Self‐pity: Exploring the links to personality, control beliefs, and anger

Abstract

Self-pity is a frequent response to stressful events. So far, however, empirical research has paid only scant attention to this subject. The present article aims at exploring personality characteristics associated with individual differences in feeling sorry for oneself. Two studies with N=141 and N=161 university students were conducted, employing multidimensional measures of personality, control beliefs, anger, loneliness, and adult attachment. With respect to personality, results showed strong associations of self-pity with neuroticism, particularly with the depression facet. With respect to control beliefs, individuals high in self-pity showed generalized externality beliefs, seeing themselves as controlled by both chance and powerful others. With respect to anger expression, self-pity was primarily related to anger-in. Strong connections with anger rumination were also found. Furthermore, individuals high in self-pity reported emotional loneliness and ambivalent-worrisome attachments. Finally, in both studies, a strong correlation with gender was found, with women reporting more self-pity reactions to stress than men. Findings are discussed with respect to how they support, extend, and qualify the previous literature on self-pity, and directions for future empirical research are pointed out.
There are a hundred ways to overcome an obstacle and one sure way not to—self-pity.
Dale Dauten, columnist

Full text

Self-Pity: Exploring the Links to Personality,
Control Beliefs, and Anger
Joachim Sto
¨ber
Martin Luther University of Halle-Wittenberg
ABSTRACT Self-pity is a frequent response to stressful events. So far,
however, empirical research has paid only scant attention to this subject.
The present article aims at exploring personality characteristics
associated with individual differences in feeling sorry for oneself. Two
studies with N5141 and N5161 university students were conducted,
employing multidimensional measures of personality, control beliefs,
anger, loneliness, and adult attachment. With respect to personality,
results showed strong associations of self-pity with neuroticism,
particularly with the depression facet. With respect to control beliefs,
individuals high in self-pity showed generalized externality beliefs, seeing
themselves as controlled by both chance and powerful others. With
respect to anger expression, self-pity was primarily related to anger-in.
Correspondence concerning this article should be addressed to Joachim Sto
¨ber,
Department of Educational Psychology, Martin Luther University of Halle-
Wittenberg, Franckesche Stiftungen, Haus 5, D-06099 Halle (Saale), Germany.
Electronic mail may be sent via Internet to stoeber@paedagogik.uni-halle.de
Parts of the results described in this article were presented at the 10th European
Conference on Personality in Krakow, Poland, in July 2000.
I would like to thank Claudia Dalbert and Uwe Wolfradt for support with the data
collection of Study 1; Wilhelm Janke for sending me the English translation of the
SVF self-pity scale and for granting permission to reproduce the scale in the present
article; Alexandra Freund and Jacqui Smith for providing me with the German short
form of the UCLA Loneliness Scale; Fritz Ostendorf and Hannelore Weber for
providing me with the preliminary German version of the NEO Personality Inventory-
Revised and the Anger-related Reactions and Goals Inventory, respectively; and
finally, Claudia Dalbert, Jutta Joormann, Hannelore Weber, and two anonymous
reviewers for helpful comments and suggestions on earlier versions of this article.
Journal of Personality 71:2, April 2003.
Copyright r2003 Blackwell Publishing.

Strong connections with anger rumination were also found. Furthermore,
individuals high in self-pity reported emotional loneliness and ambiva-
lent-worrisome attachments. Finally, in both studies, a strong correlation
with gender was found, with women reporting more self-pity reactions to
stress than men. Findings are discussed with respect to how they support,
extend, and qualify the previous literature on self-pity, and directions for
future empirical research are pointed out.
There are a hundred ways to overcome an obstacle and one sure way not
to—self-pity.
Dale Dauten, columnist
Introduction
Self-pity is a prevalent response to stressful events such as personal
failure, loss, or illness. However, psychological research has paid
only scant attention to the investigation of self-pity. The few authors
who have provided analyses of self-pity, thus far, all come from a
psychiatric or psychoanalytic background (Charmaz, 1980; Elson,
1997; Grunert, 1988; Kahn, 1965; Milrod, 1972; Wilson, 1985). Even
though their clinical observations and theoretical reflections are
helpful in providing a first approach to the subject matter, the case
studies they describe cannot substitute for systematic empirical
studies. Empirical studies on self-pity, however, are even more
scarce. Moreover, none of the empirical studies conducted so far
has focused directly on self-pity; measures of self-pity have merely
been included as one variable among many others. In most of these
studies, self-pity has been measured with the respective subscale of
the Streverarbeitungsfragebogen [Coping with Stress Questionnaire]
(Janke, Erdmann, & Kallus, 1985), a German inventory to assess
various responses to stress.
Given this background, the present article has three aims. First, I
will review the literature on self-pity, presenting observations from
the psychoanalytic and psychiatric literature and summarize the
findings of the empirical studies available. Second, I will present two
empirical studies on self-pity that further explore the links with
personality, control beliefs, and anger for which previous research
established first findings. Moreover, I will explore presumed
associations of self-pity with loneliness and adult attachment.
Specifically, in Study 1, I will show how self-pity relates to the
five-factor model of personality, as well as to control beliefs, styles
184 Sto
¨ber

of anger expression, and social and emotional loneliness; then, in
Study 2, I will follow up with a detailed analysis of how self-pity
relates to the facets of neuroticism, functional and dysfunctional
anger reactions, and different attachment styles. Finally, I will
discuss potential limitations of the two studies, integrate the present
findings with the previous literature, and point out some directions
that future studies on self-pity may take.
Psychiatric and Psychoanalytic Reflections
Pity has been defined as ‘‘sympathetic heartfelt sorrow for one that
is suffering physically or mentally or that is otherwise distressed or
unhappy’’ (Webster’s Third New International Dictionary, 1961,
p. 1726). Self-pity is pity directed toward the self. Consequently, self-
pity may be defined as a sympathetic, heartfelt sorrow for oneself
prompted by one’s own physical or mental suffering, distress, or
unhappiness. Interviews with individuals suffering from chronic
illness (Charmaz, 1980) have indicated that self-pity is often
accompanied by feelings of sadness and loss and a heightened sense
of injustice. Moreover, for a person who feels self-pity, it is
characteristic to feel envy of others who have not suffered a similar
loss or fate. This is expressed in questions like ‘‘Why not them?’’,
‘‘Why me?’’, or ‘‘What did I do to deserve this?’’, which typically
accompany the internal monologue associated with experiences of
self-pity (Charmaz, 1980; Grunert, 1988).
The experience of self-pity is not restricted to individuals suffering
from chronic illness or severe losses. Rather, it is an emotional
experience which, in all likelihood, all humans encounter occasion-
ally (Kahn, 1965). Life holds many opportunities to feel sorry for
oneself. Not only critical life events such as not getting a promotion
but also minor incidents such as being rebuffed by someone or
simply not getting enough attention may provoke feelings of self-
pity. However, like psychological research, the psychoanalytic and
psychiatric literature has paid scant attention to self-pity, with few
publications mentioning self-pity at all (Wilson, 1985). Nevertheless,
these initial efforts, based on case studies and informal clinical
observations, can serve as both a starting point and a frame of
reference for the present investigations.
Within the available literature on self-pity, there is considerable
agreement that self-pity is an emotional response that emerges in
times of stress (e.g., Elson, 1997; Kahn, 1965; Wilson, 1985). The
Self-Pity 185

propensity to react to stress by feeling sorry for oneself, however, will
show great individual differences related to certain personality
characteristics. In this respect, Kahn (1965) was the first to point
out that individual differences in neuroticism may predict who will
respond with self-pity and who will not. Based on his clinical
experience, he argued that self-pity may play a significant role in the
lives of people he called ‘‘psychoneurotic.’’ These individuals are
characterized by great self-insecurity when confronted with problems.
Moreover, they are described as people with a ‘‘thin skin’’ who are
less able to cope with stress than those who are emotionally more
stable and who thus react oversensitively in the face of situational
difficulties that others might easily brush off (Kahn, 1965). More-
over, self-pity reactions have been closely linked to individual
differences in depression. Grunert (1988), for example, argued that
self-pity plays an important role in melancholia. Moreover, Wilson
(1985) stated that he found ‘‘an underlying smoldering depression’’
(p. 183) in all his case studies of pervasive self-pity.
Self-pitying persons are characterized as likely to overindulge in
their failures, hardships, and losses, and the circumstances elicited
by these setbacks, thus becoming self-consciously preoccupied with
their own suffering (Charmaz, 1980). Nevertheless, self-pity is not an
emotional response directed exclusively towards the self. Whereas
the primary focus in self-pity may be on the self, self-pity also has a
strong interpersonal component. Quite often, self-pity is an
emotional response directed toward others with the goal of
attracting attention, empathy, or help (Kahn, 1965). In this respect,
however, it is a strategy doomed to fail. Whereas, initially, the
display of self-pity may evoke empathy from others (Milrod, 1972),
pervasive self-pity will not. On the contrary, people who show
pervasive self-pity are most likely to be rejected. Even for individuals
who suffer from chronic illness, the period of time is quite limited
during which the social environment will allow for a display of self-
pity. After a while, people are expected to accept their fate, stop
complaining, and carry on with their lives (Charmaz, 1980).
Finally, the psychiatric and psychoanalytic literature holds that
self-pity is linked to feelings of both loneliness and anger. Clinical
observations suggest that individuals who experience self-pity
usually expect more from the environment than the environment is
willing to give (Kahn, 1965). Personal relationships are perceived
as unstable and characterized by high demandingness on the part
186 Sto
¨ber

