University of Southern California Dissertations and Theses

Deconstructing the psychological components of emotional decision making and their relation to the suicide continuum

(USC Thesis Other)

Deconstructing the psychological components of emotional decision making and their relation to the suicide continuum

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content DECONSTRUCTING THE PSYCHOLOGICAL COMPONENTS OF EMOTIONAL
DECISION MAKING AND THEIR RELATION TO THE SUICIDE CONTINUUM

by

Emily B. Fine

A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(PSYCHOLOGY)

May 2009

Copyright 2009 Emily B. Fine

ii
Table of Contents
List of Tables iii
List of Figures iv
Abstract v
Chapter 1: Background and Significance 1
The Iowa Gambling Task and Emotional Decision Making 1
The Somatic Marker Hypothesis 2
Emotional Decision-Making and Suicidality 14

Chapter 2: The Present Research 24
Chapter 3: Research Design and Methods 28
Participants 28
Procedure 29
Primary Measures 29
Potential Co-Variates 33

Chapter 4: Results 41
Descriptive Statistics 41
Missing Data 43
Reliability 44
Bi-variate Associations 45
Regression Analyses 45
Additional Analyses 64

Chapter 5: Discussion 67
Coping and Suicidality 67
Greater Specificity of Autobiographical Memories Contributes to More 69
Adaptive Coping
Negative Autobiographical Memories and Suicidality: The Role of Gender 70
and Controllability of Stressors
Strong Emotional Decision-Making Provides the Cognitive Resources to 72
Ideate Suicide but Acts as a Buffer against Attempting
Limitations 73
Conclusions and Future Directions 75

Bibliography 77
Appendix 89

iii
List of Tables
Table 1. Means and Standard Deviations of Demographic and Key Variables 42
Table 2. Correlations amongst Key and Demographic Variables 47
Table 3. Hypothesis 1a 48
Table 4. Hypothesis 1b 49
Table 5. Hypothesis 1c 51
Table 6. Hypothesis 2: Regression Equation 1 52
Table 7. Hypothesis 2: Regression Equation 2 53
Table 8. Hypothesis 2: Regression Equation 3 54
Table 9. Hypothesis 2: Regression Equation 4 55
Table 10. Hypothesis 3a 58
Table 11. Hypothesis 3b 58
Table 12. Hypothesis 3c 59
Table 13. Hypothesis 3d 61
Table 14. Hypothesis 3e 62
Table 15. Hypothesis 3f 64
Table 16. Additional Analyses 65

iv
List of Figures
Figure 1. A Model of the Association amongst Emotional Decision-Making, 25
Coping, Autobiographical Memory, and Suicidality

Figure 2. Coping and Emotional Decision-Making as Mediators in the Relation 26
between Autobiographical Memory and Suicidality

Figure 3. Gender as a Moderator in the Relation between Autobiographical 26
Memory and Suicidality

Figure 4. Gender Differences in the Mediating Role of Coping and Emotional 26
Decision-Making in the Relation between Autobiographical Memory
and Suicidality

Figure 5. Gender as a Group Differences Variable in the Association 60
between Negative Autobiographical Memory and Lifetime Suicidality

Figure 6. Gender as a Group Differences Variable in the Association between 63
Negative Autobiographical Memory and Suicide Attempt Status

Figure 7. Emotional Decision Making Moderates the Relation between Suicide 66
Ideation and Attempts

v
Abstract
Research has consistently demonstrated the association between drug use and
poor emotional decision making. However, little is known about the psychological
components that contribute to the emotional decision-making process. The current study
aimed to assess two of the assumptions of the somatic marker hypothesis, which are: 1)
reasoning and decision-making depend on the availability of knowledge about situations
and options for actions and outcomes (Bechara et al., 2000), and 2) the individual must be
capable of making this knowledge explicit in order to make effective decisions. In
addition, the current study aimed to apply research on emotional decision making to
another problem of volition: suicidality. Lastly, the current study attempted to replicate
research that has found significant associations between suicidality, coping, and
autobiographical memory, but in a homeless population.
Results indicated that, like previous studies of coping and suicidality, homeless
young adult injection drug users who utilized maladaptive coping mechanisms were more
suicidal. Additionally, individuals with more specific autobiographical memories were
better able to utilize adaptive coping mechanisms. However, they displayed higher levels
of suicidality. This latter finding was found to be moderated by gender, with females
displaying higher levels of suicidality in the presence of more specific negative
autobiographical memories. Lastly, more adaptive emotional decision making predicted
higher levels of suicide ideation, which was contrary to predictions. However, with
further exploration, it was found that strong emotional decision-making acted as a buffer

vi
against attempting suicide in those who were ideating. These results have important
implications for assessment and treatment of suicidal patients.

1
Chapter 1: Background and Significance
The somatic marker hypothesis, a theory of emotional decision making (Damasio,
1995), has been applied to studies of drug users in an attempt to explain their neural
deficits and why they continue to use despite serious health, social, and work
consequences. Although the neural mechanisms involved in this theory have been
outlined, very little is known about the underlying psychological processes that compose
this theory. Fortunately, particular assumptions of this theory provide a framework for
exploring the psychological components. The current study examined possible
psychological variables that constitute the emotional decision-making process with the
potential benefits of identifying specific factors that can be targeted in therapy settings in
order to reduce poor decision-making.
The other goal of the study was to potentially extend the study of emotional
decision-making to another problem of volition: suicidality. Suicidal individuals display
decision-making deficits (e.g. Jollant et al., 2005) and emotional distress (e.g. Holden,
Kerr, Mendonca, & Velamoor, 1998). The interaction of these two variables may be
associated with the inability to “fear” the potential consequences of attempting suicide.
The current study examined the relevance of emotional decision-making and its
underlying psychological components to suicide ideation and attempts in a sample of
homeless, drug users, a population at increased risk for such behaviors.
The Iowa Gambling Task and Emotional Decision-Making
Emotional decision making deficits have commonly been assessed in the
laboratory by the Iowa Gambling Task (Bechara, Damasio, Damasio, & Anderson, 1994).

2
In the Iowa Gambling Task, participants choose from four decks of cards with different
potential monetary gains, with the goal of making the most money possible (Bechara et
al., 1994). After each selection, participants are given feedback on a computer screen
regarding their performance. Through this feedback, adaptive decision makers learn to
avoid the bad decks and to choose from the good decks, whereas maladaptive decision-
makers frequently display deficits on this task (Bechara, Damasio, Damasio, &
Anderson, 1994; Bechara, Dolan, Denburg, Hindes, Anderson, & Nathan, 2001; Bechara,
Damasio, Damasio, & Lee, 1999). The mechanisms by which these deficits may occur
have been addressed in the context of the somatic marker hypothesis neural framework,
which will be discussed next.
The Somatic Marker Hypothesis
Damasio’s somatic marker hypothesis posits that bodily feelings normally
accompany representations of the anticipated outcomes of options and as a result,
influence decision-making. In the somatic marker hypothesis, Damasio (1996) stated
that:
Somatic markers are a special instance of feelings generated from secondary
emotions. Those emotions and feelings have been connected by learning to
predict future outcomes of certain scenarios. When a negative somatic marker is
juxtaposed to a particular future outcome, the combination functions as an alarm
bell. When a positive somatic marker is juxtaposed instead, it becomes a beacon
of incentive. Thus, somatic markers are emotion-related signals, which can be
either conscious or unconscious, and are conceived as bioregulatory signals that
express themselves as emotions.

Thus, somatic markers serve as automatic devices to speed one to select the most
advantageous option-outcome pairs, and even when individuals possess the knowledge to
adequately deal with the situation at hand, advantageous decision making is not

3
guaranteed when these emotional signals are absent (Damasio, 1995). Saver and
Damasio (1991) concluded that in the absence of emotional input through activation of
somatic states, the decision making process is overwhelmed by trivial information, and
one is unable to mark the implications of a social situation with a signal that would
separate good and bad options. In other words, one has no internal goal representation
and as a result, one cannot keep a problem in perspective in relation to other goals, and
decision-making is virtually impossible.
Although the overall process of emotional decision-making has been outlined in a
theoretical framework, the factors that contribute to this process have not been
thoroughly examined. The Iowa Gambling Task assesses the final outcome of emotional
decision making, and hypotheses have been made in regards to how these decisions are
formed from a neural perspective. However, research is needed to explore the cognitive
and psychological variables that underlie effective emotional decisions. Identifying
different variables involved in the process may help to further clarify what goes awry in
the emotional decision-making process, and particular assumptions of the somatic marker
hypothesis provide clues regarding the underlying constructs that should be assessed.
The next section will discuss the assumptions involved in the somatic marker hypothesis
and how they relate to the encoding and implementation of emotional memories to be
applied in situations that require coping.
The Process and Assumptions of the Somatic Marker Hypothesis
In studying emotional decision making, it is important to examine the
assumptions and components that are inherent in the process. Bechara, Damasio, Tranel,

4
and Damasio (2005) stated that although complex laboratory tasks, such as the IGT,
succeed in capturing critical components of the decision-making process, they do not
allow for a finer resolution of the underlying processes. Thus, decomposing and defining
the variables involved in the decision-making process may be beneficial for future
research and treatment of disorders of volition.
Bechara, Damasio, Tranel, and Damasio (1997) posited that when a situation
requires decision making, there are two parallel but interacting processes that occur. In
one, the sensory representation of a situation, or the facts evoked from it, activate non-
declarative dispositional knowledge related to the individual’s previous emotional
experience for similar situations. Non-conscious signals then bias cognitive evaluation
and reasoning (the emotional component). In the other process, the representation of the
situation leads to the recall of relevant information that can be used for determining
various options and outcomes of the situation, as well as reasoning which options and
outcomes are most desirable (the working memory component). Bechara, Damasio, and
Damasio (2000) further decomposed these interacting processes by providing
assumptions of the emotional decision-making process, with two of these assumptions
holding particular relevance for the current study.
Assumption 1: Emotional Decision-Making and Autobiographical Memory
The first assumption states that reasoning and decision-making depend on the
availability of knowledge about situations and options for actions and outcomes (Bechara
et al., 2000). When asked to recall emotional memories, individuals must determine
whether a memory is relevant to a specific situation, and somatic reinforcement from

5
these memories will play a role in determining whether an appropriate memory has been
recalled. The current study aimed to demonstrate that the same individuals who have
difficulty on the Iowa Gambling Task, a measure of emotional decision making, would
also have difficulty when presented with an emotional cue word and asked to
differentiate amongst a vast store of autobiographical knowledge in order to find a life
event that matched the cue word, as the somatic reinforcement required in this task
mirrors the somatic reinforcement experienced in the emotional decision making process.
The current study focused on autobiographical memory (AM), or memory that is
concerned with the recollection of personally experienced past events. AM has been
shown to contribute to an individual's sense of self, to his or her ability to remain oriented
in the world, and to pursue goals effectively in the light of past problem solving
(Williams, 2004). Research on autobiographical memory has suggested that in order to
have acceptable autobiographical memory and cope effectively in given situations, one
must be able to recall specific memories. Specific memories involve the recall of an
event that happened on a particular day and at a particular place and details a specific
event (e.g. Last Friday, I had a nice time at Olivia’s party; Williams & Broadbent, 1986).
Two types of non-specific memories are “categorical” and “extended.” Categorical
memories are memories for events that occur repeatedly (e.g. I like parties.), whereas
extended memories involve memories for events that lasted for longer than a day (e.g. I
took a vacation to Hawaii.).
The construct of autobiographical memory was chosen in the current study for
two main reasons: 1) The first assumption of the somatic marker hypothesis posits that

6
sufficient knowledge is required for effective emotional decision making. AM is relevant
and necessary for real-world decision making and has played a demonstrated role in
coping abilities, which will be discussed in detail later; 2) When a situation arises that
requires use of existing knowledge stores, one must be able to sort through and discard
irrelevant information in the search and retrieval process. Through somatic
reinforcement, one is then able to determine what is relevant and what is not and whether
the memory matches the just noted emotion. More specifically, an alarm is needed to
indicate that a situation requires looking to AM scripts to help with the current situation.
When those AM stores are accessed, one needs the ability to confirm, through somatic
reinforcement, that this is indeed the relevant information that matches the just noted
emotion. In other words, a disruption in either of these processes will result in deficits on
the Autobiographical Memory Test, just as they result in impaired decision making.
Thus, based on the importance of knowledge stores for making emotional decisions and
the hypothesized similarity of the abilities necessary for emotional decision making and
the retrieval of specific and relevant AMs, the current study hypothesized that
overgeneral autobiographical memory would be associated with poor emotional decision
making.
Two studies exist to date that can be used as support for the hypothesis that
individuals with emotional decision-making deficits may be unable to recall specific
autobiographical memories due to an inability to produce emotional states when
considering past situations that produced the same emotion. Bechara, Damasio, and
Damasio (2003) studied a group of eight ventromedial patients to determine whether an

7
inability to re-experience emotional states associated with punishment when recalling
previous instances of punishment could be responsible for absence of anticipatory SCRs
and thus, an inability to learn from experience. In this study, participants were asked to
think about and describe situations in their lives where they felt sad, happy, angry, and
fearful. After they provided a description, they were asked to re-experience those
emotions while their physiological signs were measured. All participants were able to
recall emotional memories, but they did not produce significant SCRs. Bechara et al.
(2000) concluded that these individuals possessed a weakened ability to re-experience an
emotion from the recall of an appropriate emotional event.
In another study, participants (4 amygdala patients, 3 ventromedial patients) were
not asked to report on the details of events, but they were asked to utilize imagery to
induce an emotional experience, while having their SCRs measured (Tranel, Bechara,
Damasio, & Damasio, 1998). More specifically, subjects were asked to begin focusing
on the target emotional experience (anger, fear, or neutral), and to then give a hand signal
when they were at the height of the emotional experience. After the hand signal, SCR
recording was started, and the emotional imagery continued for 60 seconds. At the end of
the imagery of each emotional situation, the subject was asked to rate how much they
“felt” the target emotion (on a scale of 0 to 4). Both the amygdala and ventromedial
subjects reported higher subjective ratings of anger and fear during those respective
conditions, and they generated higher SCR activity during the imagery of angry and
fearful situations relative to neutral images. However, ventromedial and amygdala
patients were still impaired in generating SCRs compared to controls.

8
The cause of the lower SCRs in these two studies is unknown, but one could
propose that although the ability to experience an emotion from emotional imagery is
critical to the emotional decision-making process, it is also important to determine
whether individuals with somatic deficits can adequately conjure up memories of the
requested valence. Although it was reported that all participants were able to recall
emotional memories in the former of these two studies, there were no strict criteria for the
recall of those memories. Thus, it is quite possible that even though they were able to
recall a memory of a particular valence, that memory may have been vague, but yet was
still considered an acceptable response for the purposes of the study. If indeed the
memories recalled were too vague, they may not have had the power to produce
sufficient SCRs. Thus, their deficits may not have been grounded in the re-experiencing
of the emotion but rather in their ability to conjure up specific emotional memories,
which then impacted the re-experiencing of the emotion. Therefore, if individuals could
conjure up more specific memories, maybe they could experience the respective emotion
more intensely and produce the SCRs necessary for effective decision-making.
The current study set out to examine the ability to conjure up specific emotional
memories utilizing the Autobiographical Memory Test. I hypothesized that poorer
emotional decision making would be associated with more over-general autobiographical
memories.
Assumption 2: Emotional Decision-Making and Coping Abilities
The second assumption of the somatic marker hypothesis states that the individual
must be capable of making situation-relevant knowledge explicit in order to make

9
effective decisions (Bechara et al., 2000). In other words, people must not only possess
knowledge stores, but they must also be able to apply that knowledge in appropriate
situations. The ability to apply knowledge in particular situations can be measured
through an assessment of coping strategies. Coping has been defined as the cognitive and
behavioral activities by which a person attempts to manage specific stressful situations,
as well as the emotions that they generate (Lazarus & Folkman, 1984). Coping strategies
have been divided into two categories: approach and avoidance. Approach strategies, as
termed by Billings and Moos (1984), allow for taking appropriate action and making
changes in a situation that may make it more controllable, whereas avoidant coping
involves absence of attempts to cope. Whether approach or avoidant coping is more
adaptive depends on the specific situation at hand and the degree to which the situation is
under the individual’s control (Folkman, 1984).
Coping differs from decision making in that tasks of decision making, such as the
Iowa Gambling Task, involve correct or incorrect answers, whereas coping involves a
variety of choices, some of which are more adaptive than others. However, there is no
clear-cut, black and white choice that is correct, above and beyond the others. There are
more shades of gray in tasks of coping, and these coping choices are attached to a variety
of outcomes. In addition, particular coping choices may be beneficial for some situations
but not for others. Thus, coping choices require more complex cognitive processes and
more precise emotional reinforcement, due to the intricacies of the behaviors. Also, the
ability to weigh such a large number of choices requires greater cognitive resources for
working memory.

