scores reported on general university aged samples that have been
recruited nationally on other university campuses for loneliness research
(Peplau and Perlman, 1982). (See Appendix C for presentation of the loneli-
ness measures data.)

Cross tabulations of these three measures of loneliness provided some
interesting results. 0f the 78 students who checked ”yes" to the direct
question, ”Are you lonely?" all of these students endorsed a response on the
Likert scale asking for an indication of severity of loneliness. 0f the 107
students who checked "no" to the question, 24 students gave responses
ranging from I to 9 on the Likert scale, with ten students circling number 1
and five students circling number 2. (These subjects may have thought that
indicating a number on the Likert scale was required, and thus indicated low
numbers). However, there were still nine students who indicated that they
were not lonely, who then gave responses for 3 or above on the 10-point
Likert scale, including two scores above 6.

Of the 14 students who checked ”don't know” to the question, "Are you
lonely?" eight endorsed numbers on the Likert scale ranging from 1 to 4,
with five students circling the number 3. There were no subjects in the

”Don't know” category who responded at 6 or above on the Likert scale (one

208

standard deviation above the mean. Thus, responses of ”Yes" to the question
"Are you lonely?” correspond to expected responses on the Likert scale,
whereas responses of ”No” to the question did not result in endorsement of
the Likert scale, as one would have expected.

Cross tabulations were also investigated between responses to "Are you
lonely?" and scores on the UCLA Loneliness ScaLe. Of the 78 subjects who
stated that they were lonely, 28 scored below the mean on the UCLA Scale and
51 scored above the mean. Of the group of 51 subjects, 27 were at least one
standard deviation beyond the UCLA mean. The mean score for this group was
44.4.

0f the 107 subjects who identified themselves as not being lonely, 71
had expected scores below the mean on the UCLA Scale, but 31 subjects had
scores that were above the UCLA mean, including eight scores that were
beyond one standard deviation above the mean. The mean score for this group
was 36.1.

Scores on the UCLA Scale for subjects identifying themselves as lonely
ranged from 22-70 (out of a possible range of 20-80). Scores on the
loneliness scale for the group of subjects identifying itself as not lonely
ranged from 22 to 57.

Of the group of subjects that responded “Don't know” to the question
"Are you lonely?‘ seven subjects scored above the mean on the UCLA Scale,
including three scores beyond one standard deviation above the mean. The
scores for this small group ranged from 27 to 54, with a mean score on the
UCLA scale of 41.5. Thus, the group of subjects that identified themselves
as not knowing whether or not they were lonely had a group mean score on the
loneliness scale that was above the mean score for the general sample.

Clearly, it is seen that the question, "Are you lonely?” cannot suffice

209

alone as a measure of loneliness if one is interested in more than a merely
subjective measure of loneliness. There were 38 students in all who scored
above the mean on the UCLA Loneliness Scale that had stated that they were
either not lonely or that they did not know whether or not they were lonely.

In looking at the cross tabulations between the Likert scale and the
UCLA Loneliness Scale, one sees that there were a substantial number of
students who scored at the Likert scale mean of 4 or below, who scored above
the mean on the UCLA Scale. There were also students who identified
themselves as feeling lonely according to the Likert Scale, who did not
yield scores at or above the mean on the UCLA measure. There were 12
students in this latter category. Thus, there were students who scored low
on the Likert scale and high on the UCLA Scale, and students who scored high
(”I the Likert scale who had loneliness scores below the mean on the UCLA
Loneliness Scale. There were eight students who scored one standard
deviation above the mean on the Likert Scale who scored at or below one
standard deviation below the mean on the UCLA measure. Thus, there is quite
a spread of "reversal" between these two measures on loneliness. The Likert
scale was able to identify some subjects who perceived themselves as lonely
or very lonely that the UCLA scale did not identify as such. Likewise, the
UCLA measure identified other subjects as lonely who had themselves not
responded to such a degree to the question ”Are you lonely?" or to the
.Likert scale on experienced loneliness. Given that a majority of the ”yes”
responses had corresponding indications of loneliness on one or both of the
other two measures, and that ”no" responses were represented by a wide
variation on the other two measures, it was decided that the use of the UCLA
scale and the Likert scale would provide for the best way of defining a

subsample of lonely subjects. In other words, the ”yes" subjects would be

210

"caught" by one of the other two measures if the person was experiencing a
mild to substantial degree of loneliness, whereas ”no” subjects or ”don't
know” subjects would also be identified by one of the other two measures as
well.