of the person who experiences self-pity and who sees his or her
environment as unwilling to provide the empathy, comfort, and
support he or she demands. Consequently, a person who feels self-
pity is permanently frustrated. This permanent frustration with
others may have two consequences. First, it may lead to social
withdrawal and feelings of loneliness (Charmaz, 1980; Kahn, 1965).
Second, it may lead to feelings of aggression, hostility, and anger
(Kahn, 1965; Milrod, 1972; Wilson, 1985). However, open displays
of aggression, hostility, and anger are in conflict with the aims of
attracting empathy, support, and acknowledgment from others.
Once more, as Kahn (1965) suggested, individuals with a suscept-
ibility for self-pity often are characterized by great self-insecurity.
Thus, they may lack the self-assertiveness needed to confront others
openly. As a consequence, the direct expression of aggression and
hostility will be inhibited. Only mild forms of anger will be
expressed, whereas strong anger will be suppressed, turned inward,
or even turned against oneself (Milrod, 1972; Wilson, 1985). Under
the surface, however, the anger against others will continue to exist,
often accompanied by ruminations about retributions for the past
(Charmaz, 1980).
Empirical Findings
Empirical studies on self-pity are largely restricted to research
conducted with the Streverarbeitungsfragebogen (SVF; Janke et
al., 1985), simply because it is the only questionnaire that contains a
reliable and valid scale to assess individual differences in self-pity.
1
Overall, the SVF comprises 19 scales to assess different ways of
coping, covering a wide spectrum of behavioral and cognitive
strategies that people use to deal with stress. One of those
scales measures Selbstbemitleidung or, in English, self-pity (see
Appendix A).
1. There are two English publications that contain scales to measure ‘‘self-pity.’’
Both scales, however, have problematic aspects. First, the measure presented by
MacAndrew (1989) seems to have low face validity, as is suggested by inspection
of its items (e.g., ‘‘Several times a week I feel as if something dreadful is about to
happen’’; ‘‘I have often found people jealous of my good ideas, just because they
had not thought about them first’’). Another study (Sundberg, 1988) contains a
scale that measures self-pity, but only in combination with rejection and lack of
purpose. Moreover, all items refer to situations in which the respondent feels
lonely.
Self-Pity 187

The self-pity scale of the SVF contains six items. One item
addresses self-pity directly; the other items address characteristics
that are typical for self-pity such as ‘‘Why me?’’ questions and envy
of others who seem to fare better (e.g., Charmaz, 1980; Grunert,
1988). Factor analyses of the SVF scales have resulted in solutions
with four or six higher-order factors of coping with stress. In all
these solutions, self-pity loaded on the same higher-order factor as
rumination, self-accusation, social withdrawal, resignation, and
avoidance tendencies ( Janke et al., 1985). Thus, self-pity clearly
falls into the class of ineffective coping strategies that are more likely
to exaggerate a problem and create new difficulties than to help deal
successfully with stressful situations.
Further support for the claim that self-pity is a highly ineffective
coping strategy comes from two studies. In one study (Kro
¨ner-
Herwig, Muck, & Weich, 1988), a sample comprised of experts
(psychologists) and laypersons (engineers and university employees)
judged the effectiveness of the coping strategies presented in the
SVF. For each coping strategy, participants indicated how effective
such a reaction would be to help cope with a stressful situation and
regain psychological balance. Overall, there was considerable
agreement between experts and laypersons about the effectiveness
of the various strategies. Self-pity was judged to be one of the least
effective coping strategies. Only aggression, social withdrawal,
resignation, and drug use received lower effectiveness ratings.
Another study (Becker, 1985) investigated the relationship of the
SVF coping strategies with mental health. Mental health was
measured with a combination of L-data (life data) and Q-data
(questionnaire data). L-data were based on reports from the
participants’ physicians and included diagnoses on mental health
and interview data on social adjustment, emotional stability, and
self-actualization. Q-data were based on participants’ self-reports
and included global ratings on emotional stability and a number of
measures tapping the degree of self-actualization (e.g., satisfaction
with life, self-acceptance, openness, and purpose in life). L-data and
Q-data were subjected to a principal component analysis, and the
regression scores from the first principal component were taken as a
comprehensive measure of mental health. When this measure of
mental health was correlated with the SVF subscales, self-pity was
among the strategies that showed the highest negative correlation—
even higher than those obtained for resignation and drug use.
188 Sto
¨ber

Systematic findings on self-pity regarding possible links to
personality are largely restricted to studies including the traits
of neuroticism and extraversion ( Janke et al., 1985). Across
different samples, findings have shown that self-pity is closely
related to neuroticism, but largely unrelated to extraversion.
Correlations with neuroticism were in the range of .40 to .60, thus
supporting Kahn’s (1965) view that self-pity seems to be an
emotional response that is characteristic of ‘‘psychoneurotic’’
individuals. In addition, Janke and colleagues found self-pity to be
highly correlated with measures of depression, as well as with
sensitization, as measured with Byrne’s (1961) Repression-Sensitiza-
tion Scale. Originally, Byrne’s scale was constructed as a means to
measure individual differences in defensiveness and repression.
Psychometric research, however, produced evidence that Byrne’s
scale correlates with measures of trait anxiety in the same order of
magnitude as its own reliability. Therefore, it has been argued that
Byrne’s scale is a measure of trait anxiety rather than sensitization
(for a review, see Krohne, 1996). Consequently, the high correlation
between self-pity and sensitization may indicate that self-pity is not
only closely related to depression, but also to trait anxiety.
Apart from high correlations with neuroticism, depression,
and sensitization (trait anxiety), research has produced only a
few findings that show self-pity to have systematic associations
with other personality variables. However, there are two noteworthy
exceptions. These relate to individual differences in control beliefs
and styles of anger expression. In a study with a large community
sample ( Janke et al., 1985), self-pity showed a significant negative
correlation with locus of control as measured with Rotter’s (1966) I/
E scale, indicating that individuals with a tendency for self-pity have
a more external locus of control. Further, in a study with a large
student sample (Schwenkmezger, Hodapp, & Spielberger, 1992),
self-pity showed substantial correlations with Spielberger’s (1988)
anger expression scales, indicating that individuals with a tendency
for self-pity show higher levels of both anger-in and anger-out as
well as lower levels of anger control.
Open Questions
The available empirical findings, albeit few and unsystematic,
provide the first empirical support for some of the notions on
Self-Pity 189

self-pity proposed in the clinical literature. Some questions are still
unanswered. With respect to broad dimensions of personality,
systematic research including self-pity has been restricted to the
traits of neuroticism and extraversion, thus leaving open the
question of whether self-pity will show links with any other broad
dimensions of personality. The five-factor model of personality is
currently the dominant model for capturing broad dimensions of
personality. Its Big Five dimensions have been labeled (a)
neuroticism or, its opposite, emotional stability; (b) extraversion
or surgency; (c) agreeableness; (d) consciousness or dependability;
and (e) openness, culture, or intellect ( John, 1990). Though there
has been—and still is—considerable debate as to whether these five
dimensions constitute an adequate and comprehensive description
of personality (e.g., Block, 1995; McAdams, 1992), the five-factor
model of personality represents an established base from which one
may start to explore potential relationships between self-pity and
personality. Moreover, with measures available on both trait level
and facet level, this model allows for a hierarchical assessment of
personality (Costa & McCrae, 1995a, 1995b). In this regard, it may
be particularly useful to further explore the relationship between
self-pity and the neuroticism facets of depression and anxiety, and to
investigate further the previous findings related to depression and
sensitization.
Furthermore, the finding that self-pity is related to an external
locus of control also leaves some questions unanswered. The reason
is that Rotter’s (1966) I/E scale conceptualizes control beliefs as a
unidimensional construct with internality and externality as the
endpoints of a continuum. This unidimensionality, however, has
proven untenable, with most factor analyses of Rotter’s scale clearly
showing multidimensional solutions. In the wake of these findings,
efforts were undertaken to develop models and instruments that
took the apparent multidimensionality of control beliefs into
account. One prominent outcome of these efforts was Levenson’s
(1974, 1981) tripartite model of control beliefs. This model
distinguishes three dimensions of control: (a) internality, (b)
externality related to powerful others, (c) and externality related
to chance. The differentiation between powerful others and chance
was based on research with students who engaged in political
activism (Levenson, 1974). This research demonstrated that it makes
a great difference whether people conceive of their fate as controlled
190 Sto
¨ber

by chance or controlled by powerful others. Only in the latter case
does the potential exist for the individual to regain personal control
and initiate changes. Studies correlating measures from Levenson’s
tripartite model with Rotter’s I/E scale have shown that Rotter’s
scale mainly captures externality related to chance (e.g., Brosschot,
Gebhardt, & Godaert, 1994). Consequently, with respect to the
findings of Janke and colleagues (1985), it remains an open question
as to whether self-pity is associated only with externality related to
chance, or if it also shows associations with externality related to
powerful others or internality.
Finally, some open questions remain from the previous research
findings on self-pity and anger (Schwenkmezger et al., 1992). First,
these findings appear somewhat inconsistent, if not contradictory,
because they indicate that self-pity is closely related to nonexpres-
sion of anger (anger-in) and, at the same time, to outward expression
of anger (anger-out) and low anger control. Second, they are not in
line with observations reported in the psychoanalytic and psychiatric
literature, where there is broad agreement that self-pity is associated
with the suppression of anger. Whereas subtle expressions of anger
may be observed in self-pity, open expressions of anger and
aggression are unlikely. From these accounts, one would expect
individuals with a tendency to react with self-pity to show high levels
of anger-in, but neither high levels of anger-out nor low levels of
anger control. A potential explanation for these inconsistencies may
lie in the fact that the three anger expression scales of the Spielberger
inventory show substantial overlap; anger-out and anger control in
particular have shown substantial correlations (e.g., Schwenkmezger
et al., 1992; Spielberger, 1988). Consequently, taking this overlap
into account and exploring unique relationships between self-pity
and anger expression may produce a different pattern of correlations
that would show more internal consistency and more congruence
with the literature.
Following the same line of reasoning, it may also be important to
control for gender differences. Studies with the SVF have shown
that women report higher self-pity scores than men ( Janke et al.,
1985). The potential effect that gender differences may have on the
observed correlations has not been considered in any of the previous
studies. Consequently, it may well be the case that correlations
between self-pity and variables for which women show higher values
have been inflated because of shared variance with gender.
Self-Pity 191

