p.“
4
\ .azina. .5

.. u is?

I
n1 .

.hftaz z.
I. v‘ 5).)... Illa?
§z§ii V

:0

.-
l

r
an
. s). r
‘ ‘ ‘Fflihlivh.
H! ‘ i!
utu‘n.:l..§ .2.
El],
.

’1, 2 ’{nt
“div... 9m..." 1.....1
. hm?!
.

$x........‘.zi:i
”2.3.331; arm.
o».......,fu:ez....t
in)!" u Knit"
.. 1.1.23!-
, C‘h ‘1;l.:|.!-
ikwm... .f .n

., I

‘9 fill-Ila
Half»! uh. ‘
it. .-

 

 

 

 

 

 

 

 

.latlltlci.”

.159»! .

 

 

 

 

 

 

 

- .uis31rdunpc
,) o
u é)l$|0..l.r
. . 1:)... . El...
urn. (:96 I‘lzb.
v3.13! 5.: |.Il\0|
{zflré-EII; :31} u.
5...}: :1: x. 31.1.!» 30.). n
\v‘: I loll}.

.17
g}?!

1......)
vii!»

 

.tfl .n ., “51.30.‘ 4L...“

I. . :9 3.5:"

3:87.

grim; , , ......

153.7. 1...... c.
. : .f . ,

 

10°

”2/

This is to certify that the

dissertation entitled

APPLICATION OF ATTACHMENT THEORY
TO ADOLESCENT SUICIDALITY:
AN ETIOLOGICAL MODEL

presented by

Sondra Renee Wilen

has been accepted towards fulfillment
of the requirements for

Ph . D . degree in Psycho logy

 

 

,0
[MW

Major? professor

nae Sri'cal

 

MS U is an Affirmative Action/Equal Opportunity Institution

 

0-12771

 

LIBRARY
Michigan State
University

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.

TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE

DATE DUE

 

 

w I'fil 1 5} I?!“
W I} Q 5

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c:/CIRC/DateDue.p65-p.15

“

APPLICATION OF ATTACHMENT THEORY TO ADOLESCENT SUICIDALITY:
AN ETIOLOGICAL MODEL

By

Sondra Renee Wilen

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Department of Psychology

2002

ABSTRACT

APPLICATION OF ATTACHMENT THEORY TO ADOLESCENT SUICIDALITY:
AN ETIOLOGICAL MODEL

By

Sondra Renee Wilen

A theoretical model of adolescent suicidality based on attachment theory
that incorporated internal working models, the dynamic characteristics of interpersonal
anxiety and avoidance of negative affect, and the developmentally salient challenges of
separation/individuation and interpersonal stress was investigated through structural
equation modeling with a sample of 354 undergraduate students at a large Midwestern
university.

As predicted, high levels of family risk, including family violence and parental
rejection, were more likely to be associated with negative views of self and negative
views of others, working models that are characteristic of insecure attachment styles.
These negative models of self and others were associated with poorer
separation/individuation and greater levels of anxiety. In addition, negative views of self
were significantly related to avoidance of negative affect, although negative views of
others were not related to avoidance, as had been predicted.

Although the hypothesized dynamic characteristic of interpersonal anxiety was
significantly related to depression and suicidality, limited support was found for the role
of avoidance of negative affect; avoidance was not associated with suicidality and was

positively related to depression, which was not in the direction hypothesized. Potential

difficulties with assessing the hypothesized unconscious coping style, avoidance of
negative affect, are discussed. Additionally, the hypothesized developmentally salient
challenges of separation/individuation and interpersonal stress were not associated with
suicidality. However, as predicted, depression was positively associated with suicidality.

Post hoc modifications of the original model were performed, resulting in a more
parsimonious, better fitting model; adequate support was found for overall model fit
(CFI=.979, NFI=.970, TLI=.973, RMSEA=.O79), and the model accounted for 35% of
the variance in self-reported suicidality.

Overall, the model based on attachment theory offers a valuable alternative to
models previously utilized in the literature, which frequently have relied upon main
effects between variables such as risk factors and adolescent suicidality. The findings are
consistent with the role of attachment cognitions as mechanisms by which associated
family-related risk factors may impact the development of suicidality and suggest that
attachment theory provides an important framework from which to address the etiology

of suicidality.

Dedicated to Michael Nodine

iv

ACKNOWLEDGMENTS

I would first like to thank Dr. Anne Bogat. Dr. Bogat has served as mentor and
advocate, always watching out for my best interests and always motivating me to reach
my potential, even when I have questioned what my potential truly is. Thank you, Dr.
Bogat, for teaching me by example how to be a better researcher, and thank you for your
neverending faith in my abilities. Your guidance has been instrumental to the completion
of this work.

To Dr. Alytia Levendosky, I want to extend my appreciation for inspiring me to
grow, personally and professionally, by introducing me to psychodynamic theory and
attachment theory. My view of the world has changed dramatically as a result of your
opening my eyes to these perspectives. For that I am forever grateful.

I also wish to thank Dr. Alex VonEye, Dr. Gersh Kauffman, and Dr. Francisco
Villaruel for the time and effort each of you put into reviewing this work and helping me
shape it into what it is today.

I must also mention three special friends and sources of invaluable support. Dr.
Lisa Galasso, thank you for your companionship throughout this program. You have
been there with me and for me, sharing highs and lows, through these dynamic years.
Graduate school would have been a great deal less fulfilling without you. Jennifer
McQueeney, I thank you, as I fear I do not often enough, for believing in me and always
offering to lend an attentive ear or a thoughtful word. And, Elinor Gerock, beloved Nana,
know that without your generosity with your time and love, I never could have gotten

where I am today.

To Jennifer Trotter, I am thankful for all your hard work in helping me collect the
data for this project. You are a skilled and dedicated researcher, and I know you are well
on your way to a long, accomplished career in the field of psychology.

Finally, I thank my loving husband and beautiful daughter for all they do for me.
Brian, you are my soulmate and companion (and accomplished data collector), always a
true partner and loving father. And Ashlynn, you are a symbol of purity in my life, purity
of heart and mind of which I am in great awe. You give new meaning to all I do, you are
the motivation for me to strive to be a role model as mother and professional, and I pray
that I give back to you a fraction of the love and joy you give to me each day. Thank
you both for your love and patience during the completion of this work and throughout

our lives together. I am grateful to share my life with you.

vi

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................................ IX
LIST OF FIGURES ........................................................................................................... X
INTRODUCTION .............................................................................................................. 1
RISK FACTORS AND ADOLESCENT SUICIDALITY ................................................. 4
Family Characteristics ........................................................................................ 4

Parental Psychopathology .................................................................................. 5

Family Violence ............................................ . ..................................................... 6

General Family Disturbances ............................................................................. 6
Individual Characteristics ................................................................................... 8
METHODOLOGICAL WEAKNESSES OF PRIOR RESEARCH ................................. 11
ATTACHMENT ............................................................................................................... 14
Introduction to Attachment Theory .................................................................. 15
Psychodynamic Theory .................................................................................... 16
Attachment Theory ........................................................................................... 18

Internal Working Models of Self and Others ............................................ 20

Dynamic Characteristics ........................................................................... 22

FAMILY CHARACTERISTICS AND INSECURE ATTACHMENT ........................... 24
ATTACHMENT AND PSYCHOLOGICAL DISTRESS ................................................ 27
ATTACHMENT AND SUICIDALITY ........................................................................... 32
Factors Neglected by Past Studies Investigating Attachment and Suicidality. 36

Role of Internal Models and Resulting Dynamic Constructs ................... 36
Developmental Task of Separation/Individuation .................................... 39

Interpersonal Stressors .............................................................................. 43

Depression ................................................................................................. 44

RATIONALE .................................................................................................................... 47
HYPOTHESES ................................................................................................................. 49
METHODS ....................................................................................................................... 53
Participants ....................................................................................................... 53

Measures ........................................................................................................... 53
Procedures ........................................................................................................ 65

vii

RESULTS ......................................................................................................................... 66

Data Screening and Preparation ....................................................................... 66
Measurement Model Modifications ................................................................. 68
Evaluation of Overall Model Fit ...................................................................... 69

Post Hoc Model Modifications ......................................................................... 70
Evaluation of Specific Hypotheses ................................................................... 71
DISCUSSION ................................................................................................................... 77
Proposed Constructs ......................................................................................... 78

Family Characteristics .............................................................................. 78

Attachment, Models of Self and Models of Others .................................. 78

Avoidance of Negative Affect .................................................................. 80

Proposed and Post Hoc Models ........................................................................ 80

Proposed Pathways ........................................................................................... 81
Alternative models ........................................ V ................................................... 88
Limitations of the Current Study and Directions for Future Research ............. 89
Appendix AzTables ............................................................................................................ 95
Appendix BzFigures ......................................................................................................... 100
Appendix C :Questionnaire Packet ................................................................................... 109
REFERENCES ............................................................................................................... 153

viii

LIST OF TABLES

Table 1: Descriptive Statistics for Model Variables ......................................................... 95

Table 2: Standardized Maximum Likelihood Estimates of Measurement Model
Parameters ........................................................................................................ 96

Table 3: Standardized Maximum Likelihood Estimates of Structural Model
Parameters ........................................................................................................... 97

Table 4: Comparison of Post Hoc Nested Models ............................................................. 98

ix

Figure 1:
Figure 2:

Figure 3:

Figure 4:

Figure 5:

Figure 6:

Figure 7:

Figure 8:

LIST OF FIGURES

Model of the Etiology of Adolescent Suicidality ............................................ 100
Example of Model Previously Utilized to Explain Adolescent Suicidality 101

Example of Attachment Theory Model Previously Utilized to Explain

Adolescent Suicidality ..................................................................................... 102
Intersection of Models of Self and Others, Attachment and Individuation

and Resulting Dynamic Characteristics ......................................................... 103
Structural Model of the Etiology of Adolescent Suicidality .......................... 104
Standardized parameter estimates for altemative model proposed in

Figure 2 ........................................................................................................... 105
Standardized parameter estimates for alternative model proposed in

Figure 3 ........................................................................................................... 106
Final Post Hoc Model of the Etiology of Adolescent Suicidality .................. 107

INTRODUCTION

Suicidal ideation and suicidal behavior are prevalent among the youth of our
society. Recent studies have revealed increases in the already high incidence rates of
adolescent suicidality, which includes suicidal thoughts, plans, and attempts as well as
completed suicide (Garland & Zigler, 1993; Henry, Stephenson, Hanson, & Hargett,
1993). Suicide now accounts for 11.3 deaths per 100,000 adolescents and is the third
leading cause of adolescent death (Garland & Zigler, 1993). Among 15- to 24-year olds,
an average of 14 suicides occur every day, accounting for over 5,000 deaths each year
(Henry et al., 1993). While the suicide rate for the general population has increased by
17% since 1960, adolescent suicide increased by more than 200% (Garland & Zigler,
1993)

Unfortunately, these rates are probably underestimates. Many attempts and
completed suicides never come to the attention of the authorities or are intentionally
misclassified as “accidents” due to factors such as social stigma, religious reasons, and
insurance benefits; it is therefore difficult to report the incidence rate accurately
(Holinger, Offer, Barter, & Bell, 1994). Despite such measurement difficulties, suicide
attempts are estimated to be 5 to 20 times more prevalent than completed suicides
(Asamow, 1992).

Much of the research on adolescent suicide has focused on distinguishing
characteristics of suicidal adolescents in order to assist in the identification of those who
are at risk for suicide. As a result, although a substantial body of empirical evidence now

exists supporting an association between adolescent suicidality and family-related risk

factors (e. g., Asamow, 1992; Cohen-Sandler, Berman, & King, 1982; Krarup, Nielsen,
Rask, & Petersen, 1991), including parental psychopathology (e. g., Brent, Kolko, Allan,
& Brown, 1990; Kashani, Ezpeleta, Dandoy, Doi, & Reid, 1991; Shafi, Carrigan,
Whittinghill, & Derrick, 1985) and family violence (e.g., Husain, 1990; Spirito, Brown,
Overholser, & Fitz, 1989; Pfeffer, 1985), an understanding of the etiology of adolescent
suicidality eludes the literature, as no risk factor appears to be necessary or sufficient for
the development of suicidality. The purpose of the current investigation is to develop and
test empirically a theoretically-based model of adolescent suicidality which not only
accounts for the mechanisms underlying the previously established relationships between
family-related risk factors and adolescent suicidality but also addresses the etiology of
suicidality within a developmental context. It is proposed that a model founded in
attachment theory which also accounts for underlying mechanisms and the impact of the
age-salient developmental task of separation/individuation will provide a more
comprehensive understanding of the development of adolescent suicidality (See Figure
1).

The model proposed in Figure 1 is founded in attachment theory and addresses
methodological weaknesses of prior research which have hindered efforts not only to
identify the mechanisms by which family risk factors influence the development of
suicidality (Kazdin & Kagan, 1994) but also to understand this problem within a
developmental context. This model proposes that early family characteristics impact
early attachment style, which then produces specific combinations of positive and
negative models of self and other (Bartholomew, 1990; Bowlby, 1977). The current

investigation proposes that the resulting patterns of positive and negative views of self

and others not only produce particular levels of avoidance of negative affect and
interpersonal anxiety (Shaver & Clark, 1996) but also impact resolution of the task of
separation/individuation during adolescence (e. g., Allen, Hauser, Bell, & O’Connor,
1994; Kroger & Haslett, 1988; Salzman, 1996). Particular combinations of avoidance
and anxiety are hypothesized to influence the manifestation of both suicidality and
depression (e. g., Borst & Noam, 1993; Van der Kolk, Perry, & Herman, 1991).
Similarly, it is proposed that resolution of the task of separation/individuation impacts
both suicidality and depression, with inadequate resolution functioning as a stressor
during adolescence. F urthermore, both depression and interpersonal stressors are
hypothesized to be related to suicidality.

Therefore, this study utilizes attachment theory as a foundation from which to
investigate “common developmental pathways” (Adam, 1994, p. 276) which lead to
suicidality specifically during the developmental stage of adolescence. By incorporating
underlying mechanisms, as well as the stage salient task of separation/individuation, this
study addresses major weaknesses of past research which primarily has examined direct
causal relationships between risk factors and suicidality in adolescents. It is hypothesized
that the proposed model will offer a valuable alternative to models previously utilized to

address the development of adolescent suicidality.

RISK FACTORS AND ADOLESCENT SUICIDALITY

F_amilv Characteristics

Over thirty years ago Morrison and Collier (1969) stated that childhood and
adolescent suicidal behavior is “a symptom not only of individual upheaval but of
underlying family disruption” (Pfeffer, 1986, p. 124). Today there exists a substantial
body of empirical evidence to support the conclusion that “pervasive and long-standing
family disturbances” (Blurnenthal & Kupfer, 1990, p. 159) are associated with an
increased risk for adolescent suicide (Brent et al., 1990; Cohen-Sandler et al., 1982;
Husain, 1990; Kashani, Beck, & Burk, 1987; Krarup et al., 1991; Pfeffer, 1989; Seiden,
1984; Shafi et al., 1985). Adolescent suicide is associated with a range of familial
disruption, such as parental psychopathology, parental abuse, parental absence, antisocial
behavior, drug and alcohol abuse, and previous suicidal behavior (Shafi et al., 1985).
Parental problems precipitate as many as 50% of adolescent suicide attempts, and
adolescent self-reported lack of family support discriminates suicide attempters fi'om
nonattempters even when hopelessness and the severity of depression are controlled
(Asamow & Carlson, 1988; Brent et al., 1990; Spirito et al., 1989; Tishler, McKenry &
Morgan, 1981).

The current discussion will focus on the familial characteristics which are most
commonly associated with adolescent suicide and therefore have been the subject of
much research: parental psychopathology, family violence, and general family
disturbances. However, while a substantial body of literature has attempted to identify

particular factors which place adolescents specifically at risk for suicidality, it will be

seen that no risk factor appears to be necessary or sufficient for the development of
suicidality.
Parental Psychopathology

Research has found parental psychopathology to be a major factor contributing to
adolescent suicide (Henry et al., 1993; Kashani etal., 1987; Kashani et al., 1991; Pfeffer,
1989; Shafi et al., 1985). When suicidal patients are compared to non-suicidal depressed
patients, the parents of the suicidal patients are characterized by higher rates of affective
disorders and “an earlier onset of chronic psychiatric illness” (Brent et al., 1990, p. 587).
The severity of parental psychopathology also is significantly related to a child’s risk for
a psychiatric diagnosis, such as depression and suicidal ideation (e. g., Kashani et al.,
1987). Furthermore, a history of family suicide attempts is more common in adolescent
suicide attempters than in relevant controls (e. g., Henry et al., 1993; Krarup et al., 1991).
However, while a substantial amount of empirical evidence exists to support a
relationship between parental psychopathology and adolescent suicidality, parental
psychopathology does not appear to be either necessary or sufficient for the development
of adolescent suicidality. For example, a study conducted by Shafi et a1. (1985) found
that 60% of the adolescents who committed suicide had a parent with emotional problems
compared to 24% of a non-suicidal control group.

Kashani et a1. (1991) identified the qualities of emotional coldness, negativism,
and abuse from the mother, as perceived by the child, as factors mediating the
relationship between parental psychiatric illness and child psychopathology. Kashani et
a1. (1991) argue that the results of this study “provide further evidence of the parental

contribution to child psychopathology” (p. 572). More importantly, this study provides

information regarding possible mechanisms by which parental psychopathology affects
children’s outcomes. This will be an important area for future research.
F_amilv Violence
Another factor that is related to adolescent suicidal behavior is family violence
(Pfeffer, 1985). Both physical and sexual abuse are commonly found in the histories of _
suicidal adolescents (Husain, 1990; Pfeffer, 1989; Seiden, 1984). Additionally, a history
of family aggression is related to adolescent suicidal behavior (Spirito et al., 1989), and
suicidal and assaultive tendencies are commonly found in the parents of suicidal
adolescents (Pfeffer, 1985). In fact, the literature suggests that parental behavior
characterized by “self-directed and outwardly directed” violence is a significant risk
factor for adolescent suicidal behavior (Pfeffer, 1985, p. 224). However, it does not
appear that a history of family violence is either necessary or sufficient for the
development of adolescent suicidality.
General Famin Disturbances
The climate of the suicidal adolescent’s family has been described as hostile,
indifferent, and openly rejecting (Asamow, 1992; Curran, 1987; DeMan, Labreche-
Gauthier, & Leduc, 1993; Kosky, Silbum, & Zubrick, 1986; McIntire & Angle, 1975).
Family environment characteristics such as a lack of maternal support (DeMan etal.,
1993), disturbed child-father and child-sibling relationships (Kosky et al., 1986), severe
disciplinary techniques (ofien characterized by a lack of reasoning) (Kosky, et al., 1986;
Spirito, Brown, Overholser, & Fritz, 1989), as well as a high level of parental control and
a perceived lack of parental warmth (DeMan et al., 1993), are all associated with

increased risk for suicidality in adolescents. Other parental characteristics which are

associated with suicidality in adolescents include a tendency toward rejection, nagging,
criticism, perfectionism, withholding of approval, yelling, and a lack of nurturance
(Curran, 1989; Henry et al., 1993; Kosky et al., 1986). In fact, Godwin (1986) reports
that the cross-cultural literature available on adolescent suicide provides additional
evidence that the family environment is related to suicidality cross-culturally.

Additionally, the way that adolescents view their family environment plays an
important role in their development of suicidal tendencies (Asamow, 1992; McKenry,
Tishler, & Kelley, 1982). For example, Asamow (1992) found that children who had
attempted suicide saw their families as less expressive, less cohesive, and higher in
conflict than the non-attempter psychiatric controls. Furthermore, the scores of children
with suicide ideation, but no attempt, fell between those of the attempter and the non-
suicidal groups, lending additional support to the potential predictive value of these
variables (Asamow, 1992). While the family stress that characterizes the families of
suicidal children may be both quantitatively and qualitatively different than the stress
experienced by other children, the child’s perception of the experienced family stress and
lack of support appears to be particularly influential in the later development of
suicidality (Asamow, 1992; McKenry et al., 1982).

Another general family characteristic that is associated with adolescent suicidality
is a history of severe and chronic life stress, including the loss of significant others (e.g.,
Cohen-Sandler et al., 1982). In fact, the most common precipitants of suicidal behavior
in adolescents are interpersonal conflict with or the loss of a significant family member or
fiiend (Henry et al., 1993). Cohen-Sandler et a1. (1982) compared the life experiences of

suicidal children to those of a control sample of non-suicidal, psychiatrically hospitalized

children. The results showed that, compared to the control group, the frequency and
severity of life stress experienced by the suicidal children increased throughout their
childhood, with these children experiencing many losses such as separation from their
parents and death of a grandparent (Cohen-Sandler et al., 1982).

Although suicidality in adolescents is related to a history of an adverse family
environment, it again remains unclear as to the exact nature of this relationship (Spirito,
et al., 1989). Not only have inadequate control groups frequently been utilized in these
investigations (Kosky et al., 1986), but no individual family characteristic appears to be
necessary or sufficient for the development of suicidality. More research investigating
family characteristics and adolescent suicidality is necessary in order to gain a better
understanding of the mechanisms underlying these relationships.

Individual Characteristics

In addition to the empirical evidence that exists regarding the relationship
between family-related risk factors and adolescent suicidality, there also exists a
substantial body of empirical evidence that links individual characteristics, such as
depression and anger, to an increased risk for suicide during adolescence (Bettes &
Walker, 1986; Friedman et al., 1984; Harrington, Brendenkamp, Groothues, & Rutter,
1994; Kovacs, Goldston, & Gatsonis, 1993; Robbins & Alessi, 1985; Pfeffer, Plutchik, &
Mizruchi, 1983; Withers & Kaplan, 1987). While a complete review of this literature is
beyond the scope of the current investigation, the major individual characteristics that
have been associated with adolescent suicidality will be addressed briefly.

The individual characteristic that has been most thoroughly investigated in

relation to adolescent suicidality is depression (Spirito et al., 1989). It is well-established

that depression is associated with an increased risk for adolescent suicidality (e. g.,
Friedman etal., 1984; Harrington et al., 1994; Hollis, 1996; Kovacs et al., 1993);
however, it also is well-established that depression is not sufficient for the development
of suicidality, as a majority of depressed adolescents never evidence suicidality. As a
result, the adolescent suicide literature frequently utilizes control groups of nonsuicidal
depressed adolescents in an attempt to isolate those factors specific to suicidality (e. g.,
Brent et al., 1990; Kosky et al., 1986).

Furthermore, depression is not a necessary emetional state for the development of
suicidality (Feldman & Wilson, 1997; Hollis, 1996; Khan, 1987; Spirito, 1989; Taylor &
Stansfeld, 1984). Hollis (1997) reports that although depression was the largest single
risk factor associated with suicidal behavior in his study, only twenty-seven percent of the
suicidal adolescents were categorized as depressed. In fact, Spirito et al. (1989)
conducted a thorough review of the literature and concluded that, while depression is a
characteristic common to many suicidal adolescents, not all adolescents who are suicidal
evidence depressive symptomatology. They argue that the samples utilized to investigate
adolescent suicidality suffer from several confounding factors, resulting in an over-
representation of depressed adolescents (Spirito et al., 1989). For example, affective
disorders often are over-represented in inpatient psychiatric populations due to hospital
admission criteria, and many suicidal adolescents who do not evidence depressive
symptomatology never come to the attention of the medical community (Spirito et al.,
1989)

Similarly, Feldman and Wilson (1997) located different subsets of suicidal

adolescents and concluded that the typical method of identifying adolescents at-risk for

suicidality, which relies upon measures of depressive symptomatology, “overlooks a
majority of at-risk adolescents” (p. 75). They suggest, instead, that structural personality
variables, such as affect tolerance ("the way in which a person manages affective
arousal") and affect expression ("how an individual's emotions play a role in a
communicational matrix with important others"), be utilized to identify adolescents who
are at-risk for suicidality (F eldman & Wilson, 1997, p. 77). The identification of
subgroups as well as the utilization of structural personality variables in relation to
suicidal adolescents will be addressed more thoroughly at a later point in this manuscript.
While studies have focused primarily upon depression, hopelessness and anger
also have been associated with suicidality (Asamow, Carlson, & Guthrie, 1987; Beck,
Steer, Kovacs, & Garrison, 1985; Bettes & Walker, 1986; Kazdin etal., 1983; Pfeffer et
al., 1983). However, similar to the research regarding both family-related risk factors and
depression, neither hopelessness nor anger appear to be necessary or sufficient for the
development of suicidality during adolescence (Spirito et al., 1989). Again, it seems that
more research investigating the mechanisms underlying the relationships between risk

factors and adolescent suicidality is necessary.

10

METHODOLOGICAL WEAKNESSES OF PRIOR RESEARCH

While significant differences between the characteristics of suicidal adolescents
and control subjects have been identified, the adolescent suicide literature unfortunately
has been characterized by many of the methodological weaknesses outlined by Kazdin
and Kagan (1994) as common impediments in developmental psychopathology research.
These researchers state that developmental psychopathology research typically assumes
single pathways of dysfunction and investigates main effects between variables such as
risk factors and outcomes (Kazdin & Kagan, 1994). However, despite the identification
of a statistically significant relationship between a risk factor and an outcome, it generally
is the case that a large proportion of individuals who experienced the risk factor do not
develop the outcome of interest, implying an important relationship “between the
experience...[of the risk factor]...and some other internal or external condition (Kazdin &
Kagan, 1994, p. 37).

The etiology of adolescent suicidality remains poorly understood. While a variety
of familial risk factors are associated with adolescent suicide, frequently these variables
are not specific to suicidality, but rather place an adolescent at risk for psychopathology
in general. Similarly, while several family characteristics are associated with suicidality
during adolescence, such as parental psychopathology and poor parent-child relationships
(e.g., Henry et al., 1993; King et al., 1990), there does not appear to exist a particular risk
factor that is either necessary or sufficient for the development of suicidality in
adolescents. Only recently has the adolescent suicide literature, which has long been

characterized by assumptions of single causal pathways and outcomes, begun exploring

11

alternatives to direct effects models (e. g., Adam, Sheldon-Keller, & West, 1996; Hollis,
1996)

Kazdin and Kagan (1994) also argue that the dynamic nature of potential risk
factors needs to be taken into consideration when developing models of dysfunction in
developmental psychopathology, and different models may be necessary to explain
similar outcomes at different ages. While some of the research investigating adolescent
suicidality investigated the dynamic nature of factors (e. g., Cohen-Sandler et al., 1982),
many studies have not considered the potential impact of the particular developmental
tasks associated with the stage of adolescence.