A subsample of lonely subjects was arrived at by the criteria of having
a score of 50 or greater on the UCLA Loneliness Scale or a score of 6 or
more on the Likert scale. These are scores at one standard deviation or
above the means for the respective scales. This subsample of lonely
subjects resulted in 49 students. The mean score on the Likert scale for
this group was 4.1 and the mean score on the UCLA Loneliness Scale was 51.7.
This compares with a mean of 30.9 on the UCLA Scale for a group of nonlonely
subjects. This subsample was arrived at by grouping subjects that had a
score of 35 or less on the UCLA scale and a score of 5 or less on the Likert
scale. Thus, among the lonely sample, the mean of 5l.7 on the UCLA Loneli-
ness Scale is almost 22 points higher than the mean on the scale for a

nonlonely group.

REFERENCES

Anderson. A. A. (1980). Metlxatienal and bentemanee deficits as a

hunting gt attributignal style. Unpublished doctoral disserta-
tion. Stanford University.

Beck. A. (1976). Cognitue therapy and the emetienal disendens. New
York: International Universities Press.

Becker. E. (1974). The spectrum of loneliness. Humanjtm 10,. 237-
246.

Bedeian. A. 8.. Geagud. R. J.. 8 Zmud. R. N. (n.d.). Test-retest
reliability and internal consistency of the short form of

Coopersmith's Sel f-Esteem Inventory. Esxchnlmml Ballads; 4.1.
725-726.

Berscheid. E.. 8 Halster. E. H. (1978). WW. 2nd
ed. Menlo Park. CA: Addison-Wesley.

Blau. Z. (1973). .Qde age in a ghangjng misty. New York: New
Viewpoints.

Blom. 6.. Ed. K.. 8 Kul karni. H. R. (1984).E_mJ_agtgn.c.o_p_1ng.w_Lth

disability Emotes: Lem East Lansing: Michigan State
University.

Bowlby: J- (1973). Attachment and less. 1191. IL. Separation; Anxietx
and anger. New York: Basic Books.

Bragg. M. A. (1979). W studx ei JenelJness and .dentessien.
Unpublished doctoral dissertation. University of California at

Los Angeles.

Brennan. T.. 8. Auslander. N. (1979). MW. Prepared
for National Institute of Mental Health. Juvenile Problems Divi-
sion. Behavioral Research Institute. Grant No. ROl-M-l 289 l2-Ol.

Brickman. P.. Ryan. K.. 8. Wortman. C. B. (1975). Causal chains:
Attribution of responsibility as a function of immediate and

prior causes. Jeumalet Bersenaljtxand Seeial £531:leme 32.
420-43l.

2|]

212

Christaans. X. (l965). Study of favorable and unfavorable influences
on successful study in 200 students in a teacher's college.

museum: loo: Remedischzaocial Keck; 12(4). 155-163.

Cohen. 0. H.. 8. Obrist. P. A. (1975). Interactions between behavior

and the cardiovascular system. W m 31. 693-
706.

Col son. C. (l973). An objective-analytic approach to the cl assifica-

tion of suicidal motivation. Acta Esyghgatjca Scandinaxin 4.2.
lOS-ll3.

Coll ins. B. C.. 8. Hoyt. M. F. (l972). Personal responsibil ity for
consequences: An integration and extension of ”forced compliance"

literature. Journal of Expenimental Social Esxcholoox. a. 558-
593.

Coopersmith. s. (1984). Selfisteem Intentonies. Palo Alto:
Consul ting Psychologists Press.

Cutrona. C. E. (1982). Transition to college: Loneliness and the
process of social adjustment. In L. A. Peplau 8. D. Perlman

(Eds.). Lonelinessibsouncebookofcucnenttbeomceseanchand
hunting. New York: John Hiley 8 Sons.

Cutrona. C. E.. 8 Pepl au. L. A. (1979. September).AJ_nng1_i-_u_d1na1

study on loneliness. Paper presented to the annual meeting of
the American Psychological Association. Toronto.