10
Research on coping choices is particularly relevant for individuals with
weaknesses in emotional decision-making, given that case studies have demonstrated that
they display deficits in social, occupational, and health behaviors in their daily lives
(Harlow, 1848, 1868; Ackerly & Benton, 1948; Brickner, 1932; Welt, 1888; Eslinger &
Damasio, 1985). The deficits that they display in the laboratory may be effectively
measured in their daily lives through an assessment of coping strategies. Thus, the
current study aims to extend research on poor decision making of individuals in
laboratory settings, to their poor decisions outside the laboratory, decisions that have real-
world consequences for their personal relationships, occupations, health, etc.
Perhaps of greatest importance to this hypothesis is the fact that coping is a
process, whereas decision making is the outcome of that process, a differentiation that
supports the purpose of the current study: to understand the psychological components
and processes of emotional decision making. Coping is what an individual does—
behaviorally or cognitively—to deal with a particular situation, whereas decision making
on the IGT is a study of the outcome—the number of good cards versus bad cards
selected. Therefore, an assessment of coping will provide more detailed insight into the
process of decision making, which was one of the overarching goals of the study. I
hypothesized that these same individuals who displayed deficits on the IGT would also
make poor decisions and utilize more maladaptive coping mechanisms in real life. Thus,
individuals who implemented more maladaptive coping strategies would also exhibit
poorer emotional decision making skills.

11
Coping and autobiographical memory. Interestingly, adaptive coping may be
due, in part, to specific autobiographical memory, the other hypothesized component of
decision making in this study. For example, if an individual has experienced the loss of a
loved one, the emotions evoked from this loss will trigger autobiographical memories
associated with prior loss. Information regarding prior loss (e.g. how one coped) can be
beneficial in helping the individual deal with the present situation. However, individuals
with autobiographical memory deficits have vague stores to draw from, and as a result,
they have fewer and less specific coping mechanisms to utilize. Research has confirmed
this association.
For example, in a sample of 60 students (41 women) with a mean age of 16.98
years (SD = 0.13), Hermans, Defranc, Raes, Williams, and Eelen (2005) found that the
number of specific responses on the AMT was negatively associated with an avoidant
coping style, as assessed by the Cognitive-Behavioral Avoidance Scale (r = -.30). Thus,
more avoidance was associated with less specific memories. Similarly, Raes, Hermans,
Williams, and Eelen (2006) studied 425 first-year psychology students and found that a
more repressive coping style was associated with retrieval of less specific AMs (β = -
.30).
In addition, numerous studies have demonstrated that problem solving, a type of
coping, is associated with overgeneral autobiographical memory. For example, Kaviani,
Rahimi, Rahimi-Darabad, Kamyar, and Naghavi (2003) found that autobiographical
memory specificity was significantly positively correlated with scores on the Means-Ends
Problem Solving Task (MEPS) in 20 depressed patients with suicide ideation (10 men,

12
average age = 25.50 years; 10 women, average age = 29.70 years) and 20 without suicide
ideation (10 men, average age = 25.20 years; 10 women, average age= 28.70 years).
More specifically, the more irrelevant problem solving methods used (r = .42) and the
more ineffective their problem solving effectiveness (r = .40) the greater the
overgenerality of their autobiographical memories.
In another study by Kaviani and colleagues, they examined 20 depressed suicide
attempters (8 men and 12 women; mean age = 28.05) and 20 healthy participants (mean
age = 27.65) and found that autobiographical memory specificity was correlated with
problem solving effectiveness (r = .58) and problem solving relevancy (r = .62; Kaviani,
Rahimi-Darabad, & Naghavi, 2005).
On the other hand, Williams, Chan, Crane, Barnhofer, Eade, and Healy (2006)
studied 40 undergraduate and graduate students (25 women, 15 men; mean age = 24.44
years, SD = 3.13) and did not find significant associations between MEPS-specificity and
AMT specificity scores (r = .38). However, an insufficient sample size may have
explained the lack of significance for this finding. Nevertheless, the authors did find that
mean MEPS scores and MEPS-effectiveness were associated with AMT specificity
scores (r = .57 and r = .45, respectively).
Pollock and Williams (2001) examined suicide attempters (10 males and 14
females, age range = 21-72 years) and found that overgeneral autobiographical memory
was associated with problem solving effectiveness on the MEPS (r = .51) as well as mean
scores on the MEPS (r = .37).

13
In a study of 24 patients (15 women, mean age = 39.04 years, S.D. = 11.04, range
= 21–58), all meeting DSM-IV criteria for current major depressive disorder, Raes and
colleagues concluded that AMT specificity and MEPS effectiveness were significantly
positively correlated (r = .65; Raes, Hermans, Williams, Demyttenaere, Sabbe, Pieters, &
Eelen, 2005). Partial correlations between MEPS effectiveness and memory specificity,
adjusting for depression severity (BDI) and for latency to respond to AMT cues, were
also significant (r =.61 and r = .68, respectively). These results suggest that the
relationship between memory specificity and MEPS effectiveness cannot be explained by
mood or general unresponsiveness.
Also, Goddard, Dritschel, and Burton (1996) studied 16 outpatients meeting BDI
cut-off scores for Major Depression (12 women and 4 men, mean age = 46.6 years, SD =
6.66 years, range = 34 - 59 years) and 16 outpatients who did not meet criteria for Major
Depression (mean age = 41.9 years, S.D. = 9.7 years) to examine whether there were
differences between the two groups in types of non-specific memories retrieved and the
potential impact of those differences on stages of the problem solving process. Again,
they found that the number of specific memories correlated with the MEPS effectiveness
score (depressed, r = .54; control, r = .59). In addition, the number of specific memories
correlated significantly with mean MEPS scores for the control group (r = .53) but not the
depressed group (r = .38). The authors explored this latter finding further and found that
for over-general memories, depressed patients gave a larger proportion of categoric
responses than controls. The authors proposed that these differences in types of non-
specific memories may have lessened the strength of the association between the number

14
of specific memories and MEPS scores for the depressed group. However, they did not
examine this hypothesis specifically, and it is quite possible that their sample size simply
may not have been large enough to detect significance in the depressed group.
Lastly, Sidley, Whittaker, Calam, and Wells (1997) found results similar to the
previously detailed studies in their examination of 35 patients (21 males, 14 females;
mean age = 32.1 years) admitted to the hospital after a drug overdose. They concluded
that the total number of specific memories were positively associated with MEPS
effectiveness (r = .38). Thus, research has consistently found associations between
specific autobiographical memories and more adaptive problem solving abilities.
To adjust for the association between coping and autobiographical memory and
the predicted association between decision making and autobiographical memory,
autobiographical memory was included as a covariate in the analyses examining
associations between decision making and coping. I hypothesized that even after
adjusting for inadequate knowledge stores, deficits in emotional decision making would
still be associated with poor coping.
Emotional Decision-Making and Suicidality
Although the evidence is clear that drug users display some impairments in the
mechanisms of emotional decision making, the question remains unanswered as to
whether there are other related disorders of volition that are characterized by these
deficits and their hypothesized psychological components. In other words, can other
forms of psychopathology also be explained, in part, by deficits in emotional decision-

15
making? The current study sought out to answer this question by focusing on the
association between suicidality and emotional decision making.
Suicidality is a logical choice in potentially extending the somatic marker
hypothesis to other disorders of volition, as suicidality and drug use have demonstrated
consistent associations with each other and overlapping neuro-abnormalities. The
“Suicide Proneness Model,” a theoretical model that incorporates drug use and suicide
ideation, refers to a broad-based measure of a person’s propensity at a point in time to
engage in suicidal and other life-threatening behaviors (Lewinsohn, Langhinrichsen-
Rohling, Langford, Rohde, Seeley, & Chapman, 1995). The definition of suicidal
behavior includes subtle or non-obvious self-destructive behaviors and risk-taking
behaviors, as well as behaviors that are overtly suicidal (Menninger, 1938). Research has
suggested that suicide proneness can be conceptualized by four behavior categories, two
of which are relevant for the current study. Rohde, Noell, Ochs, and Seeley (2001)
labeled these two categories “health-related” (e.g. behaviors that are damaging to health,
such as lack of regular check-ups and use of illicit drugs) and “death-related” (e.g.
traditional suicidal and death-related behaviors, including suicidal ideation). Previous
research supports this theory in homeless populations, concluding that higher levels of
suicidality are associated with higher levels of maladaptive health-related behaviors, such
as drug use.
Drug use and suicidality have been correlated in studies of homeless youth, and
drug use has predicted suicidality in logistic regression analyses. For example, alcohol
and drug use were correlated with suicide attempts and ideation (Yoder, Hoyt, &

16
Whitbeck, 1998; Yoder, 1999; Unger, Kipke, Simon, Montgomery, & Johnson, 1997),
and IV drug use predicted suicidal ideation (Rohde et al., 2001). In addition, total days
intoxicated and total days of drug use in the previous 30 days predicted suicide attempts
in the subsequent 30 days (Desai, Liu-Mares, Dausey, & Rosenheck, 2003), and 23% of
participants reported using drugs or alcohol just prior to their attempt (Rotheram-Borus,
1993). Thus, regardless of the mechanism, there is evidence to suggest that drug use is
related to suicidal behaviors, warranting further investigation into the mechanisms and
commonalities across these two types of behaviors (Brent, Perper, Kolko, & Zelenak,
1988).
Emotional Decision-Making Deficits: A Commonality between Drug Users and Suicidal
Individuals
In addition to the association between drug use and suicidality, drug users and
suicidal individuals possess overlapping neuroanatomical and neurotransmitter
abnormalities. The frontal lobe theory of suicide posits that individuals who are suicidal
are likely to exhibit dysfunction in the prefrontal cortex. Indeed, studies have identified
low levels of—and abnormal interactions among—neurotransmitters in the PFC
(specifically, serotonin, dopamine, and norepinepherine) in suicidal individuals. For
example, several studies have identified a functional decrease in serotonergic neuronal
activity (Stanley, Virgilio, & Gershon, 1982; Stanley & Mann, 1983; Andorn, 1991) and
a higher number of serotonin binding sites in the prefrontal cortex of suicide victims
(Arora & Meltzer, 1989; Biegon & Israeli, 1988). Drug users have repeatedly
demonstrated deficits in the prefrontal regions of the brain as well, in particular, the

17
ventromedial prefrontal cortex. The prefrontal abnormalities of suicidal individuals have
likewise been localized to this area (for review see Mann, 1998). More specifically,
researchers have demonstrated reduced serotonin neuronal activity and increased
serotonin receptors in the OFC of suicide victims (Arango, Underwood, Gubbi, & Mann,
1995). In one study of brain tissue, suicide cases (n = 22) were matched with control
subjects (n = 22) on the basis of postmortem interval (PMI, ± 5 hours), age (± 5 years),
sex, season of death, side of brain and, whenever possible, race (Arango et al., 1995).
The authors found that the binding sites of suicide completers were most different from
controls in the orbital and lateral prefrontal cortex, suggesting the greatest abnormalities
in suicide victims are in these regions.
The deficits that suicidal individuals display in the ventromedial prefrontal cortex
are believed to play a role in decreased behavioral and cognitive inhibition (Robbins et
al., 1995), leading to a greater propensity to act on suicidal thoughts (Mann, 1998;
Damasio, Grabowski, Frank, Galaburda, & Damasio, 1994). Furthermore, Träskman-
Bendz and Mann (2000) postulated that the ventromedial prefrontal cortex acts as a sort
of restraint system, modulating the probability of suicidal behavior. In line with this
hypothesis, PET studies have reported an association between lower levels of
ventromedial activity and both higher suicidal intent and higher lethality of the act
(Mann, 2003). Due to the similarities in neuroanatomical deficits between drug users and
suicidal individuals and the correlation between drug use and suicidality, I hypothesized
that suicidality would be associated with emotional decision making deficits as well.

18
Suicidality and the Mechanisms of Emotional Decision-Making
Although extant research has posited that ventromedial deficits result in lack of
restraint, and subsequently, higher levels of suicide attempts, an additional hypothesis to
consider is that the choice to repeatedly consider and/or attempt suicide may involve a
faulty mechanism in the chain of emotional decision making. Say for example that an
individual has repeatedly experienced suicidal thoughts and has been weighing the pros
and cons of attempting suicide. In the somatic marker hypothesis, it has been
hypothesized that somatic markers (or emotional responses) may be triggered in two
ways: (1) primary induction, which is brought about by primary inducers (events that are
occurring in the environment and which have the properties to automatically and
involuntarily elicit a somatic response; Damasio, 1995); and (2) secondary induction,
which is driven by secondary inducers (i.e. the recall or thought that produces a somatic
response when brought to memory; Damasio, 1995). Thus, secondary inducers, i.e.
thinking about a situation, can trigger an emotional response (somatic markers) that,
ultimately, can consciously or unconsciously bias decision making regarding the situation
at hand. Thoughts of suicide should produce a negative somatic marker or alarm that
result in seeking long term memory regarding knowledge of suicide and its
consequences. When pre-existing knowledge of suicide and its consequences are
accessed, one should realize the serious consequences of such action and, as a result, veto
the idea of attempting suicide.
Although attempting suicide may seem to be an effective way to ease pain and
suffering in the moment, in the long run, the pain and suffering will likely subside, but a

19
suicide attempt could be fatal. An individual with normal emotional decision-making
would realize these potential consequences and decide against attempting and possibly
even allowing oneself to think about suicide in the future. Thus, in suicidal individuals,
the difference between those who repeatedly consider and/or attempt versus those who do
not may be based on their ability to weigh the costs and benefits of such a decision and
somatically reinforce the most adaptive choices. In individuals who are suicidal, it is
hypothesized that they are either unable to weigh the long term consequences with the
immediate benefits, and/or they do not receive negative somatic reinforcement when
considering attempting suicide.
In addition, it is possible that emotional decision-making is responsible for the
inclination to attempt suicide but not to repeatedly ideate suicide, given that suicide
ideation does not involve actually behaving disadvantageously. However, it is also
possible that the inability to veto the idea of attempting, resulting in repeated thoughts of
suicide, may also reflect a deficit in the cognitive component of the emotional decision
making process. Bechara (2004) referred to this phenomenon as perceptual (or cognitive)
impulsiveness, which reflects an inability to inhibit a recurrent thought held in working
memory. Suicide ideation may be one type of thought that susceptible individuals have
difficulty inhibiting, and thus, it is posited that this construct can also be effectively
captured by the IGT and that emotional decision-making will be associated with suicidal
ideation in addition to attempts.
To date, only one study has examined the association between types of suicidal
behaviors and scores on the Iowa Gambling Task. Jollant et al. (2005) explored group

20
differences on the IGT across Healthy Comparisons (n = 82; mean age = 38.8 years, SD
= 9.1), Affective Controls (n = 25; mean age = 40.4 years, SD = 12.2), Nonviolent
Suicide Attempters (n = 37; mean age = 42.9 years, SD = 11.1), and Violent Suicide
Attempters (n = 32; mean age = 43.2 years, SD = 12.6), using an ANOVA design. They
found that for the fifth score and the net score, the healthy comparison subjects scored
significantly better than both groups of suicide attempters, and the violent suicide
attempters scored significantly worse than the affective control subjects. In contrast, no
significant differences were observed between the nonviolent (mean = –2.7, SD = 6) and
violent attempters (mean = –3.7, SD = 4.4), between the healthy comparison (mean = 6.3,
SD = 9.6) and affective control subjects (mean = –1.8, SD = 7.5), and between the
nonviolent suicide attempters and the affective control subjects. Thus, violent suicide
attempters appeared to be a distinct population from healthy controls and affective
controls, but they did not differ from non-violent suicide attempters, demonstrating that
methods of attempting did not play a role in emotional decision-making.
However, it is important to note that, in their study, the authors did not include
individuals who had ideated but not attempted suicide. Thus, research is needed to
explore the relation between suicide ideation and emotional decision-making abilities.
Merely examining differences between ideators and attempters utilizing an ANOVA
design ignores the subtleties within groups of suicidal behaviors. More specifically, two
individuals who think about attempting suicide may demonstrate differences in the
frequency and intensity of their thoughts. For example, one individual who experienced
a romantic break-up 10 years ago may have considered suicide immediately after the