Perhaps of greatest relevance to the adolescent suicide literature is the observation
by Kazdin and Kagan (1994) that researchers in developmental psychopathology often
identify particular constructs “as the basis for the relation when in fact other constructs or
a larger set of correlated constructs could be responsible” (p. 42). They suggest that
future developmental psychopathology research should investigate the mechanisms
underlying correlated factors, as variables which are related to a particular outcome often
can be more accurately explained by an underlying condition. As mentioned previously,
there are a paucity of studies in the adolescent suicide literature that investigate possible
mechanisms which may underlie the significant but potentially spurious relationships that
have been identified between familial risk factors and adolescent suicidality. Despite the
strong association between family risk factors and suicidal behavior in adolescents,
researchers acknowledge that the relationship between these variables requires further
investigation, as the mechanism by which family factors may influence the development

of adolescent suicidality remains unclear (e. g., Asamow, 1992; Pfeffer, 1985).

12

In addition, Kazdin & Kagan (1994) argue that future research should take into
account that there exists a significant amount of variability in the way in which
individuals “cognitively process and represent ‘objective’ events” (p. 47) and that it may
be the subjective experience of events, rather than the events themselves, that is related to
their impact. Internal representations of particular events, therefore, can lead to
significantly different effects of those events and should be considered in models of
developmental psychopathology (Kazdin & Kagan, 1994). This is another area in which
the adolescent suicide literature could be improved, as previous studies have tended to
focus on the mere presence or absence of events rather than to account for differences in

the meaning of events for individuals.

13

ATTACHMENT

The extensive literature investigating the relationship between family risk factors
and adolescent suicidality contains many methodological difficulties. Proposed models
generally have failed to incorporate the dynamic nature of factors, underlying
mechanisms, and internal representations of external events. Not only have previous
studies been searching in vain for a common singular event or risk factor underlying
adolescent suicidality (Curran, 1987), but the methodological difficulties associated with
these studies have hindered our understanding of the etiology of adolescent suicidality.

In an attempt to gain a more complete understanding of this problem, the current
investigation proposes attachment theory as a foundation for building a model to explain
adolescent suicidality. Attachment theory addresses differences in the way that
individuals conceptualize the world around them. The present study proposes that the
internal working models of attachment theory provide the mechanisms by which familial
risk factors influence the development of adolescent suicidality. Attachment theory will
allow for the investigation of “common developmental pathways” (Adam, 1994, p. 276)
leading to suicidality rather than the mere examination of single causal relationships
between stressful events and adolescent suicidality. In this way, not only may a model
based in attachment theory account for the fact that a specific risk factor may result in
suicidality for some adolescents but not for others, but it also may provide the
mechanisms by which different risk factors can lead to the common outcome of
suicidality.

To support this argument, a brief overview of attachment theory and a discussion

14

of its relationship to relevant psychodynamic theory is offered. Empirical evidence
regarding the impact of characteristics of the family on the formation of attachment is
discussed. In particular, it will be seen that many risk factors that are associated with
adolescent suicidality also adversely impact attachment security. Additionally, because
empirical evidence about the relationship between attachment and psychological distress
may reveal important information regarding attachment and suicidality, previous research
in this area is reviewed as well as the few existing, yet promising, studies that have
investigated attachment and suicidality.

Introduction to Attachment Theory

Attachment, as defined by Bowlby (1977) and Ainsworth (1968), refers to
affectional bonds that human beings make to specific others that are enduring and occur
throughout life. Bowlby (1977, 1988) describes attachment behavior as that which serves
the homeostatic firnction of maintaining proximity to a caregiver, with perceptions of
safety versus danger regulating the preferred distance from the attachment figure. While
attachment behavior is the medium through which attachment can be observed, the
behavior is distinct from attachment itself, as fluctuations in the attachment behavior
elicited in different situations are not necessarily considered to be indications of changes
in the strength of the internal attachment (Ainsworth, 1968).

An enduring attachment to important others is thought to result fiom the
formation of internal working models of the self and of attachment figures, which are
constructed from past experiences and serve the function of guiding future behavior in
response to predictions of the behavior of others (Bowlby, 1988; Bretherton, 1985). In

this way, internal working models are formed representing caregivers as available or

15

unavailable and the self as worthy or unworthy of care and protection (Bowlby, 1977;
Bretherton, 1985). These internal working models, which are formed early in childhood,
produce qualitatively different internal experiences of the same external events
throughout the lifetime of different individuals. The present investigation proposes that
these internal working models are the mechanisms by which familial risk factors
influence the manifestation of suicidality during adolescence. In order to gain a more
complete understanding of the internal working models that are hypothesized by
attachment theory, a brief discussion of the related psyChodynamic theories of object
relations theory and self psychology theory is presented.
Psvchodvnamic Theorv

Object relations theory, much like the theory of attachment proposed by Bowlby
and Ainsworth, considers the quality of the relationship between the infant and the
caregiver to form the basis of the child’s internal models of relationships which continue
to significantly influence the way in which the individual relates to others throughout his
life (Bowlby, 1977). This theory presumes that the “good-enough mother,” who provides
adequate emotional and physical care for her child, and the formation of the holding
environment, in which the child is consistently protected by the caregiver, facilitates the
child’s attachment to the caregiver and promotes healthy development (Mitchell & Black,
1995).

Object relations theorists propose that infants begin to divide the world into good
and bad experiences in the earliest stages of life. When the caretaker is experienced as
bad due to the fi'ustration of the child’s needs, the child’s unconscious internal

representations of the caretaker are “split” into good and bad internal objects in an

16

attempt to maintain the positive parts of the caretaker (Cashdan, 1988). Over the course
of development, the “splitting” of experiences with caretakers results in the unconscious
“splitting” of aspects of the developing self, resulting in feelings of being lovable and
worthwhile or unlovable and worthless (Cashdan, 1988).

Pine (1990) argues that internal representations of other and self are not
considered in object relations theory to be true reflections of past and present
relationships, but rather, represent the child’s experience of both real and fantasized
relationships. The emphasis which is placed on the role of fantasy in the development of
children’s internal representations of self and other in the object relations perspective is
an important area of divergence between object relations theory and attachment theory
(Lopez, 1995).

In self psychology theory, Kohut emphasized the importance of empathy in the
parent-child relationship and proposed that the formation of the child’s positive sense of
self required parental mirroring of the child’s worth and the presence of a parental figure
whom the child could idealize (Mitchell & Black, 1995; Slade & Aber, 1992). Kohut
argued that the child experienced the caregiver as part of the self and believed that the
child’s internal models of the self grew out of this merger with as well as the empathic
responses and appropriate failures of the caregiver (Taylor, 1992). Similar to attachment
theory and object relations theory, self psychology considers early experiences to be
templates for later relationships and labels these internal models “transmuting
intemalizations” (Slade & Aber, 1992). To Kohut, psychopathology originates in early
development from deficiencies in the individual’s experience with empathic self-objects

(Taylor, 1 992).

17

Blass and Blatt (1992) argue that the primary difference between object relations
theory and self psychology is the goal which is hypothesized to predominate
psychological development. Within object relations theory, the relationship with the
caregiver is viewed as primarily serving the goal of attachment, with the goal of
separateness remaining secondary. Within self psychology, the converse is true; the
relationship with the caregiver primarily promotes the development of separateness.

Atlachment Theory

Attachment theory, while compatible with and having grown out of existing
psychodynamic theories such as object relations theory, proposes that while internal
working models of self and other are distinct concepts, they need to be understood in
relation to one another, as they signify opposite sides of the same relationship
(Bretherton, 1985). In this way, attachment theory captures Blatt and Schichman’s
(1983) conception of the development of personality, which they argue results from an
intricate transaction of two developmental paths, one of which leads to the development
of fulfilling, intimate relationships with others, the other leading to the establishment of a
positive, enduring sense of self. In attachment theory, the goals of separateness and
attachment remain inextricably connected to one another over the course of development.

Empirical findings from Ainsworth, Blehar, Waters, and Wall’s (197 8) studies not
only revealed three reliable classifications of child attachment to the caregiver, but also
illustrated the way in which the two developmental paths are intertwined. Evidence
supporting the child’s use of the attachment figure as a secure base from which
exploration and autonomy develops clarified the way in which separation of self and

autonomous exploration was affected by development of a secure attachment to the

18

caregiver. Whereas securely attached children are confident that the attachment figure
will be helpful and available if the child should encounter danger, insecurely attached
children are not, with anxious-resistant (insecure-ambivalent) children being uncertain of
their ability to rely on the caregiver and anxious-avoidant (insecure-avoidant) children
being confident that they cannot rely on the attachment figure (Bowlby, 1988).
Furthermore, Ainsworth et al.’s (1978) investigations have provided empirical support for
the relationship between the quality of caretaking over the first year of life and the
resulting child attachment classifications, as assessments of parental care were predictive
of later attachment style (Bowlby, 1988).

Although the importance of attachment to or the relationship with caregivers in
infancy and childhood has been discussed by prior psychodynamic theorists (e. g., Freud,
Mahler, Winnicott, and Kohut), the context of ethology in which Bowlby placed the
development and evolutionary function of attachment to caregivers provided a
mechanism through which the formation of connection to others and a sense of oneself as
worthy of care from caregivers would develop in all human beings (Bowlby, 1977, 1988;
Lay, Waters, Posada, & Ridgeway, 1995). In fact, Pine (1990) argues that Bowlby’s idea
of the evolutionary adaptiveness of the innate biological process of attachment to
caregivers provided a link between biology and the psychology of the object tie proposed
by psychodynamic theory. Additionally, through the application of ethological findings
to human development, such as Harlow’s (195 8) report on the effects of maternal
deprivation in rhesus monkeys, Bowlby highlighted a limitation of earlier theories which
regarded the primary motivational force behind infant attachment as the satisfaction of

biological needs, with the establishment of relationships occurring as a derivative of

19

instinctual gratification (Bretherton, 1992; Slade & Aber, 1995). In contrast, Bowlby
asserted that humans are inherently social, with an inborn capacity to form “intimate
emotional bonds” which is neither secondary to nor a derivative of the fulfillment of
biological needs (Bowlby, 1988; Bretherton, 1992; Fishler, Sperling, & Carr, 1990; Slade
& Aber, 1995).

The theory of attachment proposed by Bowlby, and the research of Ainsworth
which followed, operationalized the behavioral manifestations of the internal working
models and a stable self concept which had been proposed by others such as Winnicott
and Kohut, allowing observation and measurement of these theoretical constructs in
infants and children (Bretherton, 1992; Lay et al., 1995; Sroufe, 1986). As a result,
empirical evidence was provided by attachment researchers which indicated that infants
had an innate ability to accurately perceive, integrate, represent, and adapt to reality
(Oppenheim & Waters, 1995; Slade & Aber, 1995). Not only have classifications of
adult attachment styles been found to be similar to those found in infants and children,
but studies have also revealed that parental attachment, as assessed by the Adult
Attachment Interview (AAI), was related to the attachment styles of the children of these
adults (Bretherton, 1992; Main, Kaplan, & Cassidy, 1985). Furthermore, empirical
support for a relationship between a child’s attachment to the caregiver and the
development of the child’s sense of self exists (Cassidy, 1988; Salzman, 1996).

Internal Working Models of Self and Others

Attachment categories correspond to four quadrants representing combinations of

positive/negative views of self/other. A secure attachment is associated with positive

models of self and other. Individuals in this quadrant tend not only to be comfortable

20

with both autonomy and intimacy but also to be characterized by a positive self-esteem
and a sense of basic trust (Bartholomew, 1990; F eeney, Noller, & Hanrahan, 1994). A
preoccupied/anxious-ambivalent insecure attachment category is characterized by
negative views of self and positive views of other. Individuals in this quadrant tend to be
overdependent on others, interpersonally anxious, focused on approval, and preoccupied
with relationships (Bartholomew, 1990; Feeney et al., 1994).

Bartholomew ( 1990) argues that the consideration of positive and negative views
of self and other results in two subclassifications of avoidant attachment, dismissing and
fearful. Negative views of others and positive views of self correspond to a dismissing
avoidant attachment, whereas negative view of self and others correspond to a fearful
avoidant attachment (Bartholomew, 1990). Dismissing individuals tend to be
characterized by independence, self-reliance, a lack of trust, and an avoidance and
dismissiveness of intimacy (Bartholomew, 1990; F eeney et al., 1994). Fearful
individuals tend to be characterized by low self-esteem, interpersonal anxiety, and a lack
of trust coupled with a desire for intimate relationships and approval from others, a
combination which often leads these individuals to feel alone and experience a significant
amount of hostility and anger (Bartholomew, 1990; Feeney et al., 1994). A disorganized
attachment style does not directly correspond to one of the quadrants representing
positive/negative views of self/other but rather represents an inability to resolve past
traumatic events.

Research has revealed that specific combinations of positive/negative models of
self/other are related to gender. Shaver and Clark (1996) report that more males than

females are characterized by a positive view of self combined with a negative view of

21

other (dismissing avoidant). Also, more females than males are characterized by a
negative view of self combined with a positive view of other (pre-occupied) and by a
negative view of self and other (fearful avoidant) (Shaver & Clark, 1996). In addition,
not only have views of self/others been found to mediate the relationship between
attachment and outcomes (Kenny, Moilanen, Lomax, & Brabeck, 1993; Roberts, Gotlib,
& Kassel, 1996), but relationships between specific patterns of positive/negative views of
self/others and specific psychopathology also have been hypothesized (e. g., Camelley,
Pietromonaco, & Jaffe, 1994); these topics will be discussed in more detail at a later point
in this manuscript.
Dmamic Characteristics

In addition, Shaver and Clark (1996) argue that combinations of positive/negative
models of self/others result in differences in two dynamic characteristics: interpersonal
anxiety and avoidance of negative affect. A two-axis model of human development
proposed by Kaplan (1988) provides a useful heuristic for understanding the levels of
interpersonal anxiety and avoidance of negative affect that result from the intersection of
these constructs (i.e., Shaver & Clark, 1996). In this model, attachment and individuation
are conceptualized as orthogonal dimensions on which individuals vary; this results in a
four quadrant diagram representing four categories: attached and individuated; detached
and individuated; attached and deindividuated; and detached and deindividuated (Kaplan,
1988; See Figure 4).

Because Kaplan (1988) presents attachment as a dichotomous variable, it does not
directly correspond to Bowlby’s conception of attachment. However, Kaplan’s model

has certain heuristic advantages. His first axis describes the internal working models of

22

others proposed by Bowlby-- positive models of others are at the end of the axis
representing attachment and relatedness, while negative models of others are at the
detachment end. The second axis, representing individuation/autonomy and de-
individuation, captures the models of self proposed by attachment theory, as positive
models of self are at the end of this axis representing the achievement of individuation
and negative models of self are at the de-individuation end. The attachment categories
fall within the four quadrants representing the intersection of attachment and
individuation. Each of these quadrants produces different combinations of interpersonal
anxiety and avoidance of negative affect, as illustrated in Figure 4 (Shaver & Clark,
1996). Perhaps combinations of these two dynamic characteristics mediate the
relationship between views of self/others and symptomatology. Again, this topic will be

discussed further at a later point in this manuscript.

23

FAMILY CHARACTERISTICS AND INSECURE ATTACHMENT

As can be seen, through the pioneering work of Bowlby, Ainsworth, and others,
the impact of reality-based characteristics of the parent-child relationship on children’s
attachment and the significant impact that early parenting can have on development
throughout life has been supported empirically. Empirical evidence also has been
provided regarding the importance and endurance of internal working models in
relationships across the lifespan as well as their effects on adult caregiving and the
intergenerational transmission of attachment styles (Main et al., 1985). In addition, the
attachment literature consistently has revealed that when family characteristics, such as
parental psychopathology and family violence, adversely affect parental availability and
responsiveness, insecure attachment styles result (Cassidy & Berlin, 1994). In fact, many
of the family characteristics identified as risk factors for adolescent suicidality are known
to produce insecure attachment styles in children, and several examples will be discussed.
The current investigation proposes that, while there do not exist specific risk factors
which are necessary or sufficient for the development of suicidality, the numerous
familial risk factors which are associated with adolescent suicidality frequently impact a
“common developmental pathway” (Adam, 1994, p. 276), producing common attachment
patterns and their associated internal working models and dynamic characteristics.

As discussed previously, a history of separation from parents as well as a history
of parental completed suicide or suicide attempts consistently is associated with an
increased risk of suicidality in adolescents (Henry et al., 1993; Kienhorst, Wolters,

Diekstra, & Otte, 1987; Krarup et al., 1991). While this frequently has been understood

24

within a behavioral learning theory context, this also could be understood from an
attachment theory perspective. Parental suicidality could adversely impact the formation
of a secure attachment to caregivers, as the act of parental suicide may be the ultimate
form of abandonment.

Similarly, not only is the level of parental psychopathology present in the families
of suicidal adolescents more severe than those of control groups, including non-suicidal
depressed patients, but the onset of psychiatric illness occurs significantly earlier in the
parents of suicidal adolescents (e.g., Brent et al., 1990;1Kashani et al., 1991). Parental
psychopathology can significantly impact parental responsiveness and availability, both
of which are the basis for the formation of internal working models in attachment theory
(Ainsworth et al., 1978; Bowlby, 1977, 1988). Therefore, the relationship between severe
parental psychopathology, often present from an early age, and adolescent suicidality
may result from the formation of an insecure attachment to caregivers.

Additionally, the family environments of suicidal adolescents are significantly
more likely to be characterized by maternal emotional coldness, negativism, and abuse as
well as by a general family climate of conflict, anomie, hostility and rejection (e.g.,
Curran, 1987; Kosky et al., 1986). In addition, the way in which adolescents view their
family environment plays an important role in their development of suicidal tendencies.
Research has found that an individual’s perception of family stress and lack of support
can be influential in the etiology of suicidality (e.g., Asamow, 1992; Marttunen, Aro, &
Lonnqvist, 1993). Again, parental responsiveness characterized by qualities such as those
described above, including rejection and emotional coldness, are associated with the

formation of insecure attachment styles in children.

25

In fact, suicidality in childhood and adolescence is most commonly preceded by
significant losses (Lester, 1994), such as separation fiom parents, death of a grandparent
(Cohen-Sandler, et al., 1982), or loss of a significant family member or fiiend (Henry et
al., 1993). As attachment styles in children are associated with distinct and often
dramatic reactions to separation fiom attachment figures (Ainsworth et al., 1978), perhaps
attachment provides the mechanism through which significant losses during adolescence
result in suicidality for some adolescents but not for others, with the manifestation of
adolescent suicidality in response to a significant loss paralleling the separation reactions

of insecure children (Adam, 1994).

26

ATTACHMENT AND PSYCHOLOGICAL DISTRESS

Not only has a relationship between parenting and attachment styles in children
been established, but attachment styles consistently have been related to psychological
distress later in life. The following discussion focuses on the adverse effects of insecure
attachment styles throughout the lifespan. Specifically, the relationship between
attachment styles and general psychological distress during the latter life stages of
adolescence and adulthood is discussed in order to provide a foundation from which
attachment theory can be applied to the more specific topic of adolescent suicidality.

Although research investigating the relationship between attachment and
psychological distress in adolescents and adults is a relatively new area in the attachment
literature, associations between insecure attachment styles and problems such as
depression, low self-esteem, eating disorders, and suicidality recently have been
established (Camelley et al., 1994; Hammen et al., 1995; Heesacker & Neimeyer, 1990;
long, 1992; Kobak & Sceery, 1988; Papini & Roggrnan, 1992). Beginning with the
development of the AAI by Main and her colleagues, a means by which attachment could
be measured in older individuals became available, and classifications paralleling those
identified by the behavior of children in Ainsworth’s Strange Situation task were created
(Ainsworth et al., 1978; Main et al., 1985). The adult attachment classifications of
secure, dismissing, and preoccupied attachments are found in proportions comparable to
those found for the corresponding classifications in children (secure, avoidant, and
ambivalent, respectively; Main et al., 1985).

Using the AAI, Kobak and Sceery (1988) conducted a pioneering study which

27

investigated the relationship between attachment classifications, affect regulation, and
representations of self and others in late adolescence. This study utilized a sample of
first-year college students and provided an important addition to the attachment literature
due to both the lack of research on these issues in late adolescence and the saliency of
separation-individuation issues during the transition to college. This study did not rely
solely on the subject’s self-report but also included external sources of information from
the subjects’ fiiends.

Kobak and Sceery (1988) found significant relationships between attachment
classifications, styles of affect regulation, and representational bias. Those adolescents
who were classified as securely attached evidenced the best adjustment on all measures,
with peer reports of high ego-resiliency and low levels of hostility and anxiety, as well as
self-reports of few symptoms of distress and high levels of social competence and
support. Peers evaluated the dismissing group as more hostile than the preoccupied
group, with the converse for ratings of anxiety. Furthermore, the dismissing group
reported higher levels of loneliness and lower levels of social support than the
preoccupied group, along with levels of perceived social competence and distress
comparable to the secure group.

However, despite self-reports of low levels of negative affect, peer ratings of the
dismissing group revealed low levels of ego-resilience and high levels of anxiety and
hostility. Kobak and Sceery (1988) interpret these findings as evidence for different
styles of affect regulation resulting from attachment status. They pr0pose that the
discrepancy found between peer and self-ratings in the dismissing group could be

interpreted as a bias towards avoiding negative affect. The lower levels of perceived

28

competence and higher levels of anxiety and distress found in the preoccupied group
distinguished it from the other groups and were interpreted as tentatively providing
support for this group’s tendency to form dependent relationships in an unsuccessfirl
attempt to alleviate anxiety (Kobak & Sceery, 1988).

Relationships between attachment and psychological distress similar to those
identified by Kobak and Sceery (1988) have been found in samples of 12-year old early
adolescents (Papini & Roggman, 1992), as well as in older samples (Camelley et al.,
1994; Hammen et al., 1995). In a sample of women, Camelley et al. (1994) found that
attachment style was the best predictor of relationship functioning and that the insecure
attachment characteristics of fearful avoidance and greater preoccupation were associated
with depression. They argue that perhaps a sample of individuals exhibiting different
symptoms, such as antisocial behavior, might evidence different models of self and
others, with a predorrrinantly dismissing (positive view of self and negative view of
others) rather than fearful (negative view of self and others) avoidant style (Camelley et
aL,1994)

Hammen et al. (1995) found that interpersonal attachment cognitions in a sample
of high school seniors were predictive of depression in addition to general psychological
symptomatology. Attachment styles characterized by less trust in other’s dependability
(depend) and more anxiety resulting from fears of abandonment (anxiety) were
significantly related to changes over a one year period in the level of depression and
general psychopathology, including disorders such as substance abuse, eating disorders,
and antisocial personality disorder. Also, the attachment characteristics of anxiety and

comfort with closeness (close) both moderated the effects of interpersonal stress on

29

general psychological outcome, with those individuals characterized by insecure
attachment (high anxiety or low close) evidencing the greatest symptomatology upon
exposure to stress. Hammen et al. (1995) argue that insecure attachment cognitions are
not associated with specific symptomatology but rather with general dysfunction.
However, while the authors concluded that there does not exist a relationship between
specific attachment cognitions and symptomatology (Hammen et al., 1995), the variables
utilized to represent attachment cognitions in this study did not access models of self
adequately. In fact, the results of this study seem to lend support to the hypothesis that
attachment cognitions characterized by negative views of others are associated with
psychopathology as well as to leave open the possibility that the simultaneous
consideration of views of self may contribute important additional information.

Not only has an insecure attachment been associated with an increased risk of
general psychopathology, but researchers have also hypothesized that a secure attachment
can serve a protective firnction, particularly during exposure to stress (Papini &
Roggman, 1992). Papini and Roggrnan (1992) found that a secure attachment to parents
served a protective function in the lives of adolescents, as it was positively correlated
with perceptions of competence and negatively related to feelings of depression and
anxiety.

The mechanisms by which attachment status may influence levels of
psychological distress also have been investigated (e. g., Kenny et al., 1993; Roberts et al.,
1996). Kenny et al. (1993) found that self-concept significantly mediated the relationship
between attachment and depression in a sample of early adolescents, providing support

for the role of Bowlby’s internal working models of self in the development of

30

psychopathology. In addition, significant gender differences were found in the
magnitude of the relationship between variables, with parental attachment accounting for
48% of the variance in self concept in boys but only 28% of the variance in this construct
for girls.

Similarly, Roberts et al. (1996) investigated the mediating role of self-esteem in
the relationship between attachment and depression in a college—age sample. While adult
attachment was associated with the severity of depressive symptomatology, the
anxious/ambivalent subscale provided the only unique contribution to the prediction of
depression, and the attachment measure only explained 12% of the variance in
symptomatology. When maladaptive contingencies of worth and low self-esteem were
tested as mediators of this relationship, Roberts et al. (1996) found that these self-concept
variables ahnost entirely accounted for the effect of attachment on depression. They
report that insecure attachment is related to dysfunctional attitudes (e. g., “If I fail at my
work, then I am a failure as a person” and “I do not need the approval of other people in
order to be happy” (p. 314) which contribute to low self-esteem and result in depression.
However, in light of evidence regarding the relationship between attachment
classifications, views of self, and styles of affect regulation (i.e., Kobak & Sceery, 1988),
a significant weakness of both Roberts et al.’s (1996) and Kenny et al.’s (1993) studies
was that only the mediating role of views of self was investigated, leaving the possible

mechanisms of views of others and the resulting dynamic characteristics unexarnined.

31

ATTACHMENT AND SUICIDALITY

Overall, the attachment literature provides an impressive amount of empirical
evidence supporting Bowlby’s assertion that early parent-child relationships form the
basis of internal working models of self and others which impact individual’s functioning
throughout the lifespan. Insecure attachment styles are related to problems such as
depression, low self-esteem, and eating disorders (Camelley et al., 1994; Hammen et al.,
1995; Heesacker & Neimeyer, 1990; Kobak & Sceery, 1988; Papini & Roggman, 1992).
In addition, attachment classifications are associated with styles of affect regulation,
representational bias, and relationship functioning (Camelley et al., 1994; Hammen et al.,
1995 ; Kobak & Sceery, 1988).

The attachment literature only recently has begun to address the potential
relationship between attachment and suicidality (Adam et al., 1996; Armsden, McCauley,
Greenberg, Burke, & Mitchell, 1990; long, 1994; Van der Kolk et al., 1991). However,
the results of the few existing studies are promising. Armsden et al. (1990) found
significantly less secure attachments to parents in a group of depressed adolescents than
in control groups of non-depressed psychiatric adolescents or nonpsychiatric children of
mothers with a chronic illness. Both clinical groups evidenced less secure peer
attachments than the nonpsychiatric groups, and insecure parental attachments were
associated with more maladaptive attributional styles, the presence of separation anxiety
disorder, and a history of suicidal ideation.

long (1992) more directly investigated the relationship between an insecure

attachment style and suicidality and discovered that undergraduates with a history of

32

suicidality exhibited lower levels of attachment security and less individuation from their
parents than either a group of depressed undergraduates with no history of suicidality or a
control group. This finding is consistent with Blass and Blatt’s (1992) proposal that the
development of attachment and separateness is intertwined. Furthermore, those students
with a history of suicidality rated their parents as significantly more emotionally absent in
childhood than the other two groups. Measures of peer attachment and individuation did
not differ between the three groups. Similarly, Van der Kolk et al. (1991) also found that
adolescent suicidality was associated with a childhood history of trauma, disrupted
parental care, and abandonment as well as a lack of secure attachments.