Aims of the Present Studies
Based on these questions, the aim of the present studies was to
investigate the relationships between self-pity and personality traits,
control beliefs, and styles of anger expression in order to replicate,
extend, and possibly qualify the previous findings. To this end, the
aim of Study 1 was to locate self-pity within the five-factor model of
personality, to investigate the relationships between self-pity and
Levenson’s control beliefs, and to reinvestigate the relationships
between self-pity and styles of anger expression. Moreover, Study 1
had the goal of providing a first investigation of the relationship
between self-pity and dimensions of loneliness. From the clinical
literature and the previous research findings, it was expected that
individuals with a tendency for self-pity would show higher levels of
neuroticism, external locus of control, loneliness, anger-in, and
anger-out, as well as lower levels of anger control. Moreover,
women were expected to show higher levels of self-pity than men. All
other analyses, regarding both the relationships of self-pity with the
remaining dimensions of the five-factor model, and the associations
of the aforementioned variables after control of gender effects and
overlap between scales, were exploratory.
STUDY 1
METHOD
Participants
A sample of N5141 students (75 females, 66 males) was recruited at
the Martin-Luther University of Halle-Wittenberg. Average age was
22.6 years (SD 53.1). Participants volunteered in exchange for one
hour of extra course credit or for a lottery ticket for a chance to win
100 German marks (approximately 47 U.S. dollars).
Measures
Self-pity. Self-pity was measured with the six-item self-pity scale of
the Streverarbeitungsfragebogen ( Janke et al., 1985; see Appendix
A). Items are answered on a 5-point scale from not at all (0) to very
likely (4). Scores are computed by summing across items. In the
present sample, the scale displayed a Cronbach’s alpha of .84.
192 Sto
¨ber

Big Five personality traits. The personality traits of the five-factor
model were measured with the NEO Five Factor Inventory (NEO-
FFI; Costa & McCrae, 1992; German version: Borkenau &
Ostendorf, 1993). The NEO-FFI is a widely used, reliable, and
valid measure of the Big Five personality dimensions (see Borkenau
& Ostendorf, 1993). It comprises five 12-item scales that capture
individual differences in neuroticism (e.g., ‘‘I often feel tense and
nervous’’), extraversion (‘‘I really like talking to other people’’),
openness (‘‘I enjoy playing with theories or abstract ideas’’),
agreeableness (‘‘Most of the people I know like me’’), and
conscientiousness (‘‘I work hard in order to reach my goals’’).
Items are answered on a 5-point scale from strongly disagree (0) to
strongly agree (4). Scores are computed by averaging across items.
With alphas ranging from .72 to .87, internal consistency was
satisfactory for all scales.
Control beliefs. The control beliefs of Levenson’s (1974) tripartite
model were measured with the respective scales from the
Questionnaire on Competency and Control Beliefs (Krampen,
1991). The questionnaire contains three eight-item scales that
capture individual differences in internality (e.g., ‘‘The course of
my life is only determined by my own behaviors and efforts’’),
externality related to powerful others (‘‘The course of my life is in
many aspects determined by other persons’’), and externality related
to chance (‘‘Many things that happen in my life are determined by
chance’’). Following Krampen (1991), items are answered on a 6-
point scale from very false (–3) to very true (13) without a zero-
point. Scores are computed by summing across items. With alphas
from .73 to .75, all scales displayed satisfactory reliability. In line
with previous findings (Krampen, 1991; Levenson, 1974), powerful
others and chance were highly correlated (r5.56) and moderately
related to internality (r5–.28 and –.39, respectively).
Anger expression. Styles of anger expression were measured with the
scales from the State-Trait Anger Expression Inventory (Spielberger,
1988; German version: Schwenkmezger et al., 1992). The inventory
contains three eight-item scales capturing individual differences in
anger-in (e.g., ‘‘I could explode, but I do not let anybody notice’’),
anger-out (‘‘I lose my composure’’), and anger control (‘‘I control my
anger’’). Items are answered on a 4-point scale from almost never
Self-Pity 193

(1) to almost always (4). Scores are computed by summing
across items. Alphas ranged from .84 to .89, indicating high
reliability for all three scales. In line with previous studies
(Schwenkmezger et al., 1992; Spielberger, 1988), anger-out and
anger control were highly correlated (r5.63), while both were
unrelated to anger-in.
Loneliness. Loneliness was measured with a German short form of
the UCLA Loneliness Scale (Russell, Cutrona, Rose, & Yurko,
1984; Russell, Peplau, & Cutrona, 1980). This was developed for
inclusion in the Berlin Aging Study (see, e.g., Baltes & Smith, 1997)
2
and comprises two scales of four items each that were selected from
the UCLA Loneliness Scale to assess the two facets of loneliness
differentiated by Weiss (1973), namely emotional loneliness
(four items dealing with feelings of isolation, of being alone,
and of being secluded from contact with others; e.g., ‘‘I feel alone’’)
and social loneliness (four items asking about perceptions of
not belonging to a social group and general unavailability of
trusted others; e.g., ‘‘There are people I can openly talk to’’, reverse-
scored). Items were interspersed with NEO-FFI items. Thus, the
same 5-point answer scale used in the NEO-FFI applied, and scores
were computed by averaging across items. With alphas of .76
(emotional loneliness) and .62 (social loneliness), both scales
displayed alphas that are acceptable for research purposes
(Carmines & Zeller, 1979). In line with previous findings (e.g.,
Russell et al., 1984), emotional and social loneliness were highly
correlated (r5.51).
Preliminary Analyses
Because of the large number of statistical tests and the exploratory
nature of some analyses, an error level of .01 was adopted for all
significance tests in order to take potential inflation of type-1 error
into account. Preliminary analyses using LISREL ( Jo
¨reskog &
So
¨rbom, 1999) indicated that the correlation matrices of male and
2. For further information and evidence for the differential predictive validity of
the two scales, see, for example, Smith and Baltes (1997) or Maier and Smith
(1999).
194 Sto
¨ber

female participants did not differ significantly. Therefore, data were
collapsed across gender. Appendix B shows the zero-order correla-
tions among all measures.
RESULTS
Gender Differences and Zero-Order Correlations
First, gender differences were inspected. A one-way MANOVA,
with gender (female, male) as between-participants factor, was
performed on all measures used in Study 1. With the use of Pillai’s
criterion, results indicated a significant overall main effect of gender,
F(14, 126) 55.06, po.001. To investigate which variables showed
significant gender differences, follow-up univariate ANOVAs were
conducted. In line with the other analyses, a significance level of
po.01 was adopted to adjust for inflation of type-1 error. Results
are displayed in Table 1. As expected, female participants showed
higher levels of self-pity than male participants. Moreover, female
participants showed higher levels of extraversion and agreeableness
and lower levels of social loneliness than male participants. With
respect to control beliefs and anger expression, gender differences
were not significant.
Next, the zero-order correlations of self-pity with the variables
under investigation were inspected (see Table 2). Replicating
previous findings, self-pity showed a high correlation with neuroti-
cism while being unrelated to extraversion. In addition, the present
study indicated that self-pity also was unrelated to the traits of
openness, agreeableness, and conscientiousness. When inspecting
the relationship between self-pity and the control beliefs from
Levenson’s tripartite model, results showed that self-pity was
positively correlated with externality beliefs related to both powerful
others and chance, thus corroborating previous findings that self-
pity is associated with an external locus of control. Moreover, there
was a smaller negative correlation with internality beliefs. With
respect to styles of anger expression, the present results showed that
self-pity was related to all three expression scales of the Spielberger
inventory. Replicating previous findings, self-pity again showed
positive correlations with both anger-in and anger-out, as well as a
negative correlation with anger control. Finally, results confirmed
expectations from the literature concerning a link between self-pity
Self-Pity 195

and loneliness. However, self-pity was related only to emotional
loneliness, not to social loneliness.
Regression Analyses
To investigate which of the above relationships would still hold
when overlap between subscales and shared variance with gender
were controlled for, hierarchical regression analyses were computed
for the present sets of variables. In each regression analysis, self-pity
was the dependent variable. Gender was always entered in Step 1
with an R
2
of .08, po.001. All variables from a set were then entered
Table 1
Study 1: Gender Differences
Females Males
Measure M(SD)M(SD)F(1, 139)
Self-pity 11.25 (5.30) 8.24 (4.68) 12.64
nn
Big Five personality traits
Neuroticism 1.84 (0.72) 1.56 (0.75) 5.18
Extraversion 2.69 (0.63) 2.39 (0.60) 8.00
n
Openness 2.86 (0.61) 2.85 (0.49) 0.02
Agreeableness 2.69 (0.52) 2.30 (0.51) 19.43
nn
Conscientiousness 2.57 (0.60) 2.42 (0.60) 2.12
Control beliefs
Internality 31.42 (4.59) 32.48 (5.54) 1.56
Powerful others 25.43 (5.27) 25.82 (5.22) 0.19
Chance 25.38 (6.58) 23.36 (5.27) 3.97
Anger expression
Anger-in 16.97 (4.93) 16.83 (3.84) 0.03
Anger-out 14.85 (5.34) 12.98 (4.00) 5.40
Anger control 20.57 (4.58) 22.23 (4.60) 4.56
Loneliness
Emotional loneliness 0.90 (0.67) 1.10 (0.80) 2.69
Social loneliness 0.41 (0.39) 0.76 (0.52) 19.90
nn
Note. N 5141 (75 females, 66 males). Means of Big Five personality traits and
loneliness measures could range from 0 to 4 (see text for details).
n
po.01.
nn
po.001. Two-tailed tests.
196 Sto
¨ber

simultaneously in Step 2.
3
To investigate the unique contribution of
each variable in the prediction of self-pity, semipartial correlations
were computed. Semipartial correlations, when squared, show the
proportion of variance predicted by each variable while controlling
for the other variables in the regression (Tabachnick & Fidell, 1996).
In all regression analyses, the set of variables entered contributed
significantly to the prediction of self-pity beyond gender (see
Table 2
Study 1: Relations of Self-Pity to Big Five Personality Traits, Control
Beliefs, Anger Expression, and Loneliness
Analysis rsrDR
2
.Gender R
2
Big Five personality traits .31
nn
.40
nn
Gender .29
nn
.12
Neuroticism .59
nn
.54
nn
Extraversion .06 .11
Openness .07 .10
Agreeableness .08 .00
Conscientiousness .08 .04
Control beliefs .23
nn
.32
nn
Gender .29
nn
.24
nn
Internality .22
n
.01
Powerful others .42
nn
.21
n
Chance .48
nn
.22
n
Anger expression .18
nn
.26
nn
Gender .29
nn
.22
n
Anger-in .32
nn
.33
nn
Anger-out .32
nn
.19
n
Anger control .22
nn
.05
Loneliness .12
nn
.20
nn
Gender .29
nn
.32
nn
Emotional loneliness .30
nn
.29
nn
Social loneliness .07 .01
Note. N 5141 (75 females, 66 males). Gender was coded as 1 5female, 0 5male.
r5zero-order correlation; sr 5semipartial correlation; DR
2
Gender 5increase in R
2
with gender controlled for; R
2
5total R
2
.
n
po.01.
nn
po.001. Two-tailed tests.
3. When variables were entered stepwise, all results remained essentially the same.
Self-Pity 197

Table 2). For the five-factor personality traits, the regression
analyses did not qualify the results obtained from the zero-order
correlations, as the resulting pattern of semipartial correlations
closely mirrored that of the zero-order correlations. The same held
for loneliness. For control beliefs and anger expression, however, the
results from the regression analyses showed a different pattern. With
respect to control beliefs, internality failed to make a significant
contribution to the prediction of self-pity after controlling for
overlap with external beliefs. Moreover, the semi-partial correlations
of the two externality facets were of about equal size, thus indicating
that externality beliefs related to powerful others and chance may
contribute equally to self-pity reactions to stress. With respect to
anger expression, anger control failed to make a significant
contribution to the prediction of self-pity after controlling for
overlap with the other expression scales. Moreover, the semi-partial
correlations showed that the contribution of anger-in to the
prediction of self-pity was greater than that of anger-out.
DISCUSSION
To summarize, the results of Study 1 replicated, extended, and
qualified previous empirical findings on self-pity and provided some
support for assumptions derived from clinical observations. More-
over, the findings indicated that, when investigating potential
predictors of self-pity, it is worth including sets of variables that
allow for within-construct differentiation by looking at unique
contributions within these sets of variables. With respect to the five-
factor model of personality, Study 1 corroborated previous findings
that self-pity is closely related to neuroticism, but unrelated to
extraversion ( Janke et al., 1985). Furthermore, the near-zero
correlations of the present results with openness, agreeableness,
and conscientiousness indicate that self-pity also may be largely
unrelated to the other superfactors of the five-factor model of
personality. With respect to control beliefs, the present results
replicated and extended Janke and colleagues’ (1985) findings on
self-pity and locus of control as measured with Rotter’s I/E scale.
Using scales derived from Levenson’s (1974) tripartite model of
locus of control, and controlling conceptual and statistical overlap,
the present results indicated that self-pity is unrelated to internality,
198 Sto
¨ber

but, instead, strongly associated with externality beliefs related both
to powerful others and chance. With respect to anger expression, the
present findings corroborated the previous findings of Schwenk-
mezger and colleagues (1992), in that self-pity was associated with
high levels of both anger-in and anger-out as well as with low levels
of anger control. However, when additional regression analyses were
computed with a simultaneous investigation of the three anger
expression scales and gender, only anger-in and, to a lesser degree,
anger-out still displayed significant regression weights. Finally, the
present findings confirm observations reported in the clinical
literature that self-pity is related to loneliness. However, as the
two-dimensional conceptualization following Weiss’s (1973) typol-
ogy of loneliness showed, self-pity was related only to emotional
loneliness and not to social loneliness.
Some questions remain, however. First, the links between self-pity
and neuroticism require further investigation. Whereas the broad
dimensions of the five-factor model of personality may provide a
general orientation as to where to locate a construct, further
analyses on a subordinate level are needed. Costa and McCrae
(1992, 1995a) have developed a measure of the Big Five traits that is
comprised of facet scales. For neuroticism, their conceptualization
defines six facets: anxiety, angry hostility, depression, self-con-
sciousness, impulsivity, and vulnerability. While other models of
personality favor different facets of neuroticism (e.g., H. J. Eysenck,
1995), only the conception presented by Costa and McCrae contains
depression and anxiety on the facet level. Thus, it allows for a
simultaneous inspection of these two facets in order to investigate
the question of whether self-pity is related to both depression and
trait anxiety and, if so, to which facets unique associations exist.
Second, the links between self-pity and anger may require further
exploration. One reason is that the anger expression scales of the
Spielberger inventory refer to anger in general terms only. In
particular, they do not specify the source of anger. Research on
anger has demonstrated that anger-provoking experiences may be
classified into three different categories: personal frustrations,
interpersonal frustrations, and interpersonal exploitation (Snell,
McDonald, & Koch, 1991). In the first category, anger relates to
personal frustrations such as personal inadequacies and failures
related to unattained pursuits and goals. In the other two categories,
anger relates to interpersonal frustrations such as frustrating events
Self-Pity 199

associated with public/social aspects of the self on the one hand and
incidents associated with interpersonal exploitation and injustices on
the other. Because self-pity is related to beliefs that powerful others
are in control of one’s life, it could well be that, compared to
participants low in self-pity, participants high in self-pity have different
anger situations in mind (viz., interpersonal rather than personal
frustrations) when responding to the expression scales of the
Spielberger inventory. Moreover, anger-in and anger-out both
represent somewhat dysfunctional styles of reactions to anger. Recent
analysis of anger reactions have provided evidence that is useful to
expand the perspective on potential reactions to anger by including
functional reactions (Tangney, Hill-Barlow, Wagner, Marschall,
Borenstein, et al., 1996; Weber, Eue, Titzmann, & Freese, 1999;
Weber & Titzmann, 2001). Consequently, a further exploration of self-
pity and anger may profit from (a) being explicit about the source of
the anger-provoking experiences and (b) inspecting both functional
and dysfunctional ways of reacting to these experiences.
Third, it may be worthwhile to follow up on the finding that self-
pity is related to emotional loneliness, but not to social loneliness.
Research on the differences between social and emotional loneliness
has revealed that social loneliness is primarily related to lack of
social provisions such as social integration or reassurance of worth.
In contrast, emotional loneliness is primarily related to attachment
problems (for a review, see DiTommaso & Spinner, 1997).
Following studies with children, research on attachment has
differentiated three main styles of adult attachment—secure,
avoidant, and ambivalent—that are formed in infancy, but carry
over to close relationships across the life span (Ainsworth, 1989;
Ainsworth, Blehar, Waters, & Wall, 1978). Individuals with secure
attachments have experienced early relationships with a nurturing
adult who was sensitive to their signals of distress and available in
times of need, thus creating a basic trust in the world and the self. In
contrast, individuals with avoidant and anxious-ambivalent attach-
ments lack this experience of a ‘‘secure base.’’ Whereas secure
attachment can be conceptualized as an inner resource that may help
individuals cope with stress successfully, avoidant and ambivalent
attachment styles are considered potential risk factors for maladap-
tive coping (Mikulincer & Florian, 1997). Consequently, it may be
expected that that self-pity will be related to problematic attach-
ments as expressed in avoidant or ambivalent attachment qualities.
200 Sto
¨ber

The aims of Study 2 were thus threefold. The first objective was to
investigate which facets of neuroticism best predict reactions of self-
pity, with a particular focus being placed on depression and trait
anxiety. The second goal was to explore further the relationships
between self-pity and anger by investigating the relationship between
self-pity and both functional and dysfunctional anger reactions, with
the source of anger being restricted to interpersonal situations. The
third aim was to examine the relationship between self-pity and
adult attachment styles. As it can be assumed that self-pity is related
to one or more problematic attachment styles, Study 2 should
provide for a particularly differentiated assessment of dysfunctional
attachment qualities.
STUDY 2
METHOD
Participants
A sample of N5161 students (88 females, 73 males) was recruited at
the Martin Luther University of Halle-Wittenberg. Average age was
21.9 years (SD 52.1). All participants volunteered in exchange for a
lottery ticket for a chance to win 100 German marks (approximately
47 U.S. dollars).
Measures
Self-pity. As in Study 1, self-pity was measured with the 6-item self-
pity scale of the Streverarbeitungsfragebogen ( Janke et al., 1985).
In the present sample, Cronbach’s alpha was .78, again indicating
satisfactory reliability.
Facets of neuroticism. Facets of neuroticism were measured with the
respective scales of the revised NEO Personality Inventory (NEO-
PI-R; Costa & McCrae, 1992; German version: Ostendorf &
Angleitner, 1993). For neuroticism, the NEO-PI-R comprises six
eight-item scales that capture individual differences in anxiety (e.g., ‘‘I
am frequently concerned that things may go wrong’’), angry hostility
(‘‘People think of me as an irascible, quick-tempered person’’),
depression (‘‘Sometimes everything appears rather dark and hopeless
to me’’), self-consciousness (‘‘When among other people, I am afraid
that I may make a bad impression’’), impulsivity (‘‘I have difficulties
Self-Pity 201

in resisting my cravings’’), and vulnerability (‘‘I often feel helpless,
wishing there was someone who would solve my problems for me’’).
Items are answered on a 5-point scale from strongly disagree (1) to
strongly agree (5). Scores are computed by summing across items. With
alphas ranging from 61. to .78, all scales displayed reliabilities
acceptable for research purposes (Carmines & Zeller, 1979). As
expected, the scales showed substantial intercorrelations (mean
r5.44) with individual correlations ranging from r5.04 (between
self-consciousness and impulsivity) to r5.71 (between depression and
vulnerability). When aggregating all items to a total score for
neuroticism, the resulting score displayed an alpha of .87.
Anger reactions. Anger reactions were measured with the reactions
scales of the Anger-related Reactions and Goals Inventory (Weber
et al., 1999; Weber & Titzmann, 2001). The questionnaire contains
six four-item scales that capture individual differences in functional
and dysfunctional reactions to anger in interpersonal situations (i.e.,
situations in which another person is the source of one’s anger).
Three scales capture anger reactions that are considered
dysfunctional, namely outburst (e.g., ‘‘I explode’’), rumination
(‘‘Instead of forgetting the whole thing, I keep thinking about it’’),
and submission (‘‘I give in to avoid arguing’’); and three scales
capture anger reactions that are considered functional, namely
feedback (‘‘I tell the other person what annoys me, but without
becoming aggressive’’), noninvolvement (‘‘I try not to get angry in the
first place’’), and humor (‘‘I find the whole thing funny’’). Items are
answered on a 4-point scale from almost never (1) to almost always
(4). Scores are computed by summing across items. With alphas
ranging from .70 to .89, all scales displayed satisfactory reliability. In
line with previous findings (e.g., Weber et al., 1999), the scales
showed only moderate intercorrelations (mean |r|5.24), with
individual correlations ranging from r5–.40 (between rumination
and noninvolvement) to r5.57 (between noninvolvement and humor).
Attachment styles. Attachment styles were assessed with the
Measure of Attachment Qualities (MAQ; Carver, 1997; own
version, employing translations and backtranslations by myself
and other colleagues versed in both German and English). Whereas
most measures of adult attachment follow the classical tripartite
conceptualization of attachment (i.e., secure, avoidant, and
202 Sto
¨ber

ambivalent attachment), Carver’s measure employs a four-tiered
approach by further differentiating ambivalent attachment into two
separate factors, namely ambivalence-worry and ambivalence-
merger, thus allowing for a more differentiated analysis of
dysfunctional attachments styles. The MAQ thus comprises four
scales capturing individual differences in security (e.g., ‘‘It feels
relaxing and good to be close to someone’’), avoidance (‘‘I prefer not
to be too close to others’’), ambivalence-worry (‘‘I often worry that
my partner doesn’t really love me’’), and ambivalence-merger (‘‘I
have trouble getting others to be as close as I want them to be’’). The
scales for security, ambivalence-worry, and ambivalence-merger
each comprise three items; the scale for avoidance comprises five
items. Items are answered on a 4-point scale from I disagree with the
statement a lot (1) to I agree with the statement a lot (4). Scores are
computed by averaging across items. With alphas ranging from .69
to .74, all scales displayed satisfactory reliability. In line with
previous findings (Carver, 1997), only security and avoidance (r5–
.55) and ambivalence-worry and ambivalence-merger (r5.33) were
significantly correlated.
Preliminary Analyses
For the same reasons as in Study 1, a significance level of .01 was
adopted for all analyses. Again, analyses using LISREL ( Jo
¨reskog
&So
¨rbom, 1999) indicated that the correlation matrices of male and
female participants did not differ significantly, so data were
collapsed across gender. Appendix C shows the zero-order correla-
tions among all measures.
RESULTS
Gender Differences and Zero-Order Correlations
First, gender differences were inspected. As in Study 1, a one-way
MANOVA, with gender (female, male) as between-participants
factor, was performed on all measures simultaneously. With the use
of Pillai’s criterion, results indicated a significant overall main effect
of gender, F(17, 143) 53.74, po.001. To investigate which variables
showed significant gender differences, follow-up univariate ANO-
VAs were conducted with the significance level again set to po.01.
Results are displayed in Table 3. In line with the previous findings,
female participants again showed significantly higher levels of
Self-Pity 203

self-pity than male participants. Moreover, with respect to the facets
of neuroticism, they showed higher levels of anxiety, depression,
impulsivity, and vulnerability; with respect to anger reactions, they
showed lower levels of noninvolvement and humor; and, with
respect to attachment styles, higher levels of security.
Next, the zero-order correlations of self-pity were inspected (see
Table 4). As in Study 1, self-pity again showed a high correlation
with the personality trait of neuroticism with the correlation
between self-pity and neuroticism total score being r5.52,
po.001. When inspecting the correlations on the facet level of
Table 3
Study 2: Gender Differences
Females Males
Measure M(SD)M(SD)F(1, 159)
Self-pity 10.90 (4.81) 7.40 (3.92) 24.85
nn
Facets of neuroticism
Anxiety 24.32 (4.89) 21.09 (4.80) 17.62
nn
Angry hostility 22.41 (4.69) 20.84 (4.35) 4.79
Depression 21.42 (4.40) 19.22 (4.95) 8.93
n
Self-consciousness 23.74 (4.09) 22.84 (4.37) 1.83
Impulsivity 24.42 (4.24) 22.49 (4.30) 8.20
n
Vulnerability 20.64 (4.11) 18.70 (4.77) 7.67
n
Anger reactions
Outburst 7.70 (2.53) 7.29 (2.31) 1.17
Rumination 9.43 (2.72) 8.37 (3.20) 5.12
Submission 6.48 (2.63) 6.86 (2.22) 0.99
Feedback 10.22 (2.33) 10.02 (1.94) 0.32
Noninvolvement 7.37 (2.41) 8.45 (2.45) 8.00
n
Humor 6.92 (2.36) 8.30 (2.91) 11.06
n
Attachment styles
Security 3.59 (0.44) 3.20 (0.63) 21.39
nn
Avoidance 1.94 (0.49) 2.04 (0.53) 1.78
Ambivalence-worry 2.03 (0.71) 1.97 (0.63) 0.34
Ambivalence-merger 1.71 (0.59) 1.81 (0.49) 1.36
Note. N 5161 (88 females, 73 males). Means of attachment styles measures could
range from 1 to 4 (see text for details).
n
po.01.
nn
po.001. Two-tailed tests.
204 Sto
¨ber

neuroticism, a more differentiated picture emerged. In line with
previous findings, self-pity showed high correlations with depres-
sion. Moreover, as suggested by the findings on self-pity and
sensitization, there also was a high correlation with trait anxiety. In
addition, self-pity was substantially related to the neuroticism facets
of vulnerability, self-consciousness, and (albeit to a lesser degree)
angry hostility. In contrast, the correlation with impulsivity failed to
reach significance. When functional and dysfunctional anger
reactions in interpersonal situations were inspected, self-pity showed
Table 4
Study 2: Relations of Self-Pity to Facets of Neuroticism, Anger
Reactions, and Attachment Styles
Analysis rsrDR2.Gender R
2
Facets of neuroticism .24
nn
.37
nn
Gender .37
nn
.21
n
Anxiety .49
nn
.08
Angry hostility .26
n
–.05
Depression .53
nn
.20
n
Self-consciousness .38
nn
.04
Impulsivity .16 .03
Vulnerability .47
nn
.06
Anger reactions .20
nn
.34
nn
Gender .37
nn
.30
nn
Outburst .11 .02
Rumination .45
nn
.33
nn
Submission .17 .11
Feedback –.15 –.12
Noninvolvement –.21
n
–.02
Humor –.22
n
.02
Attachment styles .10
n
.23
nn
Gender .37
nn
.28
nn
Security .24
n
.16
Avoidance .01 .11
Ambivalence-worry .28
nn
.23
n
Ambivalence-merger .06 .03
Note. N 5161 (88 females, 73 males). Gender was coded as 1 5female, 0 5male.
r5zero-order correlation; sr 5semipartial correlation; DR
2
.Gender 5increase in
R
2
with gender controlled for; R
2
5total R
2
.
n
po.01.
nn
po.001. Two-tailed tests.
Self-Pity 205

a substantial positive correlation with rumination and moderate
negative correlations with noninvolvement and humor. Thus,
individuals with a inclination for self-pity react to anger-provoking
social situations with a pattern of heightened rumination, low
noninvolvement, and low humor. Finally, when inspecting the
correlations with attachment qualities, self-pity was related to higher
levels of ambivalence-worry. Moreover, and somewhat unexpect-
edly, self-pity showed a significant positive correlation with security.
Regression Analyses
To investigate which of the above relationships would still hold
when overlap between subscales and shared variance with gender
were controlled for, hierarchical regression analyses were computed
for each set of variables, following the method outlined in Study 1.
In each regression analysis, self-pity was the dependent variable.
Gender was always entered in step 1, this time with an R
2
of .14,
po.001. All variables from one set of measures were then entered
simultaneously in Step 2.
4
Even though the proportion of variance explained by gender was
considerably larger than in Study 1, each set of variables again
contributed significantly to the prediction of self-pity beyond gender
(see Table 4). Moreover, all regression analyses produced a pattern
of semi-partial correlations that differed markedly from that of the
respective zero-order correlations. With respect to neuroticism,
results showed that depression was the only facet that still made a
unique contribution to the prediction of self-pity after gender and
overlap between facet scales were controlled for. All other facets,
including anxiety, showed semi-partial correlations near zero. Thus,
it appears that only depression predicts differences in self-pity
beyond the common variance shared by all neuroticism facets. With
respect to functional and dysfunctional anger reactions, noninvolve-
ment and humor failed to make a significant contribution to the
prediction of self-pity after controlling for shared variance. Only
rumination still displayed a substantial semi-partial correlation,
indicating that rumination may contribute significantly to differ-
ences in self-pity beyond shared variance with other functional and
dysfunctional anger reactions and gender. With respect to attachment
4. When variables were entered stepwise, all results again remained essentially the
same.
206 Sto
¨ber