21
dissolution of the relationship but not since then. However, another individual may
suffer from a chronic illness, which makes him more likely to ideate whenever his
disease flares and makes him consider attempting more seriously, given that his disease is
life-long. Thus, the current study examined the severity of suicidal behaviors along a
continuum as they related to emotional decision-making in order to provide a more
accurate and informative assessment of differences between suicidal individuals.
Suicidality, Coping, and Autobiographical Memory
Suicidality and Coping
Interestingly, suicidality is also related to constructs described earlier that are
hypothesized to be important components of emotional decision-making. First, research
has demonstrated the association between suicidality and disengagement coping in
homeless adolescents. For example, Votta and Manion (2004) studied 70 homeless,
adolescent males and found that suicide ideation was significantly associated with
disengagement coping (f
2
= .07) but not engagement coping (f
2
= .04). In addition, the
number of past suicide attempts was significantly associated with disengagement coping
(f
2
= .07) but not engagement coping (f
2
= .07). However, the small effects may explain
lack of significance for two of these findings.
Similarly, Klee and Reid (1998) studied 200 young homeless people (mean age =
20 years; 143 males, 57 females), the largest proportion (46%) of whom became
homeless at 15 or 16 years of age. The authors found that the association between self-
medication as a coping mechanism and suicide attempts approached significance (r =

22
.38). Thus, one could infer from these results that denial coping may play a role in
suicidality in homeless young adults.
Suicidality and Autobiographical Memory
Second, research has demonstrated an association between suicidality and
autobiographical memory. For example, a study by Kuyken and Brewin (1995) provided
correlational analyses of 58 depressed patients with a mean age of 37.07 years (SD =
10.02). The authors found a significant association between number of over-general
positive autobiographical memories and number of previous suicide attempts (r = .23).
However, they did not find a significant association between number of over-general
negative autobiographical memories and number of previous suicide attempts (r = .22).
Again, however, the lack of significance in this latter finding may be due to an
insufficient sample size.
In a study by Startup, Heard, Swales, Jones, Williams, and Jones (2001), patients
with Borderline Personality Disorder displayed a negative association between AMT
scores and deliberate self-harm (r = -.47). In other words, a higher number of
overgeneral memories were associated with fewer acts of deliberate self-harm, indicating
that overgeneral memories possibly served as a protective mechanism. These results are
contrary to what has been found with suicide attempters. Thus, it may be inferred that
patients who engage in deliberate self-harm may have very different motives than suicide
attempters. Research has indicated that deliberate self-harm is used as a means to distract
from mental or physical pain in the moment (Theodoulou, Harriss, Hawton, & Bass,
2005). Rarely is the intent of these individuals to die, which is contrary to the motives of

23
suicide attempters. Accordingly, it has been suggested that overgeneral autobiographical
recall may actually protect individuals from acts of deliberate self-harm by helping them
to avoid distressing memories. Therefore, the mechanism that explains the association
between deliberate self harm and AMT versus suicide attempts and AMT may be
different, and thus, it is not surprising that the results in this study were not consistent
with what has been displayed for suicide attempters on tests of autobiographical memory.
As a result, these findings were not used as a foundation for the hypotheses in the current
study, given that I examined suicide ideation and attempts as opposed to deliberate self-
harm, which is a distinct behavior.
In addition, it is important to note that studies examining the correlation between
AM and suicidality are limited in number and breadth. More specifically, studies have
not examined a continuum of suicidal behaviors, from ideation to completion, in relation
to autobiographical memory. Therefore, additional research on overall autobiographical
memory and the suicide continuum is needed.

24
Chapter 2: The Present Research
The somatic marker hypothesis posits that, in addition to the ability to trigger
somatic states, the availability of knowledge about situations and options for actions and
outcomes (Bechara et al., 2000) are critical to the emotional decision-making process.
Also, in order to make effective decisions, an individual must be capable of making
autobiographical knowledge explicit. As an extension of this model, I hypothesized that
both autobiographical memory and coping would be critical to emotional decision
making.
In addition, I hypothesized that the somatic marker deficits of drug users may be
extended to other problems of volition as well, in particular, suicidality. Suicidal
individuals often respond impulsively to stress, and attempt without seriously considering
future consequences (e.g. the permanence of suicide, hurting loved ones, the idea that the
pain will eventually subside). Thus, it seems that they either do not experience an alarm
in response to thoughts of suicide, telling them that this is a dangerous behavior and to
avoid it, or when they consider the implications of attempting suicide, they do not
experience negative reinforcement for these consequences, preventing them from
realizing that suicidality is a behavior that needs to be reconsidered or avoided.
Therefore, I predicted that effective emotional decision making is critical to reducing
suicidality. Accordingly, the current study tested the following hypotheses:
• Hypothesis 1: Poorer emotional decision making, poorer coping, and less specific
autobiographical memory will be associated with higher levels of suicidality. (e.g.
Individuals who display more deficits in autobiographical memory, higher levels

25
of maladaptive coping—more distancing/avoidance and denial, and poorer
emotional decision making will have higher levels of suicidality; see Figure 1)
• Hypothesis 2: Coping and emotional decision-making will mediate the relation
between autobiographical memory and suicidality (see Figure 2).
• Hypothesis 3: Gender will moderate the association between autobiographical
memory specificity and suicidality (see Figure 3).
• If Hypothesis 3 is significant, an exploratory aim will be to examine the role of
gender as a moderator in the mediational model in Hypothesis 2 (see Figure 4).
Through an evaluation of the hypothesized psychological components of
emotional decision-making (autobiographical memory and coping), the associated
deficiencies in the process may become clearer. In addition, the current study has the
potential to extend this line of research on suicidality, coping, and autobiographical
memory to an understudied (and at-risk) group.
Figure 1. A model of the association amongst emotional decision making,
coping, autobiographical memory, and suicidality.

Emotional Decision
Making (IGT—Net)
Coping (WCQ—
Distancing/Avoidance)
Coping (WCQ—Denial)
Autobiographical
Memory (AMT)
Suicidality (SBQ)

26
Figure 2. Coping and emotional decision making as mediators in the relation
between autobiographical memory and suicidality.

Figure 3. Gender as a moderator in the relation between autobiographical
memory and suicidality.

Figure 4. Gender differences in the mediating role of coping and emotional decision
making in the relation between autobiographical memory and suicidality.

males

Autobiographical
Memory (AMT)
Gender
Suicidality
(SBQ)
Autobiographical
Memory (AMT)
Coping (WCQ—
Distancing/Avoidance)
Emotional Decision
Making (IGT—Net)
Suicidality
(SBQ)
Coping (WCQ—Denial)
Autobiographical
Memory (AMT)
Coping (WCQ—
Distancing/Avoidance)
Emotional Decision
Making (IGT—Net)
Suicidality
(SBQ)
Coping (WCQ—Denial)

27
Figure 4, Continued

females

Autobiographical
Memory (AMT)
Coping (WCQ—
Distancing/Avoidance)
Emotional Decision
Making (IGT—Net)
Suicidality
(SBQ)
Coping (WCQ—Denial)

28
Chapter 3: Research Design and Methods

Participants

In order to determine the approximate number of participants needed for the
current study, Cohen’s (1992) guidelines for determining an N based on effect size,
power, and statistical methods were used. It was determined that in a linear regression
model, to achieve 80% power to reject the null hypothesis that the independent effect of
one independent variable is 0 (i.e., after partialling the covariate effects), for α = .05, one-
tailed, an N between 67, for a medium-sized effect (i.e., B = .15), and 30, for a large
effect (i.e., B = .35), would be required (Cohen, 1992).
Twenty-five participants were self-selected from a larger study of approximately
125 homeless drug users from the Hollywood and Santa Monica areas, which examined
the association between ketamine use and cognitive functioning. Field interviewers
recruited these individuals to participate initially in a series of five qualitative interviews.
Participants were then scheduled for the cognitive and mental health assessments, which
were used in the current study. Twenty-nine people were additional recruits, due to the
transient nature and attrition of the initial population of 125. These 29 participants were
recruited in the same manner and according to the same criteria as the first 25.
Participation criteria for both groups included illicit drug injection at least once in the
past two years and being between the ages of 16 and 29.
The mean age of the participants was 23.56 years and the ethnic composition was
59.3% Caucasian, 11.1% Latino, 7.4% African American, 1.9% Native American, and
11.1% multi-racial. Nine point three percent of the participants did not know their ethnic

29
background. (e.g. They were adopted, etc.) Sixty-five percent were male and 35%
female. Twenty-eight percent of the sample had attempted suicide in their lifetime.
Procedure
The current study involved one wave of data collection, which began in July 2006
and concluded in August 2007. At the beginning of the assessment, participants were
handed the consent form and taken through the purpose of the study, the content of the
assessment, their rights as research subjects, and confidentiality constraints. Study
completion took approximately 4 hours, with ample breaks, and food and drink were
provided. Participants were paid $75 for completion of the study and were given a list of
referrals for mental health and drug hotlines, low-cost and free mental health clinics, and
homeless shelters. The current study received IRB approval.
Primary Measures
Assessment of Suicidality
Suicidal Behaviors Questionnaire (SBQ; Linehan, 1981)
The SBQ is a 34-item questionnaire inquiring about suicidal behaviors (ideation
and attempts) in one’s lifetime, in the past year, in the last 4 months, in the last month,
and in the last several days. The subscales utilized in the current study were: lifetime
suicidality, suicide attempts, and suicide ideation. The first question asks: “Have you
ever thought about or attempted to kill yourself in your lifetime?” Answers include: “0”
(no), “1” (It was just a passing thought), “2” (I briefly considered it, but not seriously),
“3” (I thought about it and was somewhat serious), “4” (I had a plan for killing myself
which I thought would work and seriously considered it), “5” (I attempted to kill myself,

30
but I do not think I really meant to die), and “6” (I attempted to kill myself, and I think I
really hoped to die). Cotton, Peters, and Range (1995) found the SBQ to have adequate
internal consistency (α = .80) and test-retest reliability (α = .95 over two weeks).
Assessment of Coping
Ways of Coping Questionnaire (WCQ; Folkman & Lazarus, 1985, 1988)
The WCQ, one of the most commonly used measures of coping, asks individuals
to identify a specific stressful situation and to relate their reliance on 66 coping
responses. Items are rated on a 4-point Likert scale from “0” (Does not apply or not
used) to “3” (Used a great deal), with higher scores indicating more attempts to cope, and
lower scores indicating fewer attempts to cope. Folkman and Lazarus (1985) reported
moderate to high internal consistency reliabilities for the scales, ranging from .56 to .85.
Recent studies have found that the measure is composed of four factors:
Distancing/Avoidance, Confrontation/Seeking Social Support, Problem-Focused, and
Denial, with a goodness-of-fit-index of .92 (Bouchard, Sabourin, Lussier, Wright, &
Richer, 1997). A priori, the Distancing/Avoidance, Problem-Focused, and Denial
subscales are most conceptually related to the key variables, as demonstrated by the
previously reviewed literature, so I utilized only these subscales in the data analyses.
Assessment of Autobiographical Memory
Autobiographical Memory Test (AMT; Williams & Broadbent, 1986)
The AMT consists of 12 emotional cue words (four positive, four negative, and
four neutral). Participants are asked to recall a specific event in response to each cue
word and are allowed 30 seconds for each response. Participants are given three practice

31
cue words and provided with feedback before the test begins. The cues are presented
orally by the experimenter in a fixed order with alternating positive, negative, and neutral
words (namely: nice, alone, agility, tender, insult, carry, kindness, dismal, client, sweet,
nasty, ginger). The response latency is measured starting when the cue word is presented
until the first word of the subject's response. If the first response is not a specific
memory, participants are prompted again to retrieve a specific memory (“Can you think
of a specific time—one particular episode?”). The prompting procedure is repeated until
a specific memory is retrieved or the response time of 30 seconds is exceeded. Specific
memories are defined as memories that refer to a specific event that happened on a
particular occasion and lasted less than one day. Non-specific memories are either
categoric (referring to events that occurred more often, and are difficult to date exactly,
e.g. “Listening to my favorite call-in radio show on Sunday evenings”), extended
(referring to events that last longer than one day, e.g. “My holiday in Senegal two years
ago”), semantic associate (e.g. a verbal association to the cue), no response, or same
event (a response referring to the same event already mentioned in response to a previous
cue). All verbatim responses are noted. If a subject is unable to recall a specific memory
in the time allotted, a time of 30 seconds is recorded, and the experimenter proceeds to
the next item. Most commonly, participants are given 1 point for a first response to a cue
word that is specific. Using this scoring procedure, Raes, Hermans, de Decker, Eelen and
Williams (2003) obtained a very good inter-rater agreement of 98.8% (kappa coefficient
for 0.957) for the categorization of specific vs. non-specific responses, and Kuyken and

32
Brewin (1995) and Williams and Broadbent (1986) obtained inter-rater reliabilities
between .87 and .93.
Assessment of Emotional Decision Making
Iowa Gambling Task (IGT; Bechara, Damasio, Tranel, & Anderson, 1994)
The Iowa Gambling Task is a card game that assesses the ability of participants to
evaluate immediate gains over long-term losses. The individual is instructed that the
object of the task is to accumulate as much play money as possible by picking one card at
a time from any of four decks (A, B, C, and D). The participant is permitted to take cards
from the decks in any order, and to take as much time as he or she wishes. The task ends
when the subject has gone through all 100 cards. The decks differ along three
dimensions: the immediate gain, the expected long-term gain, and the schedule of
penalties. All the cards from Decks A and B yield a larger short-term payoff ($100 per
card) than cards from Decks C and D ($50 per card). Certain cards in all of the decks also
carry a penalty, and the accumulated penalties in Decks A and B are larger than in Decks
C and D. Over the long run, continued choice from either Deck C or D leads to a net gain
($250/10 cards), whereas choice from either Deck A or B leads to a net loss (-$250/10
cards). Decks A and B, therefore, are `low yield' decks, while Decks C and D are `high
yield' decks. Even though the individual receives a reward for each card choice, he or she
cannot avoid being penalized, but can minimize the overall magnitude of the loss. The
optimal strategy, therefore, is to minimize the overall loss by avoiding the short-term
appeal of Decks A and B in favor of the slower, but ultimately larger, gain of Decks C
and D.

33
Potential Co-Variates
Assessment of Depression
Center for Epidemiological Studies—Depression Scale (CES-D; Radloff, 1977)
The CES-D is a 20-item instrument that was developed by the National Institute
of Mental Health to detect major or clinical depression in adolescents and adults in the
past week. The CES-D has 4 separate factors: Depressive Affect, Somatic Symptoms,
Positive Affect, and Interpersonal Relations. Items are rated on a 4-point Likert scale
from “1” (Rarely or none of the time—Less than 1 day) to “2” (Some or a Little of the
Time—1 - 2 days) to “3” (Occasionally or a Moderate Amount of the Time—3 - 4 days)
to “4” (Most or All of the Time—5 - 7 days). Examples of items include: 1) “I was
bothered by things that usually don’t bother me.” 2) “I did not feel like eating; my
appetite was poor.” Internal consistency for the entire CES-D was α = 0.85 for the
general population (Radloff, 1977). Items on the CES-D were summed for a total score,
in order to take into account current mood state.
Rationale for inclusion as a co-variate. The CES-D was included as a potential
covariate to account for current mood state. Research has found that higher levels of self-
reported depression scores are associated with less specific autobiographical memory
(Sampson, Kinderman, Watts, & Sembi, 2003; Brittlebank, Scott, Williams, & Ferrier,
1993; Kuyken & Brewin, 1995; Swales, Williams, & Wood, 2001; Park, Goodyer, &
Teasdale, 2002; Brewin, Reynolds, & Tata, 1999; Startup, Heard, Swales, Jones,
Williams, & Jones, 2001), poorer coping (Galaif, Nyamathi, & Stein, 1999; Littrel &
Beck, 2001; Votta & Manion, 2003; Recker Rayburn, Wenzel, Elliott, Hambarsoomians,

34
Marshall, & Tucker, 2005), higher levels of suicidality (Desai, Liu-Mares, Dausey, &
Rosenheck, 2003; Eynan, et al., 2002), and poorer emotional decision making (Dalgleish
et al., 2004; Bechara et al., 2001).
Assessment of Intellectual Ability
Wechsler Adult Intelligence Scale—Third Edition (WAIS-III; Wechsler, 1997)
The WAIS—III is a general test of intelligence that includes 14 subtests,
comprising 7 verbal (Information, Comprehension, Arithmetic, Similarities, Vocabulary,
Digit Span, Letter-Number Sequencing) and 7 performance scales (Picture Completion,
Digit Symbol – Coding, Block Design, Matrix Reasoning, Picture Arrangement, Symbol
Search, Object Assembly). The WAIS provides three scores: a Verbal IQ, a
Performance IQ, and a composite, Full-scale IQ score based on the combined scores.
However, confirmatory factor analyses suggest a four factor structure (Verbal
Comprehension: Vocabulary, Information, Similarities; Perceptual Organization: Picture
Completion, Block Design, Matrix Reasoning; Working Memory: Arithmetic, Digit
Span, and Letter-Number Sequencing; and Processing Speed: Digit Symbol-Coding and
Symbol Search).
Rationale for inclusion as a co-variate. The WAIS was included in the current
study to account for the possibility that deficits in the key variables were due to a general
deficit in retrieval or verbal ability and to thus, facilitate the interpretation of the data.
Research has found that Digit Span, a subscale of Verbal IQ, has been associated with
autobiographical memory specificity (Sampson et al., 2003), and Verbal IQ has been
associated with IGT scores (Bechara et al., 2001). IQ and overall coping with stress were