The study that most closely addresses the relationship between attachment
classifications and adolescent suicidality was conducted by Adam et al. (1996). Adam
(1994) has argued that the attachment paradigm is a usefirl framework for understanding
the apparent relationship between disturbances in the parent-child relationship early in
life and the later development of suicidality, as it accounts for the fact that “events are not
the same as experiences” (Adam, 1994, p. 288). Adam et al. (1996) found that the most
common attachment pattern associated with a history of suicidality was unresolved-
disorgarrized with preoccupied attachment. This attachment style is characterized by an
inability to maintain coherent and logical descriptions of past, potentially traumatic
experiences, including loss and abuse, indicating that the experience has remained
unresolved (Adam et al., 1996). A classification of preoccupied attachment is given
when descriptions of relationships with parents reveal overwhelming preoccupation with
these relationships or indicate that these relationships are experienced in an angry and

unobjective manner or a passive and helpless manner (Adam et al., 1996).

33

Furthermore, among adolescents receiving psychiatric treatment who did not have
a history of suicidality, a dismissing attachment was most characteristic of the male
adolescents, and an autonomous attachment was most characteristic of the female
adolescents. A dismissing attachment style is one in which the parents are either
idealized or referred to in a derogatory manner, while childhood memories are reportedly
forgotten and personal normalcy is emphasized (Adam et al., 1996). An autonomous
attachment is a secure attachment in which coherent, logical descriptions of early
childhood memories are given regarding relationships With parents and potential traumas
(Adam et al., 1996).

The demographic variables of age and sex significantly interacted with the
attachment classifications of unresolved-disorganized and preoccupied attachment (Adam
et al., 1996). Adolescent females over the age of 15 who were characterized by an
unresolved.disorganized attachment were significantly more likely to have a history of
suicidality than were adolescents with other demographic or attachment characteristics
(Adam et al., 1996). In addition, an attachment style characterized by a combination of
unresolved-disorganized attachment and preoccupied attachment significantly increased
the likelihood of having a history of suicidality, while a dismissing attachment style
significantly decreased the likelihood of having a history of suicidality (Adam et al.,
1996).

Although those adolescents in the case group did report more symptomatology
than did those in the comparison group, a specific association between suicidality and
either externalizing or internalizing symptoms did not exist (Adam et al., 1996). It

appears that specific attachment styles may be more characteristic of suicidal adolescents

34

than are specific psychopathologies, such as depression. As stated earlier, previous
research has found that while depression is commonly associated with adolescent
suicidality, many depressed adolescents do not experience suicidality, and many suicidal
adolescents are characterized by externalizing rather than internalizing symptomatology
(F eldman & Wilson, 1997; Khan, 1987; Spirito, 1989; Taylor & Stansfeld, 1984).

Perhaps of greatest significance is the finding by Adam et al. (1996) that no
significant difference in the occurrence of attachment-related trauma, such as loss, abuse,
and separations, existed based upon a history of suicidality. This seems to provide
empirical evidence to support Adam’s (1994) statement that “events are not the same as
experiences” (p. 288). It seems that individual differences in the way in which events
occurring in the environment are understood or conceptualized may be particularly
influential in the etiology of suicidality. Adam et al. (1996) argue that a model which
accounts for the way in which traumatic events have been understood and resolved will
provide more specific information regarding the relationship between trauma and
psychopathology than can be accounted for in a direct effects model which merely
documents the occurrence of a traumatic event.

While this study was an important first step towards establishing the existence of
a relationship between attachment and adolescent suicidality, it is not surprising that
attachment classifications alone did not account entirely for the development of
suicidality (i.e., Kazdin & Kagan, 1994). Adam et al. (1996) found that an unresolved-
disorganized, preoccupied attachment was strongly associated with suicidality, with 77%
of all adolescents with this attachment style having a history of suicidality. However, this

attachment style was neither necessary nor sufficient for the presence of suicidality, as

35

59% of suicidal males and 70% of suicidal females were not characterized by an
unresolved-disorganized, preoccupied attachment, and 33% of adolescent subjects with
this attachment style did not have a history of suicidality. It can be seen that a model of
the etiology of adolescent suicidality needs to consider additional factors in order to
account for this phenomenon more accurately.

_F_actors Mected by Past Studies Investigging Attachment land Suicidiity

Although the results of the few studies investigating the relationship between
attachment and suicidality are encouraging regarding the importance of this relationship,
these studies not only fail to account for the internal working models hypothesized by
attachment theory but also continue to suffer from many of the previously discussed
methodological weaknesses, including insufficient consideration of age-salient
developmental tasks (See Figure 3 for a sample model). The current investigation
proposes that constructs in addition to attachment classifications need to be considered
when addressing the etiology of suicidality during the developmental stage of
adolescence.
Role of Internlal Models and Resulting Dynamic Constructs

The review of the literature regarding the relationship between attachment and
psychopathology revealed that internal working models of self and others play an
important role in the relationship between attachment status and outcomes (e. g., Kenny et
al., 1993). However, despite evidence suggesting that internal working models may
provide the mechanisms by which attachment status influences levels of psychological
distress (e.g., Kenny et al., 1993; Roberts et al., 1996), studies that have investigated

adolescent suicidality from an attachment theory perspective have not adequately

36

incorporated views of self and others (e.g., Adam et al., 1996). Although avoidant
attachment must be broken down into two distinct categories in order to account for
internal working models of self and others (Bartholomew, 1990; Camelley et al., 1994),
this attachment classification typically has been treated as one construct in the adolescent
suicide literature.

Similarly, studies which do not account for internal working models of self and
others preclude an adequate investigation of the potential mediating role of interpersonal
anxiety and avoidance or negative affect. Although levels of either anxiety or avoidance
have been related to general symptomatology (e. g., Hammen et al., 1995; Kobak &
Sceery, 1988; Pappini & Roggman, 1992), the relationship between the combined effect
of these constructs and symptomatology such as suicidality has not been thoroughly
investigated. For example, Pappini and Roggrnan (1992) found that a secure attachment
was negatively related to anxiety, but this study did not assess avoidance. While Kobak
and Sceery (1988) not only found that different styles of affect regulation resulted from
attachment status but also identified a seemingly unconscious bias toward avoiding
negative affect, this study did not directly assess avoidance of negative affect but rather
inferred the existence of these characteristic from discrepancies found between peer and
self-ratings.

Furthermore, the two-axis model has not been applied directly to the problem of
suicidality in an empirical study, although it has been proposed as a useful fi'amework for
conceptualizing suicidality (Kaplan & Worth, 1993; see Figure 4). While it appears that
suicidality may be related to certain quadrants of this model but not to others, this has not

been directly investigated. However, the few existing studies that have explored the

37

relationship between suicidality and attachment status in adolescents are consistent with
this hypothesis (e. g., long, 1992). Suicidal behavior in adolescents has been associated
with a preoccupied attachment (quadrant 2); a dismissing attachment (quadrant 3) has
been associated with an inpatient comparison group who did not have a history of
suicidality, but were likely exhibiting significant externalizing behavior problems (Adam
et al., 1996).

Additionally, different quadrants of this model may account for different
subgroups of suicidal adolescents. As mentioned previously, Feldman and Wilson (1997)
suggest that instead of using specific risk factors, such as depression, structural
personality variables should be utilized to identify adolescents who are at-risk for
suicidality. Subgroups of suicidal adolescents have been identified based upon
characteristics such as reactivity to separation experiences (F eldman & Wilson, 1997),
personality disorders (Brent et al., 1994), and coping styles (Khan, 1987), as well as
personality structure and the resulting symptomatology exhibited in addition to
suicidality (Borst & Noam, 1993; Feldman & Wilson, 1997; Spirito et al., 1989). For
example, Van der Kolk (1991) found that self-destructive behavior has been related to
two psychological conflicts, “one centering on separation-abandonment and the other on
experiencing and expressing anger and emotional needs” (p. 1670). Similarly, two
distinct groups of suicidal inpatient adolescent females were identified by Borst and
Noam (1993). The “angry-defiant” group presented with depression, aggression, and
externalizing defense mechanisms, as well as a lower level of ego development. The
other group, labeled “self-blaming,” presented with depression and internalizing defense

mechanisms and were characterized by a higher level of ego development (Borst &

38

Noam, 1993).

Overall, studies investigating the relationship between attachment and suicidality
have not adequately accounted for the potential role either of internal working models of
self and others or of the resulting dynamic characteristics of avoidance of negative affect
and interpersonal anxiety. The current investigation proposes that additional predictive
ability would have been provided by further differentiating those individuals within the
avoidant attachment category based upon positive and negative models of self and others,
as well as by accounting for the role of internal working models and the resulting levels
of avoidance and anxiety.

Developmental_T_a_§k of Sewn/Individuation

Differences in precipitating events for suicide exist for different age groups,
perhaps resulting from the developmental stages characteristic of different ages (Lester,
1994). Because important differences exist between adolescents and adults, the variables
that distinguish a suicidal adolescent, in particular, must be investigated (Rotheram-Borus
& Trautrnan, 1988). However, while the literature suggests that suicidality during the
developmental stage of adolescence should be treated as a somewhat distinct
phenomenon, this problem typically has not been considered within a developmental
context.

Adolescence is the developmental period during which the second separation-
individuation process occurs (Bloom, 1980; BIOS, 1979; Hoffinan, 1984; Meeus, Helsen
& Vollebergh, 1996). This phase of life is a time of “radical disengagement” (Meeus et
al., 1996, p. 103), as adolescents strive for autonomy from parents. Developmental

theory has long considered adolescence to be one of the major stages in psychosocial

39

development, involving the struggle for identity versus identify diffusion (Erikson, 1959).
The successful achievement of a sense of ego identity is essential to healthy development.
In fact, Erikson (1959) considered a sense of ego identity to be necessary in order for
human beings to feel alive.

The struggle for individuality and self-identity during adolescence occurs within
an interpersonal context, and identity and intimacy are unavoidably intertwined (Damon,
1983). This period of life is a time when a radical new understanding of self is formed;
the self is separated from others while a connection to them is maintained, with the
identity crisis typically being solved in our culture between 18 and 22 years of age
(Damon, 1983). More recently, the task of separation/individuation has been
conceptualized as a multidimensional construct (Hoffinan, 1984; Hoffman & Weiss,
1987; Lopez, Campbell, & Watkins, 1986). Hoffman (1984) argues that adolescent
separation/individuation consists of four components of independence from parents:
functional, attitudinal, emotional, and conflictual.

Although at first glance, attachment and separation/individuation seem to embody
contradictory goals, models addressing the importance of both attachment and separation
have been discussed by several theorists (e.g., Blass & Blatt, 1992; Bowlby, 1988; Lester,
1991; Schultheiss & Blustein, 1994). Overall, the aims of attachment and separateness
are considered to be equally important and compatible goals. In fact, the second
separation-individuation process that occurs during adolescence has been reformulated to
address the dynamic interplay between attachment and separation/individuation (e.g.,
Grotevant & Cooper, 1985, 1986; Meeus et al., 1996). Not only does attachment to the

primary caregiver fi'om childhood remain important into adolescence, but the sense of

40

self at this stage in development is tied to attachment security (Salzman, 1996). Children
who have connected well with their parents are “most able to disengage from the parents
and most able to develop into independent individuals (individuation)” (Meeus et al.,
1996, p. 104). The rebellion that is characteristic of adolescence results from the
necessary process of detaching from parents, and a context of closeness to parents allows
this rebellious behavior and the resulting “growth toward personal independence and
autonomy” (Damon, 1983, p. 310).

Preliminary empirical evidence exists supporting the presumed relationship
between attachment and resolution of the task of separation/individuation during
adolescence (Allen et al., 1994; Kroger & Haslett, 1988; Salzman, 1996).
Separation/individuation during adolescence is significantly related to security of
attachment to the primary caregiver from childhood (Salzman, 1996), and researchers
have concluded that “structuralization of the ego evolves through the process of
attachment” (Kroger & Haslett, 1988, p. 76).

A more direct investigation of the relationship between attachment and separation
during adolescence was conducted by Allen et al. (1994), who asserted that the most
important and salient developmental task during this stage is the achievement of
separation and identity formation while simultaneously remaining positively connected to
parents. This study revealed the following:

“The exhibition of autonomous-relatedness in family interactions was strongly

related to both concurrent ego development and self-esteem in

adolescence...These findings are consistent with predictions made by attachment

theory for other stages of development in that they suggest that, in optimal

41

interactions, the parent facilitates the adolescent’s exploration of differences with

the parent (e. g., autonomy) from the secure base of a positive relationship” (p.

1 90).

Allen et al. (1994) also report that the degree of autonomy and relatedness present in the
parent-child relationship appears to be related to the models of self and other which
adolescents develop as well as to the task of identity exploration. In this way, the views
of self and others which result from early attachment relationships are related to the
degree to which separation/individuation is achieved later in life. Despite a history of
having been placed at the opposite ends of a continuum, it appears that a “state of
‘autonomous-relatedness,’ a term coined by John Bowlby, is an optimal outcome for the
adolescent-parent relationship” (Allen et al., 1994, p. 179).

In addition, the adolescent’s resolution of the task of separation from parents has
important psychosocial consequences. Erikson (1968) argued that when the crisis of
identify formation results in identity confusion, psychopathological symptoms often
result. In fact, resolution of the task of separation/individuation is related to self-esteem,
emotional outlook on life, and personal and academic adjustment (e. g., Damon, 1983;
Hoffinan, 1984). Conflictual independence from parents, defined as “freedom from
excessive guilt, anxiety, mistrust, responsibility, inhibition, resentment, and anger in
relation to the mother and father,” has been related to adolescent psychological health
(Hoffman, 1984, p. 171-172; Hoffman & Weiss, 1987). Those adolescents who have not
yet achieved self and object constancy by working through the task of
separation/individuation often find college to be a significant and difficult transitional life

task (Edward, Ruskin, & Tunini, 1991).

42

Furthermore, inadequate resolution of the task of separation/individuation has
been associated with suicidality, although this relationship remains virtually unexamined
empirically. Erikson (1968) argued that identity confusion produces an inability to see
that time and activity can result in changes in one’s life, leading adolescents to give up
due to “despair...on the part of the ego to let itself die” (p. 151). Psychoanalytic theorists
have identified aspects of both the psychology of adolescence and the psychology of
suicide which coincide, namely the experiences of object loss, loneliness, hopelessness,
and helplessness (Kienhorst, de Wilde, & Diekstra, 1996). However, despite theoretical
arguments supporting the potential influence of the developmental challenge of
separation/individuation on suicidality during adolescence, there only exists one study
that has investigated this relationship. Meyer and Phillips (1990) found that parental
alcoholism influenced the development of adolescent suicidality by impeding
separation/individuation. The current investigation proposes that inadequate resolution of
the task of separation/individuation can be a stressor for adolescents and that the
inclusion of this developmental task in models of adolescent suicidality will provide
important additional information.

Interpersonal Stressors

As mentioned previously, the most common precipitants associated with suicide
during adolescence are interpersonal conflict and loss, such as the loss of a significant
family member or fiiend (Henry et al., 1993; Lester, 1994). However, models of
adolescent suicidality often fail to incorporate the impact of this factor (e. g., Adam et al.,
1996; J ong, 1992). The current investigation proposes that the inclusion of interpersonal

stressors in a model of adolescent suicidality will provide important additional

43

information.
Depression

As mentioned previously, depression is associated with an increased risk for
adolescent suicidality, although it is neither necessary nor sufficient for the development
of adolescent suicidality. Of particular interest to the current discussion is that in addition
to the relationship between depression and suicidality which has been established, the
development of depression seems to follow many of the same pathways outlined above
regarding the development of suicidality (e. g., Camelley et al., 1994; Hammen et al.,
1995). These relationships have been addressed by much empirical research, and an
extensive review of this literature is beyond the scope of the current investigation.
However, the relationship between these constructs and adolescent depression will be
discussed briefly.

A substantial body of empirical evidence exists supporting a relationship between
family-related risk factors and adolescent depression (e. g., Kaufman, 1991; McCauley &
Meyers). Birmaher, Ryan, Williamson, & Brent (1996) conducted a thorough review of
the literature regarding childhood and adolescent depression and reported that familial
factors, such as parental psychopathology, poor maternal functioning, and family conflict,
are associated with adolescent depression. Adolescent depression also is associated with
a negative cognitive style, including negative views of self and negative attributions
(Binnaher et al., 1996). However, similar to the adolescent suicide literature, the
mechanisms underlying these relationships remain unclear (Birmaher et al., 1996).

Similar to what has been addressed above regarding adolescent suicidality,

evidence also exists supporting relationships between attachment, views of self and

44

others, and depression. As mentioned previously, there exists an association between
insecure attachment styles and depression (e. g., Camelley et al., 1994; Hammen et al.,
1995; Pappini & Roggman, 1992). Similarly, inadequate resolution of the task of
separation/individuation is related to aspects of adolescent psychological health,
including depression (e.g., Allen et al., 1994; Hoffinan & Weiss, 1987).

Furthermore, the two-axis model of attachment and individuation proposed by
Kaplan (1988), as well as the resulting dynamic characteristics discussed above, also are
consistent with the development of depression. For example, the characteristics of less
trust in other’s dependability and more anxiety resulting from fears of abandonment, have
been related to depression (e.g., Hammen et al., 1995). Blass and Blatt (1992) have
proposed two subtypes of depression: anaclitic depression, where depressed individuals
primarily focus on issues of relatedness and dependency; and introjective depression,
where depressed individuals are self-critical and primarily focus on issues of self-
definition. They argue that a primary focus on either relatedness or self-definition may
also result in “different coping or defensive styles” (Blass & Blatt, 1992, p. 421), such as
different levels of avoidance in response to negative affect. Blass and Blatt ( 1992) report
that a significant amount of empirical research has supported the distinction between
these two subtypes of depression.

Overall, it is anticipated that depression will be related to suicidality as the result
of similar developmental pathways. It is also anticipated that greater explanatory power
for the prediction of suicidality will be provided by a model which does not simply rely
upon associated factors, such as depression or family characteristics, but rather

incorporates developmental and dynamic constructs in order to address the mechanisms

45

underlying the development of both depression and suicidality during adolescence.

46

RATIONALE

The application of attachment theory to the problem of adolescent suicidality
addresses many of the weaknesses of prior research by providing potential mechanisms
by which associated family-related risk factors may impact the development of
adolescent suicidality. Although evidence exists supporting a relationship between
insecure attachment styles and psychological distress later in life (e.g., Camelley et al.,
1994; Kobak & Sceery, 1988), few studies have investigated the relationship between
attachment and suicidality. Furthermore, while the results of the few existing studies
provide preliminary support for a relationship between attachment and suicidality, these
studies not only fail to account for the internal working models hypothesized by
attachment theory but also continue to suffer from many of the previously discussed
weaknesses of developmental psychopathology research.

The current investigation proposes that constructs in addition to attachment
classifications need to be considered when addressing the etiology of suicidality during
the developmental stage of adolescence and that incorporation of views of self and others,
the dynamic characteristics of interpersonal anxiety and avoidance of negative affect, and
the developmental task of separation/individuation will provide a more comprehensive
model of adolescent suicidality. In addition, the current investigation utilizes a two-axis
model of human development proposed by Kaplan (1988) as a heuristic for understanding
the levels of interpersonal anxiety and avoidance of negative affect which result from
combinations of positive/negative views of self/others (Shaver & Clark, 1996). It is

proposed that a model incorporating underlying mechanisms as well as the salient

47

developmental task of separation/individuation will more accurately account for the

development of adolescent suicidality than have previous models.

48

HYPOTHESES

The overall model proposed in Figure l is founded in attachment theory and
incorporates potential mechanisms, as well as the developmental task of
separation/individuation, in order to account for adolescent suicidality. This model
proposes that early family risk factors directly impact the formation of an early insecure
attachment style (Path A), which is related to views of self and others as delineated by
Bartholomew (1990) (Paths B and C).

The views of self and others that result directly affect the dynamic characteristics
of avoidance of negative affect (Paths F and H) and interpersonal anxiety (Paths G and I;
Shaver & Clark, 1996; see Figure 4), as well as resolution of the stage salient task of
separation/individuation (Paths D and E; e.g., Allen et al., 1994; Kroger & Haslett, 1988;
Salzman, 1996). Resolution of the task of separation/individuation is then directly related
to both depression (Path J) and suicidality (Path K), with inadequate resolution
functioning as a stressor during adolescence. Similarly, levels of avoidance and anxiety
are directly related to both depression (Paths L and N) and suicidality (Paths M and O;
e.g., Borst & Noam, 1993; Van der Kolk et al., 1991).

In addition, depression is directly related to suicidality (Path P). Interpersonal
stressors are also directly related to suicidality (Path Q). The following hypotheses more
specifically delineate the individual relationships among variables which will be tested.
Hypothesis 1

A direct, significant positive relationship is expected between early family risk

factors and insecure attachment (Path A).

49

Hypothesis 2

Attachment style is hypothesized to have direct, significant associations with
views of self, as delineated by Bartholomew (1990; Path B). More specifically, a direct,
significant positive relationship is predicted between both secure and dismissing
attachment styles and positive views of self. A direct, significant negative relationship is
predicted between both preoccupied and fearful attachment styles and positive views of
self (See Figure 4).

Hypothesis 3

Attachment style is hypothesized to have direct, significant associations with
views of others, as delineated by Bartholomew (1990; Path C). More specifically, a
direct, significant positive relationship is predicted between both secure and preoccupied
attachment styles and positive views of others. A direct, significant negative relationship
is predicted between both dismissing and fearful attachment styles and positive views of
others (See Figure 4).

Hypothesis 4

A direct, significant positive relationship is hypothesized between positive views
of self and separation/individuation (Path D).

Hypothesis 5

A direct, significant positive relationship is hypothesized between positive views
of others and separation/individuation (Path E).

Hypothesis 6
Positive views of self and others are hypothesized to have direct, significant

negative relationships with avoidance of negative affect (Paths F and H, respectively).

50

Hypothesis 7

Positive views of self and others are hypothesized to have direct, significant
negative relationships with interpersonal anxiety (Paths G and 1, respectively).
Hypothesis 8

A direct, significant negative relationship is expected between
separation/individuation and depression (Path J).

Hypothesis 9

A direct, significant negative relationship is expected between
separation/individuation and suicidality (Path K).

Hypothesis 10

Avoidance of negative affect is predicted to have a direct, significant negative
association with depression (Path L).

Hypothesis 11

Interpersonal anxiety is predicted to have a direct, significant positive association
with depression (Path N).

Hypothesis 12
A direct, significant positive relationship is predicted between avoidance of

negative affect and suicidality (Path M).

Hypothesis 13

A direct, significant positive relationship is predicted between interpersonal
anxiety and suicidality (Path 0).

Hypothesis l4

Depression is hypothesized to have a direct, significant positive relationship with

51

suicidality (Path P).
Hypothesis 15

A direct, significant positive relationship is predicted between interpersonal

stressors and suicidality (Path Q).

Hypothesis 16

The model proposed in Figure l is hypothesized to be a better fit than the model
previously utilized to explain adolescent suicidality represented by Figure 2.

Hypothesis 17

The model proposed in Figure 1 is hypothesized to be a better fit than the
attachment theory model previously utilized to explain adolescent suicidality represented

by Figure 3.

52

METHODS

Participants

Subjects in the current investigation were undergraduate students at a large
midwestern university. With a final sample size of 354, adequate power was achieved to
investigate the model proposed in Figure l (Schurnacker & Lomax, 1996). Students were
recruited from undergraduate psychology courses in which they received class credit for
completing the questionnaires for this study.

Measures

The following measures were administered to undergraduate participants (See
Appendix C for copies of all measures).

Demogaphic Information

Brief Demogzaphic Ouestionnfl. A brief demographic questionnaire designed
by the investigator was utilized to obtain information regarding the undergraduate’s age,
gender, year in college, racial/ethnic background, parents’ marital status, parents’
educational background, and parents’ occupations.

Early Risk Factors

Family Histogg Questionnaire. A brief family history questionnaire designed by

the investigator was utilized to obtain information regarding family-related risk factors,
such as history of parental psychopathology and physical abuse. This questionnaire was
based on aspects of the family environment which the literature consistently has revealed
can adversely affect attachment (e.g., Cassidy & Berlin, 1994). This measure was

utilized to assess family risk. A total score was obtained for each of the following

53

subscales by summing the subscale items: parental psychopathology, absence of parental
warmth, and family violence.
Perception of Adult Attachment Questionnaire (PAAO; Lichenstein & Cassidy,

_19_91_)_. This questionnaire obtains information regarding participants’ early childhood
relationships with the primary caregiver. Two subscales from this measure, “rejected”
and “loved,” were utilized. These subscales have been found to have adequate reliability
and validity (Lichtenstein & Cassidy, 1991). Intemal consistency coefficient alphas of
.87 have been reported for both subscales. In addition,'test-retest correlations of .81 for
the “rejected” subscale and .86 for “loved” subscale have been found over a three week
period (Lichtenstein & Cassidy, 1991). This questionnaire contains sixty items that are
rated on a five-point Likert scale (1=strongly disagree; 5=strongly agree). Specified
items are reverse scored, and the subscale items are summed to obtain subscale scores.
Examples of items from the “rejected” subscale of this measure include: “When I was a
child my mother sometimes told me that if I was not good she would stop loving me” and
“ In childhood I knew I was low on my mother's priority list.” Examples of items from
the “loved” subscale of this measure include: “In childhood I felt like I was really
treasured by my mother” and “I remember when I was frightened as a child my mother
holding me close.”

Attachment Status

Close Relationships Questionnaire (Bartholomew, 1990; Bartholomew &

Horowitz, 1991). This questionnaire assesses attachment style utilizing Bartholomew’s
(1990) four attachment categories, which have been discussed above: secure, fearful,

preoccupied, and disrrrissing. This questionnaire is comprised of two parts. In the first

54

part, four brief paragraph descriptions are provided describing different approaches to
relationships; the participants choose which paragraph best describes their general
approach to relationships. For example, the paragraph describing a fearful attachment
style read as follows: “I want to be completely emotionally intimate with others, but I
often find that others are reluctant to get as close as I would like. I am uncomfortable
being without close relationships, but I sometimes worry that others don’t value me as
much as I value them.” In the second part, the participants indicate, using a seven-point
Likert scale (1=not at all like me; 7: very much like me), the degree to which they agree
with each of four descriptions. This questionnaire is highly related to another commonly
utilized scale of attachment, Hazan and Shaver’s (1987) scale (Brennan, Shaver, &
Tobey, 1991), and has good test-retest reliability of specific styles in samples of college
students.

Views of Self and Other

Attachment Sgle Questionnaire (ASQ) (Feeney. Noller. & Hanrahg, 1994). This

questionnaire contains forty items which describe thoughts, emotions, and behaviors
relevant to relationships. Participants rate the degree to which they agree with items on a
six-point Likert scale, ranging from totally disagree (1) to totally agree (6). Examples of
items on this measure include: “I feel confident that other people will be there for me
when I need them” and “Other people often disappoint me.” This measure provides
information regarding both the three and four categories of adult attachment proposed by
researchers such as Hazan and Shaver (1987) and Bartholomew (1990) as well as “the
basic themes of infant attachment theory” (F eeney et al., 1994, p. 133). In addition, the

ASQ was designed to be used with adolescent and young adult samples and therefore

55

does not rely on romantic relationships to assess attachment characteristics. Factor
analysis has revealed the following dimensions of the ASQ: confidence (in others;
models of self and others), discomfort with closeness (model of others), need for approval
(model of self), preoccupation with relationships (model of self), and relationships as
secondary (model of others) (Feeney et al., 1994). Internal reliabilities ranging from .76
to .84 have been reported for the individual scales, and good test-retest reliabilities have
been reported (.67 to .78) with samples of undergraduate students (F eeney et al., 1994).
In addition, Feeney et al. (1994) report that the subscales of this measure are significantly
related to the theoretically equivalent attachment category specified by other attachment
measures as well as to family functioning and personality variables.

Rosenberg Self-Esteem Scale (RSE) (Rosenberg. 1965). This measure of self-
esteem was used as an indicator of participant’s views of self. This measure is a global
measure of self-esteem for which adequate reliability and validity have been documented.
Internal consistency coefficient alphas ranging from .77 to .88 have been reported. This
questionnaire contains ten items which are rated on a five-point Likert scale (1=strongly
disagree; 5=strongly agree). Specified items are reverse scored, and the items are
summed to obtain a total score. Examples of items from this measure include: “I feel I
do not have much to be proud of” and “ I certainly feel useless at times.”
Mtion/Individaa—tiop

Psychological Saparation Inventog (PSI) (Hoffman, 1984). The PSI is a 138-
item questionnaire which assesses participants’ ability to separate fi'om their parents
psychologically. This measure utilizes a five-point Likert scale to rate aspects of

relationships to parents (1=not at all true of me; 5=very true of me), with half of the items

56

pertaining to the relationship with the mother and half pertaining to the relationship with
the father. This study utilized the conflictual independence subscale of this measure
pertaining to the mother to assess an aspect of separation/individuation that has been
associated with adolescent psychological health. Conflictual independence pertains to
“freedom from excessive guilt, anxiety, mistrust, responsibility, inhibition, resentment,
and anger in relation to the mother and father” (i.e., Hoffman, 1984, p. 171-172; Hoffman
& Weiss, 1987). Examples of items from this subscale include: “When I don’t write my
mother often enough I feel guilty” and “I feel like I am'constantly at war with my
mother.” Good test-retest reliability (.96) as well as good internal consistency (reliability
alpha coefficient = .92) have been found for the conflictual independence subscale of this
measure. In order to score this measure, the total for each subscale is summed and this
number is subtracted from the total possible score for this subscale. Higher scores
correspond to lower levels of independence.
Avoidance of Negative Affect

Defense Mechanisms Inventory (DMD (Ilrilevich & Glesrer. 1986L The DMI
assesses the adaptiveness of defensive structures in nonclinical populations. This
measure consists of ten vignettes which require participants to imagine themselves in
difficult and frustrating situations. The following is an example vignette: “You are
waiting for the bus at the edge of the road. The streets are wet and muddy after the
previous night’s rain. A motorcycle sweeps through a puddle in front of you, splashing
your clothing with mud.” After each vignette, there are four subsections for which
participants indicate how they would react: their actual reaction, the nature of their

impulsive reaction or behavioral fantasy to the situation, their thoughts in regard to the

57

situation, and their affective experience/rationale for their feelings. For each of these
subsections, subjects select from five possible solutions (representing five major clusters
of defense mechanisms) the statement which is most representative (scored a 2) and least
representative (scored a 0) of how they would react. A score of one is assigned to the
remaining three responses in each subsection. Examples of possible solutions for the
example vignette provided above include: “I wonder if that biker splashed me on
purpose” (thought) and “Wipe that biker’s face in the mud” (impulsive
reaction/behavioral fantasy). The scores for the five defense mechanisms then are
summed for the ten vignettes, resulting in a score ranging from O to 20 for each of the
five defense mechanisms subscales (Crarner, 1988).

Although the DMI assesses five clusters of defense mechanisms, this study

utilized two subscales to assess avoidance of negative affect: reversal, which pertains to

 

the production of internal responses to the perception of the threat of internal conflict in
order to avoid experiencing anxiety, including denial, repression, and reaction formation;
and turning against the object, which pertains to the experience and/or expression of
aggression towards others in order to protect the self fiom consciously experiencing the
pain of either external threats or inner conflicts.

Adequate reliability and validity have been reported for the DMI (Crarner, 1988).
Test-retest and inter-item reliability coefficients of about .78 are reported for this measure
(Crarner, 198 8). In addition, various subscales of the DMI are significantly related to
psychiatric symptomatology, physical distress, and to defensive scales on other measures,
such as the Minnesota Multiphasic Personality Inventory (See Crarner, 1988 for a more

extensive review of the DMI research). The DMI was utilized as one assessment of

58

avoidance of negative affect.

Ways of Coping Checklist (WCCL) (Folkman & Lazarus 1980' Vitaliano Russo

 

Carr, Maiuro, & Becker, 1985 ). This 42-item measure assesses the ways in which
participants cope with stressfirl situations. Vitaliano et al. (1985) revised the original
questionnaire designed by Folkman and Lazarus (1980), resulting in a version of this
questionnaire which contains a smaller number of items and scales which are more stable,
overlap less, and have higher internal consistency. The WCCL utilizes a four-point
Likert scale to assess the degree to which participants Would use a range of cognitive and
behavioral strategies to cope with stressful situations (1: would not use; 4=would use a
great deal). The following scales were identified by Vitaliano et al. (1985): problem
focused, seeks social support, blamed self, wishful thinking, and avoidance. Good
internal consistencies are reported (ranging from .73 to .88) for the WCCL. In the present
investigation, the avoidance subscale was utilized. Examples of avoidant coping
responses include: “Tried to forget the whole thing” and “I refused to believe it had
happened.”
my

Me—Trait Anxiety Inventory. Form Y (STAI) (@ielberger. 1983). The STAI
(F orm Y) is a 40-item questionnaire which utilizes a four-point Likert scale (1=almost
never; 4=almost always) to assess the degree to which participants feel that descriptions
of anxiety symptoms pertain to them both at the present moment (state anxiety) and in
general (trait anxiety). Trait anxiety as conceptualized by the STAI pertains to “relatively
stable individual differences in anxiety-proneness,” and the trait anxiety scale accesses

“differences between people in the tendency to perceive a stressfirl situation as dangerous

59

or threatening and to respond to such situations with elevations in the intensity of their
state anxiety reactions” (Spielberger, 1983, p. 5). State anxiety is conceptualized by the
STAI as current feelings of apprehension, tension, nervousness, and worry. Spielberger
(1983) argues that stronger levels of trait anxiety increase the likelihood of greater state
anxiety in response to a threatening situation. Examples of items include: “I feel
strained” (assessing state anxiety) and “I worry too much over something that really
doesn’t matter” (assessing trait anxiety).

The STAI was developed for use with high school and college undergraduate
students (Spielberger, 1983) but has also been utilized extensively with clinical samples
in both research and clinical practice (Spielberger, 1983). Good test-retest reliability
coefficients for the trait anxiety subscales are reported for college student samples,
ranging from .73 to .86, with a mean reliability of .77. As the state anxiety subscale was
designed to assess anxiety at particular moments in time, relatively low stability
coefficients for this subscale were both expected and observed, ranging from .16 to .62,
with a mean reliability of .33. Due to the transitory nature of state anxiety, alpha
coefficients of internal consistency provide more meaningful information regarding
reliability, and good Cronbach alpha coefficients are reported for both the state (ranging
from .86 to .95) and trait (ranging from .89 to .91) anxiety subscales of the STAI
(Spielberger, 1983). The STAI correlates highly with other anxiety measures (ranging
from .73 to .85), and scores on the state anxiety subscale are sensitive to “different
degrees and kinds of stress” (Spielberger, 1983, p. 44). The STAI was utilized as one
assessment of participants’ anxiety. Specified items are reverse scored, and the items in

the state and trait subsections are summed to obtain total state and trait anxiety scores,

60

respectively.

m Anxiety Inventory (BAI) (Beck, Epstein, Brown, & Steer, 1988). The BAI
contains twenty-one items pertaining to a variety of anxiety symptoms, and participants
utilize a four-point Likert scale (O=not at all; 3=severely) to rate the degree to which they
have been bothered by each of these symptoms during the past week. The items on this
measure reflect a range of cognitive, somatic, and affective symptoms of anxiety and
were chosen in order to reduce overlap with symptoms of depression (Beck et al., 1988).
Examples of symptoms assessed by the BAI include: “Fear of the worst happening” and
“Feeling hot.” All items are then summed to obtain a score representing the intensity of
self-reported anxiety.

This measure has been normed on both clinical and nonclinical samples, including
a sample of undergraduate students, and good internal consistency reliabilities (ranging
from an alpha of .92 to .94) have been reported (Beck & Steer, 1993). Studies have also
revealed that the BAI correlates more highly with other measures of anxiety (F5 1) than
measures of depression (r=.25) (Beck et al., 1988).

Dapression

Beck Depression Inventory (BDI) (Beck. 1978). This self-report questionnaire
consists of twenty-one items that ask respondents to indicate the degree to which they are
currently experiencing a variety of cognitive, somatic, motivational, and behavioral
symptoms of depression, including crying easily, appetite disturbances, and sleep
disturbances (Beck & Steer, 1984). Example items include: “I cry more now than I used
to” and “I feel I have nothing to look forward to.” Four statements, representing a range

of symptom severity, are then presented in a Likert scale fashion for each item, and

61

participants indicate the degree to which each symptom has been experienced during the
past week (O=do not feel this way; 3=strongly feel this way). For example, the choices
for the first example item presented above are as follows: “O=I don’t cry any more than
usual, 1=I cry more now than I used to, 2=I cry all the time now, 3=I used to be able to
cry, but now I can’t cry even though I want to.”

The BDI is a valid and reliable test for depression, with a reported mean
coefficient alpha of .86 (Beck & Steer, 1984; see Beck, Steer, & Garbin, 1988 for a
review). In addition, the BDI is highly related to clinical ratings of depression. Specified
items are reverse scored, and the items are summed to obtain a total score.

Suicidalig

Adult Suicigal Ideation Questionnaire (ASIQ) (Remolds, 1991 ). The ASIQ
contains twenty-five items pertaining to a variety of symptoms of suicidal ideation, and
participants utilize a seven-point Likert scale (O=I never had this thought; 6=Almost
every day) to rate the degree to which they have been bothered by each of these
symptoms during the past month. Example items include: “I wished that I had never
been born” and “I thought about what to write in a suicide note.” This measure has been
normed on a sample of 2,000 adults, including psychiatric outpatients, normal adults, and
college students. The ASIQ is a valid measure of suicidal ideation and correlates
significantly with other measures of psychological distress as well as a history of past
suicide attempts (Reynolds, 1991). In addition, good internal consistency and test-retest
reliability coefficients have been reported (ranging from .85 to .97). The items on the
ASIQ are summed to obtain a total score.

Suicide Ideation Scale (SISIIRudd. 1989). This 10-item questionnaire includes

62

questions regarding the continuum of suicidality over the past year, including past
thoughts, plans, and attempts. A five-point Likert scale is utilized to rate “how often the
subject has felt or behaved that way during the past year” (1=never or none of the time;
5=always or a great many times.” Example items include: “I have told someone I want
to kill myself” and “I have come close to taking my own life.” Rudd (1989) reports that
the SIS was designed for use with undergraduate students and is a valid and reliable self-
report measure of suicidal ideation. Internal consistency alpha coefficients ranging from
.86 to .90 have been found for this measure. Items on the SIS are summed to obtain a
total score. In addition, the items on this questionnaire were reworded to assess the
degree to which participants have “felt or behaved that way” during their lifetime. In this
way, two scores were obtained, one assessing recent suicidal behavior (within the past
year; “SUICYR”) and one assessing past suicidal behavior (prior to the last year;
“SUICLIFE”).
Interpersonal Stressors

The Inventory of College Studentg’ Recent Life Experiences (ICSRLE) (Kohn,
,L_afreniere. & Gurevich. 1990). This measure contains thirty-five descriptions of
potentially stressful events. Participants rate the degree to which they have been affected
by each of these items during the past month on a five-point Likert scale (1=not at all part
of your life; 4=very much part of your life). Examples of items include: “Conflicts with
boyfiiend/girlfriend/spouse” and “Loneliness.” This measure was designed specifically
for use with an undergraduate student population and attempts to address weaknesses of

other hassles scales.

63

Good internal reliability (0t=.89) is reported, and the ICSRLE is significantly
correlated with other measures of stress such as the Perceived Stress Scale (r=.65). Seven
subscales with adequate alpha coefficients (ranging from .47 to .80) and acceptable
intercorrelations (ranging from .17 to .49) have been identified by factor analysis:
Developmental Challenge, Time Pressure, Academic Alienation, Romantic Problems,
Assorted Annoyances, General Social Mistreatrnent, and Friendship Problems (Kohn et
al., 1990). Items are summed to obtain a total hassles score, and subscale scores are
obtained by summing the items corresponding to each Subscale. This study utilized three
subscales of this measure to assess interpersonal stressors: Romantic Problems, General
Social Mistreatrnent, and Friendship Problems.

Norbegt Social Support Questionnaire (N SSOI (Norbeck. Lindsey. & Camp
1981. 1983). This questionnaire assesses the social support network of individuals. Only
the section of the NSSQ which obtains information regarding the recent loss of important
relationships, including the number of persons lost and how much support was provided
by the people lost, will be utilized in the current investigation. The following variables
are computed fiom the participant’s responses regarding loss: recent losses, loss quantity,
loss quality, and total loss quality variable. The questions regarding loss of support have
demonstrated adequate test-retest reliability (ranging from .71 to .83), and internal
consistency intercorrelations between similar constructs have ranged from .89 to .98. The
subset of questions regarding loss from the NSSQ was utilized to assess recent stress

related to the loss of significant others.

64

Procedures

Participants completed all instruments in a single group session lasting
approximately two hours. The author and/or a trained research assistant administered the
measures to the participants. All participants provided informed consent by signing and
returning an informed consent form prior to the completion of the questionnaire packet.
The answers to all questionnaires were placed on a computer scoring sheet identifiable
only by a subject code in order to protect student anonymity. Following the collection of
all measures, subjects were given an information sheet debriefing them regarding the
general purpose of this study, as well as a referral sheet with the phone numbers of local

mental health agencies and suicide hotlines.

65

RESULTS

The hypothesized relationships proposed in the model of the etiology of
adolescent suicidality (Figure 1) were examined through structural equation modeling
(SEM). SEM is a confirmatory statistical technique that is utilized to test a theory
(Tabachnick & Fidel], 1996). Tabachnick and Fidell (1996) argue that there are a number
of advantages to utilizing SEM for hypothesis testing, including the estimation and
elimination of measurement error as well as the complete and simultaneous testing of
complex variable and factor relationships.

Data Screening and Praparation

Less than two percent of the values for any variable were initially missing. A
function available through the statistical program AMOS that allows for the computation
of full information maximum likelihood estimates in the presence of missing data was
utilized to complete the analyses; maximum likelihood estimation utilizes the covariance
matrix in all analyses with computation of full information maximum likelihood
estimates (even in the presence of missing data) through the estimation of means and
intercepts (Arbuckle, 1997).

Prior to the analyses of the hypothesized relationships among variables, the
assumptions of multivariate normality and linearity were evaluated per the steps
suggested by Tabachnick and F idell (1996). The data was scanned for univaraite and
multivariate outliers, as well as kurtosis and skewness distribution qualities (Tabachnick
& Fidell, 1996). More specifically, standardized scores were computed and examined to

identify potential urrivariate outliers. Cases with z—scores greater than 3.29 were

66

considered potential outliers and were examined more closely (Tabachnick & Fidell,
1996). Outliers were identified among the distributions of five variables. Values
adjacent to the closest non-outlier value were assigned for each outlier in the order of
original ranking, therefore maintaining the rank order of subjects for each variable (Kline,
1998; Tachnick & F idell, 1996). Less than two percent of values were altered for any
variable (mompsi- 2 variables, depress- 2 variables, famviol- 2 variables, asiqtotl- 5
variables, and suicyr- 6 variables).

In addition, Mahalanobis distance was utilized to detect potential multivariate
outliers (Tabachnick & F idell, 1996). A regression analysis was run utilizing subject ID
as the dummy dependent variable, as Tabachnick and F idell (1996) point out that
multivariate outliers among independent variables are unaffected by the dependent
variable. The criterion utilized for multivariate outliers was Mahalanobis distance at
p<.001, utilizing Chi—square with 19 degrees of freedom (19 total variables). Six cases
that were identified as multivariate outliers were deleted from the data set, as Kline
(1998) argues that such a small percentage of the total sample can be dropped with “little
concern about loss of information” (p. 88).

Furthermore, the normality of the variables were examined through investigation
of kurtosis and skewness. All variables displayed adequate kurtosis and skewness as per
the guidelines specified by Kline (1998) and therefore appeared to represent adequately
normal distributions. The final sample size utilized for the analyses was 354. Table 1
provides the descriptive statistics for the study variables as well as the variable names

utilized in the model and the directionality of the constructs.

67

Measurement Model Modifications

Prior to investigation of the structural model presented in Figure l, improvements
in the measurement model resulted from modifications based upon confirmatory factor
analysis (CF A). The standardized maximum likelihood estimates of the measurement
model parameters are represented in Table 2. The modifications made to the originally
proposed measurement model are discussed below.

The latent construct of Family Characteristics was designed to capture
characteristics of the early family environment that theliterature suggests are related to
attachment (e.g., Brent et al., 1990; Cassidy & Berlin, 1994). The original CFA revealed
that one observed variable, parental psychopathology, had a particularly large standard
error, and this variable was eliminated, resulting in a latent construct comprised of four
observed variables: family violence and absence of parental warmth from the Family
History Questionnaire and parental rejection and parental love from the PAAQ
(Lichtenstein & Cassidy, 1991). Due to the scaling on the absence of parental warmth
variable, high scores on this variable indicate low parental warmth whereas low scores
indicate greater perceived parental warmth.

The CF A conducted on the four attachment variables fiom Bartholomew’s (1990)
Close Relationships Questionnaire resulted in a negative error variance, which is an
impossible solution, and revealed two distinct underlying factors that appeared to
correspond to models of self versus models of others, per Figure 4. The results of the
CFA suggested that a separate latent construct representing attachment style did not exist
among the measures collected but rather that these measures assessed the models of self

and others that the literature argues constitute attachment style (e.g., Bartholomew, 1990;

68

 

Bowlby, 1977; Bretherton, 1985). CFAs were conducted separately for two separate
latent constructs, model of self and model of others, that also included the observed
variables assessing views of self and others from the Attachment Style Questionnaire
(ASQ; Feeney, Noller, & Hanrahan, 1994) as well as a measure of self-esteem
(Rosenberg Self-Esteem Scale; Rosenberg, 1965). Bartholomew’s fourth category
assessing a dismissing attachment style did not load significantly and was eliminated
from the analyses. High scores on models of self indicate negative views of self whereas
low scores on this construct indicate positive views of Self. High scores on models of
others indicate positive views of others whereas low scores on this construct indicate
negative views of others.

The CF A investigating the hypothesized measurement model for the latent
construct of avoidance of negative affect resulted in a negative error variance, which is an
impossible solution, and indicated that the observed variables did not assess a similar
construct. Due to the poor fit of the data to a single underlying latent construct, the
avoidance coping subscale of the WCCL, which appeared to have the greatest degree of
both face validity and empirical support, was utilized in the structural model to assess
avoidance of negative affect. Higher scores on this construct indicate greater use of an
avoidant coping style.

Evaluation of Overall Model Fit

The following overall fit indices were examined to determine whether the
observed data adequately fit the hypothesized model: Comparative Fit Index (CF I),
Tucker-Lewis Index (TLI), Bentler-Bonett Nonned Fit Index (NF 1), and Root Mean

Square Error of Approximation (RMSEA). The following guidelines were utilized to

69

evaluate overall goodness of fit: CF I, TLI, and NFI values greater than 0.90 indicating a
good-fitting model, with a value of 1.00 indicating an ideal fit; RMSEA value less than
0.10 indicating an acceptable fit, with a value of 0.00 indicating an ideal fit (Tabachnick
& F idell, 1996). Smaller chi-square statistics indicate a better fit, with significant chi-
square values indicating a poor fitting model. However, the chi-square statistic is
sensitive to sample size, and inaccurate probability levels can result from a computed chi-
square that is not actually distributed as chi-square due to a small sample size
(Tabachnick & F idell, 1996).

Marginal support was found for the overall model fit, with CFI=.963, NFI=.951,
TLI=.954, and RMSEA=.093. The Chi-square was significant (1164; df= 286; p<.000),
which does not support the overall model fit. The squared multiple correlation obtained
for the primary variable of interest, suicidality, was 0.341 , indicating that the model
accounted for 34% of the variance in the suicidality factor. See Figure 5 and Table 3 for
the standardized parameter estimates as well as squared multiple correlations for the
structural model.

P_ost Hoc Model Modificm

Post hoc modifications were performed in an attempt to develop a more
parsimonious model with a better fit (Tabachnick & Fidell, 1996). See Table 4 for the
models evaluated, Chi-square value, RMSEA, CPI, and Chi-square difference tests.
Based upon modification indices obtained through the AMOS software as well as
theoretical relevance, two paths were added (from avoidance and anxiety) to predict
interpersonal stress. Although these pathways were not originally hypothesized, a

relationship between these constructs is consistent with the literature, as perceptions of

70

interpersonal stress may be influenced by the dynamic constructs of avoidance and
anxiety. These additional pathways resulted in an improved model fit.

Next, the variable representing separation/individuation and all corresponding
pathways were deleted, as this variable was not significantly predictive of any variable.
This revision towards a more parsimonious model resulted in a better fitting model.
Similarly, an improvement in the overall model fit resulted from the deletion of the
interpersonal stress variable and all corresponding pathways, as interpersonal stress was
not significantly predictive of any variable. An improved model fit did not result from
the deletion of an additional nonsignificant parameter (avoidance predicting suicidality),
and therefore, this parameter was retained.

The final model is represented in Figure 8 and represents a more parsimonious,
better fitting model than the one originally hypothesized. Adequate support was found
for overall model fit (CFI=.979, NFI=.970, TLI=.973, RMSEA=.O79), and the model
accounted for 35% of the variance in self-reported suicidality. The relationships among

variables are addressed in more detail below within the discussion of specific hypotheses.

Evaluation of Specific Hypotheses

Hypothesis 1

A direct, significant positive relationship was expected between early family risk
factors and insecure attachment (Path A, Figure 1). However, as discovered during the
investigation of the measurement model, there did not exist a unitary construct of
“attachment,” but rather the observed variables appeared to capture the underlying
components of “models of self” and “models of others.” As a result, this hypothesized

relationship could not be investigated directly.

71

However, as the literature suggests that family risk would be negatively related to
the positive models of self and others characteristic of a secure attachment style (e. g.,
Bartholomew, 1990; Bretherton, 1992), this post-hoc hypothesis was investigated. In
fact, family risk was significantly negatively related to positive models of others
(standardized regression weight=-.454, p<.000) and significantly related to negative
views of self (standardized regression weight=.435, a: , p<.000). This indicates that high
levels of family risk, including family violence and parental rejection, were more likely to
be associated with negative views of self and negative Views of others, models that are
characteristic of insecure attachment styles.
Hypotheses 2 and 3

Attachment style was hypothesized to have direct, significant associations with
views of self, as delineated by Bartholomew (1990; Path B). In addition, attachment style
also was hypothesized to have direct, significant associations with views of others, as
delineated by Bartholomew (1990; Path C). However, as discussed above, the
investigation of the measurement model revealed that there did not exist a unitary factor
of “attachment,” but rather the observed variables appeared to capture the underlying
factors of “models of self” and “models of others.” As a result, these hypotheses could

not be investigated.

Hypothesis 4

As hypothesized, a direct, significant positive relationship was identified between
positive views of self and good separation/individuation (Path D; standardized regression
weight=-.437, p<.000). High scores on the variable representing separation/individual

indicate better separation/individuation while high scores on views of self indicate

72

negative self models. Therefore, the identified significant, negative relationship indicates
that respondents who had negative models of self were more likely to have poor
separation/individuation.
Hypothesis 5

As hypothesized, a direct, significant positive relationship was identified between
positive views of others and separation/individuation (Path E; standardized regression
weight=.l37; p<.011). This indicates that subjects who were characterized by positive
views of others were more likely to have achieved good separation/individuation from
mother.
Hypothesis 6

As hypothesized, a direct positive relationship was identified between negative
views of self and avoidance of negative affect (Path F; standardized regression
weight=.56; p<.000). This indicates that subjects who were characterized by greater
levels of avoidance of negative affect were more likely to be characterized by negative
views of self. However, the direct negative relationship hypothesized to exist between
positive views of others and avoidance of negative affect was not supported (Path H;
standardized regression weight=.099; p<.053). This indicates that avoidance of negative
affect was not significantly predicted by views of others.
Hypothesis 7

As hypothesized, a direct, negative relationship was found between positive views
of self and anxiety (Path G; standardized regression weight=.808; p<.000). High scores
on views of self reflect negative self models and high scores on the anxiety measure

reflect great levels of experienced anxiety. This indicates that respondents that identified

73

 

negative views of self were more likely to experience anxiety. Similarly, a direct,
negative relationship was identified between positive views of others and anxiety (Path 1;
standardized regression weight=-.18; p<.000), indicating that respondents that identified
negative views of others were more likely to experience anxiety.
Hypothesis 8

The direct, negative relationship hypothesized to exist between
separation/individuation and depression was not supported (Path J; standardized
regression weight=-.O6l; p<.101). This indicates that the degree to which
separation/individuation was achieved was not related to depression.
Hypothesis 9

Similarly, the direct, negative relationship hypothesized to exist between
separation/individuation and suicidality was not supported (Path K; standardized
regression weight=-.075; p<.159). This indicates that the degree to which
separation/individuation was achieved was not related to suicidality. Elimination of the
variable separation/individuation and all corresponding pathways led to a more
parsimonious, better fitting post hoc model.
Hypothesis 10

Although a significant relationship was identified between avoidance of negative
affect and depression (Path L; standardized regression weight=.l41; p<.000), this
relationship was not in the direction hypothesized. These results indicate that those
respondents who reported high levels of avoidance were more likely to report depression.