Dweck. C. 5.. 8 Reppucci. N. D. (l973). Learned helplessness and
reinforcement and responsibility in children. Journal at

Eersonalth and Social Esxcnolooxr 25.. 109-116.

Ellis. A. 0962). Reason and emotion in 9mm. New York: Lyle
Stuart.

Erikson. E. (1950). mm and Societx. New York: Norton.

Fol kman. 8.. 8 Lazarus. R. (l985). If it changes it must be a process:
Study of emotion and coping during three stages of a college

examination. Journal of EersonaJth and Social Esxcholouxr AB.
l50-l70.

Forsythe. D. R.. 8. McMillan. J. H. (1981). Attributions. affect. and
expectations: A test of Weiner's three dimensional model.
Journal of Educational Esxcboloou 13(3). 393-403.

Fromm-Rechmann. F. (l959). Loneliness. W22. l-lS.

213

Garmezy. N. (n.d.). Children under stress: Perspective on antecedents
and correlates of vulnerability and resistance to psychopathol-
ogy. In A. 1. Robin. J. Aranoff. A. M. Barclay. 8 R. A. Zucker

(Eds.). further exnlorations .in nersonality- New York-

Girodo. M. (l973). Film-induced arousal. information search and the

attribution process. Journal of Eersonality and Social Esxcbol:
ogy.. 25. 357-360.

Gordon. S. (l976). Lonely in mm. New York: Simon 8 Schuster.

Goswich. R. A.. 8Jones. N. H. (1981). Loneliness. sel f-concept. and
adjustment. Journal oi Esycnolooy. 101. 237-240.

Haan. N. (l965). Coping and defense mechanisms related to personality
inventories. Journal oi Consulting W 22. 373-378.

Haan: N. (1977). Coping and defending: Erocess of selicenuironment
organization. New York: Academic Press.

Hartog. J.. Audy. J. R.. 8 Cohe. Y. A. (Eds.). 0980). Iho anatomy of
Jonolinoss. New York: International Universities Press.

Heider. F. (1958). Inc osycholony of interpersonal relations New
York: Wiley.

Holmes. J. G... 8 Strickland. L. H. (1970). Choice. freedom and
confirmation of incentive expectancy as determinants of attitude
change. Journal of Eersonality and Social Esycholoou is. 39—45.

Horowitz. L. H.. 8 French. R. (1979). Interpersonal problems of

people who describe themselves as lonely. Journal of mulling
and Clinical Esycnolonyr 41. 762-764.

Hugdahl. K. (1980). A test of external-internal attributions in

learned helplessness. Scandinayian Journal of Benayior Ineranya.
2(2). 79-86.

Jacobs. L. E.. Berschei d. E.. 8 Nalster. E. (1971). Self-esteem and

attraction. Journal of Eersonality and Social Esucnolom 11.
84-910

James. VI. (l908). Esyohology. New York: Henry Holt.

Jones. N. H.. Freeman. J. A.. 8 Goswick. R. A. (l98l). The persistence
of loneliness: Self and other determinants. Journal of Benson:

ML 4.29 27-48.

214

Jones. W. H.. Hanson. R. 0.. 8 Smith. T. G. (1980). Loneliness and

lore. Implications for psychological and interpersonal function:
ing- Unpublished manuscript. University of Tulsa.

Jones. M. H.. Hobbs. S. A.. 8 Hockenbury. D. (198l). Loneliness and
social skills deficits. Joumal of Eersonality and Social
WI. 12: 27-48.

Kelley. H. H. (197l). Attribution in social interaction. In E. E.
Jones. D. E. Kanouse. H. H. Kelley. R. E. Nesbitt. S. Valins. 8

B. Weiner (Eds). Attrihutiom Eerceixing the cause of. hehauior.
Morristown. NJ: General Learning Press.

KU1kdf‘l’l‘lp M. Re, &B]Ofl|o G. E. (1983LW MWM
tory-_-A tool .for the eualuation of coping hehaxior of disabled
persons. East Lansing: Department of Vocational Rehabilitation.
Michigan State University.

Lamont. L. (l979). Canons shook. New York: Dutton.

Lazarus. R. S.. 8 Cohen. J. B. (1977). Ihe lays of Coping Shale.
Berkeley: Department of Psychology. University of California at
Berkeley.