35
significantly positively correlated (Masten, Burt, Roisman, Obradovic, Long, & Tellegen,
2004), and Digit Symbol Coding was significantly correlated with Self-Blame Coping on
the WCQ (Tapert, Ozyurt, Myers, & Brown, 2004). Due to the already lengthy testing
time and the frequent use of the Vocabulary and Information subtests alone as measures
of verbal ability, these two subtests were the only WAIS Verbal Comprehension indices
administered. However, because substance users are frequently found to have deficits in
Working Memory and Processing Speed, and subtests from these factors have not been
used on abbreviated versions of these tests (e.g. Wechsler Abbreviated Scale of
Intelligence), all of these subtests were administered.
Assessment of Memory
Wechsler Memory Scales—Third Edition (WMS-III; Wechsler, 1997)
The WMS—III provides subtest and composite scores that assess memory and
attention using both auditory and visual stimuli. There are eight Primary Indices
(Auditory Immediate, Visual Immediate, Immediate Memory, Auditory Delayed, Visual
Delayed, Auditory Reception Delayed, General Memory, and Working Memory), which
constitute the overall scales of Immediate Memory, General (Delayed) Memory, and
Working Memory. Internal consistency ranges from the .70s to the .90s.
Rationale for inclusion as a covariate. In order to determine that associations
between the key variables were not due mostly to variance from verbal learning and
memory deficits, specific subscales of the WMS were included as covariates. Reduced
specificity of AM has been related to poor working memory on the WMS (Raes, et al.,
2006) and poor verbal fluency on the Controlled Oral Word Association Test (COWAT;

36
Sampson, Kinderman, Watts, & Sembi, 2003). In addition, the following California
Verbal Learning Test (CVLT) subtests were significantly correlated with suicide
attempts: List A, Free Short Recall, Cued Short Recall, Free Delayed Recall, and Cued
Delayed Recall. Thus, the Working Memory subscale on the WMS and the Word Lists
subtest on the WMS, which asks participants to recall lists of words before and after a
delay and represents a similar verbal learning construct as the COWAT and CVLT, were
considered as potential covariates in all analyses.
Assessment of Cumulative Education
Wide Range Achievement Test—Third Edition (WRAT—III; Stone, Jastak, & Wilkinson,
1995)
The WRAT—III consists of 3 subtests—spelling, reading, and written
arithmetic—and is considered a test of basic academic skills for individuals ages 5 to 75.
The Reading subtest, which was utilized in the current study, requires individuals to
pronounce words out of context. Alternate form correlations of the Reading subtest (.92)
support the reliability of the measure.
Rationale for inclusion as a covariate. Years of education were correlated with
self-blame coping in a drug and alcohol using sample (Tapert et al., 2004), and other
measures of verbal learning detailed previously have correlated with the key variables.
Therefore, the Reading subtest of the WRAT was utilized as a potential covariate to
assess academic skills.

37
Assessment of Drug Use
Lifetime Use
Research on heavy drug use has consistently considered 11 or more times a month
as the cut-off for qualifying as heavy drug use (Bucholz et al., 1994) and heavy use
scores in neuropsychological testing have been formed by taking times per month
multiplied by the duration of usage (Yip & Lee, 2005). Accordingly, in the current study,
respondents were asked to circle all drugs they had tried at least once from a
comprehensive list. Then they were asked a set of questions for each drug they circled,
including items inquiring about age at heavy use, number of uses during that time,
duration of heavy use, etc. For each drug that participants had used 11 or more times in a
month, a composite was formed by multiplying the number of times that drug was used in
a month during heavy use by the number of months that they used that drug most heavily.
Drug Use in the Past 3 Months
In addition to lifetime heavy use, the number of times each drug had been used in
the past 3 months was assessed.
Drug Use in the Past 24 Hours
Also, participants were asked whether they had used each drug in the last 24
hours. Each drug was coded as a separate variable, and the number of times participants
used each drug in the past 24 hours was entered.
Rationale for inclusion as a covariate. Drug use was included as a potential
confound, given that it has negatively correlated with coping (Wagner, Myers, &
McIninch, 1999; Galaif, Nyamathi, & Stein, 1999; Nyamathi, Stein, & Brecht, 1995;

38
Nyamathi, Stein, Dixon, Longshore, & Galaif, 2003), autobiographical memory (Eiber,
Puel, & Schmitt, 1999), and emotional decision making (see studies detailed previously),
and positively correlated with suicidality (see studies detailed previously). In addition, it
was important to determine if lifetime use, recent heavy use, and current intoxication
were associated with cognitive functioning, as it is possible that drug use may have
impacted scores on the WMS, WAIS, WRAT, which may have accounted for the
association between drug use and the key variables.
Assessment of Gender Differences
Mean Differences
Autobiographical memory. In one study, gender was a significant predictor of
autobiographical memory specificity, but the authors did not conduct post-hoc analyses to
explore whether males or females displayed better recall (Sampson et al., 2003). I
hypothesized that females would have more specific autobiographical memory, as some
research has suggested that increased sensitivity to catecholamine in females may result
in augmentation of emotional memory consolidation (Altemus, 2006). However, gender
differences in autobiographical memory specificity have not been tested empirically.
Suicidality. Homeless females have been shown to be significantly more likely
than males to attempt (Desai et al., 2003) and ideate (Reifman & Windle, 1995) suicide.
The current study attempted to replicate these findings.
Emotional decision making and coping. The literature does not suggest gender
differences for the other key variables (IGT and WCQ subscales), but I conducted

39
exploratory analyses to determine whether my results are consistent with the existing
literature.
Interactions
I explored whether gender moderated the association between autobiographical
memory and suicidality.
Data Reduction
Potential Covariates and Key Variables
Demographic variables—gender and age—were entered in all the models. Other
potential covariates were examined in relation to the key variables. Potential covariates
that did not correlate with the key variables with at least an r of .20 were not included in
the analyses. Then, covariates that met this criterion were entered in the model, and if
they did not predict the outcome with at least a β of .20, they were dropped from the
model.
Drug Use and Cognitive Functioning
Analyses then explored whether cognitive covariates were associated with drug
use covariates. Drugs that were associated with cognitive deficits and key variables were
entered as covariates in the association between cognitive covariates and key outcome
variables to determine whether drug use contributed any variance to key outcome
variables above and beyond that of cognitive deficits. Drugs that contributed
significantly to key outcome variables above and beyond cognitive covariates were
included in the key analyses as covariates. Thus, regression analyses were conducted by
entering WAIS, WRAT, WMS, and CES-D scores in Step 1 and heavy and recent drug

40
use in Step 2 to determine if drug use contributed anything unique to the key variables
above and beyond the covariates in Step 1.

41
Chapter 4: Results
Descriptive Statistics
The distributions of both the key variables and the covariates were examined, and
variables with skewed distributions were log transformed. After transformations,
distributions were again examined, and variables that were still skewed were categorized
by making cut points at equal percentile intervals when appropriate. Of the key variables,
the Ways of Coping Questionnaire subscales, the Suicide Behaviors Questionnaire (SBQ)
Lifetime Suicidality subscale, and the Iowa Gambling Task (IGT) Net Score displayed
approximately normal distributions, and no transformations were needed. The Positive
and Negative totals for the Autobiographical Memory Test (AMT) and the SBQ Ideation
subscale were still skewed after transformations.
Of the covariates, CES-D, WAIS Information, WAIS Vocabulary, WAIS
Working Memory, WMS Visual Immediate, WMS Working Memory, WMS Visual
Delayed, WMS General Memory, and WRAT Arithmetic displayed approximately
normal distributions. WAIS Processing Speed, WMS Auditory Delayed, WRAT
Reading, inhalant use in the past three months, and marijuana use in the past three
months were skewed. After transformation, WAIS Processing Speed and marijuana use
in the past three months were no longer skewed, but WMS Auditory Delayed, WRAT
Reading, and inhalant use in the past three months still were.
Due to the small sample size, there was no trimming of outliers, but key variables
(AMT Negative and Suicide Ideation) and covariates (WAIS Information, WAIS
Vocabulary, WAIS Working Memory, WMS Working Memory, and WRAT Arithmetic)

42
with outliers were tucked. After transforming and tucking when appropriate, the Positive
and Negative totals for the Autobiographical Memory Test and the SBQ Ideation
subscale were still skewed but were not categorized, as the existing literature has not
examined the acceptability of doing so. Thus, none of the key variables were
categorized. Of the covariates, WMS Auditory Delayed, WRAT Reading, and inhalant
use in the past three months were still skewed and were therefore categorized. Means
and standard deviations for the key measures are presented in Table 1.
Table 1. Means and Standard Deviations of Demographic and Key Variables
Mean Standard Deviation
Age 23.56 3.40
Autobiographical Memory Test-
Positive 2.19 1.35
Autobiographical Memory Test-
Negative 2.40 1.04
Ways of Coping-Denial 6.89 3.17
Ways of Coping-
Distancing/Avoidance 18.60 7.81
Ways of Coping-Problem-
Focused 15.94 6.36
Iowa Gambling Task 1.55 32.06
Suicide Behaviors
Questionnaire-Ideation 3.19 4.09
Suicide Behaviors
Questionnaire-Lifetime
Suicidality 4.94 4.84

There were gender differences in the means of many of the key variables.
Specifically, females displayed higher levels of all types of suicidal behaviors. Thirty-
seven percent of females and 23% of males had attempted suicide in their lifetime, χ
2
(1,
N = 54) = 1.20, ns, and females had significantly higher levels of ideation, F(1, 52) =
4.05, p = .05, and lifetime suicidality, F(1, 52) = 4.01, p = .05. Females also displayed

43
higher mean levels of WCQ Distancing/Avoidance, F(1, 50) = 1.76, p = .19, and WCQ
Denial, F(1, 51) = .68, p = .41). However, males and females did not differ in their
reports of WCQ Problem-Focused Coping, F(1, 49) = .02, p = .90.
Missing Data
The following is a summary of complete data for all variables of interest (see
Appendix for complete details): Autobiographical Memory Test: 53 (98%), Ways of
Coping Questionnaire-Problem Focused subscale: 51 (94%), Ways of Coping
Questionnaire-Distancing/Avoidance subscale: 52 (96%), Ways of Coping
Questionnaire-Denial subscale: 53 (98%), Suicide Behaviors Questionnaire-Suicide
Attempts: 54 (100%), Suicide Behaviors Questionnaire-Suicide Ideation: 54 (100%),
Suicide Behaviors Questionnaire-Lifetime Suicidality: 54 (100%), and Iowa Gambling
Task: 44 (82%).
Individuals with incomplete data on the above measures did not differ
significantly from other participants on most of the demographic or key variables. The
exceptions are as follows: On the Iowa Gambling Task, individuals who did not
complete the task had significantly higher scores on AMT Positive, F(1, 51) = 4.18, p =
.05, and significantly lower scores on WCQ Distancing/Avoidance, F(1, 50) = 5.51, p =
.02, and WCQ Denial, F(1, 51) = 4.69, p = .04. For WCQ Problem Focused, the three
individuals with missing data were significantly older, F(1, 52) = 4.17, p = .05, had lower
scores on the AMT Negative, F(1, 51) = 6.26, p = .02, and higher scores on WCQ
Distancing/Avoidance, F(1, 50) = 5.51, p = .02. For WCQ Distancing/Avoidance, the
two individuals with missing data were significantly older, F(1, 52) = 4.71, p = .04, and

44
had lower scores on AMT Negative, F(1, 51) = 3.93, p = .05, than those who completed
the subscale.
No action was taken to handle incomplete data for the following reasons: 1) In
the case of the Iowa Gambling Task, the individuals did not elect to skip the task, but
rather the measure was added to the current study after the first 9 participants were run,
and 1 individual experienced a software malfunction during completion of the task. 2) In
most cases of comparing complete and incomplete data, the assumption of
homoscedasticity was violated, so significant differences should be interpreted with
caution. 3) A very small number of cases were missing on the measures, with the
exception of the Iowa Gambling Task.
Reliability
Items on the Autobiographical Memory Test were scored twice by the same
researcher with a delay of 8 to 18 months between ratings to ensure no bias from memory
of prior ratings. The first and second ratings were highly correlated for the positive items
(r = .96, p = .00), negative items (r = .98, p = .01), and total items (r = .98, p = .01),
indicating excellent intra-rater reliability.
The Suicide Behaviors Questionnaire (α = .92), the Ways of Coping
Questionnaire (α = .93), and the CES-D (α = .90) all displayed strong internal
consistencies, while the WCQ subscales had moderate to strong internal consistencies:
Distancing/Avoidance = .78, WCQ Denial = .56, and WCQ Problem-Focused = .80.

45
Bi-variate Associations
The correlations between the key variables can be found in Table 2. The
following correlations displayed a medium or large effect size: gender and lifetime
suicidality (r = .27), gender and emotional decision-making (r = -.27), positive and
negative autobiographical memory (r = .29), WCQ Denial and emotional decision-
making (r = .29), WCQ Denial and WCQ Problem-Focused (r = .60), WCQ Denial and
WCQ Distancing/Avoidance (r = .52), WCQ Distancing/Avoidance and suicide attempt
status (r = .23), WCQ Distancing/Avoidance and WCQ Problem-Focused (r = .37), WCQ
Problem-Focused and suicide ideation (r = -.26), WCQ Problem-Focused and emotional
decision-making (r = .29), emotional decision-making and suicide ideation (r = .23),
suicide ideation and suicide attempt status (r = .54), suicide ideation and lifetime
suicidality (r = .79), and lifetime suicidality and suicide attempt status (r = .82)
Regression Analyses
Hypothesis 1a: Poorer coping and less specific autobiographical memory will be
associated with higher levels of suicide ideation. Poorer emotional decision making may
be associated with higher levels of suicide ideation or not at all.
It was predicted that lower levels of Problem-Focused Coping, Negative
Autobiographical Memory, and Positive Autobiographical Memory, and higher levels of
WCQ Denial and WCQ Distancing/Avoidance would be associated with higher levels of
suicide ideation. Additionally, it was hypothesized that higher scores on the Iowa
Gambling Task would be associated with lower levels of suicide ideation or not at all.
More specifically, poor emotional decision making could put one at risk for all suicidal

46
behaviors, or it could just inhibit individuals’ ability to make good decisions in terms of
their actions and play no role in purely cognitive processes of suicidality.
In the prediction of suicide ideation, demographic variables (gender and age) were
entered in Step 1, covariates (CES-D) in Step 2, and key variables (AMT Positive, AMT
Negative, IGT, WCQ Denial, WCQ Distancing/Avoidance, and WCQ Problem-Focused)
in Step 3 of the linear regression equation. The overall model was significant (R
2
∆ = .15,
F∆ = 1.34, p = .04; see Table 3), and the direction of the results partially confirmed the
original predictions. More specifically, WCQ Denial positively predicted (β = .11, SE =
.19, p = .55)—while WCQ Problem-Focused (β = -.36, SE = .21, p = .12) negatively
predicted—suicide ideation, but not significantly.
The association between emotional decision-making and suicide ideation was in
the opposite direction predicted. More specifically, higher levels of emotional decision-
making were significantly associated with higher levels of suicide ideation (β = .37, SE =
.14, p = .03).
Also contrary to prediction, individuals with more specific levels of Negative
Autobiographical Memories possessed higher levels of suicide ideation (β = .25, SE =
.17, p = .16). However, these results were not statistically significant.
Lastly, two variables did not predict suicide ideation: WCQ
Distancing/Avoidance (β = .002, SE = .20, p = .99) and AMT Positive (β = -.01, SE = .16,
p = .96).