qualities, results showed that only ambivalence-worry still con-
tributed significantly to the prediction of self-pity after controlling
for gender and overlap among attachment scales. Thus, the
significant positive correlation between self-pity and security may
possibly be attributed to variance shared with gender, as both
variables showed substantial gender differences (see Table 3). This
was corroborated in a follow-up analysis with gender and security
entered stepwise to predict self-pity. Results showed that after
gender entered the regression (R
2
5.14, sr 5.37, po.001), security
failed to contribute further to the prediction of self-pity (DR
2
5.01,
sr 5.12, ns). In sum, from each set of variables, only one variable
made a unique contribution to the prediction of self-pity beyond
variance shared with gender and other variables, namely depression
(from facets of neuroticism), rumination (from anger reactions), and
ambivalence-worry (from attachment styles).
Finally, to investigate whether these three variables held
independent relationships to self-pity, a multiple regression analysis
was computed in which gender was entered in step 1, followed by
depression, rumination, and ambivalence-worry in step 2. Results
showed that the three variables combined were associated with an
increase in R
2
(DR
2
.Gender) of .26 resulting in a total R
2
of .40, both
pso.001. Inspecting the individual contributions, only depression
and anger rumination still displayed significant semipartial correla-
tions (sr 5.29, po.001 and sr 5.20, po.01, respectively), whereas
worry-ambivalent attachment did not (sr 5.08, ns). Thus, worry-
ambivalent attachment failed to make a unique contribution to the
prediction of self-pity after controlling for gender, depression, and
anger rumination.
DISCUSSION
Though discussion of the main findings of Study 2 will be left to the
general discussion, there are two correlations that warrant attention
at this point. The first concerns one of the facets of neuroticism,
namely the nonsignificant correlation of self-pity with impulsivity.
This correlation is of particular note because impulsivity was the
only facet of neuroticism not associated with self-pity. A potential
reason for this may be that impulsiveness is not as clearly related to
the superfactor of neuroticism as the other facets are. In a factor
Self-Pity 207

analysis with the facet scales of the five-factor model (Costa &
McCrae, 1991), impulsivity not only displayed the lowest loadings of
all neuroticism facets on the factor that represented neuroticism; it
also showed substantial positive loadings on the factor that
represented extraversion and substantial negative loadings on the
factor that represented conscientiousness. With impulsivity appar-
ently being some joint function of neuroticism, extraversion, and
low conscientiousness on the one hand, and with self-pity being
unrelated to extraversion and conscientiousness on the other (see
Study 1), the nonsignificant correlation between self-pity and
impulsivity is perhaps not too surprising.
The second correlation concerns the nonsignificant relationship of
self-pity with anger outburst. This correlation is of particular note in
relation to Study 1, where self-pity was found to be substantially
correlated with anger-out. Previous findings have indicated that the
anger outburst scale of the Anger-related Reactions and Goals
Inventory is closely related to the anger-out scale of the Spielberger
measure (Weber et al., 1999; Weber & Titzmann, 2001). However, a
close inspection of the item content of the respective scales suggests
that some of the reactions subsumed in the anger-out scale are less
direct, less aggressive, and less extreme than those described in the
outburst scale. Moreover, anger-out is unspecific as to the source of
anger, whereas the anger reactions of the Weber measure explicitly
refer to interpersonal situations. As such, the finding that self-pity is
related only to anger-out in general (Study 1), but not to angersome
outburst in social situations (Study 2), is well in line with notions
found in the literature that self-pity is related to more subtle
expressions of anger.
GENERAL DISCUSSION
Two studies were conducted to explore the links of self-pity to
personality, control beliefs, and anger, employing multidimensional
measures of personality, control beliefs, anger, loneliness, and adult
attachment. The results can be summarized as follows:
With respect to personality, results showed that self-pity was
strongly associated with neuroticism, particularly the depression
facet, but unrelated to the other dimensions of the five-factor
model of personality (i.e., extraversion, openness, agreeableness, and
208 Sto
¨ber

conscientiousness). With respect to control beliefs, self-pity showed
equally strong associations with externality beliefs related to
powerful others and externality beliefs related to chance, indicating
that individuals with a tendency for self-pity seem to have
generalized beliefs that their life is controlled by external forces.
Analyses of anger showed that self-pity was primarily related to
nonexpression of anger. This was demonstrated when analyzing
anger expression styles in general and when analyzing anger
reactions in interpersonal situations in particular. With respect to
the former, self-pity had a strong unique relationship with anger-in;
with respect to the latter, it showed a strong unique relationship with
anger rumination. Concerning loneliness, analyses indicated that
self-pity showed a substantial correlation with emotional loneliness,
but none with social loneliness. Because emotional loneliness is
closely related to attachment problems, further analyses were
conducted to explore the relationship between self-pity and adult
attachment. Results showed that self-pity was related to specific
dysfunctional attachment qualities, such that individuals with a
tendency to feel sorry for themselves indicated higher levels of
ambivalence-worry in their interpersonal relationships. Finally, in
both studies, a strong correlation with gender was found, with
women reporting more self-pity reactions to stress than men.
Integration of Findings
The pervasive gender effect may indicate that self-pity is a stress
response that is more prevalent in women than in men. Moreover,
this gender effect does not seem to be restricted to adulthood, but is
already apparent in early adolescence, as demonstrated in a Maltese
survey on bullying in schools (Borg, 1998). In this nationwide
survey, students aged 9 to 14 years were asked how they responded
emotionally to being bullied by other students. Of all emotional
reactions under study, self-pity displayed the largest gender
difference. Whereas only 28% of boy victims reported feelings of
self-pity after being bullied, 46% of girls victims reported such
feelings. Thus, gender differences in coping by means of self-pity
seem to have an early adolescent onset, comparable to that observed
for ruminative coping. Research on gender differences in depression
has found that women more often rely on ruminative coping when
facing loss and failure than men do (Nolen-Hoeksema, 1991, 1995).
Self-Pity 209

This gender difference starts to emerge in early adolescence, when
boys tend to choose more active, aggressive coping strategies for
dealing with adversities, whereas girls tend to choose more passive,
emotional coping strategies and ruminate more about these
adversities. It has been suggested (Broderick, 1998; Nolen-Hoekse-
ma, 1991) that sex-role socialization may play an important role in
differentially shaping children’s repertoire for handling stress, loss,
and negative affect. Parents’ expectations of which emotional
reactions are appropriate for boys and which for girls will lead to
encouragement of some reactions, and sanctions for others. Thus,
parents may shape masculine and feminine ways of reacting to stress
and encourage gender-related differences in the expression and
suppression of certain emotional reactions. Self-pity, like rumina-
tion, seems to be a typical feminine response to stress in this regard.
The present findings provide empirical support for some of the
notions about self-pity that have been put forward in the psychiatric
and psychoanalytic literature on the basis of informal observations
and unsystematic case studies. Moreover, they replicate, extend, and
qualify previous empirical findings on self-pity. First, the analyses of
relations with social and emotional loneliness corroborated anec-
dotal findings that self-pity is related to loneliness (Kahn, 1965).
Moreover, the finding that self-pity is related only to emotional
loneliness lends support to the view that individuals who experience
self-pity, even if they are not actually socially isolated, may
nevertheless feel emotionally isolated (Charmaz, 1980). Emotional
loneliness has been related to dysfunctional attachments. Analysis of
the relationships between attachment qualities and self-pity indeed
showed self-pity to have the expected relationship with dysfunc-
tional attachments: Individuals with a disposition towards self-pity
indicated ambivalent-worrisome attachment, suggesting that these
individuals show increased levels of fear of not being loved and
worry about being abandoned. In particular, people with anxious-
ambivalent attachments have been shown to deal with stress in a
hypervigilant manner. Unable to suppress negative emotions, they
cannot detach from inner pain. Instead, they direct attention toward
their distress and toward mentally ruminating on negative thoughts,
memories, and affects (Mikulincer & Florian, 1997).
With respect to control beliefs, the present findings show that the
two externality dimensions defined by Levenson (1974)—chance and
powerful others—play an equally strong role in predicting self-pity
210 Sto
¨ber

reactions to stress. According to Levenson (1974, 1981), externality
beliefs related to powerful others may motivate individuals to initiate
actions with the goal of regaining power and control over one’s life. In
contrast, externality beliefs related to chance are likely to lead to
amotivation and inactivity. Consequently, with the two facets of
external control beliefs being equally strong predictors of self-pity,
individuals with a propensity to feel sorry for themselves may find
themselves in a state of emotional inertia and fatigue caused by a
deadlock between competing action tendencies: to confront others in
order to change their fate (activism) or to give up (passivity).
However, it seems rather unlikely that individuals with a leaning
towards self-pity will confront others directly (Kahn, 1965). This
also was reflected in the findings on self-pity and anger. With respect
to styles of anger expression, regression analyses indicated that self-
pity was primarily correlated with anger-in, and only to a lesser
degree with anger-out. Moreover, follow-up analyses on the links
between self-pity and anger reactions in interpersonal situations
showed a strong and unique association of self-pity with ruminative
responses to anger. This pattern of findings (viz. large correlations
with anger-in and anger rumination and a smaller correlation with
anger-out) dovetails nicely with the clinical observations on self-pity
that hold self-pity to be closely related to feelings of anger and
hostility. Most of the time, direct expression of these feelings will be
avoided (Milrod, 1972). Instead of venting their anger, individuals
who experience self-pity will keep a lid on their angry feelings and
keep their anger in, while at the same time ruminating about
potential injustices suffered and fantasizing about possible retribu-
tions (Charmaz, 1980).
Moreover, corroborating the views of Kahn (1965) and replicat-
ing the findings of Janke and colleagues (1985), the present findings
demonstrate that self-pity is a stress response that is closely related
to individual differences in the trait of neuroticism. Research on the
relationship between the five-factor personality traits and coping has
found neuroticism to be the strongest and most persistent predictor
of dysfunctional coping mechanisms. In the face of stress,
individuals with high neuroticism scores tend to refrain from active,
problem-focused coping strategies. Instead, they rely on emotion-
focused and ineffective forms of coping, such as wishful thinking,
escapist fantasies, denial, self-blame, avoidance, passivity, and
withdrawal (for reviews, see O’Brien & DeLongis, 1996; Watson &
Self-Pity 211