Lazarus. R. 8.. 8 Fol kman. S. (1984).mwmm.
New York: Springer.

Lowenthal. M. F.. Thurner. M.. 8 Chiri boga. D. (l976).£oui:s_tagesof
.LLie. San Francisco: Jossey-Bass.

Lubin. 8. (l98l). Manual for the Depression Adioctilte Checklists. San
Diego: Educational and Industrial Testing Service.

Lubi n. B.. Nathan. M. M.. 8 Nathan. R. G. (n.d.). Wot
response formats iorthe Depression Adiectiue Checklists.

Mahler. M.. Pine. F.. 8 Bergnan. A. (l975). Ihe osychologioal birth of
the human infant. New York: Basic Books.

McCormack. S. H.. 8 Kahn. A. (l980. May). flehayiocalohanaeterietios
of lonely and nonlonely college students Paper Presented at the
Midwestern Psychological Association. St. Louis. ND.

Meninger. K. (l954). Regulatory devices of the ego under major stress.
Journal of Esychoanalysisi 35. 412—420.

Michel a. J. L.. Peplau. L. A.. 8 Weeks. D. B. (l980. July). .Eeooeiyeo

dimensions and consequences of attributions for loneliness.
Unpublished manuscript. University of California at Los Angeles.

215

Mijuskovic. F. (l980). Loneliness: An interdisciplinary approach. In
J. Hartog. J. R. Audy. 8 Y. A. Cohe (Eds.). Anatomy ofloneli:
ness. pp. 65-95. New York: International Universities Press.

Moore. J. A.. 8 Sermat. V. (1974). Relationship between loneliness and
interpersonal relationships. Canadiannounselori 8. 84-89.

Moustakas. C. (l96l). Loneliness. Englewood Cliffs. NJ: Prentice-
Hall.

Murphy. L. B. (1974). Coping. vulnerability and resilience in
childhood. In G. B. Coelho. D. A. Hamburg. 8 J. E. Adams (Eds.).

Coping and adaptation. New York: Basic Books.

Nerriano. V. J.. 8Gross. W. F. (1976). Loneliness and locus of

control for alcoholic males. Journal of Clinical Esychology. 32.
479-484.

Obrist. P. A. (1981). Cardioxascular patchoohxsiologn A perspectiue.

New York : Plenum.

Ostrov. E.. 8 Offer. D. (l980). Loneliness and the adolescent. In J.

Hartog. J. R. Audy. 8 Y. A. Cone. Anatomyoiloneliness. New
York: International Universities Press.

Paloutzian. R.. 8 Ellison. C. (1982). Loneliness. spiritual well-
being. and the quality of life. In L. A. Peplau 8 D. Perlman

(Eds.). Lonelinessnsourcehookoicurrenttheorxrresearchland
oractioe. New York: John Wiley 8 Sons.

Pannalee. P.. 8 Werner. C. (1978). Lonely losers; Stereotypes of

single dwellers. Eersonality and Social Esychology Bulletim A.
292-295.

Pearl ine. L. I.. 8 Schooler. C. (1978). The structure of coping.
JournaloihealthandSocialBehayiorrla. 2-2l.

Pelletier. K. (1977). Mind as healen mind as slayer. New York:
Delta.

Pepl au. L. A.. 8 Perl man. D. (1982).].onelinessahsoureehoohof
currenttheom researchrandpractice. New York: John Wiley 8

Sons.

Perl man. D.. Gerson. A. C.. 8Spinner. B. (1978). Loneliness among
senior citizens. An enpi rical report. Essence; 2. 239-248.

216

Piaget. J. H952). Ihe origins of intelligence in children. New York:
International Universities Press.

Pittman. T. S. (1975). Attributions of arousal as a mediator in

dissonance reduction. Journal of. Experimental Social Esxchologyr
11. 53-63.

Pl att. J. J.. 8 Spivack. G. (1975). Manual .for the W problem;
soloing orocodura. Philadelphia: Hahnemann Medical College 8
Hospital.

P01 husky: N- (1979). on loneliness; A program .for social pork.
Northanpton. MA: Smith College.

Revenson. T. (1981. December). Coping with loneliness: The impact of

causal attributions. Eersonality and Social Esxchology Bulletim
1(4). 565-571.