Hypothesis 11

As hypothesized, a direct, positive relationship was found between anxiety and

74

depression (Path N; standardized regression weight=. 720; p<.000). This indicates that
those respondents who reported high levels of anxiety were more likely to report

depression.

Hypothesis 12

The direct, positive relationship hypothesized to exist between avoidance of
negative affect and suicidality was not supported (Path M; standardized regression
weight=.063; p<.27l). This indicates that those individuals who reported greater levels
of avoidance were not more likely to report suicidality.

Hypothesis 13

As hypothesized, a direct, positive relationship was found between anxiety and
suicidality (Path 0; standardized regression weight=.247; p<.013). This indicates that
respondents who reported more anxiety were also more likely to report greater
suicidality.

Hypothesis 14

As hypothesized, a direct, positive relationship was found between depression and
suicidality (Path P; standardized regression weight=.289; p<.002). This indicates that
respondents who reported more depression were also more likely to report greater
suicidality.

Hypothesis 15

The direct, positive relationship hypothesized to exist between interpersonal
stressors and suicidality was not supported (Path Q; standardized regression weight=.038;
p<.469). This indicates that those greater levels of interpersonal stress were not

associated with suicidality. Elimination of the latent construct interpersonal stressors led

75

to a more parsimonious, better fitting post hoc model.
Hypothesis 16

Although the two models could not be directly compared, the fit indices for the
post hoc modified model represented in Figure 8 appear to indicate a better fit than the
model previously utilized in the literature to explain adolescent suicidality represented by
Figure 2 (CFI=.946; NFI=.943; TLI=.874; Chi-square=168.54, df= 9, p<.000;
RMSEA=.224) which appears to fit poorly. In addition, the model accounted for a
slightly smaller proportion of the variance in suicidality (33%) than did the post hoc
model represented in Figure 8. See Figure 6 for the standardized parameter estimates and
squared multiple correlations for this model.
Hypothejsis 17

Although the two models could not be directly compared, the fit indices for the
post hoc modified model represented in Figure 8 appear to indicate a better fit than the
model previously utilized in the literature to explain adolescent suicidality represented by
Figure 3 (CF I=.985; NFI=982; TLI=.968; Chi-square=73.751, df= 13, p<.000;
RMSEA=.115). Furthermore, the model represented in Figure 3 only accounted for 5%
of the variance in suicidality. See Figure 7 for the standardized parameter estimates and

squared multiple correlations for this model.

76

DISCUSSION

The present investigation presented a theoretical model of adolescent suicidality
based on attachment theory that offers a valuable alternative to models previously utilized
in the literature, which frequently have relied upon main effects between variables such
as risk factors and outcomes. The findings are consistent with the role of attachment
cognitions as mechanisms by which previously associated family-related risk factors may
impact the development of suicidality.

Consistent with prior research, the present investigation found that no risk factor
was necessary or sufficient for the development of suicidality. Rather, as hypothesized,
the results of this study are consistent with a “common developmental pathway” (Adam,
1994, p. 276) by which familial risk factors, such as family violence and parental
rejection, may produce insecure attachment patterns and their associated internal working
models and dynamic characteristics. In fact, the present investigation addresses a
significant weakness of prior research by offering the internal working models of
attachment theory as potential mechanisms by which environmental events can affect
outcomes, such as suicidality.

The present investigation found that the post hoc modified model, which
accounted for internal representations and dynamic responses, successfirlly accounted for
35% of the variance in adolescent suicidality and adequately fit the observed data.
Overall, the results of this study suggest that attachment theory provides an important
framework from which to address the etiology of suicidality. The results regarding

individual constructs and pathways (Figure 1), as well as alternative models (Figures 2

77

and 3), are discussed in more detail below. For those constructs that are not discussed in
the following section, the observed variables adequately loaded on the hypothesized
latent construct, and the proposed measurement model did not require modification.
Proposed Con_structs

Family Characteristics

The construct “family characteristics” was designed to capture attributes of the
family environment that the literature has associated with the development of child
attachment to the caregiver (e. g., Ainsworth, et al., 1978; Cassidy & Berlin, 1994). This
construct initially included five aspects of the family environment: parental
psychopathology, family violence, absence of parental warmth, parental rejection, and
parental love. However, CFA revealed that parental psychopathology may not be
accessing a similar underlying construct, as the other variables appeared to be measuring
perceived parental love/warmth versus rejection. Although the literature suggests that
parental psychopathology can significantly impact parental responsiveness, availability,
and warmth (e. g., Cassidy & Berlin, 1994), the statistically significant but low order
correlation found in the present investigation between parental psychopathology and the
variable parental warmth (r=.2979; p<.000) suggests that there does exist some variability
in this relationship. As a result, parental psychopathology was excluded from the
construct “family characteristics.”

Attachment Models of Self an_d Modelgf Otheg

 

Rather than capturing a single “attachment” construct, the results of the analyses
revealed two latent constructs that appeared to represent the underlying models of self

and others that the literature suggests constitute attachment style (e. g., Bartholomew,

78

1990; Bowlby, 1977, Bretherton, 1985). Attachment theory proposes that an attachment
style results from the formation of internal working models of the self and of attachment
figures, which are constructed from past experiences and serve the firnction of guiding
future behavior in response to predictions of the behavior of others (Bowlby, 1988;
Bretherton, 1985).

Given the present investigation’s reliance upon self-report measures, it may not
have been possible to access “attachment” independent of the underlying models of self
and other. In fact, an extensive literature relies upon interview data to classify individuals
into attachment categories (e. g., Main et al., 1985), with the quality of the responses
regarding early relationships providing invaluable information that cannot be obtained
merely from the content of the responses. The attachment literature also suggests that
models of self and others provide the mechanisms by which attachment status influences
levels of psychological distress (e. g., Kenny et al., 1993; Roberts et al., 1996). For
example, Roberts et al. (1996) found that self-concept variables primarily accounted for
the effect of attachment on depression. It therefore appears that the underlying models of
self and others provide important information regarding the impact of attachment.

Although the attachment literature historically has relied upon three attachment
categories (e. g., Ainsworth, et al., 1978), Bartholomew (1990) argues that the
consideration of positive and negative views of self and others results in two
subclassifications of avoidant attachment, dismissing and fearful. Sufficient statistical
support for inclusion of Bartholomew’s fourth attachment category, a dismissing avoidant
attachment style, was not found, and this variable was eliminated from the present

analyses.

79

Avoidance of Nagam/e Affect

Investigation of the hypothesized four-factor model of “avoidance of negative
affect” that incorporated variables from both the Defense Mechanisms Inventory (DMI;
Ihilevich & Gleser, 1986) and the Ways of Coping Checklist (WCCL; Vitaliano, et al.,
1985) revealed that the observed variables in the present investigation did not assess a
similar construct. As a result, the avoidance c0ping subscale of the WCCL, which
appeared to have the greatest degree of both face validity (e. g., “Tried to forget the whole
thing” and “I refused to believe it had happened”) and empirical support regarding the
intended construct (e. g., Vitaliano et al., 1985), was utilized in the structural model to
assess avoidance of negative affect. However, it remains unclear whether self-report
measures can sufficiently access avoidance of negative affect, as this construct is intended
to measure an unconscious coping style, and external sources of information may be
necessary (e.g., Kobak & Sceery, 1988).

Proposed and Pofist Hoc Models

The construct modifications discussed above were incorporated into the model
proposed in Figure 1, and SEM was utilized to investigate the overall fit of the
hypothesized model. Marginal support was found for the overall model fit, and the
model accounted for 34% of the variance in the primary variable of interest, suicidality.

Post hoc modifications of the original model (Figure 1) resulted in a more
parsimonious model with adequate fit (Figure 8). Although two paths initially were
added (from avoidance and anxiety) to predict interpersonal stress, these paths ultimately
were eliminated when the non-significant pathway from interpersonal stress to suicidality

was deleted. In addition, the variable separation/individuation and the corresponding

80

pathways were deleted due to this construct’s lack of predictive value. This post hoc
model accounted for 35% of the variance in the primary variable of interest, suicidality
(see Figure 8).

Overall, these results provide support for the attachment paradigm as a useful
framework for understanding the relationship between disturbances in the parent-child
relationship and the development of suicidality (Adam, 1994). Furthermore, these
findings are consistent with a “common developmental pathway” (Adam, 1994, p. 276)
through which familial risk factors, such as those previously associated with adolescent
suicidality, may produce similar insecure attachment patterns.

Proposed PathwapLs

The hypothesized relationship between family risk and attachment style could not
be investigated directly in the current study due to the lack of a unitary “attachment”
construct discussed above. However, based upon the literature (e. g., Bartholomew, 1990;
Bretherton, 1992), the relationship between family risk and negative models of self/others
was investigated post-hoe. The results of this analysis confirmed that negative models of
self and others, which are characteristic of insecure attachment styles, are more likely to
be associated with family risk factors, such as family violence and parental rejection.
This finding is consistent with the attachment literature, which finds that a lack of
parental availability and responsiveness is related to the development of insecure
attachment styles (e. g., Cassidy & Berlin, 1994). In addition, the present study
established the importance of incorporating models of self and others rather than relying
exclusively upon one or the other (e.g., Hammen et al., 1995; Kenny et al., 1993; Roberts

et al., 1996).

81

Furthermore, the identified relationship between family risk and models of
self/others provided important information regarding potential mechanisms underlying
previously identified relationships between risk factors, such as perceived family stress
and lack of support, and suicidality (e.g., Asamow, 1992; Marttunen et al., 1993; Nielsen
et al., 1991). Additionally, internal models of self/others take into consideration
differences in the way that events are understood. In this way, the present investigation’s
use of attachment cognitions to account for the way that traumatic events have been
understood and resolved appears to have successfully provided important information
regarding the relationship between family risk and psychopathology that could not be
provided by recording the mere occurrence of an event (Adam et al., 1996).

Although internal models of self/others provided important information regarding
mechanisms by which family risk factors may impact the development of adolescent
suicidality and depression, one weakness of the present investigation is that it is not clear
what protects some individuals from developing negative models of self or other, despite
being placed at risk. The inclusion of additional potential risk and protective factors may
provide important information regarding the relationship between family risk and the
resulting internal models of self and others. In addition, investigation of the impact of
gender on the relationship between risk and specific combinations of positive/negative
models of self/others may provide important additional information (Adam et al., 1996).
Furthermore, a longitudinal study beginning prenatally could provide important
information regarding these developmental pathways that could not be provided by this

cross-sectional study.

82

As hypothesized, significant relationships were found between good
separation/individuation and positive models of both self and others. This finding
suggests that attachment theory may be a useful framework from which to consider
previously identified relationships between separation/individuation and other factors,
such as self-esteem, emotional outlook on life, and personal and academic adjustment
(e.g., Damon, 1983; Hoffman, 1984). However, no support was found regarding the
hypothesized role of poor separation/individuation as a stressor for adolescents, as
separation/individuation was not significantly related to depression or suicidality. In
addition, the elimination of this variable led to a more parsimonious, better fitting model.

This finding is inconsistent with prior research that found a relationship between
poor separation/individuation and adolescent suicidality (e. g., Meyer & Phillips, 1990).
It may be that the present investigation did not sufficiently assess this developmental
task. Hoffman (1984) argues that adolescent separation/individuation consists of four
components of independence from parents: functional, attitudinal, emotional, and
conflictual; however, the current study only utilized a single aspect of ’
separation/individuation, conflictual independence. Another possibility is that, for this
college-age sample of adolescents, the task of separation/individuation was no longer
salient enough to significantly impact the development of psychopathology. Perhaps a
study that investigates the impact of separation/individuation in a younger sample of
adolescents may find that difficulty with resolution of this developmental task is more
strongly related to suicidality.

The current investigation also proposed that views of self and others were related

to avoidance of negative affect and interpersonal anxiety (Shaver & Clark, 1996) and that

83

these dynamic characteristics would influence the manifestation of both suicidality and
depression (e.g., Borst & Noam, 1993; Van der Kolk et al., 1991). The results revealed
that negative views of self were strongly related to higher levels of anxiety (standardized
regression weight= .81). This finding is consistent with attachment theory and Figure 4,
as individuals characterized by a secure attachment (positive views of self/others)
frequently are described as having high self-esteem and low levels of interpersonal
anxiety (e. g., Shaver & Clark, 1996), whereas individuals characterized by
preoccupied/anxious ambivalent (negative views of self/positive views of others) or
fearful avoidant (negative views of self/others) insecure attachment styles are described
as having low self-esteem and high levels of interpersonal anxiety (Bartholomew, 1990;
F eeney et al., 1994; Shaver & Clark, 1996). Similarly, a negative view of self was
moderately related to avoidance (standardized regression weight= .56), which also is
consistent with Figure 4; although negative views of self are characteristic of both
preoccupied/anxious ambivalent and fearful avoidant insecure attachment styles,
avoidance of negative affect is hypothesized to vary from low to high with these
attachment categories, respectively (e. g., Shaver & Clark, 1996; see Figure 4).

In addition, as hypothesized, anxiety was strongly predictive of both depression
and suicidality. This finding is consistent with previous research that has identified
strong rates of comorbidity between disorders such as depression and anxiety (e. g.,
Goodyer, Herbert, Secher, & Pearson, 1997; Kendall, Brady, & Verduin, 2001).
Furthermore, this suggests that those attachment categories characterized by high levels
of anxiety (i.e., anxious ambivalent and fearful avoidant attachment) are more likely to

experience both depression and suicidality (e.g., Shaver & Clark, 1996; see Figure 4).

84

Overall, these findings are consistent with the hypothesized role of attachment cognitions
in the etiology of anxiety, depression, and suicidality. However, the cross-sectional
design of the present investigation, as well as the use of SEM, precluded any definitive
conclusions regarding developmental processes as well as causal relationships among
variables.

Negative views of others were mildly related to anxiety (standardized regression
weight=.18), a finding that also is consistent with Figure 4; although negative views of
other are characteristic of both dismissing avoidant and fearful avoidant insecure
attachment styles, interpersonal anxiety is hypothesized to vary from low to high with
these attachment categories, respectively (e.g., Shaver & Clark, 1996; see Figure 4).
However, the results of the present investigation did not support a relationship between
negative views of others and avoidance of negative affect, a finding that is not consistent
with the hypothesized relationships represented in Figure 4. In addition, not only was the
statistically significant relationship found between avoidance of negative affect and
depression (standardized regression weight=.14) not in the direction hypothesized, with
those subjects who reported avoidance of negative affect being more likely to report
greater levels of depression, but avoidance of negative affect was not related to
suicidality.

One possible explanation for the lack of predictive value of the construct
“avoidance of negative affect” in the present investigation may be the previously
discussed measurement model difficulties experienced with this construct. In fact, it may
be that the hypothesized unconscious coping mechanism of “avoidance” can not be

accessed with self-report instruments (e. g., Kobak & Sceery, 1988). Another possibility

85

is that models of others and the hypothesized resulting levels of avoidance of negative
affect (see Figure 4) may be more strongly related to externalizing behavior problems
while models of self and the associated levels of anxiety may be more strongly related to
internalizing behavior problems such as those discussed in the present investigation. This
would be consistent with Kobak and Sceery’s (1988) argument that different styles of
affect regulation result from attachment status as well as with several previous studies
that have found depression and suicidality to be associated specifically with
anxious/ambivalent and fearful avoidant attachment styles (e.g., Adam et al., 1996;
Camelley et al., 1994; Kobak & Sceery, 1988).

Furthermore, Camelley et al.(1994) proposed that specific attachment patterns
may be related to specific psychopathologies, with a sample of individuals exhibiting
symptoms other than depression, such as antisocial behavior, perhaps evidencing a
predominantly dismissing, rather than fearful, avoidant attachment style. In fact, Adam et
al. (1996) found that a dismissing attachment style was most characteristic of male
adolescents and a secure attachment was most characteristic of female adolescents who
were receiving psychiatric treatment but did not have a history of suicidality. Perhaps
important information regarding the mechanisms underlying the well-established
relationship between attributions of hostile intent to others and children’s engagement in
aggressive behavior towards others (e. g., Dodge, Price, Bachorowski, & Newman, 1990)
can be obtained through the application of attachment theory and the associated models
of others to the etiology of externalizing behavior problems.

The results of the current study revealed, as predicted, that the presence of

depression was neither a necessary nor sufficient condition for the presence of suicidality

86

(standardized regression weight=.29). This is consistent with Adam et al.’s (1996)
argument that attachment classifications provide more predictive information regarding
suicidality than does a single form of psychopathology, such as depression. As
hypothesized, it appears that depression may be associated with suicidality due to similar
developmental influences, with the development of depression following many of the
same pathways as the development of suicidality.

Although the most common precipitants associated with suicide during
adolescence are interpersonal conflict and loss (e.g., Henry et al., 1993; Lester, 1994), the
inclusion of interpersonal stressors in this model of adolescent suicidality did not provide
significant additional information, and elimination of this variable resulted in a more
parsimonious, better fitting model. It may be that the construct of “stressors” was
represented inadequately in the current investigation; not only did the cross sectional
design of the present study preclude the investigation of the impact of stress over time,
but numerous potential stressors, such as sexual identity and religiosity issues, were not
included in the model.

In addition, it may be that the model utilized in the present investigation did not
adequately represent the relationship between interpersonal loss and suicidality. As it had
been proposed that the manifestation of adolescent suicidality in response to a significant
loss parallels the separation reactions of insecure children, perhaps a model incorporating
the interaction between models of self/others and interpersonal stress would have more
accurately represented the mechanisms by which significant losses during adolescence
result in suicidality for some adolescents but not for others. This would be consistent

with the hypothesized role of internal working models as the mechanism by which

87

qualitatively different internal experiences of the same external events are generated for
different individuals. In fact, attachment characteristics have been found to moderate the
effects of interpersonal stress on general psychological outcome, with those individuals
characterized by insecure attachment evidencing the greatest symptomatology upon
exposure to stress (Hammen et al., 1995).

Alternative models

Although the post hoc model could not be directly compared to the alternative
models, the fit indices for the post hoc modified model (Figure 8) indicate an adequate fit
while the fit indices for the models previously utilized in the literature to explain
adolescent suicidality (see Figures 6 and 7) reveal relatively poor fit. In addition, the
present investigation’s model accounted for a greater proportion of the variance in
suicidality than did the models represented in Figure 2 ( 33%) or Figure 3 (5%) (see also
Figures 6 and 7).

As hypothesized, greater explanatory power for the prediction of suicidality
appears to be provided by a model that does not simply rely upon associated factors, such
as depression or family characteristics, but rather incorporates developmental and
dynamic constructs to address the mechanisms underlying the development of depression
and suicidality during adolescence. It appears that a model which accounts for the way in
which events have been understood and resolved more accurately represents the
relationship between family risk factors and adolescent suicidality than do models which
merely document the occurrence of risk factors or the presence of associated factors, such

as depression (e.g., Adam et al., 1996).

88

Limitations of the Current Study and Directions for Future Research

As the current study relied upon measures of suicidal thoughts, it remains unclear
the extent to which the identified significant relationships among variables remain
meaningful when suicidal behavior is utilized as the primary variable of interest.
Similarly, the use of a sample comprised entirely of college students necessarily limited
the generalizability of these findings to more severely impaired, psychiatric populations
of suicidal adolescents. Future studies that utilize samples more representative of the
spectrum of potential suicidality, including psychiatric‘inpatient subjects hospitalized for
suicidal behavior, will provide further information regarding the degree to which similar
mechanisms operate with regard to suicidal behavior.

In addition, the sample utilized in the present investigation, which
disproportionately consisted of women (78%) and Caucasians (85%) between the ages of
eighteen and twenty-five, potentially limits the generalizability of these findings to males
and other age-groups, as well as to the ethnic/racial backgrounds that were inadequately
represented in this sample (8% Afiican-American; 3% Asian American; 3%
Hispanic/Latino; 0.5% American Indian; 2% Multiracial). Future studies that investigate
these relationships in more diverse populations not only will provide important
information regarding the role of attachment in the etiology of suicidality in the general
population but also may reveal differential relationships among these variables in various
populations (e. g., Adam et al., 1996; Shaver & Clark, 1996).

For example, despite a disproportionately small sample of male subjects (n=77;
22%), post-hoe investigation of the relationship between attachment and gender revealed

that a dismissing avoidant attachment style (attachs4) was significantly correlated with

89

male gender (r=. 129; p<.016), a finding that is consistent with prior research (e.g., Shaver
& Clark, 1996). It remains unclear whether the measure of a dismissing avoidant
attachment style (attachs4), which did not load significantly during the investigation of
the measurement model within this largely female (78%) sample, would have in fact
contributed significant information in a more representative sample or during separate
analyses of the model for males and females. In addition, ratings indicating that
interpersonal relationships and approval from others are not of primary importance were
significantly related to male gender (relat2nd: r=.186, p<.000; approval: r=-.l39,
p<.012), as was better separation/individuation (mompsi: F. 126, p<.019). Furthermore,
post-hoe analyses revealed that the dynamic characteristics of anxiety and avoidance were
significantly related to gender, with higher levels of these characteristics associated with
female gender (bsianx: r=.-.l30, p<.016; avoidanc: r= -.111, p<.038). These results
further suggest that separate models for males and females may be necessary to more
accurately capture salient stressors and developmental pathways.

As discussed previously, the cross-sectional design of the current investigation, as
well as the use of SEM, precluded an investigation of the development of internal
working models of self and others as well as the drawing of definitive conclusions
regarding causal relationships among variables. A longitudinal investigation, beginning
prenatally, would provide important information regarding developmental processes.
Additionally, this would allow for the investigation of the impact of additional risk and
protective factors as well as the potential protective function of a secure attachment given
exposure to risk over time (e.g., Papini & Roggman, 1992). However, it should be noted

that SEM, while valuable for simultaneously investigating complex variable and factor

90

relationships, can only indicate whether the hypothesized model is compatible with the
observed data and can not eliminate the possibility of alternative models, the existence of
which must always be considered (Tabachnick & F idell, 1996).

Throughout the previous discussion of several hypothesized constructs, it was
noted that the present investigation’s exclusive reliance upon self-report data may have
limited this study’s ability to accurately capture various theoretical constructs. For
example, investigation of the proposed measurement model for the construct “avoidance
of negative affect” revealed that the various variables that had been utilized in an attempt
to capture this underlying construct did not appear to be accessing similar constructs. As
a result, it remained unclear whether the construct of avoidance utilized in the
investigation of the structural model adequately captured the intended dynamic
characteristic of “avoidance of negative affect,” which necessarily limited the extent to
which conclusions could be drawn regarding the model parameters related to this
variable. It will be important for future studies to utilize additional sources of
information, including interview data (e. g., AAI; Main et al., 1985) and external sources
of information (such as peer ratings; e.g., Kobak & Sceery, 1988), to gain better access to
hypothesized, unconscious mechanisms.

Overall, the results of this study support the usefulness of attachment theory as a
heuristic with which to understand the etiology of suicidality in adolescents. In addition,
the developmental pathways represented in the proposed model accounted for a
significant amount of the variance in measures of both anxiety (76%) and depression
(68%), suggesting that the etiology of these intemalizing symptoms may share similar

developmental pathways. Furthermore, as discussed more extensively above, the

91

application of attachment theory to other problem behaviors and manifestations of
psychological distress, such as externalizing behavior disorders, also may provide
important information regarding the mechanisms underlying previously identified
relationships between variables, such as attributions of hostile intent to others and
children’s engagement in aggressive behavior towards others (e. g., Dodge et al., 1990).
Future studies based in attachment theory that investigate “common developmental
pathways” (Adam, 1994, p. 276) leading to a variety of psychopathologies not only may
provide important information regarding the etiology ofthese disorders but also may

assist in prevention and intervention efforts.