47
Table 2. Correlations Amongst Key and Demographic Variables

SBQ
Attempts
SBQ
Lifetime
Suicidality
SBQ
Ideation IGT
WCQ
Problem-
Focused
WCQ
Distancing-
Avoidance
WCQ
Denial
AMT
Negative
AMT
Positive Gender Age
Age 0.06 -0.02 -0.05 0.12 0.14 0.00 -0.03 -0.22 0.01 -0.16 1.00
Gender 0.15 *0.27 0.13 *-0.27 -0.02 0.18 0.12 0.03 -0.01 1.00
AMT Positive 0.01 0.08 0.17 -0.03 -0.18 0.09 -0.15 *0.29 1.00
AMT Negative 0.08 0.01 0.10 -0.09 0.11 0.15 -0.10 1.00
WCQ Denial 0.17 0.12 -0.08 *0.29 **0.60 **0.52 1.00
WCQ
Distancing/Avoidance *0.23 0.17 0.14 0.21 **0.37 1.00
WCQ Problem-Focused -0.01 -0.15 *-0.26 *0.29 1.00
IGT -0.06 0.07 0.23 1.00
SBQ Ideation **0.54 **0.79 1.00
SBQ Lifetime Suicidality **0.82 1.00
SBQ Attempts 1.00

* p < .05
** p < .01

48

Table 3. Hypothesis 1a
Suicide Ideation
β S.E.
Demographic Variables (Step 1)
age 0.12 0.15
gender 0.20 0.17
Co-variates (Step 2)
depression (CES-D) 0.33 0.22
Key Variables (Step 3)
AMT Positive -0.01 0.16
AMT Negative 0.25 0.17
WCQ Problem-Focused -0.36 0.21
WCQ Denial 0.11 0.19
WCQ Distancing/Avoidance 0.002 0.20
Emotional Decision Making (IGT) *0.37 0.14

* p < .05

Hypothesis 1b: Poorer emotional decision making, poorer coping, and less specific
autobiographical memory will be associated with higher levels of lifetime suicidality.
It was predicted that lower levels of WCQ Problem-Focused Coping, less specific
Positive and Negative Autobiographical Memory, poorer Iowa Gambling Task
performance, and higher levels of WCQ Denial and WCQ Distancing/Avoidance would
be associated with higher levels of lifetime suicidality. Demographic variables (gender
and age) were entered in Step 1, and key variables (AMT Positive, AMT Negative, IGT,
WCQ Denial, WCQ Distancing/Avoidance, and WCQ Problem-Focused) were entered in
Step 2 of the linear regression equation. There were no significant covariates. The
overall model was significant (R
2
∆ = .22, F∆ = 1.87, p = .05), and the direction of the
results partially confirmed the original predictions (see Table 4).

49

Table 4. Hypothesis 1b
Lifetime Suicidality
β S.E.
Demographic Variables (Step 1)
age -0.01 0.10
gender 0.29 0.16
Key Variables (Step 2)
AMT Positive 0.16 0.16
AMT Negative 0.09 0.16
WCQ Problem-Focused -0.26 0.18
WCQ Denial 0.27 0.19
WCQ Distancing/Avoidance 0.25 0.17
Emotional Decision Making (IGT) 0.12 0.12

In line with predictions, WCQ Denial (β = .27, SE = .19, p = .16) and WCQ
Distancing/Avoidance (β = .25, SE = .17, p = .15) positively predicted—while WCQ
Problem-Focused (β = -.26, SE = .18, p = .18) negatively predicted—lifetime suicidality.
However, these results were not significant.
In accord with the latter of the two proposed hypotheses for the association
between emotional decision making and lifetime suicidality, higher scores on the Iowa
Gambling Task predicted higher levels of lifetime suicidality (β = .12, SE = .12, p = .49),
albeit not significantly.
Contrary to prediction, more specific Positive Autobiographical Memory
predicted higher levels of lifetime suicidality (β = .16, SE = .16, p = .31), while Negative
Autobiographical Memory did not predict lifetime suicidality (β = .09, SE = .16, p = .57).

50

Hypothesis 1c: Poorer emotional decision making, poorer coping, and less specific
autobiographical memory will be associated with increased risk for suicide attempts.
It was predicted that lower scores on the Iowa Gambling Task, WCQ Problem-
Focused, and Positive and Negative Autobiographical Memory, and higher scores on
WCQ Denial and WCQ Distancing/Avoidance would predict increased risk for suicide
attempts. Demographic variables (gender and age) were entered in Step 1 and key
variables (AMT Positive, AMT Negative, IGT, WCQ Denial, WCQ
Distancing/Avoidance, and WCQ Problem-Focused) were entered in Step 2 of the
logistic regression equation. There were no significant covariates. The individual
relations between these variables and the outcome variable are presented in Table 5 and
described in subsequent paragraphs, but the overall model was not significant (χ
2
= 11.43,
p = .18).
The result for the prediction of suicide attempts by WCQ Denial was consistent
with hypotheses. More specifically, higher levels of WCQ Denial (B = .29, SE = .21, OR
= 1.33, p = .17) predicted increased risk for suicide attempts. However, this finding was
not significant.
The results for the association between autobiographical memory and suicide
attempts were contrary to prediction, demonstrating that individuals with more specific
levels of AMT Negative (B = .37, SE = .42, OR = 1.45, p = .38) and AMT Positive (B =
.15, SE = .34, OR = 1.16, p = .67) were at increased risk for suicide attempts.

51

Lastly, WCQ Problem-Focused (B = -.05, SE = .09, OR = .95, p = .95), WCQ
Distancing/Avoidance (B = .09, SE = .08, OR = 1.10, p = .21), and Iowa Gambling Task
performance (B = -.01, SE = .02, OR = .99, p = .99) did not predict suicide attempts.
Table 5. Hypothesis 1c
Suicide Attempt Status
B S.E. OR
Demographic Variables (Step 1)
age 0.09 0.13 1.09
gender 0.54 0.96 1.72
Key Variables (Step 2)
AMT Positive 0.15 0.34 1.16
AMT Negative 0.37 0.42 1.45
WCQ Problem-Focused -0.05 0.09 0.95
WCQ Denial 0.29 0.21 1.33
WCQ Distancing/Avoidance 0.09 0.08 1.10
Emotional Decision Making (IGT) -0.01 0.02 0.99

Hypothesis 2: Coping and emotional decision-making will mediate the relation between
autobiographical memory and suicidality.
Four criteria necessary for testing mediation were examined (Baron & Kenny,
1986): 1) the predictor must be significantly associated with the hypothesized mediator,
2) the predictor must be significantly associated with the dependent variable, 3) the
mediator must be significantly associated with the dependent variable, and 4) the impact
of the predictor on the dependent variable must be less after controlling for the mediator.
If these criteria were met, the significance of the indirect effect was calculated
(Holmbeck, 2002).
Separate regression equations were run for each mediating variable as predicted
by AMT Positive and AMT Negative. In the first equation, demographic variables (age

52

and gender) were entered in Step 1, significant covariates (WMS Visual Immediate,
WMS Auditory Delayed, WMS Working Memory, WRAT Arithmetic, categorical
lifetime heavy GHB use, categorical lifetime heavy PCP use, ever used LSD, past 24
hour alcohol use, and past 24 hour speed use) in Step 2, and predictor variables (AMT
Positive and AMT Negative) in Step 3, in the prediction of emotional decision making
(R
2
∆ = .001, F∆ = .06, p = .001; see Table 6).
Table 6. Hypothesis 2: Regression Equation 1

Emotional Decision-Making
(IGT)
β S.E.
Demographic Variables (Step 1)
age *0.39 0.17
gender -0.02 0.13
Co-variates (Step 2)
WMS Visual Immediate 0.10 0.18
WMS Auditory Delayed 0.11 0.17
WMS Working Memory -0.26 0.17
WRAT Arithmetic *0.39 0.17
categorical lifetime heavy GHB use *0.27 0.13
categorical lifetime heavy PCP use -0.24 0.15
ever used LSD *0.31 0.12
past 24 hour alcohol use 0.16 0.14
past 24 hour speed use *-0.39 0.13
Key Variables (Step 3)
AMT Positive -0.01 0.26
AMT Negative 0.01 0.15

* p < .05

In the second equation, demographic variables (age and gender) were entered in
Step 1, significant covariates (WAIS Working Memory, WMS Visual Delayed, WMS
General Memory, WMS Working Memory, and ever used ketamine) in Step 2, and

53

predictor variables (AMT Positive and AMT Negative) in Step 3, in the prediction of
WCQ Denial (R
2
∆ = .05, F∆ = 1.64, p = .21; see Table 7).
Table 7. Hypothesis 2: Regression Equation 2
WCQ Denial
β S.E.
Demographic Variables (Step 1)
age -0.23 0.15
gender 0.17 0.13
Co-variates (Step 2)
WAIS Working Memory 0.29 0.22
WMS Visual Delayed -0.42 0.24
WMS General Memory 0.26 0.21
WMS Working Memory *-0.41 0.22
ever used ketamine **0.44 0.13
Key Variables (Step 3)
AMT Positive -0.21 0.14
AMT Negative -0.05 0.14

* p < .05
** p < .01

In the third equation, demographic variables (age and gender) were entered in
Step 1, significant covariates (CES-D, WMS Auditory Delayed, WRAT Arithmetic,
categorical lifetime heavy alcohol use, categorical lifetime heavy inhalant use, and
categorical lifetime heavy marijuana use) in Step 2, and predictor variables (AMT
Positive and AMT Negative) in Step 3, in the prediction of WCQ Distancing/Avoidance
(R
2
∆ = .004, F∆ = .22, p = .81; see Table 8).
Lastly, in the prediction of WCQ Problem Focused (R
2
∆ = .06, F∆ = 2.50, p =
.10; see Table 9), demographic variables (age and gender) were entered in Step 1,
significant covariates (CES-D, WAIS Information, WAIS Processing Speed, WMS

54

Visual Delayed, WMS General Memory, WRAT Reading, and ever used ketamine) in
Step 2, and AMT Positive and AMT Negative in Step 3.
Table 8. Hypothesis 2: Regression Equation 3

WCQ
Distancing/Avoidance
β S.E.
Demographic Variables (Step 1)
age 0.10 0.11
gender -0.10 0.12
Co-variates (Step 2)
CES-D **0.51 0.11
WMS Auditory Delayed **0.36 0.11
WRAT Arithmetic -0.21 0.12
categorical lifetime heavy
alcohol use *-0.25 0.12
categorical lifetime heavy
inhalant use *0.27 0.11
categorical lifetime heavy
marijuana use 0.22 0.11
Key Variables (Step 3)
AMT Positive -0.07 0.13
AMT Negative 0.06 0.11

* p < .05
** p < .01

Results indicated that AMT Negative significantly predicted the mediator, WCQ
Problem Focused (β = .28, SE = .13, p = .03). More specifically, higher levels of AMT
Negative were associated with more Problem Focused Coping. In addition, AMT
Positive predicted WCQ Denial (β = -.21, SE = .14, p = .13), albeit not significantly. The
direction of these findings is consistent with hypotheses.

55

Contrary to predictions, more specific levels of positive autobiographical
memories were associated with less WCQ Problem Focused (β = -.13, SE = .13, p = .33),
but again, not significantly.
Table 9. Hypothesis 2: Regression Equation 4
WCQ Problem-Focused
β S.E.
Demographic Variables (Step 1)
age 0.19 0.13
gender 0.21 0.14
Co-variates (Step 2)
CES-D **-0.35 0.13
WAIS Information -0.24 0.13
WAIS Processing Speed **-0.56 0.14
WMS Visual Delayed *-0.44 0.18
WMS General Memory **0.80 0.20
WRAT Reading *0.29 0.14
ever used ketamine **0.32 0.12
Key Variables (Step 3)
AMT Positive -0.13 0.13
AMT Negative *0.28 0.13

* p < .05
** p < .01

Several analyses demonstrated less than small effects. More specifically, AMT
Positive did not predict emotional decision making on the Iowa Gambling Task (β = -.04,
SE = .26, p = .75) or Distancing/Avoidance on the WCQ (β = -.07, SE = .13, p = .58).
AMT Negative did not predict emotional decision-making (β = .001, SE = .15, p = .99),
WCQ Denial (β = -.05, SE = .14, p = .72) or WCQ Distancing Avoidance (β = .06, SE =
.11, p = .59).

56

For the second criteria for testing mediation—the prediction of the outcome
variable by the predictor variable—none of the associations were significant (see
Hypothesis 1 results), but the prediction of suicide ideation by AMT Negative
demonstrated a medium sized effect.
The results of examining the third criteria—the mediator must be significantly
associated with the dependent variable—were described previously (see Hypothesis 1
results), and numerous predictions were significant and/or demonstrated a medium sized
effect: Emotional decision-making and suicide ideation, WCQ Problem-Focused and
suicide ideation, WCQ Denial and lifetime suicidality, WCQ Distancing/Avoidance and
lifetime suicidality, and WCQ Problem-Focused and lifetime suicidality.
For step four, AMT Negative (β = -.12, SE = .16, p = .46) predicted lifetime
suicidality in the hypothesized direction, albeit not significantly, while AMT Positive (β
= .24, SE = .16, p = .21) predicted lifetime suicidality in the opposite direction predicted.
AMT Positive (β = -.02, SE = .16, p = .92) and AMT Negative (β = .06, SE = .17,
p = .70) did not predict suicide ideation after adjusting for the mediating variables and
covariates detailed previously. In the prediction of suicide attempts, AMT Positive (B =
.15, SE = .34, OR = 1.16, p = .67) and AMT Negative (B = .37, SE = .42, OR = 1.45, p =
.38) were not significant either.
Not all of the conditions necessary to determine mediation were satisfied, and
therefore, the indirect effect was not calculated.

57

Aim 3: Gender will moderate the association between autobiographical memory and
suicidality.
There were no a priori predictions as to how gender would moderate the relation
between autobiographical memory and suicidality, but these relations were explored due
to previously concluded mean gender differences in both of the variables (Sampson et al.,
2003; Desai et al., 2003; Reifman & Windle, 1995).
Aim 3a: Gender will moderate the association between negative autobiographical
memory and suicide ideation.
In the prediction of suicide ideation, demographic variables (age and gender) were
entered in Step 1, significant covariates (CES-D and WRAT Arithmetic) in Step 2, AMT
Positive in Step 3, AMT Negative in Step 4, and the interaction between AMT Negative
and gender in Step 5. The overall model was significant (R
2
∆ = .01, F∆ = .45, p = .001),
but gender did not moderate the relation between AMT Negative and suicide ideation (β
= .09, SE = .15, p = .51; see Table 10).
Aim 3b: Gender will moderate the relation between positive autobiographical memory
and suicide ideation.
In the prediction of suicide ideation, demographic variables (age and gender) were
entered in Step 1, significant covariates (CES-D, WRAT Arithmetic, and categorical
lifetime heavy inhalant use) in Step 2, AMT Negative in Step 3, AMT Positive in Step 4,
and the interaction between AMT Positive and gender in Step 5. The overall model was
significant (R
2
∆ = .004, F∆ = .27, p = .01), but gender did not moderate the relation
between AMT Positive and suicide ideation (β = -.08, SE = .15, p = .60; see Table 11).