Hubbard, 1996). Inasmuch as the present findings show that
neuroticism is highly predictive of self-pity responses to stress, they
add a further facet to this picture by pointing out that self-pity is
another dysfunctional response to stress that is prototypical for
individuals high in neuroticism.
Finally, in line with the clinical literature and previous findings,
the present findings show that self-pity is closely related to
depression, even when common variance with gender and other
facets of neuroticism are controlled for. In comparison, the
significant relationship with trait anxiety was reduced to values
around zero after controlling for common variance with the other
facets of neuroticism. In the same vein, the relationship between self-
pity and worry-ambivalence in attachment was reduced to
nonsignificance after controlling for common variance with depres-
sion and angersome rumination. As worry is the cognitive
component of anxiety (M. W. Eysenck, 1992), both findings
may indicate that self-pity is a stress response that is specifi-
cally related to depression, rather than to anxiety. Thus, self-pity
may be an important characteristic that researchers in abnormal
psychology need to attend to when looking for specifics that
differentiate depression from anxiety (e.g., Beck, Brown, Steer,
Eidelson, & Riskind, 1987; Nitschke, Heller, Imig, McDonald, &
Miller, 2001).
Limitations and Future Directions
The limitations of the present studies mainly pertain to three points.
First, some of the analyses were exploratory. This applies in
particular to the regression analyses. Consequently, their results may
be considered preliminary. Moreover, the self-pity scale of the
Streverarbeitungsfragebogen ( Janke et al., 1985) only captures
relatively mild and fleeting forms of self-pity, and both of the present
studies were conducted with university students. Therefore, it
remains an open question as to whether the chronic and pervasive
forms of self-pity described in the psychoanalytic and psychiatric
literature will show the same pattern of relationships as self-pity
responses to stress observed in normal student samples. Never-
theless, the present findings are well in line with theoretical
reflections, clinical observations, and empirical findings from the
previous literature. One can thus be fairly confident that the overall
212 Sto
¨ber

picture presented here will hold in further replications. Second, it
remains unclear whether the pervasive gender differences in the
present studies represent differences in the experience of self-pity, or
merely differences in reporting self-pity. Charmaz (1980), for
example, holds that reporting self-pity seemed to be easier for
women than for men. Consequently, the present gender differences
could possibly be caused by men underreporting the frequency of
self-pity. Third, the findings are cross-sectional. Consequently, they
do not allow for any inferences about causal relationships among
the variables under study. Still, in view of the hierarchical status of
the present variables, it seems reasonable to assume that those
variables representing global personality traits or stable trait-like
characteristics (neuroticism, personal beliefs, and attachment styles)
are more likely to be responsible for differences in situation-specific
reactions (viz. self-pity reactions to stress) than vice versa. The po-
tential causal relationships between self-pity and the other variables
under study (anger expression, anger reactions, and loneliness),
however, remain uncertain, as it is perfectly possible that they re-
present concurrent processes to self-pity or consequences of self-pity.
Future research on self-pity may therefore profit from replicating
the present findings, additionally employing longitudinal designs in
order to investigate the direction of the potential influences among
variables and to establish further evidence on the causes for
individual differences in self-pity. In addition to identifying the
characteristics that may dispose individuals to feeling sorry for
themselves, future research may also benefit from closer investiga-
tion of the effects of self-pity. A future goal here may be to arrive at
a functional analysis of self-pity responses to stress in a similar way
as has been achieved for ruminative coping with depression. It may
be valuable to employ studies with experimental manipulations of
self-pity and to extend the investigation of self-pity beyond student
samples to more chronic and pervasive cases of self-pity as reported
in the clinical literature (e.g., Charmaz, 1980). Moreover, future
research may investigate how self-pity is different from, or similar
to, other emotion-focused coping responses to stress. Finally, as was
recently suggested by Funder (2000), it may be of advantage to
integrate the present trait-oriented approach with a social-cognitive
approach to personality in order to investigate how self-pity, both
through processes of self-regulation and through processes related
to the interaction of person and environment, may become a
Self-Pity 213

prominent reaction to stress in some people some of the time and in
others most of the time.
CONCLUSIONS
Self-pity has been much neglected in psychological research despite
suggestions that it may be a frequent response not only to major
failures and losses but also to minor problems and everyday hassles.
Building on informal observations and unsystematic findings, the
present studies attempted a first systematic investigation of self-pity
responses to stress, linking these to personality, personal beliefs, and
anger, with additional analyses including loneliness and attachment
styles. The strong associations with neuroticism and depression,
generalized externality beliefs, anger-in and anger rumination, as
well as emotional loneliness and ambivalent-worrisome attachments
corroborate previous contentions of self-pity as a highly ineffective,
if not self-damaging way of coping with stress (Grunert, 1991) that,
rather than resolving distressing situations, may instead amplify the
associated states of distress. Moreover, self-pity seems to contain a
strong interpersonal component, including not only feelings and
fears of loneliness, but also feelings of envy, blame, anger, and
hostility directed towards others. While the present studies provide a
first systematic pattern of findings, further research is needed to
establish a clearer picture of the precipitants, concomitants, and
consequences of self-pity. In this endeavor, gender differences and
the potential reasons for the higher prevalence of self-pity in women
may remain an important focus.
APPENDIX A
The Self-Pity Scale of the Streverarbeitungsfragebogen (SVF;
Janke et al., 1985; English translation: W. Janke, personal
communication, March 26, 2001)
When I feel upset by something or somebody, or when something has
thrown me off balancey
yI feel a little sorry for myself.
yI envy others to whom such things don’t happen.
yI have the feeling that luck is never on my side.
yI can’t understand why I am always the one who has bad luck.
yI think that bad things always seem to happen to me.
yI ask myself why this had to happen to me of all people.
214 Sto
¨ber

APPENDIX B
Study 1: Zero-Order Correlations Among Measures
Measure 1 2 3 4 5 6 7 8 9 10 11 12 13
1. Self-pity —
2. Neuroticism .59 —
3. Extraversion .06 .34 —
4. Openness .07 .00 .25 —
5. Agreeableness .08 .01 .26 .04 —
6. Conscientiousness .08 .27 .25 .10 .25 —
7. Internality .22 .35 .25 .10 .12 .13 —
8. Powerful others .42 .41 .38 .14 .12 .01 .28 —
9. Chance .48 .34 .23 .11 .02 .06 .39 .56 —
10. Anger-in .32 .41 .32 .16 .06 .13 .27 .33 .26 —
11. Anger-out .32 .20 .03 .11 .26 .16 .01 .15 .14 .05 —
12. Anger control .22 .21 .05 .05 .16 .15 .07 .02 .17 .13 .63 —
13. Emotional loneliness .30 .52 .46 .04 .11 .26 .22 .34 .21 .32 .05 .05 —
14. Social loneliness .07 .16 .52 .26 .32 .17 .05 .31 .22 .20 .03 .06 .51
Note. N 5141 (75 females, 66 males). Measures 2-6: Big Five personality traits; 7-9: control beliefs; 10-12: anger expression; 13-14:
loneliness.
Correlations of |r|X.22 are significant at po.01 and correlations of |r|X.28 at po.001. Two-tailed tests.

APPENDIX C
Study 2: Zero-Order Correlations Among Measures
Measure 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
1. Self-pity —
2. Anxiety .49 —
3. Angry hostility .26 .48 —
4. Depression .53 .63 .44 —
5. Self-consciousness .38 .64 .36 .58 —
6. Impulsivity .16 .23 .42 .14 .04 —
7. Vulnerability .47 .68 .41 .71 .55 .22 —
8. Outburst .11 .22 .55 .16 .13 .31 .24 —
9. Rumination .45 .45 .39 .46 .47 .20 .36 .18 —
10. Submission .17 .15 .07 .25 .33 .09 .25 .21 .09 —
11. Feedback .15 .22 .04 .30 .35 .01 .31 .11 .02 .32 —
12. Noninvolvement .21 .31 .43 .29 .30 .31 .28 .35 .38 .25 .10 —
13. Humor .22 .40 .33 .34 .35 .25 .33 .23 .40 .15 .08 .57 —
14. Security .24 .06 .05 .08 .01 .25 .06 .08 .18 .18 .17 .26 .17 —
15. Avoidance .01 .00 .04 .07 .21 .21 .04 .09 .06 .24 .24 .30 .17 .55 —
16. Ambivalence-worry .28 .30 .23 .33 .23 .11 .30 .01 .29 .01 .12 .26 .29 .01 .08 —
17. Ambivalence-merger .06 .10 .17 .15 .15 .04 .17 .01 .02 .17 .02 .05 .04 .09 .02 .33
Note. N 5161 (88 females, 73 males). Measures 2-7: facets of neuroticism; 8-13: anger reactions; 14-17: attachment styles. Correlations of
|r|X.20 are significant at po.01 and correlations of |r|X.26 at po.001. Two-tailed tests.

REFERENCES
Ainsworth, M. D. S. (1989). Attachments beyond infancy. American Psychologist,
44, 709–716.
Ainsworth, M. S., Blehar, M. C., Waters, E., & Wall, S. (1978). Patterns of
attachment: A psychological study of the strange situation. Potomac, MD:
Erlbaum.
Baltes, P. B., & Smith, J.(Eds.). (1997). The Berlin Aging Study [Special section].
Psychology and Aging,12, 395–472.
Beck, A. T., Brown, G., Steer, R. A., Eidelson, J. I., & Riskind, J. H. (1987).
Differentiating anxiety and depression: A test of the cognitive content-
specificity hypothesis. Journal of Abnormal Psychology,96, 179–183.
Becker, P. (1985). Bewa
¨ltigungsverhalten und seelische Gesundheit [Coping
behavior and mental health]. Zeitschrift fu
¨r Klinische Psychologie,14, 169–184.
Block, J. (1995). A contrarian view of the five-factor approach to personality
description. Psychological Bulletin,117, 187–215.
Borg, M. G. (1998). The emotional reactions of school bullies and their victims.
Educational Psychology,18, 433–444.
Borkenau, P., & Ostendorf, F. (1993). NEO-Fu
¨nf-Faktoren Inventar (NEO-FFI)
[NEO Five Factor Inventory (NEO-FFI)].Go
¨ttingen, Germany: Hogrefe.
Broderick, P. C. (1998). Early adolescent gender differences in the use of
ruminative and distracting coping strategies. Journal of Early Adolescence,18,
173–191.
Brosschot, J. F., Gebhardt, W. A., & Godaert, G. L. R. (1994). Internal, powerful
others and chance locus of control: Relationships with personality, coping,
stress and health. Personality and Individual Differences,16, 839–852.
Byrne, D. (1961). The Repression-Sensitization Scale: Rationale, reliability, and
validity. Journal of Personality,29, 334–349.
Carmines, E. G., & Zeller, R. A. (1979). Reliability and validity assessment.
Beverly Hills, CA: Sage.
Carver, C. S. (1997). Adult attachment and personality: Converging evidence and
a new measure. Personality and Social Psychology Bulletin,23, 865–883.
Charmaz, K. (1980). The social construction of self-pity in the chronically ill.
Studies in Symbolic Interaction,3, 123–145.
Costa, P. T. Jr., & McCrae, R. R. (1991). Trait psychology comes of age. In T. B.
Sonderegger (Ed.), Nebraska Symposium on Motivation (Vol. 39, Psychology
and aging, pp. 169–204). Lincoln, NE: University of Nebraska Press.
Costa, P. T. Jr., & McCrae, R. R. (1992). Revised NEO Personality Inventory
(NEO PI-R) and NEO Five Factor Inventory: Professional Manual. Odessa,
FL: Psychological Assessment Resources.
Costa, P. T. Jr., & McCrae, R. R. (1995a). Domains and facets: Hierarchical
personality assessment using the Revised NEO Personality Inventory. Journal
of Personality Assessment,64, 21–50.
Costa, P. T. Jr., & McCrae, R. R. (1995b). Solid ground in the wetlands of
personality: A reply to Block. Psychological Bulletin,117, 216–220.
Self-Pity 217

DiTommaso, E., & Spinner, B. (1997). Social and emotional loneliness: A re-
examination of Weiss’ typology of loneliness. Personality and Individual
Differences,22, 417–427.
Elson, M. (1997). Self-pity, dependence, manipulation, and exploitation: A view
from self-psychology. Annual of Psychoanalysis,25, 5–16.
Eysenck, H. J. (1995). The Eysenck Personality Profiler and Eysenck’s theory of
personality. London, UK: Corporate Assessment Network.
Eysenck, M. W. (1992). Anxiety: The cognitive perspective. Hove, UK: Erlbaum.
Funder, D. C. (2000). Personality. Annual Review of Psychology,52, 197–221.
Grunert, U. (1988). Der Selbstdialog im Selbstmitleid [The self-dialogue in self-
pity]. Psyche,42, 602–627.
Grunert, U. (1991). Selbstmitleid ist immer selbstscha
¨digend [Self-pity is always
self-damaging]. Psychologie heute,18(5), 21–26.
Janke, W., Erdmann, G., & Kallus, W. (1985). Streverarbeitungsfragebogen
(SVF) nach W. Janke, G. Erdmann und W. Boucsein [Coping with Stress
Questionnaire after W. Janke, G. Erdmann, and W Boucsein]. Go
¨ttingen,
Germany: Hogrefe.
John, O P (1990). The ‘‘Big Five’’ factor taxonomy: Dimensions of personality in
the natural language and in questionnaires. In L. A. Pervin (Ed.), Handbook of
personality: Theory and research (pp. 66–100). New York: Guilford.
Jo
¨reskog, K., & So
¨rbom, D. (1999). LISREL 8.3 [Computer software].Chicago,
IL: Scientific Software International.
Kahn, E. (1965). Self-pity. American Journal of Psychiatry,122, 447–451.
Krampen, G. (1991). Fragebogen zu Kompetenz- und Kontrollu
¨berzeugungen
(FKK) [Questionnaire on Competency and Control Beliefs (QCC)]. Go
¨ttin-
gen, Germany: Hogrefe.
Krohne, H. W. (1996). Repression-Sensitization. In M. Amelang (Ed.),
Enzyklopa
¨die der Psychologie. Differentielle Psychologie und Perso
¨nlichkeits-
forschung, Band 3: Temperaments- und Perso
¨nlichkeitsunterschiede [Encyclo-
pedia of psychology. Differential psychology and personality research. Vol. 3:
Individual differences in temperament and personality] (pp. 153–184).
Go
¨ttingen, Germany: Hogrefe.
Kro
¨ner-Herwig, B., Muck, C., & Weich, K.-W. (1988). Bewertung der Effizienz
von Bewa
¨ltigungsverhalten am Beispiel der Streverarbeitungsmanahmen aus
dem SVF [Evaluation of the effectiveness of coping behaviors: The example of
the coping-with-stress strategies from the SVF]. Zeitschrift fu
¨r Differentielle
und Diagnostische Psychologie,9, 295–307.
Levenson, H. (1974). Activism and powerful others: Distinctions within the
concept of internal-external control. Journal of Personality Assessment,38,
377–383.
Levenson, H. (1981). Differentiating among internality, powerful others, and
chance. In H. M. Lefcourt (Ed.), Research within the locus of control construct
(Vol. 1 pp. 15–63). New York: Academic Press.
MacAndrew, C. (1989). Factors associated with the problem-engendering use of
substances by young men. Journal of Studies on Alcohol,50, 552–556.
Maier, H., & Smith, J. (1999). Psychological predictors of mortality in old age.
Journal of Gerontology: Psychological Sciences,54B, 44–54.
218 Sto
¨ber

McAdams, D. P. (1992). The five-factor model in personality: A critical appraisal.
Journal of Personality,60, 329–361.
Mikulincer, M., & Florian, V. (1997). The relationship between adult attachment
styles and emotional and cognitive reactions to stressful events. In J. A.
Simpson, & W. S. Rholes (Eds.), Attachment theory and close relationships (pp.
143–165). New York: Guilford.
Milrod, D. (1972). Self-pity, self-comforting, and the superego. In R. S. Eissler
(Ed.), The psychoanalytic study of the child (Vol. 27, pp. 505–528). New York:
Quadrangle.
Nitschke, J. B., Heller, W., Imig, J. C., McDonald, R. P., & Miller, G. A. (2001).
Distinguishing dimensions of anxiety and depression. Cognitive Therapy and
Research,25, 1–22.
Nolen-Hoeksema, S. (1991). Responses to depression and their effects on the
duration of depressive episodes. Journal of Abnormal Psychology,100, 569–
582.
Nolen-Hoeksema, S. (1995). Gender differences in coping with depression across
the lifespan. Depression,3, 81–90.
O’Brien, T. B., & DeLongis, A. (1996). The interactional context of problem-,
emotion-, and relationship-focused coping: The role of the big five personality
factors. Journal of Personality,64, 775–813.
Ostendorf, F., & Angleitner, A. (1993). Deutsche Bearbeitung des NEO
Personality Inventory-Revised (NEO-PI-R) nach Costa und McCrae (1992)
[German version of the NEO Personality Inventory-Revised (NEO-PI-R) after
Costa and McCrae (1992)]. Unpublished manuscript, University of Bielefeld,
Germany.
Rotter, J. B. (1966). Generalized expectancies for internal versus external control
of reinforcement. Psychological Monographs,80, No. 609.
Russell, D., Cutrona, C. E., Rose, J., & Yurko, K. (1984). Social and emotional
loneliness: An examination of Weiss’s typology of loneliness. Journal of
Personality and Social Psychology,46, 1313–1321.
Russell, D., Peplau, L. A., & Cutrona, C. E. (1980). The revised UCLA
Loneliness Scale: Concurrent and discriminant validity evidence. Journal of
Personality and Social Psychology,39, 472–480.
Schwenkmezger, P., Hodapp, V., & Spielberger, C. D. (1992). Das State-Trait-
A
¨rgerausdrucks-Inventar STAXI [The State-Trait Anger Expression Inventory
STAXI]. Bern, Switzerland: Huber.
Smith, J., & Baltes, P. B. (1997). Profiles of psychological functioning in the old
and oldest old. Psychology and Aging,12, 458–472.
Snell, W. E. Jr., McDonald, K., & Koch, W. R. (1991). Anger provoking
experiences: A multidimensional scaling analysis. Personality and Individual
Differences,12, 1095–1104.
Spielberger, C. D. (1988). State-Trait Anger Expression Inventory (STAXI):
Research edition. Odessa, FL: Psychological Assessment Resources.
Sundberg, C. P. (1988). Loneliness: Sexual and racial differences in college
freshman. Journal of College Student Development,29, 298–305.
Tabachnick, B. G., & Fidell, L. S. (1996). Using multivariate statistics (3rd ed.).
New York: HarperCollins.
Self-Pity 219

Tangney, J. P., Hill-Barlow, D., Wagner, P. E., Marschall, D. E., Borenstein, J.
K., Sanftner, J., Mohr, T., & Gramzow, R. (1996). Assessing individual
differences in constructive versus destructive responses to anger across the
lifespan. Journal of Personality and Social Psychology,70, 780–796.
Watson, D., & Hubbard, B. (1996). Adaptational style and dispositional
structure: Coping in the context of the five-factor model. Journal of
Personality,64, 737–774.
Weber, H., Eue, N., Titzmann, P., & Freese, A. (1999). Der Fragebogen zum
Verhalten bei A
¨rger (FVA
¨)[Anger-related Reactions and Goals Inventory
(ARGI)]. Unpublished manuscript, Ernst Moritz Arndt University of
Greifswald, Germany.
Weber, H., & Titzmann, P. (2001, September). Anger-related reactions and goals:
A two-dimensional anger inventory. Paper presented at the 6th European
Conference on Psychological Assessment, Aachen, Germany.
Webster’s Third New International Dictionary. (1961). Springfield, MA: Merriam.
Weiss, R. S. (1973). Loneliness: The experience of emotional and social isolation.
Cambridge, MA: MIT Press.
Wilson, S. L. (1985). The self-pity response: A reconsideration. In A. Goldberg
(Ed.), Progress in self-psychology (Vol. 1, pp. 178–190). New York: Guilford
Press.
220 Sto
¨ber