92

APPENDICES

93

APPENDIX A

TABLES

94

Table 1: Descriptive Statistics for Model Variables (n=354)

 

 

Latent Observed Variables (variable name) M SQ Mia Mag
Construct
Family Risk High scores = greater familial risk/rejection
Absence of parental warmth (parwarm) 12.2 4.5 6 30
Family violence (famviol) 1.5 2.0 0 11
Parental love (lovedpa) 25.1 4.7 9 30
Parental rejection (rejectpa) 183 6.9 11 42
Models of High scores = positive view of others
Others Secure attachment (attachsl) 4.8 1.7 1 7
Avoidant attachment (attacth) 4.3 2.0 1 7
Discomfort with closeness (discomft) 32.8 9.0 11 58
Confidence in other (confiden) 35.1 5.7 14 48
Relationships as secondary (relat2nd) - 15.8 4.4 7 27
Models of High scores = negative view of self
Self Anxious/ambivalent attachment (attachs3) 3.2 1.8 1 7
Low self-esteem (esteem) 17.9 5.9 10 40
Need for approval (approval) 21.8 6.1 7 39
Preoccupation with relationships (preoccup) 28.4 6.8 12 46
Anxiety High scores = greater anxiety
Trait anxiety (traitanx) 40.1 11.1 20 77
State anxiety (stateanx) 36.4 11.3 20 73
Anxiety symptoms (bsianx) .73 .76 0 4.33
Avoidance High scores = greater avoidance
of Negative Avoidant coping style (avoidanc) 14.5 6.7 0 33
Affect
Separation/ High scores =better separation/individuation
Individuation Conflictual independence (mompsi) 78.2 16.4 26 100
Depression High scores = greater depression
Depression (depress) 8.5 6.6 O 33
Interpersonal High scores = greater stress
Stress General social mistreatment stress 11.0 3.7 6 23
(stressmt)
Total loss of support past year (totlloss) 4.1 3.3 0 l3
Friendship problems stress (stressfp) 5.8 2.1 3 12
Romantic problems stress (stressnn) 6.5 2.7 3 12
Suicidality High scores = greater suicidality
Total suicidality during past month 12.6 14.8 0 72
(asiqtotl)
Suicidality during past year (suicyr) 12.1 4.5 10 33
Suicidality during the lifetime (suiclife) 15.5 7.7 10 47

95

Table 2

Standardized Maximum Likelihood Estimates of Measurement Model Parameters

 

 

Squared
Latent Observed Variables (variable name) Standardized p Multiple
Construct parameter Correla-
estimate tlfl
F amily Risk High scores =greater familial risk/rejection -
Absence of parental warmth (parwarm) .80 .000 .64
Family violence (famviol) .49 .000 .24
Parental love (lovedpa) -.89 .000 .78
Parental rejection (rejectpa) .86 .000 .73
Models of High scores = positive view of others
Others Secure attachment (attachsl) .69 .000 .48
Avoidant attachment (attachs2) -.62 .000 .39
Discomfort with closeness (discomft) -.86 .000 .73
Confidence in other (confiden) .74 .000 .55
Relationships as secondary (relat2nd) -.48 .000 .23
Models of High scores = negative view of self
Self Anxious/ambivalent attachment (attachs3) .46 .000 .22
Low self-esteem (esteem) .71 .000 .50
Need for approval (approval) .78 .000 .60
Preoccupation with relationships .71 .000 .51
(preoccup)
Anxiety High scores = greater anxiety
Trait anxiety (traitanx) .93 .000 .86
State anxiety (stateanx) .74 .000 .54
Anxiety symptoms (bsianx) .63 .000 .40
Interpersonal High scores = greater stress
Stress General social mistreatment stress .72 .000 .52
(stressmt)
Total loss of support past year (totlloss) .43 .000 .19
Friendship problems stress (stressfp) .83 .000 .70
Romantic problems stress (stressrrn) .51 .000 .26
Suicidality High scores = greater suicidality
Total suicidality during past month .90 .000 .82
(asiqtotl)
Suicidality during past year (suicyr) .84 .000 .71
Suicidality during the lifetime (suiclife) .75 .000 .56

96

Table 3

Standardized Maximum Likelihood Estimates of Structural Model Parameters

 

 

Variable Pathway Standardized Squared
Parameter I! mLtipLe

From To estimate W

Family Risk
Models of others -.454 .000
Models of self .435 .000

Models of Others .207
Anxiety -. 176 .000
Avoidance -.099 .053
Separation/Individuation . 1 37 .01 1

Models of Self .189
Anxiety .808 .000
Avoidance .557 .000
Separation/Individuation -.437 .000

Avoidance . .342
Depression .141 .000
Suicidality .063 .271

Anxiety .739
Depression .720 .000
Suicidality .247 .013

Separation/

Individuation .233
Depression -.06l . 101
Suicidality -.075 . 159

Depression .684
Suicidality .289 .002 .26

Interpersonal

Stress
Suicidality .038 .469

97

Table 4: Comparison of Post Hoc Nested Models

 

 

AXZ
X2 df RMSEA CFI p-value

Model

1) Hypothesized Model 1164 286 .093 .963

2) 2 paths added- Interpersonal Stress 940 284 .081 .972 .001
predicted by Anxiety and Avoidance

3) Separation/Individuation variable 821 263 .078 .975 .001
and corresponding paths deleted

4) Interpersonal stress variable and 578 180 .079 .979 .001
corresponding paths deleted

5) Deletion of nonsignificant parameter- 579 181 .079 .979 ns

Suicidality predicted by avoidance

98

APPENDIX B

FIGURES

99

8%on
Raoflofioafi

 

qoamoaom

socmsEficfi
Eocmbwmom

 

9850
m0 @505

   
    
      

\ E08533.
m

 

_ semi

100

:ommmocmofl

 

“93:25 Enema—03.. Emaxm 9 BEES 230305 382 we oEEaxm

08205 333m

Qofioflaonozmm
3:25am

   
   

N 25mm

101

Eofinomcxx 05035 BEBE Via Egan

 

 

gwmomdm “5822?. :84me 8 BEES 390505 _ovo—z >132; Eoanmt< mo ofiadxm

m seam

102

.68: has a. seam Ea .83: aoeozosam Ana: Eng 3 8885 2088 Boa Baa... _

A

as

Z

E08588 :82on
528m
03332
550 mo

 

FHNEV NVZQMENANNRNZN 32

HZHEEU<HHD

E

august“ «5296
mfimmflfimmc

 

zofign—Eezwme 382
328m :

Hangouts
“soaking msocam
\woisooooa

o>uawoz

 

“Goa—hogan 058m

“Summon

how we _ocoz
HZHEEU<H>H<

can #7

ZO—H4Dngmn—Z—

HUNEKV Nx—Nh VGNZ
£6 NUZVQNQS V

32

 

 

103

Figure 5- Structural Model of the Etiology of Adolescent Suicidality: Standardized
parameter estimates; squared multiple correlations in bold italics (total variance)

   

 

 

 

 

-45
famchar ’
’ model_others ‘
.21 ” ¢
.43
.23
-.10
.19 -.44 mompsi
model self -
- .56 a '18.

 

V

ns
@ .81 07 ns -.06
avoidance "
'34 \X .14 a
CFI= .963

NFI= .951 depress
TLI= .954 . 74
RMSEA= .093
Chi-square= 1164

d% 286

p_<.OOO

 

 

 

 

 

 

 

 

 

.72 .68

it

.29

04 ”S a
' >

interpersonal_stress

 

All pathways p<.001 unless otherwise indicated
*p<.05 **p<.01 ns= not significant

104

Figure 6: Standardized parameter estimates for alternative model proposed in Figure 2
squared multiple correlations in bold italics (total variance accounted for)

 

 

parpsyc

 

asiqtotl

 

 

 

suicyr

 

suiclife

 

 

 

 

 

 

 

 

 

 

 

famviol
CF I= .946

NFI= .943

TLI= .874
RMSEA= .224
Chi-square= 168.54
df= 9

p<.000

 

 

 

 

 

 

depress

 

 

All pathways significant at p<.001 unless otherwise

indicated

105

Figure 7: Standardized parameter estimates for alternative model proposed in Figure 3,
squared multiple correlations in bold italics (total variance accounted for)

.39

 

 

asiqto suicy suiclif

.87
I .77

parpsy

parwar

 

 

 

 

 

 

 

 

 

 

 

   

 

 

 

 

famvi

 

 

 

 

CFI= .985 attachs

NFI= .982
TLI= .968

 

 

 

RMSEA= .1 15
Chi-square= 73.75
df= 13

p<.000

All pathways significant at p<.001 unless
otherwise indicated

*p<.05 **p<.01 ns= not significant

106

Figure 8- Final Post Hoc Model of the Etiology of Adolescent Suicidality: Standardized
parameter estimates with squared multiple correlations in bold italics (total
variance accounted for)

-.44
famchar
’ model_others

 

 

.20
.40 P4
-.10 '
.16
model self -
.56 a '18
V

 

o
avoidance
e

CFI: .979

NFI= .970
TLI= .973 - 76 depress
RMSEA= .079
Chi-square= 578.51
d% 180
p<.000

 

 

 

 

 

 

 

 

.75 .68

.30

it

.29

All pathways p<.001 unless otherwise indicated . 3 5
*Q<,05 **p<.01 ns= not significant

107

APPENDIX C

QUESTIONNAIRE PACKET

108

Attachment, Relationships, and Feelings

This questionnaire packet contains questions about how you have been feeling recently,
events that have happened to you, your current relationships with others, and your
memories of your childhood. This will take a total of approximately two hours.

In order to assist in keeping all information strictly confidential, please DO NOT place
your name or student number on the computer scoring sheets or the questionnaire packet.

Please place _afi answers to the questionnaires on the computer scoring sheets. The only
exception to this is on the first questionnaire, where you will be asked to specify general
information regarding parental occupation in the space provided.

PLEASE pay close attention to the question numbering as you complete this
questionnaire packet. There will be reminders along the way to help you remember to
make sure that the question you are answering corresponds to the appropriate space on
the answer sheet.

Also, please pay attention to the time frame specified for each questionnaire (for example,
some questionnaires ask about the past month, others ask about the past year).

If you have ANY questions, please do not hesitate to ask the administrator. Thank you
for participating!

109

Demographic Questionnaire

0n computer scoring sheet #1, please blacken the number that corresponds to the
appropriate answer. Where necessary, write in answers in the spaces provided.

1. Your sex: 1) female 2) male

 

2. Yourage: 1) 17 2) 18 3) 19 4) 20 5) 21
6) 22 7) 23 8) 24 9) 25 10) Other:
specify—
3. Your year in college: 1) freshman 2) sophomore 3) junior
4) senior 5) other
4. Your ethnic/racial background: 1) African-American 4) Asian-American
2) Caucasian/White 5) Hispanic/Latino
3) American Indian 6) Multiracial
7) Other:
specify
5. Father’s education: 1) grades 1, 2, 3, 4, 5, or 6 Graduate degree:

2) grades 7, 8, 9,10,11, or 12 7) MA

 

 

3) some college 8) PhD.
4) AAdegree 9) Law
5) BA/BS 10) MD
6) some grad school 11) Other: specify
Father’s occupation:
(Please be specific)

110

6. Mother’s education: 1) grades 1, 2, 3, 4, 5, or 6
2) grades 7, 8, 9,10,11, or 12
3) some college
4) AA degree
5) BA/BS

6) some grad school

Mother’s occupation:

Graduate degree:
7) MA

8) PhD.

9) Law

10) MD

1 1) Other: specify

 

 

(Please be specific)

7. Parent’s marital status: 1=married 2= divorced 3= separated 4: never

married 5= widowed

111

CLOSE RELATIONSHIP QUESTIONNAIRE

This questionnaire applies to all emotionally close relationships, not just romantic ones.
Following are descriptions of four general relationship styles that people often report.
You should be on #8 on computer scoring sheet #1.

8. Pick the single alternative that best describes you or is closest to the way you
are.

1) It is relatively easy for me to become emotionally close to others. I am
comfortable depending on others and having others depend on me. I don't worry
about being alone or having others not accept me.

2) I am somewhat uncomfortable getting close to others. I want emotionally close
relationships, but I find it difficult to trust others completely, or to depend on
them. I sometimes worry that I will be hurt if I allow myself to become too close
to others.

3) I want to be completely emotionally intimate with others, but I often find that
others are reluctant to get as close as I would like. I am uncomfortable being
without close relationships, but I sometimes worry that others don't value me as
much as I value them.

4) I am comfortable without close emotional relationships. It is very important to me
to feel independent and self-sufficient, and I prefer not to depend on others or
have other depend on me.

Now, for numbers 9 thru 12, please rate (on a scale of l to 7) each of the four relationship
styles described above according to the extent to which each description corresponds to
your general relationship style.

 

Not at all Somewhat Very much
like me like me like me
9. Style 1. 1 2 3 4 5 6 7
10. Style 2. 1 2 3 4 5 6 7
11. Style 3. 1 2 3 4 5 6 7

rzswm4 1 2 3 4 5 6 7

112

SELF EVALUATION QUESTIONNAIRE

DIRECTIONS: A number of statements which people have used to describe themselves
are given below. Read each statement and then blacken in the appropriate circle for each
statement to indicate (on a scale of 1 to 4) how yor_r feel right now, that is, at this
moment. There are no right or wrong answers. Do not spend too much time on any one
statement but give the answer which seems to describe your present feelings best.

1=not at all 2=somewhat =moderately so =very much so

Right now:

13. Ifeel calm

14. I feel secure

15. I am tense

16. I feel strained

17. I feel at ease

18. I feel upset

19. I am presently worrying over possible misfortunes
20. I feel satisfied

21. I feel frightened
22. I feel comfortable
23. I feel self-confident
24. I feel nervous

25. I am jittery

26. I feel indecisive
27. I am relaxed

28. I feel content

29. I am worried

30. I feel confused

31. I feel steady

32. I feel pleasant

113

SELF EVALUATION QUESTIONNAIRE

DIRECTIONS: A number of statements which people have used to describe themselves
are given below. Read each statement and then blacken in the appropriate circle for each
statement to indicate (on a scale of 1 to 4) _how yoragenerallv feel There are no right or
wrong answers. Do not spend too much time on any one statement but give the answer
which seems to describe how you generally feel.

l=not at all 2=somewhat 3=moderately so 4=very much so

Generally:

33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.

I feel pleasant

I feel nervous and restless

I feel satisfied with myself

I wish I could be as happy as others seem to be

I feel like a failure

I feel rested

I am "calm, cool, and collected"

I feel that difficulties are piling up so that I cannot overcome them
I worry too much over something that really doesn't matter

I am happy

I have disturbing thoughts

I lack self-confidence

I feel secure

I make decisions easily

I feel inadequate

I am content

Some unimportant thought runs through my mind and bothers me
I take disappointments so keenly that I can't put them out of my mind
I am a steady person

I get in a state of tension or turmoil as I think over my recent concerns and interests

114

Rosenberg Scale
You should be on #53 on computer scoring sheet #1 .

Please blacken in the number that best describes your response to each question
according to this scale:

Strongly Agree Agree Disagree Strongly Disagree
1 2 3 4

53. I feel that I am a person of worth, at least on an equal plane with others.
54. I feel that I have a number of good qualities.

55. All in all, I am inclined to feel that I am a failure.

56. I am able to do things as well as most other pe0ple.

57. I feel I do not have much to be proud of.

58. I take a positive attitude toward myself.

59. On the whole, I am satisfied with myself.

60. I wish I would have more respect for myself.

61. I certainly feel useless at times.

62. At times, I think I am no good at all.

115

Instructions: The following list of statements describes different aspects of students'
relationships with their mother. In many cases, the primary female parent-
figure/caregiver is the "mother". If you consider someone else to be your primary female
parent-figure, please respond to the questions that refer to “mother” with that person in
mind. Imagine a scale ranging from 1 to 5 that tells how well each statement applies to
you. Please blacken in the appropriate number from "1" (Not at all true of me) to "5"
(Very true of me). If the statement does not apply, enter "1 ". Please be completely
honest. Your answers are entirely confidential and will be useful only if they accurately
describe you.

Not at all true of me Very true of me
1 2 3 4 5

63. Sometimes my mother is a burden to me.

64. I feel like I am constantly at war with my mother. _

65. I blame my mother for many of the problems I have.

66. I wish I could trust my mother more.

67. I have to be careful not to hurt my mother's feelings.

68. I sometimes feel like I'm being punished by my mother.

69. I wish my mother wasn't so overprotective.

70. I wish my mother wouldn't try to manipulate me.

71. I wish my mother wouldn't try to make fun of me.

72. I feel that I have obligations to my mother that I wish I didn't have.
73. My mother expects too much from me.

74. I wish I could st0p lying to my mother.

75. I often wish that my mother would treat me more like an adult.
76. I am often angry at my mother.

77. I hate it when my mother makes suggestions about what I do.
78. Even when my mother has a good idea I refuse to listen to it because she made it.
79. I wish my mother wouldn’t try to get me to take sides with her.
80. I argue with my mother over little things.

81. My mother is sometimes a source of embarrassment to me.

82. I am sometimes ashamed of my mother.

83. I get angry when my mother criticizes me.

84. When I don’t write my mother often enough I feel guilty.

116

Not at all true of me Very true of me
1 2 3 4 5

85. I feel uncomfortable keeping things from my mother.
86. I often have to make decisions for my mother.

87. I sometimes resent it when my mother tells me what to do.

117

ASQ
You should be on #88 on computer scoring sheet #1 .

Show how much you agree with each of the following items by rating them
on this scale:

Totally Strongly Slightly Slightly Strongly Totally
Disagree Disagree Disagree Agree Agree Agree

1 2 3 4 5 6

88. Overall, 1 am a worthwhile person.

89. I am easier to get to know than most people.

90. I feel confident that other people will be there when I need them.
91. I prefer to depend on myself rather than other people.

92. I prefer to keep to myself.

93. To ask for help is to admit that you're a failure.

94. People's worth should be judged by what they achieve.

95. Achieving things is more important than building relationships.
96. Doing your best is more important than getting on with others.
97. If you've got a job to do, you should do it no matter who gets hurt.
98. It's important to me that others like me.

99. It's important to me to avoid doing things that others won’t like.

100.1 find it hard to make a decision unless I know what other people think.
101.My relationships with others are generally superficial.

102.Sometimes I think I am no good at all.

103.1 find it hard to trust other people.

104.1 find it difficult to depend on others.

105.1 find that others are reluctant to get as close as I would like.

106.1 find it relatively easy to get close to other people.

107.1 find it easy to trust others.

108.1 feel comfortable depending on other people.

118

Totally Strongly Slightly Slightly Strongly Totally
Disagree Disagree Disagree Agree Agree Agree

1 2 3 4 5 6

109.1 worry that others won't care about me as much as I care about them.
110.1 worry about people getting too close.

111.1 worry that I won't measure up to other people.

112.1 have mixed feelings about being close to others.

113.While I want to get close to others, I feel uneasy about it.

114.1 wonder why people would want to be involved with me.

115.It's very important to me to have a close relationship.

116.1 worry 3 lot about my relationships.

117.1 wonder how I would cope without someone to love me.

1181 feel confident about relating to others.

119.1 often feel left out or alone.

120.1 often worry that I do not really fit in with other people.

121 .Other people have their own problems, so I don't bother them with mine.
122.When I talk over my problems with others, I generally feel ashamed or foolish.
123.1 am too busy with other activities to put much time into relationships.

124.If something is bothering me, others are generally aware and concerned.

125.1 am confident that other people will like and respect me.

126.1 get frustrated when others are not available when I need them.

127.0ther people often disappoint me.

119

BDI

You should be on #128 on computer scoring sheet #1 .

Instructions: This questionnaire consists of 21 groups of statements. After reading each
group of statements carefirlly, blacken in the number (1, 2, 3, or 4) corresponding to the
one statement in each ggoup which best describes the way you have been feeling the
PAST WEEK, including today. If several statements within a group seem to apply
equally well, mark each one. Be sure to read all the statements in each group before
making your choice.

128.

129.

130.

131.

132.

133.

134.

#DJNr-r JXUJNH #WNH hWNH AWN-— #WNH

huh)!—

I do not feel sad.

I feel sad.

I am sad all the time and I can't snap out of it
I am so sad or unhappy that I can't stand it.

I am not particularly discouraged about the future.

I feel discouraged about the future.

I feel I have nothing to look forward to.

I feel that the firture is hopeless and that things cannot improve.

I do not feel like a failure.

I feel I have failed more than the average person

As I look back on my life, all I can see is a lot of failures.
I feel I am a complete failure as a person.

I get as much satisfaction out of things as I used to.
I don't enjoy things the way I used to.

I don't get real satisfaction out of anything anymore.
I am dissatisfied or bored with everything.

I don't feel particularly guilty.

I feel guilty a good part of the time.
I feel quite guilty most of the time.
I feel guilty all the time.

I don't feel I am being punished.
I feel I may be punished.

I expect to be punished.

I feel I am being punished.

I don't feel disappointed in myself.
I am disappointed in myself.

I am disgusted with myself.

I hate myself.

120

135.

136.

137.

138.

139.

140.

141.

142.

ADJNH thr—b AWNF-d th'd «bWNH th—l

_a

hWNI—t

 

I don't feel I am any worse than anybody else.

I am critical of myself for my weaknesses or mistakes.
I blame myself all the time for my faults.

I blame myself for everything bad that happens.

I don't have any thoughts of killing myself.

I have thoughts of killing myself, but I wouldn’t carry them out.
I would like to kill myself.

I would kill myself if I had the chance.

I don't cry any more than usual.

I cry more now than I used to.

I cry all the time now.

I used to be able to cry, but now I can't cry even though I want to.

I am no more initated now than I ever am.

I get annoyed or irritated more easily than I used to.

I feel initated all the time now.

I don't get irritated at all by the things that used to irritate me.

I have not lost interest in other people.

I am less interested in other people than I used be.
I have lost most of my interest in other people.

I have lost all of my interest in other people.

I make decisions about as well as I ever could.

I put off making decisions more than I used to.

I have greater difficulty in making decisions than before.
I can't make decisions at all anymore.

I don't feel I look any worse than I used to.

I am worried that I am looking old or unattractive.

I feel that there are permanent changes in my appearance that make me
look unattractive.

I believe that I look ugly.

I can work about as well as before.

It takes an extra effort to get started at doing something.
I have to push myself very hard to do anything.

I can't do any work at all.

121

143.

144.

145.

146.

147.

148.

149.

AWNH #MNI-r AWN—i

AWN.—

I can sleep as well as usual.

I don't sleep as well as I used to.

I wake up 1-2 hours earlier than usual and find it hard to get back to sleep.
I wake up several hours earlier than I used to and cannot get back to sleep.

I don't get more tired than usual.

I get tired more easily than I used to.

I get tired from doing almost anything.
I am too tired to do anything.

My appetite is no worse than usual.

My appetite is not as good as it used to be
My appetite is much worse now.

I have no appetite at all anymore.

I haven't lost much weight, if any, lately.
I have lost more than 5 pounds.

I have lost more than 10 pounds.

I have lost more than 15 pounds.

I am purposely trying to lose weight by eating less.

p—r

AWNH

1=Yes; 2=No

I am no more worried about my health than usual.

I am worried about physical problems such as aches and pains; or upset
stomach; or constipation.

I am very worried about physical problems and it's hard to think of much
else.

I am so worried about my physical problems that I cannot think about
anything else.

I have not noticed any recent change in my interest in sex.
I am less interested in sex than I used to be.

I am much less interested in sex now.

I have lost interest in sex completely.

PLEASE GO TO COMPUTER SHEET i2

122

 

THOUGHTS ABOUT MY LIFE

Listed below are sentences which describe thoughts that people sometimes have. Read
each sentence carefully and decide which of these thoaghts you have had in the PAST
MONTH. For each thought, blacken the number corresponding to the answer that best
describes your own thoughts. There are no right or wrong answers so answer each
sentence as openly and honestly as possible. Be sure to answer each sentence. Do NOT

leave any sentences blank.

This thought was in my mind:

Almost Couple of About Once Couple of Times
Every Day Times a Week a Week a Month
1 2 3 ~ 4
About Once I Had This Thought Before I Never Had
a Month But Not in the Past Month This Thought
5 6 7

Please begin with #1 on computer scoring sheet #2.

I thought it would be better if I was not alive

p—d

I thought about killing myself
I thought about how I would kill myself
I thought about when I would kill myself
I thought about what to write in a suicide note
I thought about telling people I plan to kill myself
I thought that people would be happier if I was not around
I thought about how people would feel if I killed myself
I wished I were dead
. I thought about how easy it would be to end it all

PWSQMP‘NN

O

. I thought that killing myself would solve my problems
. I thought that others would be better off if I was dead

. I wished I had the nerve to kill myself

, I wished that I had never been born

, I thought that if I had the chance I would kill myself

i-‘I-‘h—II—tr-I
Mitt-uteri

123

 

Almost Couple of About Once
Every Day Times a Week a Week
1 2 3
About Once I Had This Thought Before
a Month But Not in the Past Month
5 6
16. I thought about ways people kill themselves

17.
18.
19.
20.
21.
22.
23.
24.
25.

I thought about killing myself, but would not do it
I thought about having a bad accident
I thought that life was not worth living

I thought that my life was too rotten to continue

Couple of Times
a Month
4

I Never Had
This Thought
7

I thought that the only way to be noticed was to kill myself

I thought that if I killed myself people would realize I was worth caring about

I thought that no one cared if I lived or died
I wondered if I had the nerve to kill myself

I thought that if things did not get better I would kill myself.

124

Life Events Scale

Following is a list of experiences which many students have some time or other. Please
indicate for each experience how much it has been a part of your life over the pas_t

month.
Not at all Only slightly Distinctly Very Much
part of my life part of my life part of my life part of my life
1 2 3 4

26. Conflicts with boyfiiend’s/girlfriend’s/spouse’s family
27. Being let down or disappointed by fiiends

28. Conflict with professor(s)

29. Social rejection

30. Too many things to do at once

31. Being taken for granted

32. Financial conflicts with family members

33. Having your trust betrayed by a friend

34. Separation from people you care about

35. Having your contributions overlooked

36. Struggling to meet your own academic standards
37. Being taken advantage of

38. Not enough leisure time

39. Struggling to meet the academic standards of others
40. A lot of responsibilities

41. Dissatisfaction with school

42. Decisions about intimate relationship(s)

43. Not enough time to meet your obligations

44. Conflict with friends/family about your drug or alcohol use
45. Dissatisfaction with your mathematical ability

46. Important decisions about your future career

47. Financial burdens

48. Dissatisfaction with your reading ability

125

Not at all Only slightly Distinctly Very Much
part of my life part of my life part of my life part of my life
1 2 3 4
49. Important decisions about your education
50. Loneliness
51. Lower grades than you hoped for
52. Conflict with teaching assistant(s)
53. Not enough time for sleep
54. Conflicts with your family
55. Heavy demands from extracurricular activities
56. Finding courses too demanding
57. Conflicts with friends
58. Hard effort to get ahead
59. Poor health of a friend
60. Disliking your studies
61. Getting “ripped off” or cheated in the purchase of services
62. Dissatisfaction with your drug/alcohol use
63. Social conflicts over smoking
64. Difficulties with transportation
65. Disliking fellow student(s)
66. Conflicts with boyfi'iend/girlfiiend/spouse
67. Dissatisfaction with your ability at written expression
68. Interruptions of your school work
69. Social isolation
70. Long waits to get service (e.g., at banks, stores, etc.)
71. Being ignored
72. Dissatisfaction with your physical appearance
7 3. Finding course(s) uninteresting
7 4. Gossip concerning someone you care about
75. Failing to get expected job
7 6. Dissatisfaction with your athletic skills

126

Loss of Important Relationships

You should be on #77 of computer sheet #2

77. During the PAST YEAR, have you lost any important relationships due to moving, a
job change, divorce or separation, death, or some other reason? 1=Yes 2=No

78. Please indicate the number of persons who are no longer available to you.
1: 1 person 2= 2 people 3: 3 people 4= 4 people 5= 5 people
6= 6 people 7: 7 people 8= 8 people 9: not applicable/did not lose any

important relationships

79. Overall, how much of your support was provided by these people who are no longer
available to you?

1= none at all/not applicable 2= a little 3= a moderate amount

4= quite a bit 5=a great deal

127

 

 

EARLY CHILDHOOD RELATIONSHIP QUESTIONNAIRE

The majority of the following statements refer to your aarly childhood relationship with
your mother (when you were approximately 3 to 8 years old). In most cases the principal
care-giver is the "mother". If someone else was the principal person responsible for your
care in childhood, please respond to the questions which refer to “mother” with that
person in mind.

A few of the questions have two parts. For example "when I caused trouble as a child I
knew my mother would forgive me". Some people might feel like they never caused
trouble as a child, however, they consider their mother very forgiving. How then do they
answer? Only answer AGREE or STRONGLY AGREE if you agree with both parts of
the statement. If you agree with only one part of the statement answer NEUTRAL. If
you disagree with both parts of the statement answer DISAGREE or STRONGLY
DISAGREE. '

STRONGLY DISAGREE NEUTRAL AGREE STRONGLY

DISAGREE NEITHER DISAGREE AGREE
NOR AGREE
1 2 3 4 5

Please blacken in the appropriate circle on the computer sheet.

You should be on # 80 of computer sheet #2.

80. In childhood I felt like I was really treasured by my mother.

81. In childhood I sometimes felt like my mother was really lonely when I was not with
her.

82. My mother was not very affectionate.

83. When I was a young child and little things went wrong I did not feel sure I could
count on my mother to take care of me.