58

Table 10. Aim 3a
Suicide Ideation
β S.E.
Demographic Variables (Step 1)
age 0.13 0.15
gender 0.12 0.13
Co-variates (Step 2)
depression (CES-D) **0.41 0.14
WRAT Arithmetic **0.39 0.15
Key Variable (Step 3)
AMT Positive 0.03 0.13
Key Variable (Step 4)
AMT Negative -0.02 0.16
Interaction (Step 5)
AMT Negative x gender 0.09 0.15

** p < .01

Table 11. Aim 3b
Suicide Ideation
β S.E.
Demographic Variables (Step 1)
age 0.18 0.15
gender -0.03 0.16
Co-variates (Step 2)
depression (CES-D) **0.39 0.14
WRAT Arithmetic *0.31 0.14
categorical lifetime heavy
inhalant use 0.25 0.13
Key Variable (Step 3)
AMT Negative 0.02 0.18
Key Variable (Step 4)
AMT Positive 0.18 0.15
Interaction (Step 5)
AMT Positive x gender -0.08 0.15

* p < .05
** p < .01

59

Aim 3c: Gender will moderate the relation between negative autobiographical memory
and lifetime suicidality.
In the prediction of lifetime suicidality, demographic variables (age and gender)
were entered in Step 1, significant covariates (CES-D and inhalant use in the past 24
hours) in Step 2, AMT Positive in Step 3, AMT Negative in Step 4, and the interaction
between AMT Negative and gender in Step 5. The overall model was significant (R
2
∆ =
.06, F∆ = 4.92, p = .001), and gender moderated the relation between AMT Negative and
lifetime suicidality (β = .31, SE = .14, p = .03; see Table 12).
Table 12. Aim 3c
Lifetime Suicidality
β S.E.
Demographic Variables (Step 1)
age 0.01 0.12
gender 0.20 0.13
Co-variates (Step 2)
depression (CES-D) **0.42 0.12
inhalant use in the past 24 hours **0.36 0.12
Key Variable (Step 3)
AMT Positive -0.05 0.12
Key Variable (Step 4)
AMT Negative *-0.38 0.15
Interaction (Step 5)
AMT Negative x gender *0.31 0.14

* p < .05
** p < .01

Post-hoc probing of the model was performed by running regressions that
included conditional moderator variables (Holmbeck, 2002). Regression lines were then
plotted based on the resulting regression equations that included simple slopes and y-

60

intercepts. The graphs of these equations (see Figure 5) demonstrated that the association
between negative autobiographical memory and lifetime suicidality was inversely related
for males (β = -.38, SE = .15, p = .01). Specifically, more detailed autobiographical
memories were associated with lower levels of lifetime suicidality. The direction of this
association is consistent with the existing literature for male and female samples.
However, for females, the association was positive (β = .23, SE = .22, p = .30) but not
significant.
Figure 5. Gender As A Group Differences Variable in the Association
Between Negative Autobiographical Memory and Lifetime Suicidality
0
1
2
3
4
5
6
1 SD Below the Mean for
AMT Negative
1 SD Above the Mean for
AMT Negative
Lifetime Suicidality
males
females

Aim 3d: Gender will moderate the relation between positive autobiographical memory
and lifetime suicidality.
In the prediction of lifetime suicidality, demographic variables (age and gender)
were entered in Step 1, significant covariates (CES-D, past 3 month marijuana use, and

61

categorical lifetime heavy inhalant use) in Step 2, AMT Negative in Step 3, AMT
Positive in Step 4, and the interaction between AMT Positive and gender in Step 5. The
overall model was significant (R
2
∆ = .003, F∆ = .23, p = .003), but gender did not
moderate the relation between AMT Positive and lifetime suicidality (β = -.06, SE = .15,
p = .70; see Table 13).
Table 13. Aim 3d
Lifetime Suicidality
β S.E.
Demographic Variables (Step 1)
age 0.07 0.13
gender -0.01 0.19
Co-variates (Step 2)
depression (CES-D) **0.46 0.14
past 3 month marijuana use *-0.34 0.13
categorical lifetime heavy inhalant
use 0.22 0.14
Key Variable (Step 3)
AMT Negative -0.09 0.14
Key Variable (Step 4)
AMT Positive 0.10 0.15
Interaction (Step 5)
AMT Positive x gender -0.06 0.15

* p < .05
** p < .01

Aim 3e: Gender will moderate the relation between negative autobiographical memory
and suicide attempt status.
In the prediction of suicide attempts, demographic variables (age and gender)
were entered in Step 1, significant covariates (CES-D and WAIS Vocabulary) in Step 2,
AMT Positive in Step 3, AMT Negative in Step 4, and the interaction between AMT

62

Negative and gender in Step 5. The overall model was significant (χ
2
= 14.04, df = 7, p =
.05), and gender moderated the relation between AMT Negative and suicide attempt
status (B = 1.54, SE = .99, OR = 4.67, p = .12; see Table 14).
Table 14. Aim 3e
Suicide Attempt Status
B S.E. OR
Demographic Variables (Step 1)
age 0.09 0.14 1.09
gender -1.25 1.14 0.29
Covariates (Step 2)
CES-D 0.12 0.05 *1.13
WAIS Vocabulary 0.37 0.25 1.44
Covariate (Step 3)
AMT Positive 0.43 0.40 1.54
Key Variable (Step 4)
AMT Negative -0.75 0.65 0.47
Interaction (Step 5)
AMT Negative x Gender 1.54 0.99 4.67

* p < .05

Post-hoc probing of the model was performed by running regressions that
included conditional moderator variables (Holmbeck, 2002). Regression lines were then
plotted based on the resulting regression equations that included simple slopes and y-
intercepts. The graphs of these equations (see Figure 6) demonstrate that the association
between negative autobiographical memory and suicide attempts was inversely related
for males (B = -.75, SE = .65, OR = .47, p = .25). Specifically, more detailed
autobiographical memories were associated with a decreased risk for suicide attempts.
The direction of this association is consistent with the existing literature for male and
female samples. However, for females, the association was positive (B = .79, SE = .80,

63

OR = 2.19, p = .33). Specifically, more detailed negative autobiographical memories
were associated with an increased risk for suicide attempts. This finding was not
significant with the current N but demonstrates a large effect size.
Figure 6. Gender As A Group Differences Variable in the Association
Between Negative Autobiographical Memory and Suicide Attempt
Status
-10
-9
-8
-7
-6
-5
-4
-3
-2
-1
0
1 SD Below the Mean for
AMT Negative
1 SD Above the Mean for
AMT Negative
Attempts
males
females

Aim 3f: Gender will moderate the relation between positive autobiographical memory
and suicide attempt status.
In the prediction of suicide attempts (χ
2
= 14.97, df = 8, p = .06), demographic
variables (age and gender) were entered in Step 1, significant covariates (CES-D, WAIS
Vocabulary, WRAT Reading, categorical lifetime heavy alcohol use) in Step 2, AMT
Negative in Step 3, AMT Positive in Step 4, and the interaction between AMT Positive
and gender in Step 5. The interaction between gender and AMT Positive in the

64

prediction of lifetime suicidality was not significant (B = -.29, SE = .80, OR = .75, p =
.72; see Table 15).
Table 15. Aim 3f
Suicide Attempt Status
B S.E. OR
Demographic Variables (Step 1)
age 0.14 0.16 1.15
gender -0.56 1.06 0.57
Covariates (Step 2)
CES-D 0.09 0.05 1.09
WAIS Vocabulary 0.36 0.24 1.43
WRAT Reading 0.22 0.16 1.24
Categorical Lifetime Heavy
Alcohol Use 2.72 1.68 15.20
Covariate (Step 3)
AMT Negative -0.35 0.50 0.71
Key Variable (Step 4)
AMT Positive 0.95 0.56 2.59
Interaction (Step 5)
AMT Positive x Gender -0.20 0.84 0.82

Aim 4: Gender will moderate the mediating model: Coping and emotional decision-
making mediate the relation between autobiographical memory and suicidality.
The mediational model described previously was not significant and thus, the
moderating-mediating model was not examined.
Additional Analyses
Hypothesis: Suicide ideation and emotional decision-making will interact to predict
suicide attempt status.
It was predicted that suicide ideation and emotional decision-making would
interact in the prediction of suicide attempt status, with stronger emotional decision

65

making skills acting as a buffer against attempting suicide in those who were ideating.
This was not a planned analysis but rather conducted to follow-up on the finding that
more adaptive emotional decision-making predicted higher levels of suicide ideation, not
lower. In the logistic regression equation predicting suicide attempt status, gender and
age were entered in Step 1, IGT and suicide ideation in Step 2, and the interaction
between IGT and suicide ideation in Step 3. The results supported this hypothesis (B = -
.01, SE = .004, OR = .99, p = .12) but not significantly (see Table 16). More specifically,
post-hoc probing (Holmbeck, 2002) demonstrated that individuals with higher levels of
suicide ideation were at lower risk for attempting if they had higher levels of emotional
decision-making (see Figure 7).
Table 16. Additional analyses
Suicide Attempt Status
B SE OR
Demographic Variables (Step 1)
age 0.11 0.14 1.11
gender 0.76 1.05 2.13
Key Variables (Step 2)
Emotional Decision Making (IGT) 0.003 0.02 1.00
Suicide Ideation 0.45 0.16 **1.57
Interaction (Step 3)
Emotional Decision Making (IGT)
x Suicide Ideation -0.01 0.00 0.99

** p < .01

66

Figure 7. Emotional Decision Making Moderates the Association
between Suicide Ideation and Probability of Suicide Attempt
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
-3 0 3
Suicide Ideation
Prob (Suicide Attempt)
low
med
high

67

Chapter 5: Discussion
This study examined the association amongst components of cognitive
functioning (coping, emotional decision making, and autobiographical memory) and
suicidality (ideation, lifetime, and attempts). The results and their significance are
described below.
Coping and Suicidality
Results for the association between coping and suicidality were consistent with
other studies of homeless populations reviewed previously (e.g. Klee & Reid, 1998;
Votta & Manion, 2004). More maladaptive coping strategies were associated with higher
levels of suicidality. Specifically, higher levels of denial coping were significantly
associated with higher levels of lifetime suicide ideation with a small effect and lifetime
suicidality and increased risk for suicide attempts with a medium effect. Additionally,
distancing/avoidance coping was also associated with lifetime suicidality with a medium
effect size. Specifically, greater distancing/avoidance was associated with lower levels of
lifetime suicidality. The constructs of denial and distancing/avoidance examined in the
current study may be considered similar to the types of coping studied previously in
homeless, suicidal populations: self-medication (Klee & Reid, 1998) and disengagement
(Votta & Manion, 2004). One interpretation that can be formed from these and previous
results is that although self-medication, a form of denial of one’s problems, could
potentially be associated with a greater ability to cope and a reduction in suicidality
initially, it may be associated with long term problems that one cannot deny (e.g. going to
prison, having their children taken from them, etc.). Through the creation of

68

uncontrollable stressors and the possession of inadequate coping mechanisms, individuals
may be at increased risk for suicidality. Additionally, it is possible that substance users
may be more suicidal even in the absence of additional stressors because their substance
use prevents them from being able to cope effectively.
Higher levels of problem-focused coping were moderately associated with lower
levels of both suicide ideation and lifetime suicidality. This finding is consistent with
studies of housed populations (Langhinrichsen-Rohling, O'Brien, Klibert, Arata, &
Bowers, 2006) but inconsistent with a recent study of homeless young adults (median age
= 20 years), which found that problem-focused coping was not associated with suicide
ideation or attempts (Kidd & Carroll, 2007). However, the authors of the latter study
only utilized two items from the WCQ Problem-Focused subscale, which may have
altered their results.
Another explanation for the differing results in the current study compared to the
Kidd and Carroll (2007) study is that participants may not have experienced comparable
numbers of uncontrollable stressors. Homeless individuals typically face a high number
of stressors that are outside of their control, and therefore, utilization of problem-focused
coping, which is usually most effective in situations where one has the power to
implement change, may not be as helpful. Future research should examine the role of
controllability of stressors as a moderator of this relation. It is likely that homeless
individuals who experience more uncontrollable stressors may actually be more suicidal
when they attempt to engage in problem-focused coping.

69
Greater Specificity of Autobiographical Memories Contributes to More Adaptive Coping
Consistent with predictions, more specific autobiographical memories were
associated with more adaptive coping. Higher levels of negative autobiographical
memory were significantly associated with higher levels of problem-focused coping.
These results are consistent with previous studies examining problem solving abilities
and autobiographical memory. Specifically, Kaviani et al. (2003 and 2005) found that
more irrelevant and ineffective problem solving methods were associated with greater
overgenerality of autobiographical memories.
Additionally, more specific positive autobiographical memory was associated
with lower levels of denial coping with a medium effect. This finding is consistent with
Hermans et al. (2005) and Raes’ et al. (2006) results, found among older adolescent
students, demonstrating that more avoidant and more repressive coping, respectively,
were associated with less specific autobiographical memory. Thus, the current study
extends this finding to a homeless population of late adolescents and young adults.
It is possible that the ability to remember the details and outcomes of past
experiences allows individuals to adjust coping strategies in future situations as needed.
Additionally, specific positive life memories may be associated with the motivation or
self-efficacy to face subsequent situations. Thus, having specific knowledge of situations
that turned out favorably may predict the use of adaptive coping strategies in the future.

70

Negative Autobiographical Memories and Suicidality: The Role of Gender and
Controllability of Stressors
It was predicted that more specific negative autobiographical memory would be
associated with lower levels of suicidality. However, the reverse was found: more
specific negative autobiographical memory was moderately associated with higher levels
of suicide ideation and increased risk for suicide attempts. Perhaps the simplest
explanation for this finding is that in this population of young adult drug abusers, greater
access to negative autobiographical memories may have served as a reminder of their
poor life choices, which was associated with increased hopelessness and helplessness and
subsequent suicidality.
Another explanation is that autobiographical memories were only effective in
decreasing suicidality if individuals had the cognitive resources to make use of them. If
they did not have the cognitive capacity to both keep in mind past experiences and then
consider what could be done in the present situation, they would not have been able to
use the specific memories in an adaptive way.
A related possibility is that one must have experienced life circumstances that
were changeable in order for negative autobiographical memories to predict lower levels
of suicidality. Homeless populations often face more stressors, such as drug addiction
(Prigerson, Desai, Liu-Mares, & Rosenheck, 2003; Havens, Sherman, Sapun, &
Strathdee, 2006), barriers to services and dependent children (Buckner, Bassuk, & Zima,
1993; Flynn, 1997; Ingram, Corning, & Schmidt, 1996; Merves, 1992; Nyamathi,
Bennett, Leake, & Chen, 1995; Stein & Gelberg, 1995, 1997), and weaker and more

71

frequently disrupted social networks (Bassuk & Rosenberg, 1988; Wood, Valdez,
Hayashi, & Shen, 1990) and have less control over alleviating those stressors compared
to housed populations, which may result in increased hopelessness, helplessness, and
suicidality. This lack of control over stressors may be especially true for females, who
are often dependent on others for protection (e.g. 84% of the homeless population in
families are women, while only 23% of homeless singles are women; “Key Data
Concerning Homeless Persons in America,” 2004), when these significant others are
often the source of violence, discord, and unhappiness in their lives (Recker Rayburn,
Wenzel, Elliott, Hambarsoomians, Marshall, & Tucker, 2005).
To test the possibility that specific negative autobiographical memory was a
detriment to females—who may have less control over their life circumstances—gender
as a moderator of the association between negative autobiographical memory and
suicidality was examined. Results indicated that gender moderated the relation between
negative autobiographical memories and suicide attempts and negative autobiographical
memories and lifetime suicidality but not negative autobiographical memories and
suicide ideation. Post hoc probing indicated that for females, more specific negative
autobiographical memory was associated with increased risk for suicide attempts and
higher levels of lifetime suicidality. The association was the opposite for males: More
specific negative autobiographical memory was associated with decreased risk for suicide
attempts and lower levels of lifetime suicidality. These results indicate that having
specific negative autobiographical memories may actually be a detriment for homeless
females, who have may less control over their life circumstances. Further research is

72

needed to determine the role of gender differences in controllability of stressors in the
relation between negative autobiographical memory and suicidality.
One potential explanation for why gender did not moderate the relation between
negative autobiographical memory and suicide ideation is that both males and females
may have already been thinking about suicide for various reasons, but having specific
negative memories contributed to females taking action (attempting suicide), potentially
because they did not have control over their life stressors.
Strong Emotional Decision-Making Provides the Cognitive Resources to Ideate Suicide
but Acts as a Buffer against Attempting
One significant association was inconsistent with the initial prediction. It was
hypothesized that better emotional decision-making skills would either be associated with
lower levels of suicide ideation, or the two variables would not be associated at all.
However, in this study, higher levels of emotional decision-making were actually
associated with higher levels of suicide ideation with a medium-sized effect. One
potential explanation for this unexpected finding is that better emotional decision-making
provides individuals who are at-risk for suicidality with the cognitive resources to
consider attempting. Nevertheless, possessing more cognitive resources also acts as a
protective mechanism against actually attempting suicide because individuals are
effectively able to weigh the costs and benefits of doing so, ultimately controlling their
behavior and avoiding an attempt. Additional analyses suggested this was the case:
suicide ideation and emotional decision-making interacted with a moderate effect on
suicide attempt status, albeit not significantly. Specifically, individuals who were

73

ideating suicide were less likely to attempt if they had strong emotional decision-making
abilities. On the contrary, individuals who were ideating suicide but had poor emotional
decision-making abilities were at greater risk for attempting. Therefore, the results are
suggestive that strong emotional decision making skills acted as a buffer in preventing
suicide attempts in individuals who were at-risk (e.g. ideating suicide). This finding
suggests that an adequate somatic marker is associated with a decreased risk of
attempting suicide in those who are ideating, just as it is associated with a decreased risk
of using drugs in those who have used in the past. These results have important
implications for informing suicide risk assessment: the people who are at highest risk for
attempting are those who are thinking about suicide but do not have the emotional
decision-making capacity to prevent themselves from acting on those impulses.
Limitations
There are several limitations of the current study, which are described below.
Measurement
Shared method variance may have played a role in the results pertaining to one of
the hypotheses. The association between coping and suicidality may have been due, in
part, to shared-method variance, given that both the SBQ and WCQ are questionnaires.
However, all of the other analyses involved use of different methods of assessment (e.g.
interviews, computer tasks, and questionnaires), reducing the likelihood that observed
associations were due to shared method variance. Thus, overall, the study aimed to
reduce shared-method variance whenever possible and did an excellent job of achieving
this objective.