Rubenstein. C.. 8 Shaver. P. (1982). The experience of loneliness. In
L. A. Peplau 8 o. Perlman (Eds.). LonelinesslAsourceboekof
current theory... research and practice. New York: John Wiley &

Sons.

Russell. D.. Cutrona. C.. Jayne. R.. 8 Yurko. K. (1984). Social and
emotional loneliness: An examination of Weiss's typology of

loneliness. Journal oiEersonalityandSocialEsycholong in.
1313-1321.

Russell. D.. Peplau. L. A.. 8 Cutrona. C. E. (1980). The Revised UCLA
Loneliness Scale: Concurrent and discriminant validity evidence.

JournaloifiersonalityandSocialEsychologyiafiw).

Russell. D.. Peplau. L. A.. 8 Ferguson. M. L. (1980). Developing a

measure of loneliness. Journal of Eersonality Assessment. .42.
290-294.

Schacter. 8.. 8 Singer. J. (1962). Cognitive. social and physiological

determinants of emotional states. Esychologioal Beyieu 59,. 379-
399.

Schank. R.. 8Abe1 son. R. (1977).§oniotsiolans;goal_s._andunoer_-_
standing. New York: John Wiley 8 Sons.

Sel igman. M. E. P. (1975). flelolessness. San Francisco: W. H.
Freenan.

Seligman. M. E. P.. Abramson. L.. Semmel. A.. 8 von Baeyer. C. (1979).

Depressive attributional style. Journal of Abnormal Esxohologyr
5.813). 242-247.

217

Seligman. M. E. P.. Peterson. C.. Semmel. A.. von Baeyer. C..
Abramson. L.. 8 Metalsky. G. (1980). IheAttrihutional

Stile
Questionnaire. Psychology Department. University of Pennsyl-
vania.

Sel man. R. (1971). The relation of role-taking to the development of
moral Judgnent in children. Child Deyelopmenti 42. 79-92.

Sermat. V. (1980). Some situational and personality correlates of
loneliness. In J. Hartog. J. R. Audy. 8 Y. A. Cohe (Eds.).

Anatomy of loneliness. New York: International Universities
Press.

Solano. C. H.. 8 Batten. P. G. (1979).].onelinessandohiecti1eself;

disclosure in an acouaintanceship exercise Unpublished manu-
script. Wake Forest University.

Storms. M. D.. 8 Nisbett. R. E. (1970). Insomnia and the attribution

$500955. JournaloiEersonalityandSocialEsychologyrlo. 319-

Strong. S. R. (1973). Systematic causality in counseling. Applications

to theory. practice and research. Causality and lollies; .11. 143-
151.

Sullivan. H. (1953). lbs interpersonal theory of psychiatry. New
York: W. W. Norton.

Vaillant. G. E. (1977). Adaptationtolite. Boston: Little. Brown.

Vash. C. L. (1981). Iheosychologyof disability. New York: Springer.

Weiner. B. (1974). Achieuement motixation and attribution theori-
Morri stcwn. NJ: General Learning Press.

Weiner. B.. Russell. D.. 8 Lerman. D. (1979). The cognition-emotion
process in achievement related contexts. Journal of. Eersonality.
and Social Esychologxi 31. 1211-1220.

"6155. R- 8. (1973). Loneliness: Ihe experience of emotional and
social isolation. mmbridge. MA: MIT Press.

White. R. W. (1974). Strategies of adaptation: An attempt at syste-
matic description. In G. Coel ho. D. Hamburg. 8J. Adams (Eds.).

Coping and adaptation. New York: Basic Books.

Wright: 3. A- (1960). thsical disability: A psychological approach.
New York: Harper 8 Row.

 

218

Wrigit. B. A. (1980). Developing constructive views of life with a
disability. Rehabilitation Literature. 51(11-12). 274-279.

Young. J. E. (1980). Cognitive therapy and loneliness. In R.
Bedrosian. G. Emery. 8 S. Hollon (Eds.). Ney directions in
cognitlxe therapy; A case hook. New York: Gruilford Press.

Zimbardo. P. G. (1978). Shyness. New York: Harcourt-Brace-Jovanovich.

Zuckerman. M.. 8 Lubin. B. (1965). Multiple Afloat W Cheek
List; Manual. San Diego: Educational and Industrial Testing
Service.