84. As a child I couldn't stand being separated from my mother.

85. My mother can make me feel really good but when she is not nice to me she can
really tear me apart.

86. In my family of origin we don't make a show of expressing our feelings. We prefer
keeping feelings to ourselves.

87. Neither my mother nor myself are perfect but somehow we made it through my
childhood.

88. I remember when I was frightened as a child my mother holding me close.

128

STRONGLY DISAGREE NEUTRAL AGREE STRONGLY

DISAGREE NEITHER DISAGREE AGREE
NOR AGREE
1 2 3 4 5

89. When I was a child my mother sometimes told me that if I was not good she
would stop loving me.

90. My mother is selfishly caught up in herself to the exclusion of everybody else.

91. My family was not particularly intimate, but this has never bothered me.

92. It's hard for me to remember my early relationship with my mother in any detail.

93. In childhood I sometimes felt that my mother and I were so alike that I didn't
know where she ended and I began.

94. If anything happened to my mother I wonder if I could survive it.

95. I remember as a child feeling a desire to protect my mother.

96. Even though I went through rough times with my mother during my childhood,
somewhere along the line I managed to let go of the majority of those angry, hurt
feelings.

97. In childhood I knew I was low on my mother's priority list.

98. My mother was an all-around excellent mother.

99. No one gets under my skin like my mother.

100. As a child I never thought separations fiom my parents were any big deal.

101. I often felt responsible for my mother's welfare.

102. In childhood my mother sometimes threatened to leave me or to send me away it I
wasn't good.

103. To this day my mother has no clue who I am or what I am all about.

104. Even with all our past difficulties, I realize my mother did the best for me that she
could.

105. I have forgotten what most of my early childhood was like.

106. I always knew my mother was there for me; no matter what I could depend on

her.

129

STRONGLY DISAGREE NEUTRAL AGREE STRONGLY

DISAGREE NEITHER DISAGREE AGREE
NOR AGREE
1 2 3 4 5

107. There are times when I feel like shaking my mother and saying "wake up and see
me for who I am".

108. In childhood I often had the impression that my mother was not listening to me.
She often tuned me out.

109. During my childhood I sometimes felt like I was my mother's whole life.

1 10. My mother and I are more accepting of each other's differences, than we have
been in the past.

111. When I was young I often feared something dreadful would happen to my mother
or father.

1 12. I remember my mother telling me that I didn't pay enough attention to her or love
her enough.

113. I often take my mother's opinion about me to heart and lose sight of my own
opinion about myself.

1 14. My mother is a real nag.

115. My mother and I were so alike we often could finish each others sentences.

1 16. I think people put too much emphasis on the mother/child relationship.

117. I remember very little about my early childhood (ages three to seven).

118. The concept of the loving, supporting mother is pure myth.

119. My relationship with my mother has gone through major changes over the course
of my childhood and adolescence.

120. Even as an adult I sometimes feel like I will never dig myself out fi'om under my
mother's influence.

121. As a child I sometimes got the feeling that without me my mother would have
fallen apart.

122. I couldn't have asked for a better mother.

130

STRONGLY DISAGREE NEUTRAL AGREE STRONGLY

DISAGREE NEITHER DISAGREE AGREE
NOR AGREE
1 2 3 4 5

123. If my mother was not fair to me as a child I realize now it was because she was
dealing with her own problems.

124. If something really bad happened to me in childhood I did not feel I could count
on my mother to support me.

125. When I was a child I sometimes got the feeling that my mother wished I was
never born.

126. I remember when I was a child feeling scared that one or both of my parents
would die unexpectedly.

127. My mother can devastate me with her criticisms.

128. In childhood my mother often told me she was sacrificing herself for me.

129. I don't think my early childhood relationship with my mother has any significant
influence on who I am today or my present relationships.

130. My mother was always there for me when I needed her.

131. When I acted bad as a child my mother would, at times, threaten to send me away.

132. I never felt like my mother gave me enough attention.

133. For all our past problems my mother and I can still enjoy a good laugh together.

134. During my childhood my mother would often turn to me and tell me lots of things
that upset and bothered her.

135. In childhood I often worried about my mother's state of health.

136. I find it difficult to remember my early childhood.

137. My mother was a perfect mother.

138. My mother's issues are still interfering with my life.

139. When I think back to my early childhood experiences I discover things about

myself and my parents that I've never considered before.

131

Family History Questionnaire

These questions ask about your family history. For each question, blacken the number
that best describes how you feel. Although some of these questions are of a personal
nature, please try to answer as honestly as you can. Your answers will be kept

confidential.
140. Has your mother ever suffered from alcoholism, depression, or anything else that
might be considered a psychological problem? 1=Yes 2= No
141. If yes, have these difficulties ever affected your relationship with hag?
l= never/not applicable 2= rarely 3= sometimes
4=often 5= very often
142. Has your father ever suffered from alcoholism, depression, or anything else that
might be considered a psychological problem? 1=Yes 2= No
143. If yes, have these difficulties ever affected your relationship with hi_m_?
l= never/not applicable 2= rarely 3= sometimes
4=often 5= very often
144. Has your mother ever attempted suicide?
1= never 2: once 3: several times
4= many times 5= committed suicide
145. Has your father ever attempted suicide?
1= never 2= once 3= several times
4= many times 5= committed suicide
146. Have either of your parents had serious problems with the “law”?
1= never 2= rarely 3= sometimes 4=often 5= very often
147. Were you brought up by your parents? 1=Yes 2= No
148. What was your relationship with your mother like when you were younger?
l= very poor 2= poor 3= ok 4= good 5= excellent
149. What was your relationship with your father like when you were younger?

1= very poor 2= poor 3= ok 4= good 5= excellent

PLEASE GO TO COMPUTER SHEET {2;

132

Please begin with #1 on computer scoring sheet #3.

l.

10.

11.

What is your relationship with your mother like now?

l= very poor 2= poor 3: ok 4= good 5: excellent

6= no contact 7= she is deceased

What is your relationship with your father like now?

1: very poor 2= poor 3: ok 4= good 5= excellent

6= no contact 7= she is deceased

Did you ever witness physical violence between your parents while growing up?
1: never 2= rarely 3= sometimes 4=often 5= very often

Were you ever physically abused while growing up?
1= never 2= rarely 3= sometimes ‘ 4=often 5= very often

Were you ever verbally/emotionally abused while growing up?
1= never 2: rarely 3= sometimes 4=often 5= very often

When you were growing up, were you ever able to confide in your parents?
l= never 2: rarely 3= sometimes 4=often 5= very often

When you were growing up, did you ever feel like your parents understood you?
l= never 2= rarely 3= sometimes 4=often 5= very often

When you were growing up, did you ever feel loved and respected by your parents?
1= never 2= rarely 3= sometimes 4=often 5= very often

Now, are you ever able to confide in your parents?
l= never 2= rarely 3= sometimes 4=often 5= very often

Now, do you ever feel like your parents understand you?
l= never 2= rarely 3= sometimes 4=often 5= very often

Now, do you ever feel loved and respected by your parents?
1= never 2= rarely 3= sometimes 4=often 5= very often

133

Thoughts and Feelings

Using the scale below, please indicate how often during the PAST YEAR you have
felt or behaved in the following ways:

Never/ Always/
None of A great
the time many times

1 2 3 4 5

In the past year:

12.
13.
14.
15.
16.
17.
18.
19.
20.
21.

I have been thinking of ways to kill myself.

I have told someone I want to kill myself.

I believed my life will end in suicide.

1 have made attempts to kill myself.

I feel life just isn’t worth living.

Life is so bad I feel like giving up.

I just wish my life would end.

It would be better for everyone involved if I were to die.

I feel there is no solution to my problems other than taking my life.

I have come close to taking my own life.

134

N__9_w, please indicate how often In your LIFETIME (prior to the past year) you have

felt or behaved In the following ways:

Never/ Always/

None of A great

the time many times
1 2 3 4 5

In your lifetime (prior to the past year):

22. I thought of ways to kill myself.

23. I told someone I wanted to kill myself.

24. I believed my life would end in suicide.

25. I made attempts to kill myself.

26. I felt life just wasn’t worth living.

27. Life was so bad I felt like giving up.

28. I just wished my life would end.

29. It thought it would be better for everyone involved if I were to die.
30. I felt there was no solution to my problems other than. taking my life.

31. I came close to taking my own life.

135

Instructions: The following list of statements describes different aspects of students'
relationships with their father. In many cases, the primary male parent-figure/caregiver is
the "father". If you consider someone else to be your primary male parent-figure, please
respond to the questions that refer to “father” with that person in mind. Imagine a scale
ranging fiom l to 5 that tells how well each statement applies to you. Please blacken in
the appropriate number from "1" (Not at all true of me) to "5" (Very true of me). If the
statement does not apply, enter "1". Please be completely honest. Your answers are
entirely confidential and will be useful only if they accurately describe you.

Not at all true of me Very true of me
1 2 3 4 5

32. Sometimes my father is a burden to me.

33. I feel like I am constantly at war with my father.

34. I blame my father for many of the problems I have.

35. I wish I could trust my father more.

36. I have to be careful not to hurt my father's feelings.

37. I sometimes feel like I'm being punished by my father.

38. I wish my father wasn't so overprotective.

39. I wish my father wouldn't try to manipulate me.

40. I wish my father wouldn't try to make fun of me.

41. I feel that I have obligations to my father that I wish I didn't have.

42. My father expects too much from me.

43. I wish I could stop lying to my father.

44. I often wish that my father would treat me more like an adult.

45. I am often angry at my father.

46. I hate it when my father makes suggestions about what I do.

47. Even when my father has a good idea I refuse to listen to it because he made it.

48. I wish my father wouldn’t try to get me to take sides with him.

49. I argue with my father over little things.

50. My father is sometimes a source of embarrassment to me.

51. I am sometimes ashamed of my father.

52. I get angry when my father criticizes me.

5 3. When I don’t write my father often enough I feel guilty.

136

Not at all true of me Very true of me
1 2 3 4 5

54. I feel uncomfortable keeping things from my father.
55. I often have to make decisions for my father.

56. I sometimes resent it when my father tells me what to do.

137

BSI
The following is a list of problems people sometimes have. Please read each one
carefully, and blacken the circle that best describes HOW MUCH THAT PROBLEM
HAS DISTRESSED OR BOTHERED YOU DURING THE PAST 7 DAYS
INCLUDING TODAY. Blacken the circle for only one number for each problem and do
not skip any items.

Not at All A Little Moderately Quite A Extremely
Bit Bit
1 2 3 4 5

You should be on #5 7 of computer sheet #3.

During the past 7 days, including today, how much were you distressed by:

57. Nervousness or shakiness inside

58. F aintness or dizziness

5 9. The idea that someone else can control your thoughts
60. Feeling others are to blame for most of your troubles
61. Trouble remembering things

62. Feeling easily annoyed or irritated

63. Pains in heart or chest

64. Feeling afi'aid in open spaces or on the streets

65. Thoughts of ending your life

66. Feeling that most people cannot be trusted

67. Poor appetite

68. Suddenly scared for no reason

69. Temper outbursts that you could not control

70. Feeling lonely even when you are with people

71. Feeling blocked in getting things done

72. Feeling lonely

73. Feeling blue

74. Feeling no interest in things

7 5. Feeling fearfirl

76. Your feelings being easily hurt

138

Not at All A Little Moderately Quite A Extremely

Bit Bit
1 2 3 4 5

During the past 7 days, including today, how much were you distressed by:

77.
78.
79.
80.
81.
82.
83.
84.
85.
86.
87.
88.
89.
90.
91.
92.
93.
94.
95.
96.
97.
98.
99.

Feeling that people are unfriendly or dislike you
Feeling inferior to others

Nausea or upset stomach

Feeling that you are watched or talked about by others
Trouble falling asleep

Having to check and double-check what you do
Difficulty making decisions

Feeling afraid to travel on buses, subways, or trains
Trouble getting your breath

Hot or cold spells

Having to avoid certain things, places, or activities because they fiighten you
Your mind going blank

Numbness or tingling in parts of your body

The idea that you should be punished for your sins
Feeling hopeless about the future

Trouble concentrating

Feeling weak in parts of your body

Feeling tense or keyed up

Thoughts of death or dying

Having urges to beat, injure, or harm someone
Having urges to break or smash things

Feelings very self-conscious with others

Feeling uneasy in crowds, such as shopping or at a movie

139

Not at All A Little Moderately Quite A Extremely
Bit Bit
1 2 3 4 5
During the past 7 days, including today, how much were you distressed by:
100.Spells of terror or panic
101.Getting into frequent arguments
102.Feeling nervous when you are left alone
103.0thers not giving you proper credit for your achievements
104.Feeling so restless you couldn’t sit still
105.Feelings of worthlessness
106.Feeling that people will take advantage of you if you let them
107.Feelings of guilt

108.The idea that something is wrong with your mind

140

Ways of Copipg Checklist

The following items describe ways people sometimes cope in difficult situations. Please
rate each of these items for how likely you would be to use it (on a scale of 1 to 4) in a
stressful situation, such as those listed in the stressful Life Events Scale you filled out
earlier today.

Would Would use
Not use a great deal
1 2 3 4

You should be on #109 of computer sheet #3.

109.Change something so things would turn out all right.

110.Accept sympathy and understanding from someone.

111.Keep my feelings to myself.

112.Accept the next best thing to what I wanted.

ll3.Ask someone I respect for advice and follow it.

114.Try not to burn my bridges behind me, but leave things open somewhat.

115.1 know what has to done, so I will double my efforts and try harder to make things
work.

116.Think about fantastic or unreal things (like perfect revenge or finding a million
dollars) that make me feel better.

117.Just take things one step at a time.

118.Wish that I could change what had happened.

119.Wish I was a stronger person--more optimistic and forceful.

120.Have fantasies or wishes about how things might turn out.

121.Change something about myself so I could deal with the situation better.

122.Try not to act too hastily or follow my own hunch.

123.Get professional help and do what they recommend.

124.Keep others from knowing how bad things are.

125.Come up with a couple of different solutions to the problem.

126.Try to forget the whole thing.

127.Realize I brought the problem on myself.

141

Would Would use
Not use a great deal
1 2 3 4

128.Hope a miracle would happen.

129.Daydream or imagine a better time or place than the one I was in.

130.Go on as if nothing had happened.

131.Bargain or compromise to get something positive from the situation.

132.Refuse to believe it had happened.

133.Accept my strong feelings, but not let them interfere with other things too much.
134.Ta1k to someone to find out about the situation.

135.F eel bad that I couldn't avoid the problem.

136.Concentrate on something good that could come out of the whole thing.
l37.Make a plan of action and follow it.

138.Blame myself.

139.Stand my ground and fight for what I wanted.

140.Ta1k to someone about how I was feeling.

141.Wish I could change the way I felt.

142.People's worth should be judged by what they achieve.

143.Talk to someone who could do something about the problem.

144.Wish the situation would go away or somehow be finished.

145.Avoid being with people in general.

146.Sleep more than usual.

147 .Try to make myself feel better by eating, drinking, smoking, or taking medications.
148.Get mad at the people or things that caused the problem.
149.Change or grow as a person in a good way.

150.Criticize or lecture myself.

PLEASE GO TO COMPUTER SHEET EA

142

Influential Relationship Questionnaire

Please begin with #1 on computer scoring sheet #4.

1.

Has there ever been someone in your life (besides your parents) who you felt helped
you to relate better to other people or feel better about yourself?
1= Yes 2= No

If yes, did this person ever have a significant influence on your life?
1: never/not applicable 2= rarely 3= sometimes 4=often 5: very often

Has there ever been someone in your life (besides your parents) who you felt _11u_rt_ the
way you feel about yourself or relate to other people? l= Ye52= No

If yes, did this person ever have a significant influence on your life?
1: never/not applicable 2: rarely 3= sometimes 4=often 5= very often

Have you ever been in therapy? 1= Yes 2= No

If yes, did you ever find therapy helpful
l= never/not applicable 2= rarely 3= sometimes 4=often 5= very often

If yes, have you ever found yourself in a situation where you thought about what your
therapist would say or suggest?

l= never/not applicable 2= rarely 3= sometimes =often 5: very often
Have you ever looked up to someone (besides your parents) and tried to be more like
them or considered them to be a mentor? 1= Yes 2= No

If yes, did this person ever have a significant influence on your life?
1: never/not applicable 2= rarely 3= sometimes 4=often 5: very often

143

SITUATIONAL SURVEY
INSTRUCTIONS: Please read these instructions carefirlly.

On each of the next 7 pages is a short story. Following each story are 4 questions with a
choice of 5 answers for each. The four questions relate to the following four kinds of
behavior: actual behavior, thoughts, feelings, and impulsive behavior in fantasy. Of the
four, it is only actual behavior which is outwardly expressed; the other three take place
only in the privacy of one's mind and, therefore, have no external repercussions.

What we want you to do is to select the pn_e answer of the five which you think is the
alps; representative of how you would react, and mark the number corresponding to that
answer on the computer answer sheet by darkening the space marked three (3) next to that
number. Then select the % answer you think is M representative of how you would
react and mark it by darkening the space marked one (1) next to that number. The ath_er
three responses should be marked as two (2).

For example, let us assume that out of the five possible answers to a question (e. g.,
numbers 66, 67, 68, 69, 70), response number 67 is the one you consider mas;
representative of the way you would react, and response number 70 is the l_ea_s_t
representative. In this example, you would mark a three (3) next to number 67, a one ( l )
next to number 70, and a two (2) next to numbers 66, 68, and 69.

Read a_ll the five answers following the question before you make your selections. In
marking your answers on the computer sheet, be sure that the number of the answer
agrees with the number on the computer sheet.

 

144

You are waiting for the bus at the edge of the road. The streets are wet and muddy
after the previous night's rain. A motorcycle sweeps through a puddle in fi'ont of you,
splashing your clothing with mud.

For each question below:
1= Least representative of you 3= Most representative of you
The other remaining responses should be marked as 2.

What would your ACTUAL reaction be?

10. I would try and remember the biker's face so I could find him later.
11. I'd wipe myself off with a smile.

12. I'd yell Obscenities after the biker.

13. I'd scold myself for not having at least worn a raincoat.

14. I'd shrug it off since things like that happen all the time.

What would mu IMPULSIVELY (in fa_nt__asyl want to do?

15. Wipe that biker's face in the mud.

16. Tell the police about the biker since he probably does this all the time.
17. Kick myself for standing so close to the edge of the road.

18. Let the biker know that I really didn't care that he splashed me.

19. Let the biker know that bystanders also have rights.

What THOUGHT might occur to you?

20. Why do I always get myself into things like this?

21. To hell with that biker!

22. I'm sure that basically that biker is a nice person.

23. You can expect something like this to happen on wet days.
24. I wonder if that biker splashed me on purpose.

How would you FEEL and why?

25. Satisfied, after all, it could have been worse.

26. Depressed, because of my bad luck.

27. Like shrugging my shoulders, because a person can't let things like that bother him.
28. Resentment, because the biker was so careless and mean.

29. Furious, that the motorcyclist got me dirty.

145

You are spending your vacation visiting an old friend who has moved with his parents to
another town. He invites you to go with him to a fair given that weekend at the
community clubhouse. Shortly after you arrive, he accepts an invitation to go out in a
canoe with another friend, leaving you with a group of strangers to whom you have
barely been introduced. They talk with you, but while some of them have canoes, for
some reason no one asks you to go canoeing. Your friend, on the other hand, seems to be
very popular that day. He looks as if he is having a wonderful time. As he paddles past,
he calls out to you, "Why don't you go out in one of the canoes?"

For each question below:
1= Least representative of you 3= Most representative of you
The other remaining responses should be marked as 2.

What would [our ACTUAL reaction be?

30. I'd say sarcastically, "I'm not canoeing because I'd “rather watch you.”

31. I'd tell him that I really didn't feel like going out in a canoe.

32. I'd wonder what's wrong with me.

33. I'd tell him that it's easier to get to know his fiiends by talking to them than it would
be by going out in a canoe.

34. I'd get up and leave because he apparently wants to embarrass me.

What would you IMPULSIVELY (in fantasy) want to do?

35. Assure him that I am perfectly satisfied and happy, so he won't worry.

36. I'd like to punch him in the nose.

37. Point out that you cannot expect to be everybody's friend on your first day in a
strange place.

38. Tell him that now I know what sort of a “friend" he really is.

39. I'd like to sink into the ground and disappear.

_Wflu THOUGHT might occur to you?

40. He has it in for me.

41. I should never have come here in the first place.

42. I'm glad my fiiend is enjoying himself.

43. Something like this can't be avoided in a place where you don't know the crowd.
44. I'll make him sorry for his behavior.

How would you FEEL and why?

45. Upset, because I was so unsuccessful.

46. Furious at him for the embarrassment.

47. Resigned, because this is the kind of situation every newcomer must put up with once
in a while.

48. Angry to find that my friend is so disloyal.

49. Grateful, for having had such a pleasant day.

146

On your way to school, you are hurrying through a narrow street lined with tall buildings.
Suddenly a piece of brick comes crashing down from a roof where some repairmen are
working. The brick bounces off the sidewalk, bruising your leg.

For each question below:
1= Least representative of you 3= Most representative of you
The other remaining responses should be marked as 2.

What would your ACTUAL fiction be?

50. I'd tell the repairmen I was going to get my parents after them.
51. I'd be mad at myself for having such bad luck.

52. I'd hurry on so I wouldn't be late for school.

53. I'd continue on my way, happy that nothing worse had happened.
54. I'd try to discover who those irresponsible people were.

What would you IMPULSIVELY (in fantasy) want to do?

55. Remind the repairmen that they should be more careful.

56. Make sure the repairmen knew that nothing serious had happened.
57. Give them a piece of my mind.

58. Kick myself for not having watched where I was going.

59. See to it that those careless workers pay for their negligence.

What THOUGHT might occur to you?

60. Those repairmen don't know how to do their job right.
61. I'm lucky that I wasn't badly hurt.

62. Damn those men!

63. Why do such things always happen to me?

64. One can't be too careful these days.

How would you FEEL and why?

65. Angry, because I was hurt?

66. Furious, because I could have been killed by their carelessness.
67. Cahn, because one should always be able to keep one's cool.
68. Upset by my bad luck.

69. Thankful that I'd gotten away with no more than a scratch.

147

You and a school friend are competing for president of your class. Although both
of your chances seem about equal, your friend has been in school longer and is therefore
more popular. Recently, however, you have had a party at which everyone in the class has
had a good time. You are sure you are now very well-liked by all of them. However, your
friend wins as president.

For each question below:
1: Least representative of you 3= Most representative of you
The other remaining responses should be marked as 2.

What would your ACTUAL reaption be?

70. I'd try to find out which persons in the class didn't vote for me and get even.

71. I'd do my best to continue to behave as I did before the election, as a true fiiend
should. '

72. I'd accept the outcome as proof that I wouldn't have made as good a president as my
friend.

73. I'd refuse to cooperate with the new president.

74. I'd congratulate my fiiend on this victory.

What would You IMPULSIVELY (in fan—tasy) want to do?

75. Support the results as a good citizen should.

76. Kick myself for ever running for president, when I knew I wasn't as good as the other
candidate.

77. Show the class how mistaken they had been in voting for such an incompetent person.

78. Help my friend to be a good president.

79. Break the neck of each and every member of the class who voted against me.

What THOUGHT nright occur to you?

80. I guess I just don't have what it takes to be president.

81. I probably wouldn't like being president as much as I though I would.
82. There certainly is something fishy about the class's decision.

83. You can't let a failure get you down.

84. Who cares about the future of this class, anyway!

 

How would you FEEL and whE?

85. Happy that I still have my old fiiends.

86. Upset because my defeat is known throughout the school.
87. Furious at the class because of their treatment of me.

88. Shrug it off, because that's the way the cookie crumbles.

89. Angry, because I have been the victim of an unfair decision.

148

You and John, one of your best friends, are playing catch. Unexpectedly Tammy,
another friend from the next street, drops over. You introduce Tammy to John and you all
play ball together. You are supposed to go to the amusement park with Tammy in two
days and you are really looking forward to it. The day that you are supposed to go,
Tammy calls up and says she can't go with you because she has to go over to her
grandmother's. You decide to join some other friends going to the amusement park. At
the amusement park you see Tammy and John riding the roller coaster.

For each question below:
1= Least representative of you 3= Most representative of you
The other remaining responses should be marked as 2.

What woulg your ACTUAL reaction be?

90. I'd snub them because I'm sure they'd try to pretend that they didn't see me.
91. I'd greet them politely as a civilized person should.

92. I'd curse them under my breath.

93. I'd tell them that I was glad that they had become fiiends.

94. I'd go home and sulk in my room.

What would you IMPULSIVELY (in fantasy) want to do?

95. Hide somewhere in order to avoid facing them.

96. Punch them in the nose.

97. Show them that I didn't mind that they were together.

98. Ask John if stealing your fiiend is the only way he knows of getting to the amusement
park.

99. Show that you understand why they became friends.

What THOUGHT might occur to you?

100.Naturally Tammy liked John since he is so much better looking than I am.
101.Getting what you want can cause you to be disloyal to a fiiend.

102.They certainly are a pair of double-crossers.

103.1 hope they get what they deserve.

104.They really seem to get along well together.

How would you FEEL and why?

105.Pleased that my friends get along so well.

106.Upset, because I shouldn't have been so trusting

107.Shrug it off because one has to take things like this in one’s stride.
108.Really mad because they lied to me.

109.Furious at them, because of what happened.

149

You have a paper route on which there are many people. You have to work very
hard because you need the money and jobs are scarce. It is your responsibility to make
sure everything runs smoothly. You have a classmate who helps you deliver the papers.
Recently many people have been complaining about not getting their papers. You know
you have been careful in doing your job, so you decide to fire your helper. That same day
your boss from the paper drops over at your house. Without letting you explain, your boss
says the paper route is being taken away from you because you are careless. Your
assistant is assigned your job and you are now in the position of helper.

For each question below:
1= Least representative of you 3= Most representative of you
The other remaining responses should be marked as 2.

What would your ACTUAL reaction be?

110.I'd be a good sport about it, since the boss is only doing his job.

lll.I'd blame the boss for having made up his mind against me even before the visit.
112.I'd be thankful to get rid of such a tough job.

113.I'd look for a chance to make things hard for the assistant.

114.I'd blame myself for not being good enough for the job.

What would you IMPULSIVELY (in farlasy) want to do?

115.Congratulate my assistant on getting the paper route.

116.Try to find out if the boss from the paper and my helper had worked together to fix it
so I would lose the paper route.

117.Tell my assistant to go to hell.

118.I'd like to kill myself for not having done something about my helper sooner.

119.I'd like to quit, but I don't know where I could get another job.

What THOUGHT might occur to you?

120.The boss deserves a screwball for a paper carrier.

121.To sell papers you have to have the right person in the right job.

122.There is no doubt that this was just an excuse to get rid of me.

123.1'm really lucky that I only lost my job and didn't have to pay for papers not
delivered.

124.How could I be so dumb?

How would you FEEL and why?

125.Resentful, because the boss had it in for me.

126ng, at my assistant for getting my job.

127.Pleased that nothing worse had happened.

128.Upset that I am a failure.

129.Resigned, after all, you have to be satisfied with having done all that you can.

150

You have spent the last two summers working in a bicycle repair shop. At the
time you started you had a choice between bicycle repair or working with your father.
You preferred the other job despite your father's advice. Now that the repair shop has
closed, you find that there are no other jobs for the smnmer. You can either go to work
with your father or you can do odd jobs. You would like to repair the bikes of kids in the
neighborhood but you don't have the necessary tools. After a great deal of hesitation, you
decide to ask your dad to put up the money. After listening to your idea, he reminds you
that he wanted you to work for him instead of at the repair shop. The he tells you, "I'm
not prepared to throw away my hard-earned money on your crazy schemes. It's time you
started helping me in my business."

For each question below:
1: Least representative of you 3= Most representative of you
The remaining three responses should be marked as 2.

What would your ACTUAL reaction be?

130.I'd accept his offer since everyone depends on everyone else in this world.
131.I'd admit to him that maybe he could lose his money if something went wrong.
132.I'd tell him off very strongly.

133.1'd tell him that I'd always thought that he had a grudge against me.

134.1'd thank him for still wanting me to work for him.

What would you IMPULSIVELY (in fan_ta_sy) want to do?

135.60 to work for him and make him happy.

136.Give up trying and end it all.

137.Take my father's offer since jobs don't grow on trees.

138.Let him know what a rrriser everyone thinks he is.

139.Tell him that I wouldn't work for him if he were the last person on earth.

What THOUGHT mig1_rt occur to you?
140.He'll get what's coming to him one day.

141.You have to stand on your own two feet since your family won't always be around.
142.Why was I so stupid as to bring the subject up.

143.1 must admit that my father is acting for my own good.

144.This proves what I've suspected all along, that my father has never believed in me.

How would you FEEL and why?

145.Angry, because he doesn't want me to succeed on my own.
146.Grateful for his offer of a job.

147.Resentful that he is hurting me.

148.Resigned, since you can't have everything your own way all the time.
149.Hopeless, because my father won't help me.

151

THANK YOU FOR COMPLETING THESE
QUESTIONNAIRES!!!

Please make sure you pick up a debriefing sheet explaining more about this study before
you leave. Also, please remember that there will be an oral debriefing explaining more
about this study as soon as everyone participating in this group session has finished the
questionnaires.

152

REFERENCES

Adam, K. S. (1994). Suicidal behavior and attachment: A developmental model.
In M. B. Sperling and W. H. Bennan (Eds), Attachment in adults: Clinical and
developmental perspectives (pp. 275-298). New York: The Guilford Press.

Adam, K. S., Sheldon-Keller, A. E., & West, M. (1996). Attachment organization
and history of suicidal behavior in clinical adolescents. Journal of Consulting and
Clinicaa Psychology. 64(2). 264-272.

Ainsworth, M. (1968). Object relations, dependency, and attachment: A
theoretical review of the infant-mother relationship. Child Development. 40. 969-1025.

Ainsworth, M., Blehar, Waters, & Wall (1978). Paiema of Attaphment: A
psychological study of the strange situation. Hillsdale, NJ: Lawrence Erlbaum
Associates, Publishers.

Allen, J. P., Hauser, S. T., Bell, K. L., & O'Connor, T. G. (1994). Longitudinal
assessment of autonomy and relatedness in adolescent family interactions as predictors of
adolescent ego development and self esteem. Child Development, 65(1), 179-194.

Arbuckle, J. (1997). Amos Users’ Guide Version 3.6. SmallWaters Corporation:
Chicago, Il.

Armsden, G. C., McCauley, E., Greenberg, M. T., Burke, P. M., & Mitchell, J. R.
(1990). Parent and peer attachment in early adolescent depression. Journal of Abnormal

Child Psychology, 18(6), 683-697.

Asamow, J. R. & Carlson, G. (1988). Suicide attempts in preadolescent child
psychiatric inpatients. Suicide and Life-Threatening Behavior. 18. 129-136.

Asamow, J. R. (1992). Suicidal Ideation and Attempts During Middle Childhood:
Associations With Perceived Family Stress and Depression Among Child Psychiatric
Inpatients. Journal of Clinical Child Psychology, 21:1, 35-40.

Asamow, J. R., Carlson, G. A., Guthrie, D. (1987). Coping strategies, self
perceptions, hopelessness, and perceived family environments in depressed and suicidal

children. Journal of Consulting and Clinical Psychology, 55(3 ), 361-366.

Bartholomew, K. (1990). Avoidance of intimacy: An attachment perspective.

Journal of Social and Personal Relationships, 7(2), 147-178.

Bartholomew, K., Horowitz, L. M. (1991). Attachment styles among young

153

adults: A test of a four category model. Journal of Personality and Social Psychology,
61(2), 226-244.

Beck, A. T., & Steer, R. A. (1984). Internal consistencies of the original and
revised Beck Depression Inventory. loam] of Clinical Psychology, 40(6), 1365-1367.

Beck, A. T., Epstein, N., Brown, G., & Steer, R. A. (1988). An inventory for
measuring clinical anxiety: Psychometric properties. Journal of Consulting and Clinical
Psychology. 56(6). 893-897.

Beck, A. T., Steer, R. A., & Garbin, M. G. (1988). Psychometric properties of the
Beck Depression Inventory: Twenty-five years of evaluation. Clinical Paychology
Review. 8(1). 77-100.

Beck, A. T., Steer, R. A., Garbin, M. G. (1988). Psychometric properties of the
Beck Depression Inventory: Twenty-five years of evaluation. Clinipal Psycholcgy
Review 8 77-100.

 

Beck, A. T., Steer, R. A., Kovacs, M., & Garrison, B. (1985). Hopelessness and
eventual suicide: A 10 year prospective study of patients hospitalized with suicidal
ideation. American Journal of Psychiatry. 142(5). 559-563.

Bettes, B. A., & Walker, E. (1986). Symptoms associated with suicidal behavior
in childhood and adolescence. Journal of Abnormal Child Psychology. 14(4). 591-604.

Binnaher, B., Ryan, N. D., Williamson, D. E., & Brent, D. A. (1996). Childhood
and adolescent depression: A review of the past 10 years, Part 1. Journal of the Amean
Academy of Child and Adolescent Psychiat_ry, 35(11), 1427-1439.

Blass, R. B., & Blatt, S. J. (1992). Attachment and Separateness: A theoretical
context for the integration of object relations theory with self psychology.

Blatt, S. J., & Schichman, S. (1983). Two primary configurations of
psychopathology. In L. Goldberger & D. Shapiro (Ed.), Psychoanalysis and
Contemporary Thougm (pp. 187-254). Chicago: University of Chicago Press.

Bloom, M.V. (1980). Adolescent parental separation. New York: Gardener
Press.

Blos, P. (1979). The adolescent passage. New York: International Universities
Press.

Blumenthal, S., & Kupfer, D. (Eds), (1990). Suicide Over the Life Cycle- Risk

Factors, Assessmentaand Treatment of Suicidal Patients, Washington DC: American
Psychiatric Press Inc.

154

Borst, S. R., & Noam, G. G. (1993). Developmental psychopathology in suicidal
and nonsuicidal adolescent girls. Journal of the American Academy of Child and
Adolescent Psychiag, 32(3), 501-508.

Bowlby, J. (1977). The making and breaking of affectional bonds 1.: Aetiology
and psychopathology in the light of attachment theory. Briti§_h J ouraal of Psychiafl,
130, 201 -21 O.

Bowlby, J. (1988). A Secure Base: Parent-child Attaa3hment and Healthy Humg
Development. New York: Basic Books.

Brennan, K. A., Shaver, P. R., & Tobey, A. E. (1991). Attachment styles, gender
and parental problem drinking. Journal of SociL and Personal Relationships, 8(4), 451-
466. ’

Brent, D. A., Kolko, D. J., Allan, B. S., & Brown, B. A. (1990). Suicidality in
Affectively Disordered Adolescent Inpatients. Journal of the Americaa Afldemy of
Child and Adolescent Psychiatry. 29:4. 586-593.

Brent, D. A., Perper, J. A., Moritz, G., Baugher, M., Schweers, J ., & Roth, C.
(1994). Suicide in Affectively Ill Adolescents: A Case-control Study. Journal of
Affective DisorderLB L 193-202.

Bretherton, I. (1985). Attachment theory: Retrospect and prospect. Monoggaphs
of the Societygfor Research in Child Development. 50 (1-2, Serial No. 209), 3-38.

Bretherton, I. (1992). The origins of attachment theory: John Bowlby and Mary
Ainsworth. Developmental Psychology, 28(5), 759-775.

Camelley, K. B., Pietromonaco, P. R., & Jaffe, K. (1994). Depression, working
models of others, and relationship functioning. Journal of Personality and Social

Psychology, 66(1), 127-140.

Cashdan, S. (1988). Object Relations Therapy. New York: W. W. Norton &
Company.

Cassidy, J. (1988). Child—mother attachment and the self in six-year olds. Child
Development, 59, 121-134.

Cassidy, J ., & Berlin, L. J. (1994). The insecure/ambivalent pattern of
attachment: Theory and research. Child Development, 65(4), 971-981.

Cohen-Sandler, R., Berman, A. L., & King, R. A. (1982). Life stress and
symptomatology: Determinants of suicidal behavior in children. Journal of the

155

American Academy of Child Psychiata'y, 21, 178-186.

Crarner, P. (1988). The Defense Mechanism Inventory: A review of research

and discussion of the scales. Journal of Personalig Assessment, 52(1), 142-164.

Curran, D. K. (1987). Adolescent Suicidal Behavior. Washington, D. C.:
Hemisphere Publishing Corporation.

Damon, W. (1983). Social and personally development : Infancy through

adolescence New York: W.W. Norton.

 

DeMan, A. F., Labreche-Gauthier, & Leduc, C. P. (1993). Parent-Child
Relationships and Suicidal Ideation in F rench-Canadian Adolescents. Journal of Genetic
Psychology, 154:1. 17-23.

Dodge, K., Price, J ., Bachorowski, J ., & Newman, J. (1990). Hostile attributional
biases in severely aggressive adolescents. Journal of Abnormal Psychology, 99, 385-392.

Edward, J ., Ruskin, N., & Turrini, P. (1991). Separation/individuation : theory
and application, New York : Gardner Press.

Erikson, E. (1959). Identity and the Life Cycle. NY: International Universities
Press.

Erikson, E. (1968). Identity: Youth and Crisis. NY: W.W. Norton.

Feeney, J. A., Noller, P., & Hanrahan, M. (1994). Assessing adult attachment.
In M. B. Sperling and W. H. Berman (Eds.), Attachment in adults: Clinical and
developmental perspectives (pp. 128-152). New York: The Guilford Press.

Feldman, M., & Wilson, A. (1997). Adolescent suicidality in urban minorities
and its relationship to conduct disorders, depression, and separation anxiety. J oumal of
the American rfldemy of Child afl Adolescent Psychiat_ry, 36(1), 75-84.

Fishler, P. H., Sperling, M. B., & Carr, A. C. (1990). Assessment of adult
relatedness: A review of empirical findings from object relations and attachment

theories. Journal of Personalig Assessment, 55(3 & 4), 499-520.

Folkman, S., & Lazarus, R. S. (1980). An analysis of coping in a middle aged
community sample. Journal of Health and Social Behavior, 21(3 ), 219-239.

Friedman, R. C., et al. (1984). Family history of illness in the seriously suicidal
adolescent: A life cycle approach. American Journal of Onhopsycmm. 54(3). 390-397.

Garland, A. F. & Zigler, E. (1993). Adolescent Suicide Prevention: Current

156

Research and Social Policy Implications. American Psychologist, 48:2, 169-182.

Goodyer, 1., Herbert, J ., Secher, S., & Pearson, J. (1997). Short-term outcome of
major depression: Comorbidity and severity of presentation as predictors of persistent
disorder. Journal of the American Academmf Childrenjand Adolescent Psychiatg,
36(2), 179-187.

Grotevant, H. D., & Cooper, C. R. (1985). Patterns of interaction in family
relationships and the development of identity exploration in adolescence. Child
Development. 56(2). 415-428.

Grotevant, H. D., & Cooper, C. R. (1986). Individuation in family relationships:
A perspective on individual differences in the development of identity and role taking
skill in adolescence. Human Development. 29(2). 82-100.

Hammen, C. L., Burge, D., Daley, S. E., Davila, J ., Paley, B., & Rudolph, K. D.
(1995). Interpersonal attachment cognitions and prediction of symptomatic responses to
interpersonal stress. Journal of Abnormal Psychology, 104(3), 436-443.

Harrington, R, Brendenkamp, D., Groothues, C., & Rutter, M. (1994). Adult
outcomes of childhood and adolescent depression: IH. Links with suicidal behaviours.
Journal of Child Psychology and Psychiag and Allied Disciplines, 35(7), 1309-1319.

Hazan, C., & Shaver, P. (1987). Romantic love conceptualized as an attachment
process. Journal of Personality and Social Psychology, 52, 511-524.

Heesacker, R. S., & Neimeyer, G. J. (1990). Assessing object relations and social
cognitive correlates of eating disorder. Joumfl of Coun_seling Psychology, 37(4), 419-
426.

Henry, C. S., Stephenson, A. L., Hanson, M. F., & Hargett, W. (1993).
Adolescent Suicide and Families: An Ecological Approach. Adoleacence. 28:110. 291-
308.

Hoffrnan, J. A. (1984). Psychological separation of late adolescents from their
parents. Journal of Coun_seling Psychology, 31(2), 170-178.

Hoffman, J. A., & Weiss, B. (1987). Family dynamics and presenting problems
in college students. Journal of Counseling Psychology, 34(2), 157-163.

Holinger, P. C., Offer, D., Barter, J. T., & Bell, C. C. (1994). Suicide and
Homicide Among Adolescenta, New York: The Guilford Press.

Hollis, C. (1996). Depression, family environment, and adolescent suicidal
behavior. Journal of the American Academy of Child and Adolescent Psychiag, 35(5),

157

622-630.

Husain, S. A. (1990). Current Perspective on the Role of Psychosocial Factors in
Adolescent Suicide. Psychiatric Annals, 20, 122-127.

Ihilevich, D., & Gleser, G. C. (1982). Evaluating mental health proggams : the
proggess evaluation, Lexington, Mass: Lexington Books.

Jong, M. L. de (1992). Attachment, individuation, and risk of suicide in late
adolescence. JouraaI of Youth and Adolescence, 21(3 ), 357-373.

Kaplan, K. J. (1988). Teaching individuals to live together. Transactional

Analysis Journal, 18(3), 220-230.

Kaplan, K. J ., & Worth, S. A. (1993). Individuation attachment and suicide
trajectory: A developmental guide for the clinician. Omega: Journal of Death and Dfl'pg,
27(3 ), 207-237.

Kashani, J. H., Beck, N. C., & Burk, J. P. (1987). Predictors of Psychopathology
in Children of Patients with Major Affective Disorders. Canadian Journal of Psychim
3_2_, 287-290.

 

Kashani, J. H., Ezpeleta, L., Dandoy, A. C., Doi, S., & Reid, J. C. (1991).
Psychiatric Disorders in Children and Adolescents: The Contribution of the Child’s
Temperament and the Parents’ Psychopathology and Attitudes. Canadian Journal of
Psychiary. 36. 569-573.

Kaufman, J. (1991). Depressive disorders in maltreated children. Journal of the
Ameriaan Academy of Child and Adolescent Psychiagg, 30(2), 257-265.

Kazdin, A. E., & Kagan, J. (1994). Models of dysfunction in deveIOpmental
psychopathology. Clinical Psychology: Science and Practice, 1(1), 35-52.

 

Kazdin, A. E., et al. (1983). Hopelessness, depression, and suicidal intent among
psychiatrically disturbed inpatient children. Journal of Consulting and Clinical

Psychology, 51(4), 504-510.

Kendall, P., Brady, E., & Verduin, T. (2001). Comorbidity in childhood anxiety
disorders and treatment outcome. J oumaL of the American Academy of Children and

Adolescent Psychiatgy, 40(7), 787-794.

Kenny, M. E., Moilanen, D. L., Lomax, R., & Brabeck, M. M. (1993).
Contributions of parental attachments to view of self and depressive symptoms among
early adolescents. Journal of Early Adolescence, 13(4), 408-430.

158

Khan, A. U. (1987). Heterogeneity of suicidal adolescents. Joumal of the
American Acaaemy of Child and Adolescent Psychiafl, 26(1): 92 96.

Kienhorst, C. W., de Wilde, & Diekstra, R. F. (1996). In L. Verhofstadt-Denive,
I. Kienhorst, & C. Braet (Eds.), Conflict and Development in Adolescence, Leiden
University: DSWO Press.

Kienhorst, C. W., Wolters, W. H., Diekstra, R. F. & Otte, E. (1987). A study of
the frequency of suicidal behavior in children aged 5 to 14. Journal of Child Psychology
and Psychiatry and Allied Disciplines. 28(1). 153-165.

King, C. A., Raskin, A. Gdowski, C. L., & Butkus, M. (1990). Psychosocial
factors associated with urban adolescent female suicide attempts. Journal of the
American Academy of Child aad Adolescent Psychiatry. 29(2). 289-294.

Kline, R. (1998). Principles and Practice of Structural Equation Modeling.
Guilford Press: NY, NY.

Kobak, R. R., & Sceery, A. (1988). Attachment in late adolescence: Working
models, affect regulation, and representation of self and others. Child Development. 59,

135-146.

Kohn, P. M.., Lafreniere, K., & Gurevich, M. (1990). The Inventory of College
Student's Recent Life Experiences: A decontaminated hassles scale for a special
population. Journal of Behgioral Medicine. 13(6). 619-630.

Kosky, R., Silbum, S., & Zubrick, S. (1986). Symptomatic Depression and
Suicidal Ideation: A Comparative Study with 628 Children. The Joum_al of Nervous aaa

Mental Disease. 174:9. 523-528.

Kovacs, M., Goldston, D., & Gatsonis, C. (1993). Suicidal behaviors and
childhood onset depressive disorders: A longitudinal investigation. Journal of the
American Academy of Child and Adolescent Psychiatay, 32(1), 8-20.

Krarup, G., Nielsen, B., Rask, P., Petersen, P. (1991). Childhood Experiences and
Repeated Suicidal Behavior. Acta Psychiatrica Scandinavica, 83, 16-19.

Kroger, J ., & Haslett, S. J. (1988). Separation individuation and ego identity
status in late adolescence: A two year longitudinal study. Journal of Youth afli

Adolescence, 17(1), 59-79.

Lay, K., Waters, E., Posada, G., & Ridgeway, D. (1995). Attachment security,
affect regulation, and defensive responses to mood induction. Monoggphs of the Society
for Research in Child Development, 60 (2-3, Serial No. 244), 179-196.

159

Lester, D. (1991). Depression and fears of individuation and attachment.

Transactional Analysis Journal, 21(4), 218-219.

Lester, D. (1994). Are there unique features of suicide in adults of different ages
and developmental stages? Omega: Journal of Death and Dya'ng, 29(4), 337-348.

Lichtenstein, J ., & Cassidy, J. (1991, April). The Inventormf Adult Attachment
(INV AA): Validation of a new mea_sure of adult attachment. Paper presented at the
biennial meeting of the Society for Research in Child Development, Seattle, WA.

Lopez, F. G. (1995). Contemporary attachment theory: An introduction with
implications for counseling psychology. The Counseling Psychologist, 23(3 ), 395-415.

Lopez, F ., Campbell, V., & Watkins, C. (1986). Depression, psychological
separation, and college adjustment: An investigation of sex differences. Journal of
Counseling Psychology, 33, 52-56.

Main, M., Kaplan, N., & Cassidy, J. (1985). Security in infancy, childhood, and
adulthood: A move to the level of representation. Monographs of the Society for
Resemh in Child Development. 50 (1-2, Serial No. 209), 66-106.

Marttunen, M. J ., Aro, H. M., & Lonnqvist, J. K. (1993). Precipitant Stressors in
Adolescent Suicide. Journal of the American Academy of ChiLand Adolescent

Psychiam, 32(6), 1178-1183.

McIntire, M. S., & Angle, C. R. (1975). Evaluation of suicide risk in
adolescents. Journal of Family Practice. 2(5), 339-341.

McKenry, P.C., Tishler, C.L., & Kelley, C. (1982). Adolescent Suicide: A
Comparison of Attempters and Nonattempters in an Emergency Room Population.
Clinical Pediag'cs. 21 :5. 266-270.

Meeus, Helsen, & Vollebergh. (1996). In L. Verhofstadt-Denive, I. Kienhorst, &
C. Braet (Eds), Conflict and Development in Adolescence, Leiden University: DSWO
Press.

Meyer, D., & Phillips, W. (1990). No safe place: Parental alcoholism and
adolescent suicide. American Journal of Psychotherapy. 44(4). 552-562.

Mitchell, J ., Varley, C., & McCauley, E. (1988). Depression in children and
adolescents. Children: Health Care. 16(4). 290-293.

Mitchell, S. A., & Black, M. J. (1995). Freud and Beyond. New York: Basic
Books.

160

Morrison, G. C., & Collier, J. G. (1969). Family treatment approaches to suicidal
children and adolescents. Journal of the American Acalemy of Child Psychiag, 8(1),
140-153.

 

Norbeck, J. S., Lindsey, A. M., & Carrieri, V. L. (1981). The development of an
instrument to measure social support. NursinaResearch, 30(5), 264-269.

Norbeck, J. S., Lindsey, A. M., & Carrieri, V. L. (1983). Further development of
the Norbeck Social Support Questionnaire: Normative data and validity testing. Nursing
Research. 32(1). 4-9.

Oppenheim, D., & Waters, H. S. (1995). Narrative processes and attachment
representations: Issues of development and assessment. Monoggaphs of the Society for
Research in Child Development. 60 (2-3, Serial No. 244), 197-215.

Papini, D. R., & Roggman, L. A. (1992). Adolescent perceived attachment to
parents in relation to competence, depression, and anxiety: A longitudinal study. Journal
of Early Adolescence. 12(4). 420-440.

Pfeffer, C. R., Plutchik, R., & Mizruchi, M. S. (1983). Suicidal and assaultive
behavior in children: Classification, measurement, and interrelations. American Journal
of Psychiatry. 140(2). 154-157.

Pfeffer, C. R. (1985). Self-Destructive Behavior in Children and Adolescents.
Psychiatric Clinics of North America, 8:2. 215-225.

Pfeffer, C. R. (1986). The Suicidal Child, New York: The Guilford Press.

Pfeffer, C. R. (1989). Suicide Among Youth: Per_spectives on Risk and
Prevention, Washington DC: American Psychiatric Press Inc.

Pine, F. (1990). Drive, Ego, Object, and Self: A Synthesis for Clinical Worla
New York: Basic Books.

Reynolds, W. M. (1991). Psychometric characteristics of the Adult Suicidal
Ideation Questionnaire in college students. Jouraal of Personaliiy Assessment, 56(2),
289-307.

Robbins, D. R., & Alessi, N. E. (1985). Depressive symptoms and suicidal
behavior in adolescents. American Journal of Psychiat_ry, 142(5), 588-592.

Roberts, J. E., Gotlib, I. H., & Kassel, J. D. (1996). Adult attachment security and
symptoms of depression: The mediating roles of dysfirnctional attitudes and low self-

esteem. Joumflf Personality and Social Psychology, 70(2), 310-320.

 

161

Rotheram-Borus, M. J. & Trautrnan, P. D. (1988). Hopelessness, Depression, and
Suicidal Intent among Adolescent Suicide Attempters. Journal of the American Academy
of Child and Adolescent Psychiag, 27(6), 700-704.

Rudd, M. D. (1989). The prevalence of suicidal ideation among college students.
Suicide and Life Threatening Behavior, 19(2), 173-183.

Salzman, J. P. (1996). Primary attachment in female adolescents: Association
with depression, self-esteem, and maternal identification. Psychiatry. 59. 20-33.

Schultheiss, D. P., & Blustein, D. L. (1994). Contributions of family relationship
factors to the identity formation process. Journal of Counselingand Development. 73(2).
159-166.

Schumacker, R. E., & Lomax, R. G. (1996). A Beginner's Guide to Structural
Equation Modeling. Mahwah, NJ: Lawrence Erlbaum Associates.

Seiden, R. H. (1984). Death in the West: a regional analysis of the youthful
suicide rate. Western Journal of Medicine. 140 ( 6). 969-976.

Shafii, M., Carrigan, S., Whittinghill, J ., & Derrick, A. (1985). Psychological
autopsy of completed suicide in children and adolescents. American Journal of
Psychiatry. 142(9). 1061-1064.

Shaver, P. R., & Clark, C. L. (1996). F orrns of adult romantic attachment and
their cognitive and emotional underpinnings. In G. G. Noam & K. W. Fischer (Eds),

Development and vulnerability in close relationships. The Jean Piaget syr_nposium series.
(pp. 29-58). Mahwah, NJ: Lawrence Erlbaum Associates, Inc.

Slade, A., & Aber, J. L. (1992). Attachments, Drives, and Development:
Conflicts and Convergences in Theory. In J. W. Barron, M. N. Eagle, & D. L. Wolitzky
(Ed.), Interfzae of Psychoanalysis and Psychology (pp. 154-185). Washington, DC:
American Psychological Association Press.

Spirito, A., Brown, L., Overholser, J ., & Fitz, G. (1989). Attempted Suicide in
Adolescence: A Review and Critique of the Literature. Clinical Psychology Review, 9,
335-362.

Sroufe, L. A. (1986). Appraisal: Bowlby’s contribution to psychoanalytic theory
and developmental psychology; attachment, separation, loss. Journal of Child

Psychology and Psychiag, 27 (6), 841-849.

Tabachnick, B., & F idell, L. (1996). Using Multivariate Statistics (3" ed.). New
York: HarperCollins College Publishers.

162

Taylor, E., 81. Stansfeld, S. (1984). Children who poison themselves I. A clinical
comparison with psychiatric controls. British Journal of Psychiat_ry, 145, 127-132.

Taylor, G. J. (1992). Psychosomatics and Self-Regulation. In J. W. Barron, M.

N. Eagle, & D. L. Wolitzky (Ed.), Intefice of Psychoanalysis and Psychology (pp. 489-
500). Washington, DC: American Psychological Association Press.

Tishler, C. L., McKenry, P. C., & Morgan, K. C. (1981). Adolescent suicide
attempts: Some significant factors. Suicide and Life-Threatening Behflr. 11. 86-92.

Van der Kolk, B. A., Perry, J. C., & Herman, J. L. (1991). Childhood origins of
self destructive behavior. Ameriaan Journaof Psychiatry. 148( 12). 1665-1671.

Withers, L. E., Kaplan, D. W. (1987). Adolescents who attempt suicide: A

retrospective clinical chart review of hospitalized patients. Professional Psychology:
Regarca and Practice. 18(4). 391-393.

163

IIIIIIIIIIIIIIIIIIIIIIIIIIIIII

it!)1111111))!!!I)\ll\lllllllii‘ljhaii|

3 02334