74

Another limitation is that the participants’ scores on the key variables may have
been influenced by other factors such as drug use, intellectual ability, memory,
depression, and verbal abilities. The inclusion of these constructs as potential covariates
helped to adjust for alternative explanations for the relations between the key variables.
Power
The power analyses for the current study were based on linear regression
(Hypothesis 1) and did not really address Hypotheses 2 - 4. In order to detect significant
results for the moderating models, an N twice the size of that for Hypothesis 1 was
needed. The research reviewed previously found medium to large effect sizes for the
association between the variables in the moderating model (autobiographical memory and
suicidality). Indeed one of the moderating models was significant despite the sample
size. For the non-significant models, sample size may have played a role since small to
medium sized effects were found. The number of available participants was constrained
by the difficulty in recruiting this population and conducting extensive testing.
Inferences Based on Results
Because the current study examined a homeless population, the results may not be
generalizable to other populations. In addition, it is possible that drug use may have
altered the relations amongst the key variables. For example, the association between
specific autobiographical memories and higher levels of suicidality may have been
partially explained by higher levels of drug use. These potential issues were addressed
by adjusting for drug use in the appropriate analyses, and by examining the existing

75

literature to determine the extent to which drug use has played a role in the association
amongst key variables in the past.
Conclusions and Future Directions
The results of the current study suggest that substance users may have decreased
cognitive capacity, which may explain their difficulty on tasks that require working
memory. Limited cognitive resources may negatively impact their ability to cope and
problem solve and increases their risk for suicide attempts, even when they remember
specifically how they behaved in past situations. Future research should specifically
examine substance abusers’ ability to perform problem solving tasks when various
degrees of working memory are taxed.
In addition, gender played an important role in moderating the relation between
autobiographical memory and suicidality. One potential explanation is that homeless
females have less control over life stressors, and as a result, recall of situations that are
outside of their control may just increase frustration and suicidality. Future research
should examine the role of controllability of stressors in the relation between
autobiographical memory and suicidality. It is also important to consider that, in this
population, adaptive attempts to cope may not be reinforced due to the lack of
controllability, or perceived lack of controllability, of life stressors.
These results have important implications for assessment and treatment of suicidal
patients. Individuals who are ideating suicide should not only be assessed for suicidal
desire, plans, and intent (Joiner et al., 2007), but also the controllability of their life
stressors and type of coping mechanisms employed (if any). It is important to note that

76

not all coping is good coping if the techniques employed do not match the situation at
hand. Specifically, if individuals are trying to utilize problem-focused coping in
situations that are uncontrollable, they may feel more frustrated when they discover they
cannot implement change in their life situations. Therefore, treatment should focus on
developing appropriate coping techniques for the situation at hand. Similarly, discussing
adaptive emotional decision making plans may be of use to individuals who do not have
the ability to weigh the costs and benefits of attempting suicide when in crisis.

77

Bibliography
Ackerly, S.S., & Benton, A.L. (1948). Report of a case of bilateral frontal lobe defect.
Proceedings of the Association for Research in Nervous and Mental Disease, 27,
479–504.

Altemus, M. (2006). Sex differences in depression and anxiety disorders: Potential
biological determinants. Hormones and Behavior, 50, 4, 534-538.

Andorn, A.C. (1991). The low affinity component of [-3H]clonidine binding is absent in
the prefrontal cortex of presumptive suicide victims. Brain Research, 568, 1-2,
276-278.

Arango, V., Underwood, M.D., Gubbi, A.V. & Mann, J.J. (1995). Localized alterations
in pre- and postsynaptic serotonin binding sites in the ventrolateral prefrontal
cortex of suicide victims. Brain Research, 688, 121–133.

Arora, R.C. & Meltzer, H.Y. (1989). Serotonergic measures in the brains of suicide
victims: 5-HT-sub-2 binding sites in the frontal cortex of suicide victims and
control subjects. American Journal of Psychiatry, 146, 6, 730-736.

Bartzokis, G., Lu, P.H., Beckson, M., Rapoport, R., Grant, S., Wiseman, E.J., et al.
(2000). Abstinence from cocaine reduces high-risk responses on a gambling task.
Neuropsychopharmacology, 22, 1, 102-103.

Bassuk, E.L., & Rosenberg, L. (1988). Why does family homelessness occur? A case-
control study. American Journal of Public Health, 78, 783–788.

Bechara, A., Damasio, A.R., Damasio, H., & Anderson, S.W. (1994). Insensitivity to
future consequences following damage to human prefrontal cortex. Cognition,
50, 7-15.

Bechara, A., Damasio, H., Tranel, D., and Damasio, A.R. (1997). Deciding
advantageously before knowing the advantageous strategy. Science, 275, 1293-
1295.

Bechara, A., Damasio, H., & Damasio, A.R. (2000). Emotion, decision making, and the
orbitofrontal cortex. Cerebral Cortex, 10, 295-307.

Bechara, A., Damasio, H., & Damasio, A.R. (2003). The amygdala in brain function:
Basic and Clinical Approaches. Annals of the New York Academy of Sciences,
985, 356-369.

78

Bechara, A. (2004). The role of emotion in decision-making: Evidence from
neurological patients with orbitofrontal damage. Brain and Cognition, 55, 30-40.

Bechara, A., Dolan, S., Denburg, N., Hindes, A., Anderson, S.W., & Nathan, P.E. (2001).
Decision-making deficits, linked to a dysfunctional ventromedial prefrontal
cortex, revealed in alcohol and stimulant abusers. Neuropsychologia, 39, 376-
389.

Bechara, A., Dolan, S., & Hindes, A. (2002). Decision-making and addiction (part II):
myopia for the future or hypersensitivity to reward? Neuropsychologia, 40, 1690-
1705.

Bechara, A., Damasio, H., Damasio, A.R., & Lee, G.P. (1999). Different contributions of
the human amygdala and ventromedial prefrontal cortex to decision-making. The
Journal of Neuroscience, 19, 13, 5473–5481.

Bechara, A. & Martin, E.M. (2004). Impaired decision making related to working
memory deficits in individuals with substance addictions. Neuropsychology, 18,
1, 152-162.

Bechara, A. & Damasio, H. (2002). Decision-making and addiction (part I): impaired
activation of somatic states in substance dependent individuals when pondering
decisions with negative future consequences. Neuropsychologia, 40, 1675–1689.

Bechara, A., Damasio, H., Tranel, D., & Damasio, A.R. (2005). The Iowa Gambling
Task and the somatic marker hypothesis: Some questions and answers. Trends in
Cognitive Sciences, 9, 4, 159-162.

Biegon, A. & Israeli, M. (1988). Regionally selective increases in !b-adrenergic receptor
density in the brains of suicide victims. Brain Research, 442, 1, 199-203.

Billings, A.G. & Moos, R.H. (1984). Coping, stress, and social resources among adults
with unipolar depression. Journal of Personality and Social Psychology, 46, 4,
877-891.

Breiter, H.C., Gollub, R.L., Weisskoff, R.M., Kennedy, D.N., Makris, N., Berke, J.D., et
al. (1997). Acute effects of cocaine on human brain activity and emotion.
Neuron, 19, 591-611.

Brent, D., Perper, J.A., Kolko, D.J., & Zelenak, J.P. (1988). The psychological autopsy:
Methodological considerations for the study of adolescent suicide. Journal of the
American Academy of Child & Adolescent Psychiatry, 27, 3, 362-366.

79

Brewin, C.R., Reynolds, M., & Tata, P. (1999). Autobiographical memory processes and
the course of depression. Journal of Abnormal Psychology, 108, 3, 511-517.

Brickner, R.M. (1932). An interpretation of frontal lobe function based upon the study of
a case of partial bilateral frontal lobectomy. Localization of function in the
cerebral cortex. Proceedings of the Association for Research in Nervous and
Mental Disease (Baltimore), 13, 259–351.

Brittlebank, A.D., Scott, J., Williams, J.M.G., & Ferrier, I.N. (1993). Autobiographical
memory in depression: State or trait marker? British Journal of Psychiatry, 162,
118-121.

Bucholz, K.K., Cadoret, R., Cloninger, C.R., Dinwiddie, S.H., et al. (1994). A new,
semi-structured psychiatric interview for use in genetic linkage studies: A report
on the reliability of the SSAGA. Journal of Studies on Alcohol, 55, 2, 149-158.

Buckner, J.C., Bassuk, E.L., & Zima, B.T. (1993). Mental health issues affecting
homeless women: Implications for intervention. American Journal of
Orthopsychiatry, 67, 385–399.

Burt, M. (2001). Helping America’s homeless, 47. The Urban Institute Press.

Childress, A.R., Mozley, P.D., McElgin, W., Fitzgerald, J., Reivich, M., & O’Brien, C.P.
(1999). Limbic activation during cue-induced cocaine craving. American
Journal of Psychiatry, 156, 11-18.

Clements, K., Gleghorn, A., Garcia, D., Katz, M., & Marx, R. (1997). A risk profile of
street youth in Northern California: Implications for gender-specific human
immunodeficiency virus prevention. Journal of Adolescent Health, 20, 5, 343-
353.

Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 1, 155-159.

Cotton, C.R., Peters, D.K., & Range, L.M. (1995). Psychometric properties of the
suicidal behaviors questionnaire. Death Studies, 19, 4, 391-397.

Dalgleish, T., Yiend, J., Bramham, J., Teasdale, J.D., Ogilvie, A.D., Malhi, G., et al.
(2004). Neuropsychological processing associated with recovery from depression
after stereotactic subcaudate tractotomy. American Journal of Psychiatry, 161,
1913-1916.

80

Damasio, A.R. (1995). On some functions of the human prefrontal cortex. In J.
Grafman, K.J. Holyoak, & F. Boller, (Eds), Structure and functions of the human
prefrontal cortex. Annals of the New York academy of sciences, 769 (pp. 241-
251). New York: New York Academy of Sciences.

Damasio, H., Bechara, A., Tranel, D., & Damasio, A.R. (1997). Double dissociation of
emotional conditioning and emotional imagery relative to the amygdala and right
somatosensory cortex. Society for Neuroscience Abstract, 23, 1318.

Damasio, H., Grabowski, T., Frank, R., Galaburda, A.M., & Damasio, A.R. (1994). The
return of Phineas Gage: Clues about the brain from the skull of a famous patient.
Science, 264, 5162, 1102-1105.

Department of Health and Human Services. (1992, March). Alcohol, drug, and mental
health services for homeless individuals. Retrieved January 11, 2007 from
http://oig.hhs.gov/oei/reports/oei-05-91-00062.pdf

Desai, R.A., Liu-Mares, W., Dausey, D.J., & Rosenheck, R.A. (2003). Suicidal ideation
and suicide attempts in a sample of homeless people with mental illness. The
Journal of Nervous and Mental Disease, 191, 6, 365-371.

DiClemente, C.C. (2003). Addiction and change: How addictions develop and addicted
people recover. Guilford substance abuse series. New York: Guilford Press.

Durkheim, E. (1966). Suicide: a study in sociology. New York: The Free Press.

Eiber, R., Puel, M., & Schmitt, L. (1999). Substance abuse, autobiographical memory
and depression. L'Encéphale, 25, 6, 549-557.

Ensign, J. (1998). Health issues of homeless youth. Journal of Social Distress and the
Homeless, 7, 3, 159-174.

Ernst, M., Grant, S.J., London, E.D., Contoreggi, C.S., Kimes, A.S., & Spurgeon, L.
(2003). Decision making in adolescents with behavior disorders and adults with
substance abuse. American Journal of Psychiatry, 160, 33–40.

Eslinger, P.J., & Damasio, A.R. (1985). Severe disturbance of higher cognition after
bilateral frontal lobe ablation: patient EVR. Neurology, 35, 1731–1741.

Eynan, R., Langley, J., Tolomiczenko, G., Rhodes, A.E., Links, P., Wasylenki, D., et al.
(2002). The association between homelessness and suicidal ideation and
behaviors: Results of a cross-sectional survey. Suicide and Life Threatening
Behavior, 32, 4, 418-427.

81

Folkman, S. (1984). Personal control and stress and coping processes. A theoretical
analysis. Journal of Personality and Social Psychology, 46, 4, 839-852.

Folkman, S. & Lazarus, R.S. (1985). If it changes, it must be a process: Study of
emotion and coping during three stages of a college examination. Journal of
Personality and Social Psychology, 48, 150-170.

Folkman, S., Lazarus, R.S., Dunkel-Schetter, C., DeLongis, A., & Gruen, R.J. (1986).
Dynamics of a stressful encounter: Cognitive appraisal, coping, and encounter
outcomes. Journal of Personality and Social Psychology, 50, 992-1003.

Folkman, S. & Lazarus, R.S. (1988). Ways of Coping Questionnaire: Research edition.
Palo Alto, CA: Consulting Psychologists Press.

Flynn, L. (1997). The health practices of homeless women: A causal model. Nursing
Research, 46, 72–77.

Galaif, E.R., Nyamathi, A.M., & Stein, J.A. (1999). Psychosocial predictors of current
drug use, drug problems, and physical dependence in homeless women. Addictive
Behaviors, 24, 6, 801–814.

Goddard, L., Dritschel, B., & Burton, A. (1996). Role of autobiographical memory in
social problem solving and depression. Journal of Abnormal Psychology, 105, 4,
609-616.

Grant, S., London, E.D., Newlin, D.B., Villemagne, V.L., Liu, X., Contereggi, C., et al.
(1996). Activation of memory circuits during cue-elicited cocaine craving.
Proceedings of the National Academy of Science USA, 93, 12040-12045.

Grant, S.J., Bonson, K.R., Contoreggi, C.C., & London, E.D. (1999). Activation of the
ventromedial prefrontal cortex correlates with gambling task performance: a
FDG-PET study. Abstract for the Society for Neuroscience, 25, 1551.

Grant, S.J., Contoreggi, C.C., & London, E.D. (2000). Drug abusers show impaired
performance in a laboratory test of decision making. Neuropsychologia, 38, 8,
1180-1187.

Harlow, J.M. (1848). Passage of an iron bar through the head. Boston Medical and
Surgical Journal, 39, 389–393.

Harlow, J.M. (1868). Recovery from the passage of an iron bar through the head.
Publications of the Massachusetts Medical Society, 2, 327–347.

82

Hermans, Defranc, Raes, Williams, and Eelen (2005). Reduced autobiographical
memory specificity as an avoidant coping style. British Journal of Clinical
Psychology, 44, 583–589.

Holden, R.R., Kerr, P.S., Mendonca, J.D., & Velamoor, V.R. (1998). Are some motives
more linked to suicide proneness than others? Journal of Clinical Psychology, 54,
569–576.

Ingram, K.M., Corning, A.F., & Schmidt, L.D. (1996). The relationship of victimization
experiences to psychological well-being among homeless women and low-income
housed women. Journal of Counseling Psychology, 43, 218–227.

Joiner, T., Kalafat, J., Draper, J., Stokes, H., Knudson, M., Berman, A.L., et al. (2007).
Establishing standards for the assessment of suicide risk among callers to the
National Suicide Prevention Lifeline. Suicide and Life-Threatening Behavior, 37,
3, 353-365.

Jollant, F., Bellivier, F., Leboyer, M., Astruc, B., Torres, S., Verdier, R., et al. (2005).
Impaired decision-making in suicide attempters. American Journal of Psychiatry,
162, 304–310.

Kaviani, H., Rahimi, M., Rahimi-Darabad, P., Kamyar, K., & Naghavi, H. (2003). How
autobiographical memory deficits affect problem-solving in depressed patients.
Acta Medica Iranica, 41, 3, 194-198.

Kaviani, H., Rahimi-Darabad, P., & Naghavi, H.R. (2005). Autobiographical memory
retrieval and problem-solving deficits of Iranian depressed patients attempting
suicide. Journal of Psychopathology and Behavioral Assessment, 27, 1, 39-44.

Key Data Concerning Homeless Persons in America. (July 2004). National Law Center
on Homelessness and Poverty.
http://www.nlchp.org/FA_HAPIA/HomelessPersonsinAmerica.pdf

Kidd, S.A. & Carroll, M.R. (2007). Coping and suicidality among homeless youth.
Journal of Adolescence, 30, 2, 283-296.

Klee, H., & Reid, P. (1998). Drug use among the young homeless: coping through self-
medication. Health, 2, 2, 115–134.

Kuyken, W., & Brewin, C.R. (1995). Autobiographical memory functioning in
depression and reports of early abuse. Journal of Abnormal Psychology, 104, 4,
585-591.

83

Lamers, C.T.J., Bechara, A., Rizzo, M., & Ramaekers, J.G. (2006). Cognitive function
and mood in MDMA/THC users, THC users, and non-drug using controls.
Journal of Psychopharmacology, 20, 2, 302-311.

Langhinrichsen-Rohling, J., O'Brien, N., Klibert, J., Arata, C., & Bowers, D. (2006).
Gender specific associations among suicide proneness and coping strategies in
college men and women. In M. Landow (Ed.), College students: Mental health
and coping strategies (pp. 247-260). New York: Nova Science Publishers.

Lazarus, R.S. & Folkman, S. (1984). Stress, appraisal, and coping. New York: Springer.

Lewinsohn, P.M., Langhinrichsen-Rohling, J., Langford, R., Rohde, P., Seeley, J.R., &
Chapman, J. (1995). The life attitudes schedule: A scale to assess adolescent life-
enhancing and life-threatening behaviors. Suicide and Life-Threatening Behavior,
25, 4, 458-474.

Linehan, M. (1981). A social-behavioral analysis of suicide and parasuicide: Implications
for clinical assessment and treatment. In H. Glazer & J. Clarkin (Eds.),
Depression: Behavioral and directive intervention strategies. New York: Garland
Press.

Littrel, J., & Beck, E. (2001). Predictors of depression in a sample of African-American
homeless men: Identifying effective coping strategies given varying levels of
daily stressors. Community Mental Health Journal, 37, 1, 15-30.

Mann, J.J. (1998). The neurobiology of suicide. Nature Medicine, 4, 1, 25-30.

Mann, J.J. (2003). Neurobiology of suicidal behavior. Nature Reviews Neuroscience, 4,
819-828.

Masten, A.S., Burt, K.B., Roisman, G.I., Obradovic, J.O., Long, J.D., & Tellegen, A.
(2004). Resources and resilience in the transition to adulthood: Continuity and
change. Development and Psychopathology, 16, 1071-1094.

Menninger, K.A. (1938). Man against himself. Oxford, England: Harcourt, Brace.

Merves, E.S. (1992). Homeless women: Beyond the bag lady myth. In M. J. Robertson &
M. Greenblatt (Eds.), Homelessness: A national perspective. Topics in social
psychiatry (pp. 229–244). New York: Plenum Press.

Nolen-Hoeksema, S. (1994). An interactive model for the emergence of gender
differences in depression in adolescence. Journal of Research on Adolescence.
Special Issue: Affective processes in adolescence, 4, 4, 519-534.

84

Nyamathi, A., Bennett, C., Leake, B., & Chen, S. (1995). Social support among
impoverished women. Nursing Research, 44, 376–378.

Nyamathi, A., Stein, J.A., & Brecht, M. (1995). Psychosocial predictors of AIDS risk
behavior and drug use behavior in homeless and drug addicted women of color.
Health Psychology, 14, 3, 265-273.

Nyamathi, A., Stein, J.A., Dixon, E., Longshore, D., & Galaif, E. (2003). Predicting
positive attitudes about quitting drug and alcohol use among homeless women.
Psychology of Addictive Behaviors, 17, 1, 32–41.

Owen, A.M., Downes, J.J., Sahakian, B.J., Polkey, C.E., & Robbins, T.W. (1990).
Planning and spatial working memory following frontal lobe lesions in man.
Neuropsychologia, 28, 10, 1021–1034.

Park, R.J., Goodyer, I.M., & Teasdale, J.D. (2002). Categoric overgeneral
autobiographical memory in adolescents with major depressive disorder.
Psychological Medicine, 32, 267-276.

Petry, N.M., Bickel, W.K., & Arnett, M. (1998). Shortened time horizons and
insensitivity to future consequences in heroin addicts. Addiction, 93, 5, 729-738.

Pollock, L.R. & Williams, J.M.G. (2001). Effective problem solving in suicide
attempters depends on specific autobiographical recall. Suicide and Life
Threatening Behavior, 31, 4, 386-396.

Radloff, L.S. (1977). The CES-D scale: A self-report depression scale for research in the
general population. Applied Psychological Measurement, 1, 385-401.

Raes, F., Hermans, D., de Decker, A., Eelen, P., & Williams, J.M.G. (2003).
Autobiographical memory specificity and affect regulation: An experimental
approach. Emotion, 3, 2, 201-206.

Raes, F., Hermans, D., Williams, J.M.G., Demyttenaere, K., Sabbe, B., Pieters, G., et al.
(2005). Reduced specificity of autobiographical memory: A mediator between
rumination and ineffective social problem-solving in major depression? Journal
of Affective Disorders, 87, 331-335.

Raes, F., Hermans, D., Williams, J.M.G., & Eelen, P. (2006). Reduced autobiographical
memory specificity and affect regulation. Cognition & Emotion. Special Issue:
Autobiographical memory specificity and psychopathology, 20, 3-4, 402-429.

85

Raes, F., Hermans, D., Williams, J.M.G., Demyttenaere, K., Sabbe, B., Pieters, G., &
Eelen, P. (2006). Is overgeneral autobiographical memory an isolated memory
phenomenon in major depression? Memory, 14, 5, 584-594.

Recker Rayburn, N., Wenzel, S.L., Elliott, M.N., Hambarsoomians, K., Marshall, G.N.,
& Tucker, J.S. (2005). Trauma, depression, coping, and mental health service
seeking among impoverished women. Journal of Consulting and Clinical
Psychology, 73, 4, 667–677.

Reifman, A. & Windle, M. (1995). Adolescent suicidal behaviors as a function of
depression, hopelessness, alcohol use, and social support: A longitudinal
investigation. American Journal of Community Psychology, 23, 3, 329-354.

Ringwalt, C.L., Greene, J.M., Robertson, M., & McPheeters, M. (1998). The prevalence
of homelessness among adolescents in the United States. American Journal of
Public Health, 88, 9, 1325-1329.

Robbins, T.W., Shallice, T., Burgess, P.W., James, M., Rogers, R.D., Warburton, E., et
al. (1995). Selective impairments in self-ordered working memory in a patient
with a unilateral striatal lesion. Neurocase, 1, 3, 217-230.

Rohde, P., Noell, J., Ochs, L., & Seeley, J.R. (2001). Depression, suicidal ideation and
STD-related risk in homeless older adolescents. Journal of Adolescence, 24, 447–
460.

Rosenheck, R.A. & Koegel, P. (1993). Characteristics of veterans and non-veterans in
three samples of homeless men. Hospital and Community Psychiatry, 44, 858–
863.

Rotheram-Borus, M.J. (1993). Suicidal behavior and risk factors among runaway youths.
The American Journal of Psychiatry, 150, 1, 103-107.

Salit, S.A., Kuhn, E.M., & Hartz A.J. (1998). Hospitalization costs associated with
homelessness in New York City. New England Journal of Medicine, 338, 1734–
1740.

Sampson, M.J., Kinderman, P., Watts, S., & Sembi, S. (2003). Psychopathology and
autobiographical memory in stroke and non-stroke hospitalized patients.
International Journal of Geriatric Psychiatry, 18, 23-32.

Saver, J.L. & Damasio, A.R. (1991). Preserved access and processing of social
knowledge in a patient with acquired sociopathy due to ventromedial frontal
damage. Neuropsychologia, 29, 12, 1241-1249.

86

Sidley, G.L.,Whitaker, K., Calam, R.M., & Wells, A. (1997). The relationship between
problem-solving and autobiographical memory in parasuicide patients.
Behavioural and Cognitive Psychotherapy, 25, 2, 195-202.

Spotswood, S. (April 2006). More Funding, Grass-Roots Effort Needed To Serve
Homeless Veterans. U.S. Medicine,
http://www.usmedicine.com/article.cfm?articleID=1291&issueID=86.

Stanley, M., Virgilio, J., & Gershon, S. (1982). Tritiated imipramine binding sites are
decreased in the frontal cortex of suicides. Science, 216, 4552, 1337-1339.

Stapleton, J.M., Morgan, M.J., Phillips, R.L., Wong, D.F., Yung, B.C., Shaya, E.K., et al.
(1995). Cerebral glucose utilization in polysubstance abuse.
Neuropsychopharmacology, 13, 21-31.

Startup, M., Heard, H., Swales, M., Jones, B., Williams, J.M.G., & Jones, R.S.P. (2001).
Autobiographical memory and parasuicide in borderline personality disorder.
British Journal of Clinical Psychology, 40, 113-120.

Stein, J.A., & Gelberg, L. (1995). Homeless men and women: Differential associations
among substance abuse, psychosocial factors, and severity of homelessness.
Experimental and Clinical Psychopharmacology, 3, 75–86.

Stein, J.A., & Gelberg, L. (1997). Comparability and representativeness of clinical
homeless, community homeless, and domiciled clinic samples: Physical and
mental health, substance use, and health services utilization. Health Psychology,
16, 155–162.

Stone, M.H., Jastak, S., & Wilkinson, G. (1995). Wide Range Achievement Test WRAT-
3, administration manual. Wilmington, Delaware: Wide Range, Inc.

Susser, E., Struening, E.L., & Conover, S. (1989). Psychiatric problems in homeless
men. Lifetime psychosis, substance use, current distress in new arrivals at New
York City shelters. Archives of General Psychiatry, 46, 845–850.

Swales, M.A., Williams, J.M.G., & Wood, P. (2001). Specificity of autobiographical
memory and mood disturbance in adolescents. Cognition & Emotion, 15, 3, 321-
331.

Tapert, S.F., Ozyurt, S.S., Myers, M.G., & Brown, S.A. (2004). Neurocognitive ability
in adults coping with alcohol and drug relapse temptations. The American
Journal of Drug and Alcohol Abuse, 30, 2, 445-460.

87

Theodoulou, M., Harriss, L., Hawton, K., & Bass, C. (2005). Pain and deliberate self-
harm: An important association. Journal of Psychosomatic Research, 58, 4, 317-
320.

Tranel, D., Bechara, A., Damasio, H., & Damasio, A.R. (1998). Neural correlates of
emotional imagery. International Journal of Psychophysiology, 30, 107.

Träskman-Bendz, L. & Mann, J.J. (2000). Biological aspects of suicidal behaviour. In
K. Hawton & K. van Heeringen (eds), The international handbook of suicide and
attempted suicide (pp. 65-77). New York: John Wiley and Sons Ltd.

Unger, J.B., Kipke, M.D., Simon, T.R., Montgomery, S.B., & Johnson, C.J. (1997).
Homeless youths and young adults in Los Angeles: Prevalence of mental health
problems and the relationship between mental health and substance abuse
disorders. American Journal of Community Psychology, 25, 3, 371-394.

Volkow, N.D., Chang, L., Wang, G.J., Fowler, J.S., Ding, Y.S., Sedler, M., et al. (2001).
Low level of brain dopamine D2 receptors in methamphetamine abusers:
association with metabolism in the orbitofrontal cortex. American Journal of
Psychiatry, 158, 2015–2021.

Votta, E. & Manion, I. (2003). Factors in the psychological adjustment of homeless
adolescent males: the role of coping style. Journal of the American Academy of
Child and Adolescent Psychiatry, 42, 7, 778-785.

Votta, E. & Manion, I. (2004). Suicide, high-risk behaviors, and coping style in
homeless adolescent males’ adjustment. Journal of Adolescent Health, 34, 237–
243.

Wagner, E.F., Myers, M.G., & McIninch, J.L. (1999). Stress-coping and temptation-
coping as predictors of adolescent substance use. Addictive Behaviors. Special
Issue: Addictions in special populations, 24, 6, 769-779.

Wang, G.J., Volkow, N.D., Fowler, J.S., Cervany, P., Hitzemann, R.J., Pappas, N.R., et
al. (1999). Regional brain metabolic activation during craving elicited by recall
of previous drug experiences. Life Sciences, 64, 775-784.

Wechsler, D. (1997). Wechsler Adult Intelligence Scale—3rd Edition (WAIS-3®). San
Antonio, TX: The Psychological Corporation.

Wechsler, D. (1945/1997) Wechsler Memory Scale®--3rd Edition (WMS-III). San
Antonio, TX: The Psychological Corporation.

88

Welt, L. (1888). Uber charaktervaranderungen des menschen infoldge von lasionen des
stirnhirns. Deutsch Archives of Klinical Medicine, 42, 339–390.

Whitlow, C.T., Liguori, A., Livengood, L.B., Hart, S.L., Mussat-Whitlow, B.J., Lamborn,
C.M., et al. (2004). Long-term heavy marijuana users make costly decisions on a
gambling task. Drug and Alcohol Dependence, 76, 1, 107-111.

Whitney, P., Jameson, T., & Hinson, J.M. (2004). Impulsiveness and executive control
of working memory. Personality and Individual Differences, 37, 2, 417-428

Williams, J.M.G., & Broadbent, K. (1986). Autobiographical memories in suicide
attempters. Journal of Abnormal Psychology, 95, 144-149.

Williams, J.M.G. (2004). Experimental cognitive psychology and clinical practice:
Autobiographical memory as a paradigm case. In J. Yiend (Ed.), Cognition,
emotion, and psychopathology. Cambridge, UK: Cambridge University Press.

Williams, J.M.G., Chan, S., Crane, C., Barnhofer, T., Eade, J., & Healy, H. (2006).
Retrieval of autobiographical memories: The mechanisms and consequences of
truncated search. Cognition and Emotion, 20, 3-4, 351-382.

Wood, D., Valdez, B., Hayashi, T., & Shen, A. (1990). Homeless and housed families in
Los Angeles: A study comparing demographic, economic, and family function
characteristics. American Journal of Public Health, 80, 1049–1052.

Yip, J.T.H. & Lee, T.M.C. (2005). Effect of ecstasy use on neuropsychological function:
A study in Hong Kong. Psychopharmacology, 179, 3, 620-628.

Yoder, K.A., Hoyt, D.R., & Whitbeck, L.B. (1998). Suicidal behavior among homeless
and runaway adolescents. Journal of Youth and Adolescence, 27, 6, 753-771.

Yoder, K.A. (1999). Comparing suicide attempters, suicide ideators, and nonsuicidal
homeless and runaway adolescents. Suicide and Life-Threatening Behavior, 29,
1, 25-36.

89

Appendix
Positive and Negative Autobiographical Memory
Complete data was available for 53 of 54 participants. One individual said she
did not want to think about the past and refused to complete the measure.
Iowa Gambling Task
Complete data was available for 44 of 54 participants. The Iowa Gambling Task
was added to the current study after the first 9 participants were run, and 1 individual
experienced a software malfunction during completion of the task.
Ways of Coping Questionnaire—Problem Focused
Complete data was available for 51 of 54 participants. The subscale total could
not be calculated for 3 respondents due to skipped items on the questionnaire.
Ways of Coping Questionnaire—Distancing/Avoidance
Complete data was available for 52 of 54 participants. The subscale total could
not be calculated for 2 respondents due to skipped items on the questionnaire.
Ways of Coping Questionnaire—Denial
Complete data was available for 53 of 54 participants. The subscale total could
not be calculated for 1 respondent due to skipped items on the questionnaire.
Suicide Behaviors Questionnaire—Suicide Attempts
Complete data was available for all participants.
Suicide Behaviors Questionnaire—Suicide Ideation
Complete data was available for all participants.

90

Suicide Behaviors Questionnaire—Lifetime Suicidality
Complete data was available for all participants.

Abstract (if available)

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Discussion during treatment decision-making predicts emotional adjustment in prostate cancer patients

PDF

Behabioral and neural evidence of state-like variance in intertemporal decisions

PDF

The role of emotions: gender differences in different displayed emotions and witness credibility

PDF

The interpersonal effect of emotion in decision-making and social dilemmas

PDF

The link between the expression of negative emotion and future well-being: the role of disclosure modality in an East Asian American population

PDF

Principal implementation of Education Code 215: pupil and student suicide prevention policies in southern California public middle schools

PDF

The acute impact of glucose and sucralose on food decisions and brain responses to visual food cues

PDF

The influences of anxiety, coping, and social support on physical functioning among heart failure patients

PDF

Insula activity during safe-sex decision-making in sexually risky men suggests negative urgency and fear of rejection drives risky sexual behavior

PDF

Health-related quality of life correlates of the treatment decision making process of newly diagnosed prostate cancer patients

PDF

Family environment as a moderator of the association between theory of mind and social functioning in people with schizophrenia

PDF

Understanding emotional regulation and mood of young adults in the context of homelessness using geographic ecological momentary assessment

PDF

Minority stress and intersectionality: how are identity, discrimination, and biological stress related to depression and suicidality among multiple minority subgroups?

PDF

A neuropsychological exploration of low-SES adolescents’ life goals and their motives

PDF

Heart, brain, and breath: studies on the neuromodulation of interoceptive systems

PDF

Adolescent conduct problems and substance use: an examination of the risk pathway across the transition to high school

PDF

The neuroscience of ambivalent and ambiguous feelings

Asset Metadata

Creator Fine, Emily B. (author)

Core Title Deconstructing the psychological components of emotional decision making and their relation to the suicide continuum

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Psychology

Publication Date 04/26/2009

Defense Date 12/11/2008

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag autobiographical memory,coping,homeless population,Iowa Gambling Task,neuropsychology,OAI-PMH Harvest,suicidality

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Bechara, Antoine (committee chair), Prescott, Carol A. (committee chair), Huey, Stanley (committee member), Manis, Franklin R. (committee member), Parker-Dominguez, Tyan (committee member)

Creator Email efine@usc.edu,stanfordcardinal13@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m2113

Unique identifier UC1105529

Identifier etd-Fine-2682 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-223062 (legacy record id),usctheses-m2113 (legacy record id)

Legacy Identifier etd-Fine-2682.pdf

Dmrecord 223062

Document Type Dissertation

Rights Fine, Emily B.

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu