University of Southern California Dissertations and Theses

Towards designing mental health interventions: integrating interpersonal communication and technology

(USC Thesis Other)

Towards designing mental health interventions: integrating interpersonal communication and technology

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

Towards Designing Mental Health Interventions:
Integrating Interpersonal Communication and Technology

by
Liyuan Wang

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(COMMUNICATION)
December 2021

Copyright 2021 Liyuan Wang

ii

Dedication
To my family and Lynn Carol Miller.

iii

Acknowledgements
Foremost, my gratitude goes to my advisor Dr. Lynn Carol Miller. Without her
unwavering care, support, and encouragement, I would not accomplish this dissertation and
complete my Ph.D. I cannot recall, in the past five years, how many days and nights Lynn and I
spent together discussing, revising, and refining each of my research projects. I also cannot
enumerate, on this one-pager acknowledgment, her advice and inspiration that have helped me
embark on a career in academia. I still remember the first time I met her and started talking about
the meaning of research. She asked: "what about changing the world by taking small steps?" And
that is what we have been doing together and will continue to do so.
I am grateful to my committee members. I am thankful to Dr. Valente for inspiring me to
pursue a line of research at the intersection of health and network science. I can never forget
when Dr. Valente walked into my qualification defense in person despite the summer
thunderstorm and the fact that he was supposed to join the defense remotely. I am also thankful
for Dr. Lindsay Young for her support and collaboration that helped me get started on a line of
research that focuses on social networks and computation methods. Moreover, I will also never
forget Dr. Adam Leventhal, whose support and expertise in clinical psychology helped me
navigate through the most challenging scenario I had to face at USC.
The completion of this dissertation also has roots in support from family members,
friends., and co-authors. I am fortunate to have my dad and mom: Guojun Wang and Yuhua
Song, who unconditionally support me on every decision I have made in pursuing a career in
academia. I also thank my friends who make this five-year journey adventurous: Dr. Yiqi Li,
Steffie Kim, Nan YuanFeixue, Hyemin Kim, Dr. Yusi Xu, and Donna Kim. I am especially

iv

grateful to Dr. Nathan Walter, a dear friend and a collaborator who always supports and inspires.
I also want to acknowledge my co-authors in the past five years here: Dr. Sheila Murphy, Dr.
John Christensen, Dr. Ben Smith, Dr. Steve Read, Dr. David Jeong, Dr. Carlos Godoy, Dr. Paul
Appleby, Mingxuan Liu, Dr. Traci Gillig, Yunwen Wang, and Lipei Tang, who have taught me
the meaning of being a good collaborator. Indeed, it takes a village for me to accomplish my
academic goals. I will continue my career by carrying with me the goodwill and support of
faculty and staff members: Sarah Holterman, Anne Marie Campian, Calvin Cao, and Dr. Hector
Amaya, Dr. Peter Monge, Dr. Ken Sereno, and Dr. Carmen Lee, who safeguarded every step I
took at Annenberg. Now, as my academic journey continues, I can't wait to work with Dr. Leslie
Clark and Mona Desai, who laid the strongest safety net for me to pursue my academic career
during the precarious time of COVID-19.
At times, we also owe it to those who intentionally or unintentionally block our way. At
least, I owe it to those voices like "she cannot not make it to a PhD," "her research is not
interesting," and "she cannot do that as a communication researcher." Those voices have urged
me to go this far and to continue proving them wrong.

v

Table of Contents
Dedication ............................................................................................................................... ii
Acknowledgements ................................................................................................................ iii
List of Tables ........................................................................................................................ vii
List of Figures ...................................................................................................................... viii
Abstract ................................................................................................................................. ix
CHAPTER ONE: INTRODUCTION ..................................................................................... 1
CHAPTER TWO: THE VICIOUS CIRCLE OF RUMINATION AND NEGATIVE
COMMUNICATION—A Mediation Model ............................................................. 5
CHAPTER THREE: HOPE FROM ENABLING TECHNOLOGIES: EFFICACY OF
JUST-IN-TIME, ADAPTIVE INTERVENTIONS TO REDUCE RUMINATION27
CHAPTER FOUR: INTERATIVE NARRATIVE JITAI .................................................... 38
References ............................................................................................................................. 55
Appendices A: Tables ........................................................................................................... 75
Table 1. Chapter Two: Characteristics of Participant..............................................................75
Table 2. Chapter Two: Differences in depression outcomes by clinical diagnosis .......................77
Table3. Chapter Two: Correlation Tables of Variables of Interests (Spearman’s rho) ..................78
Table 4. Chapter Two: Direct and Indirect Effects of Parallel Mediation Analysis on Depression
Mitigation ........................................................................................................................79
Table 5. Chapter Two: Direct and Indirect Effects of Parallel Mediation Analysis on Depression
Exacerbation ....................................................................................................................80
Table 6. Chapter Three: JITIAI and WLC-control Studies Included (k=9) .................................81
Table7. Chapter Three: JITIAI and non-JITAI Studies Included (k=21) ...................................82
Table 8. Chapter Three: Studies eligible to extract pre/post changes (k=13) ..............................84
Table 9. Chapter Three: Forest Plots of JITAI studies by outcomes of interests ..........................86
Table 10. Chapter 4. Participants’ Information ......................................................................87
Table 11. Chapter 4. Participants’ Information The count of rumination episodes and aver age
duration (in minutes) of each ruminative episode in week 1-2.5 (time 1), and week 2.5- (time2) of
the intervention .................................................................................................................88
Table 12. Within-individual fit indices in GIMME (WLC and JITAI comparison) ......................89
Table 13 Sample Path Analysis within Individual 1 (WLC condition).......................................91

vi

Table 14 Sample Path Analysis within Individual 9 (JITAI condition) ......................................92
Appendices B: Figures ......................................................................................................... 93
Figure 1. Chapter Two: Interpersonal Communication Mediates Trait Rumination and Depression
Outcomes.........................................................................................................................93
Figure 2. Chapter Two: Statistical Models for Analysis in Chapter Two....................................94
Figure 3. Chapter Two: Results from model testing ...............................................................97
Figure 4. Chapter Two: Interaction Effects of depression diagnosis and trait rumination on
perceiving hurtfulness ........................................................................................................98
Figure 5. Chapter Two: Interaction Effects of depression diagnosis and trait rumination on social
support ............................................................................................................................99
Figure 6. Chapter Two: Interaction Effects of depression diagnosis and trait rumination on
depression exacerbation.................................................................................................... 100
Figure 7. Chapter Two: Interaction Effects of depression diagnosis and trait rumination on
depression mitigation ....................................................................................................... 101
Figure 8. Chapter Three: PRISMA chart for paper inclusion ................................................. 102
Figure 10. between group effects (WLC vs. JITAI).............................................................. 103
Figure 11. contemporaneous individual effects within the control condition ............................ 104
Figure 12. contemporaneous individual effects within JITAI conditions.................................. 105
Appendices C: JITAI Materials for Project Three ............................................................. 106
Week 1 ......................................................................................................................... 106
Week 2 ......................................................................................................................... 108
Week 3 ......................................................................................................................... 111
Appendices D: JITAI Materials for Project Three ............................................................. 114
Week 1 ......................................................................................................................... 114
Week 2 ......................................................................................................................... 116
Week 3 ......................................................................................................................... 120

vii

List of Tables

Table 1. Chapter Two: Characteristics of Participant..............................................................75
Table 2. Chapter Two: Differences in depression outcomes by clinical diagnosis .......................77
Table3. Chapter Two: Correlation Tables of Variables of Interests (Spearman’s rho) ..................78
Table 4. Chapter Two: Direct and Indirect Effects of Parallel Mediation Analysis on Depression
Mitigation ........................................................................................................................79
Table 5. Chapter Two: Direct and Indirect Effects of Parallel Mediation Analysis on Depression
Exacerbation ....................................................................................................................80
Table 6. Chapter Three: JITIAI and WLC-control Studies Included (k=9) .................................81
Table7. Chapter Three: JITIAI and non-JITAI Studies Included (k=21) ...................................82
Table 8. Chapter Three: Studies eligible to extract pre/post changes (k=13) ..............................84
Table 9. Chapter Three: Forest Plots of JITAI studies by outcomes of interests ..........................86
Table 10. Chapter 4. Participants’ Information ......................................................................87
Table 11. Chapter 4. Participants’ Information The count of rumination episodes and aver age
duration (in minutes) of each ruminative episode in week 1-2.5 (time 1), and week 2.5- (time2) of
the intervention .................................................................................................................88
Table 12. Within-individual fit indices in GIMME (WLC and JITAI comparison) ......................89
Table 13 Sample Path Analysis within Individual 1 (WLC condition).......................................91
Table 14 Sample Path Analysis within Individual 9 (JITAI condition) ......................................92

viii

List of Figures
Figure 1. Chapter Two: Interpersonal Communication Mediates Trait Rumination and Depression
Outcomes.........................................................................................................................93
Figure 2. Chapter Two: Statistical Models for Analysis in Chapter Two....................................94
Figure 3. Chapter Two: Results from model testing ...............................................................97
Figure 4. Chapter Two: Interaction Effects of depression diagnosis and trait rumination on
perceiving hurtfulness ........................................................................................................98
Figure 5. Chapter Two: Interaction Effects of depression diagnosis and trait rumination on social
support ............................................................................................................................99
Figure 6. Chapter Two: Interaction Effects of depression diagnosis and trait rumination on
depression exacerbation.................................................................................................... 100
Figure 7. Chapter Two: Interaction Effects of depression diagnosis and trait rumination on
depression mitigation ....................................................................................................... 101
Figure 8. Chapter Three: PRISMA chart for paper inclusion ................................................. 102
Figure 10. between group effects (WLC vs. JITAI).............................................................. 103
Figure 11. contemporaneous individual effects within the control condition ............................ 104
Figure 12. contemporaneous individual effects within JITAI conditions.................................. 105

ix

Abstract
Chapter One and Two introduces the vicious cycle of rumination, negative communication, and
depression. Drawing on literature from both interpersonal communication and depression, this
chapter provides an overview of current knowledge regarding rumination, how it is linked to
depression, and how interpersonal factors (e.g., perceived hurtfulness; perceived support) may
serve as mediators in the persistent rumination to depression link affording potential targets for
multiple intervention components. After delineating the link between rumination and depression,
there are three main aspects of this chapter, first, it further explicates the role of rumination in the
context of interpersonal events. Specifically, there is a focus on interpersonal communication
(e.g., the perceived level of hurtfulness from conflicts and the role of perceived support from
confidant networks). And, it discusses the role of depression diagnosis as a moderator of
communication patterns and their mediating role from rumination to depression outcomes.
Second, this project proposes a theoretical model, which tests the mediation effects of
interpersonal factors, and the moderated mediation effects of depression diagnosis in depression
mitigation and exacerbation. Finally, it delineates the importance of reducing rumination in
reducing interpersonal stress, communication problems, and depression.
Chapter Three. This chapter looks at how technology-enriched interventions can inform
the reduction and reframing of interpersonal rumination. To do so, this chapter reports on the
results of a meta-analysis (k=33) that focuses on the efficacy of JITAI. Specifically, this report
takes a close look at JITAIs with behaviors of interest in mental health and interpersonal
competence (k=6). As results have shown, JITAI could be a powerful intervention design to
improve mental well-being. Moreover, this report also provides a detailed discussion regarding

x

tailoring methods to inform technologically enhanced interventions. It also seeks to answer
questions regarding what should be tailored, the frequency of tailoring, and the kinds of feedback
(messages) that can be provided.
Chapter Four. Chapter four continues to provide a critical review of JITAI. Here, we
specifically examine the possibilities of using JITAI in reducing interpersonal rumination and
depression using it as a pilot for a randomized clinical trial. This chapter first focuses on some
of the drawbacks of JITAI especially regarding participants’ burden and boredom and how
narratives can help reduce those complications. After all, JITAI have to rely more on self-reports
provided by participants multiple times throughout the intervention period to provide tailored
feedback. This could be very taxing and burdensome for participants (Nahum-Shani, Hekler, &
Spruijt-Metz, 2015). Drawing from theories of entertainment education (Moyer-Gusé, 2008;
Slater & Rouner, 2002), narrative persuasion (Cohen, 2001; Green & Brock, 2000), and
interactive narratives (Green & Jenkins, 2014), this work argues that interactive narrative
elements may be the key to reduce participants’ burden and boredom for more efficacious JITAI.
Chapter Five. This chapter will be a short summary of the three dissertation projects.

1

CHAPTER ONE: INTRODUCTION
Depression is one of today’s most prevalent chronic adverse health outcomes (Santini et
al., 2015). Recent reports from CDC have shown that, in the US, a staggering 8.1% percent of
persons aged 20 and over had been living with depression for at least two weeks (CDC, 2016).
Also, the US has 6 -7% of full-time workers who are experiencing major depressive disorders
(MDD, Birnbaum et al., 2010). Depression is very debilitating and costly. The burden that comes
along with depression is not just psychological, it is associated with numerous long-term
disabilities (e.g., ischemic heart disease, injuries, inflammation, low immunity): This pattern is
found globally (Ferrari et al., 2013; Prince et al., 2007). Various meta-analyses have also
documented the link between depression, suicide ideation, suicide attempts, and even death
(Dong et al., 2018; Ribeiro et al., 2016, 2018; Y. Wang et al., 2008). Notably, 15% of clinically
depressed individuals died by suicide (Santini et al., 2015). Furthermore, the economic impact of
MDD by 2010 is estimated to be $210 billion per year, with a financial burden increase of 21.5%
since 2005 (P. E. Greenberg et al., 2015).
Interpersonal rumination (Cloven & Roloff, 1993; Roloff, Reznik, Miller, & Johnson,
2015), defined as recurrently thinking about previous unpleasant interpersonal events, is often a
major contributor to depression (Lyubomirsky & Nolen-Hoeksema, 1995; Nolen-Hoeksema et
al., 2008). While various theories (e.g., of interpersonal functioning; depression) concur that
ruminating about negative experiences can lead to a further deterioration in communication
behaviors and depression symptoms (Christopoulos et al., 2017; Cloven & Roloff, 1993; Joiner,
2000; Joiner & Metalsky, 1995) very few empirical studies (Witvliet, Knoll, Hinman, &
DeYoung, 2010) have found ways to reduce rumination. Here, in this dissertation, the goal is to

2

better understand interpersonal rumination and potential targets for reducing and reframing
rumination. Therefore, this dissertation consists of three projects, aiming to: (1) identify
interpersonal factors (e.g., perceived hurtfulness; perceived support) that contribute to (and may
mediate) the persistent rumination to depression link affording potential targets for multiple
intervention components, (2) evaluate innovative means to deliver interventions in reducing the
type of negative rumination that results from difficult interpersonal encounters, and (3) further
explore interpersonal factors that can reduce negative rumination.
Project one evaluated: (1) the extent to which unpleasant interpersonal events (perceived
hurtfulness) affect depression outcomes, (2) the role of interpersonal relationships (perceived
support) in reducing depression outcomes and (3) the role of perceived hurtfulness and support
as interpersonal mediators of the link between persistent (trait) rumination and depression
outcomes (see Figures 1, 2a, 2b, and 2c) using a mediational approach in a cross-section data set
of clinically depressed (or not) individuals with depression diagnosis as moderator (W). Project
two, using a meta-analytic approach, examines the efficacy of JITAI as a potential means (or
delivery mechanism for a rumination intervention) to provide in-the-moment support that is
tailored to individual participants. The studies in that meta-analysis include those that examine
the effectiveness of technological enriched interventions in delivering interventions targeting
mental health outcomes. Therefore, this affords a “first look” at the feasibility of using JITAI as
a delivery mechanism in targeting the link between negative interpersonal experiences and
rumination to reduce depression outcomes. This meta-analysis (Project 2) made it apparent that
although effective, JITAI needed to incorporate more persuasive techniques to engage
participants in these daily, often burdensome interventions. Project three, a registered pilot

3

randomized controlled trial (RCT), builds on the two earlier projects, with the goal of leveraging
knowledge of interpersonal factors in depression (and depression mitigation) and the viability of
mobile technologies in assessing and intervening in in-the-moment rumination of unpleasant
interpersonal encounters. Here, to enhance participant engagement in JITAI, the narrative
persuasion theory and research literature is reviewed. Of special interest is the work involving
interactive narratives that shows promise of persuading and engaging participants in JITAI
interventions aimed at reducing rumination. In addition, in this third project, there is a further
systematic review of the relevant literature in interpersonal communication and psychology with
the goal of specifying potential causes of interpersonal rumination, potential effective
intervention components (i.e., Cognitive Behavioral Therapy (CBT, REF), and the links
(including via rumination) to depression. Drawing on previous evidence regarding the
effectiveness of narrative approaches in presenting intervention components in exerting
perspective taking, this third project examines the feasibility and efficacy (i.e., initial promise of
this approach) using a randomized controlled trial, to compare participants rumination reduction
over time in an interactive narrative JITAI condition (that incorporates CBT components),
compared to control arms. The RCT clinical pilot trial conducted here in the dissertation
provides a formative foundation for additional research to apply for NIH grant funding to assess
the intervention’s effectiveness (e.g., in a larger national study) in a future RCT in reducing
interpersonal rumination with a larger sample size (for adequate power), and at scale. Lastly,
this third dissertation pilot project also aims to explore additional intervention components in
order to provide further means to reduce rumination. Leveraging knowledge gleaned from

4

project 1, the third project also used a social network approach that examines the link between
social network support structures and reduction of depressive symptoms of participants.

5

CHAPTER TWO: THE VICIOUS CIRCLE OF RUMINATION AND NEGATIVE
COMMUNICATION—A Mediation Model
1
“In the United States, there are over 17.3 million adults (7.1% of the adult population)
who have experienced a major depressive episode in the past 12 months(SAMHSA, 2018),
making depression one of today’s most prevalent chronic adverse health outcomes (Santini et al.,
2015). Using almost any metric, depression is both debilitating and costly. More than a
psychological burden, depression has numerous long-term health consequences, including heart
disease, injuries, inflammation, and lowered immunity functioning, and has been linked to
suicide and suicide ideation(Dong et al., 2018; Ribeiro et al., 2018) . Further, the total economic
burden resulting from disability, morbidity, and mortality of depression is estimated to be $210
billion per year(M. T. Greenberg et al., 2001).”
Indeed, there is a strong need for researchers to examine further the psychological
mechanism that contributes to depressive outcomes. To do so, in this chapter, the first project
draws from interpersonal theories that can account for the link between rumination and
depression outcomes. This project offers two sets of results.
Rumination and its link to depression
A key intrapersonal factor contributing to depression is one’s ruminative responses (i.e.
trait rumination) towards their immediate social interactions(Lyubomirsky & Nolen-Hoeksema,
1995; Nolen-Hoeksema et al., 2008). From time to time, people may find themselves

1 A p ro p o rtio n o f th is p ro ject h as b een u n der rev iew. Po s sib le repetitio n s are m ark ed in q u otatio n m arks .

6

subconsciously revisit and recurrently thinking about some unpleased interpersonal events. For
example, after an interpersonal dispute, one may have ongoing thoughts like “Why did this
happen to me”, “What was that person thinking of me,” and “Why couldn’t I handle this better. ”
Excessive ruminative process could lead to a “dangerous” path: it deviates one’s general sense-
making activities by persistently fixating one’s thought s on negative affects through a biased
attribution of responsibilities regarding the unpleasant dispute (Camper et al., 1988). Indeed,
according to the overarching definition of rumination, or depressive rumination response, those
thinking styles are defined as “the tendency to passively and repetitively focus on the experience
of negative moods, as well as their causes and consequences (Flynn et al., 2010, p. 456).”
Rumination is different from general worrying/anxiousness, even though both can
involve negative moods and recurrent thinking. Watkins and colleagues (2013) used a non-
clinical sample and compared how differently participants rate their worries and ruminations.
Their results suggested that rumination is different from worry in at least two ways. First,
rumination is mostly about recurrently thinking about events that happened in the past. This is
not the case with worrying: people can and always worry about future events. Second,
rumination has a strong basis in problems that happened. That is, while one can vicariously
worry about unreal things, one usually ruminates and deliberates upon what has already
happened in real life. Rumination is also different from obsession. Obsessions are “intrusive,
repetitive thoughts, images or impulses that are unacceptable and/or unwanted and give rise to
subjective resistance...the necessary and sufficient conditions … are intrusiveness, internal
attribution, unwontedness and difficulty to control (Rachman, 1998, p. 251).” Indeed, obsessive
thoughts have been more irrational and less realistic than ruminative thoughts(Wahl et al., 2011).

7

The link between ruminative thinking and co-occurring social impairment on one’s
psychological well-being has long been associated with the onset of depression. Theories that
look at onset of depression (Michl et al., 2013; Nolen-Hoeksema, 1991; Nolen-Hoeksema et al.,
2008) from both the cognitive and interpersonal stress perspectives (Joiner, 2000; Joiner &
Metalsky, 1995) have argued that rumination over unpleasant experiences is one major reason
for the increase of interpersonal conflicts and the lack of relationship satisfaction. Studying the
population of individuals who are vulnerable to depression, researchers have found that those
people are more likely to reflect upon and dwell on their negative affective experiences (Flynn et
al., 2010). Subsequently, they are more likely to experience emotional and psychological
maladjustment in their daily social interactions(Nolen-Hoeksema, 1991). Within various studies
of depression, the aftermath of depressive rumination is usually documented from an
intrapersonal perspective. For example, rumination has been found to increase individuals’
biased interpretation of unpleasant events (Lyubomirsky & Nolen-Hoeksema, 1995), and usually
elicits negative autobiographical memories (Lyubomirsky et al., 1998). Nolen-Hoeksema and
colleagues (1991) also postulate that ruminative thinking exacerbates and prolongs depression.
Indeed, this view was supported in a series of meta-analyses (Johnson & Whisman, 2013;
Olatunji et al., 2013; Rood et al., 2009). Indeed, here the first hypothesis draws on this
established link: (H
1
) there is a significant link between individual ruminative responses and
subsequent depressive symptoms.

8

In What Way Can Interpersonal Communication Hurt?—Perceived Level of Hurtfulness
from Conflicts
Expecting a link between rumination and depression in clinical studies, researchers have
examined the complications regarding rumination and how interpersonal relationship qualities
and experiences may feed into rumination: Indeed, there is an extensive literature examining
serial arguments and how they are linked to rumination (Bevan & Sparks, 2014; Cloven &
Roloff, 1991, 1993; Roloff et al., 2015). Also referred to as “mulling” by interpersonal scholars,
ruminations after a conflict, or interpersonal rumination, was studied as “prolonged thinking or
mulling about a problem (Cloven & Roloff, 1991, p. 136). ” Literature from serial argument can
provide an in-depth look into the attribution biases that explain the rumination and relationship
quality link. That is, rumination over previous conflicts is essentially a sense-making activity that
involves determining the causes of the disagreement and the role each party played in the
previous encounter (Cloven & Roloff, 1991; Showers, 1988). Indeed, after a disagreement,
people mull about who is responsible for the conflict. Attributions of one’s motives and actions
are easily distorted by the hedonic relevance of these attributions for a given interactant: offenses
and negative affect felt during the conflict can be easily attributed to the (other) actor (Cloven &
Roloff, 1991; Jones & Davis, 1965). Therefore, in mulling over a conflict, one usually ends up
ascribing blame to one’s interaction partner rather than to oneself (Cloven & Roloff, 1993, 1993;
Newman & Langer, 1981). In such cases, prolonged rumination over who was responsible
predominately focuses on assigning blame to the other (rather than self), which can only further
exacerbate the conflict.

9

Similarly, theories in depression and interpersonal relationships also posit a similar cycle
of increasing rumination and subsequent interpersonal impairment (Joiner, 2000). Ruminative
responses towards unpleasant events usually fixate one’s attention on dysphoric affect and
emotions. Therefore, rumination can serve as the “cognitive motor” that drives the occurrence of
problematic interpersonal behaviors: In addition to increased blaming of the other, rumination
may also make one focus more on one’s own inadequacy such that one starts to seek negative
feedback or ask for excessive reassurance about who they are or about the relationship. Both
seeking negative feedback and excessive reassurance have been perceived as part of the
hopelessness and pessimistic communication patterns that are likely to elicit further interpersonal
rejection and conflict (Joiner & Lonigan, 2000; Joiner & Metalsky, 1995). Indeed, as previous
research has shown, ruminators who fixate on previous disagreements were found to be less
effective in solving hypothetic interpersonal problems presented to them in a lab setting
(Lyubomirsky & Nolen-Hoeksema, 1995). Similarly, following up with a non-clinical sample for
9 months, Flynn and colleagues (2010) also found that depressive rumination significantly
predicts participants’ interpersonal stress.
A major consequence is that this elevated level of hurtfulness can make one withdraw
from one’s daily activities (Dimidjian et al., 2006; Kasch et al., 2002). Indeed, individuals,
motivated by fear of experiencing more hurtful events induced by their interpersonal
ruminations, may engage in a series avoidance behaviors to protect themselves from exposure to
any interpersonal activities: This pattern is part of what has been called a depression avoidant
system (Dimidjian et al., 2006; Kasch et al., 2002). Behaviors motivated by a depression
avoidance system can include consciously avoiding social interactions and activities for fear of

10

experiencing negative outcomes. Even though social interactions and activities could help one
gain important rewards like social support and belongingness, avoidance motivations can play a
significant role by inhibiting such behaviors. Indeed, in avoiding elevated feeling of hurtfulness
from interpersonal rumination, individuals may engage in maladaptive behaviors and
problematic problem-solving behaviors, which in turn may be serve to further maintain or
exacerbate depression (Pettit & Joiner, 2006). Therefore, hypothesis 2 is proposed:
(H
2
) Hurtful experiences maybe one interpersonal factor that mediates the relationship between
trait rumination and depression avoidance behaviors.
How Can Interpersonal Communication Heal? Perceived Support from Confidants
“Although for clinicians, depression is typically defined as a persistent intra-psychic
disorder in which individuals feel sad and hopeless and derive little pleasure or have little
interest in activities, it is often influenced by the social contexts in which an individual is
embedded (Bergman & Haley, 2009; Durkheim, 1951). Personal networks — i.e., the set of
social relationships immediately surrounding a focal individual — receive considerable attention
in this regard, most notably for their provision of support that protects individuals from
becoming depressed or helps them cope more effectively with its symptoms (Lin et al., 1986)..
Confidants are those who have strong social relationships (i.e., confidants) to with whom one
can talk about important matters (Hall et al., 2020; Litwin & Stoeckel, 2014; Marsden, 1987,
2011). The companionship and support offered by confidants are integral to an individual’s
ability to avoid and/or attenuate feelings of distress and depression. While no studies have
specifically looked at the extent to which a confidant can improve depression, literatures
focusing on social support has shown that, with a long history, the amount of social supports

11

from confidants can be a key to depression mitigation(Gariépy et al., 2016; Rueger et al., 2016) .
Social support, in this study, defined as the extent to which one perceives that friends, family
members, and others provide various types of psychological and other support when needed,
consistently predicts physical and mental health outcomes (Uchino et al., 2018). Indeed,
researchers have consistently found that a have a large network of confidants who offers social
support social support can usually make one more resilient towards daily stressors, therefore
protecting one from depression. Indeed, the amount and quality of perceived social support that a
individual receives is, in part, derivative of the arrangement of social ties around them (Berkman
et al., 2000). Accordingly, studies examining the structural characteristics of personal support
networks in relation to psychological well-being and depression have traditionally crystalized
around the role of personal network size. The amount of personal confidants that a person can
potentially communicate with and receive support from(Hall et al., 2020), is perhaps the most
fundamental and intuitive structural indicator of how much social support an individual can
potentially receive (Lin et al., 1999). The essential take away from this body of work is that
fewer confidants and fewer close relationships are routinely linked to depressive symptoms
(Barnett & Gotlib, 1988; Kawachi & Berkman, 2001).
Several mechanisms by which the number of confidants influences psychological health
have been posited. In one formulation, the degree of an individual’s social connectedness
produces positive psychological states like a sense of purpose, belonging, and security, as well as
feelings of self-worth and self-competence. These positive states of mind, in turn, may benefit
downstream mental health through an increased motivation to engage in self-care (e.g., engaging

12

in activities that offer personal benefit) or by increasing an individual’s cognitive capability to
handle stress (Cohen & Wills, 1985; Kawachi & Berkman, 2001) .
Structural theories of social capital (Lin et al., 1999) suggest another route through which
the number of confidants offering social support positively impacts psychological well-being
— i.e., through the resources and opportunities that supporters provide. From this perspective,
the size of one’s support (i.e. number of confidants) is indicative of the amount of resources one
has available to prevent and/or cope with depression and the adverse life events that can cause or
exacerbate depression downstream (Lin et al., 1999), for example, resources in the form of
informal mental health care and advice. Further, individuals with larger support networks may
have more opportunities to engage in beneficial social activities that mitigate or distract from
feelings of depression. Taken together, we expect a positive relationship between the size of an
individual’s confidant network and their engagement in depression mitigation behaviors.
One’s rumination patterns and their perception of social support could impact depression
mitigation as well. While social support can also buffer negative affects when one engages in
negative rumination (Puterman et al., 2010), individual differences in depressive rumination
contribute to one’s perceptions of social support. On one hand, ind ividuals with a strong
tendency to repeatedly reflect and engage in their negative experiences may be motivated to seek
more social support (Flynn et al., 2010; Nolen-Hoeksema & Davis, 1999) . Indeed, a number of
studies have shown that for those high in rumination, their perceived social support usually
attenuates the unpleasant emotional experiences induced by their rumination (Gerteis &
Schwerdtfeger, 2016; Lee, 2019; Puterman et al., 2010). On the other hand, not being able to
secure the level of social support one needs to confront one’s negative rumination, one could

13

experience elevated levels of depression and state ruminative responses (Flynn et al., 2010;
Nolen-Hoeksema & Davis, 1999; Pettit & Joiner, 2006). That is, when perceived social support
is low, subsequent ruminative and depressive symptoms could also be elevated (Flynn et al.,
2010). This leads to the third hypothesis.
(H
3
) Perceived social support may be one protective factor that mediates the relationship
between trait rumination and depression.”
Can Depression Diagnosis Distinguish Patterns of Interpersonal Communication?
While extant theories explain the links between interpersonal communication and
depression, very few of them provide a causal prediction. Indeed, it is tough to provide evidence
of such associations. Is it experiencing an elevated level of interpersonal conflict that leads to
rumination and depression. No study so far has examined the differences in rumination patterns
and perceptions of interpersonal conflicts between clinically depressed and non-depressed
samples. Therefore, we ask the following research question.
RQ
1
: What are the differences in rumination patterns and depressive symptoms between
clinically depressed and non-depressed samples?
Other questions include:
RQ
2:
Is the persistent (trait) rumination to depression link mediated by the two interpersonal
processes (i.e., perceived hurtfulness; perceived support): We consider them concurrently in
parallel mediation, control each for the other.
RQ
3
: Are these links from persistent (trait) rumination to depression through interpersonal
processes moderated by depression status?

14

Figure 1 is the theoretical model of the current project, although the goal is to examine in
separate analyses the mediated link from rumination to each of the depression outcomes.
[Figure 1]
Methods
Study Design and Sample
“Data were collected in August 2019 and in September 2019 as part of an online survey
study designed to investigate the relationship between depression and features of personal
networks. Participants were recruited via a volunteering website managed by National Institute
of Health for clinical trials: Researchmatch.org. This website (ResearchMatch.org), with a
national web-based registry of more than 150,000 patients looking for new treatment of their
disease/condition, allows researchers to access volunteers with existing health conditions who
are considering participating in research studies or clinical trials (Harris et al., 2012, 2012; Pulley
et al., 2018). ResearchMatch.org was to cater the public desire to find recruiting research trials
for their conditions and the need of researchers to translate their new therapies and treatment.
This website offers a registry that is freely available for anyone to use and to enter their goals for
trial participation, medical condition, and geographical information. ResearchMatch.org could be
an ideal site for our research because depression and anxiety were the top conditions that were
searched by its volunteers (Pulley et al., 2018). By the year of 2020, Researchmatch.org
documented 150, 364 registered volunteers, 870 projects, and 491 academic publications(for
details, see Researchmatch.org.). Candidate participants were eligible if they were (1) 18 years of
age or older, and self-reported either 2) a diagnosis of clinical depression or 2) no diagnosis of
any mental health disorder. Participants’ depression diagnosis was screened both by

15

Researchmatch.org and their subsequent self-reports upon their agreement to participate in our
study. That is, participants needed to establish their own profile on Researchmatch.org in order
to be contacted by researchers. In this study, we use the filter provided by the website and send
inquiries to participants regarding their willingness to participate in our study. We contacted
those participants (who reported to have been diagnosed with depression or not) through
Researchmatch.org.
After initial contact was made, we further screened to see if the participant’s diagnosis
was current. Those who (1) did not specify a date of diagnosis, or (2) did not specify which
depression disorders they had were excluded. Power analysis showed that to detect a small level
of difference, at least 63 people per group was needed. Our sampling procedures resulted in a
sample of 1002 adults (620 non-depressed and 382 clinically depressed) with no missing data,
ensuring sufficient power to detect a difference. Upon completion of the online survey,
participants were entered into a lottery to win a $50 gift card. The study was approved by the
Institutional Review Board of authors’ research institution, and all participants provided consent.
Table 1 shows detailed information of participants.
[Table1]
Primary Measures
Depression Diagnosis. Participants who identified with a diagnosis of depression
received a code for diagnosis of 1, those who do not have a depression diagnosis received a code
of 0.
Rumination Response Scale. Individual differences in their responses towards their
depressive feelings (i.e. rumination patterns) were measured by the Rumination Response Scale

16

(Nolen-Hoeksema, 1991). RRS consists 22 items that measure a dimension of ruminative
patterns (depression, brooding, and reflection) and provide a trait-like measure of persistent
rumination responses. In the current study, participants rate on a 1 to 7 scale (1=not at all like
me, 7=extremely like me) items like, “I think about how alone I feel,” or “ what am I doing to
deserve this?” This scale showed excellent reliability, Cronbach’s α =0.95.
Perceived Conflict and Hurtfulness. Participants’ perception towards their recent
conflict was measured by the counts of their recent hurtful events in the past month. Participants
were asked to report on whether they have hurtful feelings (1=yes, 0=no) on a list of
interpersonal conflicts (e.g. argument with relational partner, friends, or family members,
disagreement with co-workers, etc.). Higher scores indicates that participants perceived higher
level of hurtfulness from their interpersonal conflicts.
Depression Outcomes. The degree to which participants engaged in depression
mitigation activities was measured using items from the Behavioral Activation for Depression
Scale (BADS, (Kanter et al., 2006; Manos et al., 2011). The BADS scale consists of 9 items that
represent weekly behavior thought to underlie depression. Each item on the scale is measured on
a seven-point scale ranging from 1 (not at all) to 7 (completely). This scale has been widely used
to (1) track depressed patients’ engagement in behaviors that may improve/exacerbate their
depressive symptoms , (2) monitor their responsiveness in terms of Depression Behavioral
Activation Therapies, and (3) to evaluate their overall depressive symptoms in the past
week(Kanter et al., 2007; Manos et al., 2011).
The scale consists of two dimensions. The first dimension encompasses f our items
(Cronbach’s α =0.85) that measure avoidance of rumination behaviors- a set of behaviors

17

indicating depression exacerbation outcomes (e.g., “I kept trying to think of ways to solve a
problem but never tried any of the solutions” or “I spent a long time thinking over and over about
my problems”). The other dimension (Cronbach’s α =0.85) has five items that measure activation
behaviors- a set of behaviors indicating depression mitigation behaviors, which are activities
that promote positive personal benefits (e.g., “I did things that are enjoyable” or “I was an active
person and accomplished the goals I set out to do.”).
Generating Personal Confidant Networks and Measure of Level of Perceived
Support. A central component of the online survey assessment was the enumeration and
description of amount of confidants from which participants can draw social supports. To this
end, we used a name generator — i.e., an instrument used to identify the individuals (or alters)
who are in a focal individual’s (or ego’s) personal network (Marsden, 1987). Name generators
are the workhorses of personal network data collection and have been used to capture a variety of
personal networks including social support networks (Cornwell et al., 2008; McPherson et al.,
2006; Wellman & Wortley, 1989) and core discussion networks (Marsden, 1987; McPherson et
al., 2006) The confidant name generator elicited up to six confidants in the past 3 months
(including a romantic partner if one was reported by the participant) using the prompt, “Please
name the people to whom you talk about your interpersonal problems (e.g., unpleasant social
encounters, upsetting co-workers, relationship trouble, etc.).” Generally, a nomination of 5 -6 is
the recommended “balance” to avoid the redundancy of having to name too many nominations
(over reporting those who are not confidants) and the failure to capture the actual peer social
network structure by naming too few links(Michell, 2000). Therefore, the name generator

18

provides the total amount of confidants from who participants can draw support as an indicator
of perceived social supports.
Statistical Analysis
To address the first RQ
1
, a series of t-tests was used to assess differences between
clinically depressed and non-depressed individuals in terms of the structure of their depression
outcomes, perceived hurtfulness, and perceived social support. Then, the PROCESS macro in
SPSS was used to estimate the proposed theoretical model (Hayes, 2013; Hayes, 2018). We used
Model 8 for estimation (see the legend on Figure 2a). Model 8 can simultaneously evaluate the
mediating role of interpersonal factors (including more than one mediator at a time in parallel)
and concurrently the moderating role of depression diagnosis. Specifically, trait rumination was
entered as the independent variable (X), the two mediators were entered as mediators (M
i
), and
depression exacerbation/mitigation behaviors were entered as dependent variables (Y).
Depression diagnosis was the moderator (W). A parallel mediator model was planned and used
here to tease out the independent role of each of the mediators in mediating the link between
persistent (trait) rumination and each of the depression outcomes (that is, exacerbation and
mitigation outcomes were analyzed in separate PROCESS models). 2 Furthermore, age, gender,
income, and marital status were entered as covariates. However, none of these covariates were
significant, therefore, they were not retained when reporting mediation results. All network

2 Parallel mediation models (possible with Model 8 in PROCESS) afford the examination of a mediator while holding constant
the other mediator(s). Ordinarily, with multiple mediators in the model (in parallel), researchers assess correlations among
mediators to avoid multicollinearity (Hayes, 2013, p. 157), but the correlation between these variables is virtually zero (see Table
3).

19

measures were computed via the Statnet and igraph packages in R. All other descriptive and
statistical analyses were performed using SPSS v. 26 and the PROCESS macro v3.0 (Hayes,
2013) .
[Figure 2]
Results
Descriptive Results.
A summary of sociodemographic characteristics of the sample, stratified by depression
status, is shown in Table 1. Results show that depressed and non-depressed participants were
comparable in demographics, with majorities in both subsamples being White/Non-Hispanic,
earning less than $60,000 per year, and female. Participants in both samples were on average
37.4 years of age (SD=12.79). There are also no significant differences between the two samples
regarding age, years of education, and race/ethnicity. Given the gender bias in our samples (82%
of depressed participants and 79% of non-depressed participants were female) and known
associations between gender and depression, we performed a series of t-tests (not shown) to
compare differences in DVs scores between females and male. While, there were no significant
differences by gender on any of the DVs, interpretation of our findings should be made with this
caveat in mind.

[INSERT TABLE 1]
Differences in Depression Outcomes
Table 2 provides the raw scores of each variable of interest separately for the clinically
depressed and non-clinically depressed samples. Results indicate that in general, depressed

20

participants reported less social support ( number of confidants on average), M = 4.16, SD =
1.72) than non-depressed participants (M = 4.20, SD = 1.63). However, this difference is not
statistically significant (t = -.33, p > 0.50). The current sample indicates an intriguing trend with
depressed participants self-reporting more positive depression-related outcomes. While, there
were also no significant differences (t = -1.22, p > 0.50) in depression mitigation behaviors,
depressed participants self-reported more of such behavior (M = 4.36, SD = 1.34) than non-
depressed (M = 4.26, SD = 1.38) participants. And, depressed participants also self-reported
significantly (t=0.59, p<.05) less depression exacerbation behavior (M = 3.77, SD = 0.38) than
non-depressed participants (M = 3.83, SD = 1.44). However, it is worth pointing out that
depressed participants self-reported significantly higher trait rumination scores (depressed
participants, M = 4.51, SD = 1.12; non-depressed participants, M = 3.72, SD = 1.47 ) and
experienced a higher level of hurtful experiences (depressed participants, M = 9.10, SD = 2.00;
non-depressed participants, M = 2.63, SD = 2.15). Table 3 also shows the zero-order correlations
among the variables of interest.
[TABLE 2]
Mediation Model Results
Results of the serial mediation models (separately for each depression outcome) are
shown in Tables 4 and 5. And Figure 3 is a legend showing the direction and signs of the
association. Keep in mind that all scores in the current study have been recoded such that high
scores indicate high levels of each measurement (i.e. trait rumination, perceived social support,
perceived hurtfulness, depression exacerbation, and depression mitigation). Depression diagnosis
was dichotomously coded with the higher value (1 versus 0) indicating a depression diagnosis.

21

And a positive sign indicates a positive relationship and a negative sign indicate a negatively
relationship.
Because the pattern of these findings for the variables of interest (i.e., role of diagnosis,
rumination, and their interaction) was similar across the two depression dependent variables,
these results for separate depression outcomes are presented together. These variables of interest
are all significantly predictive of depression outcomes (see Tables 4 and 5). First (see Table 4),
with a depression outcome of depression mitigation (indicated with a Y in Table 4), persistent
(trait-like) rumination (X to Y) ( 𝑩 = -.28., SE=.04, 95% CI:[-.36, -.21]), diagnosis (W to Y) ( 𝑩
= -3.37, SE=.37, 95% CI:[-4.09, -2.69]), and their interaction (W*X to Y) ( 𝑩 = .82, SE=.07,
95% CI:[.71, 1.00]) were all significant predictors. Similarly (see Table 5), with depression
exacerbation as the outcome, persistent (trait-like) rumination (X to Y) ( 𝑩 = -.29., SE=.04, 95%
CI:[.36, .21]), diagnosis (W) ( 𝑩 = -3.38., SE=.36, 95% CI:[4.10, 2.67]), and their interaction
(W*X) ( 𝑩 = -.76, SE=.07, 95% CI:[.43, .22]) are significant predictors. Below, these
interactions will be discussed in more detail.
Across the two different depression outcomes, the values from each of the variables of
interest to a given mediator (i.e., perceived hurtfulness (Table 4 or Table 5, X to M
i
) or social
support (Table 4 or Table 5, X to M
ii
)) will be the same. Let’s begin with the link from trait
rumination (X) to perceived hurtfulness (M
i
; X to M
i
): That value is significant ( 𝑩 =.42, SE=.06,
p<.01). Now, let’s examine the link from trait rumination (X) to perceived social support: That
value is statistically significant (B= -.10, SE=.05, p < .05). For Diagnosis (W) to perceived
hurtfulness (M
i
; W to M
i
) that value is statistically significant (B= 9.53, SE=.50, p < .01),
indicating that individuals diagnosed with depression perceived more hurtfulness. Diagnosis (W)

22

was not significantly predictive of social support (M
ii;
W
to
M
ii;
); (B= -.72, SE=.41, ns). The
interaction between trait rumination and diagnosis (X*W) in predicting to each mediator was
significant (B= -.76, SE=.11 p < .01 for perceived hurtfulness; B= .19, SE=.09 p < .05 for social
support); these interaction effects are discussed separately in more detail below.
Now let’s consider each of the mediators and their links to each of the depression
outcomes. For depression mitigation (Table 4), the social support, but not the perceived
hurtfulness mediator significantly predicted this depression indicator ( 𝑩 =.07, SE=.02, p<.05).
For depression exacerbation, the same pattern emerged with social support a significant predictor
of the depression outcome ( 𝑩 =.06, SE=.06, p<.05).
[Figure 3]
[Table 4 &5]
Interaction Effects
From Variables of Interest to Mediators. While trait rumination was not found to be a
significant predictor of perceived social support ( 𝑩 =-.72, SE=.41, p>.05), in examining
interaction effects with depression status (Table 4 and 5) a significant interaction (reported
above) was detected. In Figure 4, we explored that interaction. For depressed participants, trait
rumination was positively associated with perceived social support, В=.10, SE=.08, 95% CI (-
.06, .26) but this was not statistically significant. For non-depressed participants, trait rumination
negatively predicts perceived social support, В=-.09, SE=.05, 95% CI (-.18,.01), but by itself this
was also not significant. But, note that these B’s differ in direction, resulting in the significant
interaction.

23

Similarly, we also detected interaction effects of depression diagnosis on perceived
hurtfulness. For depressed participants, trait rumination was positively associated with perceived
hurtfulness, В=.41, SE=.06, 95% CI (.30,.53). For non-depressed participants, trait rumination
negatively predicted perceived hurtfulness, В=-.35, SE=.10, 95% CI (-.53,-.16).
From Variables of Interest to Depression Outcomes. The interaction effect of depression
diagnosis and trait rumination was also significant for both depression outcomes. For depression
exacerbation as the outcome, for depressed participants, trait rumination was positively
associated with depression exacerbation, В=.55, SE=.04, 95% CI (.48,.62) For non-depressed
participants, trait rumination negatively predicts depression exacerbation, В=-.34, SE=.06, 95%
CI (-.48,-.24). We also observed significant interaction effects on depression mitigation: For
depressed participants, trait rumination was negatively associated with depression mitigation,
В=-.27, SE=.04, 95% CI (-.35,-.19); for non-depressed participants, trait rumination negatively
predicts depression exacerbation, В=-.57, SE=.06, 95% CI (.45,.69). Figure 4 through 7
describes the interaction effects.
[Figure 4-7]

Discussion
This study investigated how a highly intrapersonal process, rumination, is associated with
interpersonal outcomes and how those interpersonal outcomes predict to depression outcomes.
To do so, we tested a series of mediation models, which suggest that interpersonal
communication factors such as perceived social support and hurtfulness contribute to the link
between trait rumination and depression outcomes. Our results were relatively straightforward.

24

High trait rumination is associated with experiencing more hurtful interpersonal events and
having fewer confidants from whom one could draw support. In the meantime, more hurtful
experiences significantly predict more depression exacerbation behaviors, but not more
depression mitigation behaviors. And more social support is associated with more depression
mitigation behavior, but not depression exacerbation behavior.
We also found that participants diagnosed with depression, in the current study, reported
an alleviated level of hurtfulness from their daily encounters and more depression exacerbation
behaviors than those who are not diagnosed with depression. And our analysis on interaction
effects shows that diagnosed participants’ hurtful experiences and relatively low levels of
perceived social support were both predicted by high trait rumination scores. Additionally, we
also find the intriguing reversing effects of trait rumination on those who are not depressed. That
is, for participants who are not depressed, a higher trait trumination score can indicate less
perceived hurtfulness and more social supports. This was also consistent with studies that
examines, at times, rumination could be a means of positive reflective thinking for people who
do not have mental health concerns (Afifi et al., 2013). However, as this is a cross-sectional
study that only collected data at one time point, we cannot differentiate between two alternatives
here: (1) Are individuals’ existing high trait rumination scores responsible for their diagnosis of
depression or (2) Are higher trait rumination scores the result of depression? And, depressed
individuals perceptions towards their depression may be affected by those rumination patterns.
Finally, it worth pointing out that consistent with previous studies, social support here
was found to be a protective factor against depression. Indeed, in our study, we used a network
measure that directly assesses one’s existing level of confidants as an indic ator of social support.

25

And our results have suggested those who nominated more confidants in their lives are usually
the ones who demonstrate more depression mitigation behaviors. As our results show, simply
having a confidant network was a significant predictor of one’s engagement in depression
mitigation behaviors.
It is also while pointing out that we did not find a statistically significant difference
between depressed and non-depressed individuals regarding network size in our sample.
However, this result is not inconsistent with prior work that has identified associations between
social isolation (lack of social connections) and depression(Franck et al., 2016). Also, from those
participants, we found network size played a significant role in mitigating depression, which is
consistent with previous studies that focus on the size of networks in protecting one against
depression (Child & Lawton, 2020; Santini et al., 2015). Our results on network size, therefore,
have positive implications: when in time of need, clinically depressed people may have a
sufficient confidant network upon which to rely, at least at the onset of their depression. And the
number of confidants one has in one’s social network matters in helping one seek behaviors that
can contribute to the improvement of one’s depression. Indeed, it is necessary for researchers to
look into ways that can help clinically depressed individuals in maintaining or expanding their
confidant networks for the sake of improving their depressive symptoms.
Limitations and Future Directions
“Even though this study has shown the importance of confidant networks in depression, it
has some limitations. First, because of the limits of the design, it was neither possible to collect
changes in members within one’s confidant networks over time nor possible to collect changes in
participants’ depression outcomes. Therefore, results of this study can only be interpreted as

26

associations rather than as providing evidence of a causal relationship between depression
mitigating behaviors and changes in one’s interpersonal relationships. In a subsequent study, it
would be even more meaningful to examine changes in network structures over time from onset
of depression to mitigation of depression-related behaviors and recovery. Second, because of the
voluntary nature of this study, most of our participants were female, therefore, the results of this
study may not generalize to the social networks of men who are clinically depressed (or not).
Subsequent analyses could take a further look at the role of gender by recruiting more male
participants. Moreover, because ResearchMatch.org operates on a voluntary basis, our sample
may be subjected to the same self-selection bias. There is a possibility that results of the study
cannot be generalized to those who are severely depressed or socially isolated. And this
population may require more rigorous observational studies than the self-reported methods we
have used in this study.”

27

CHAPTER THREE: HOPE FROM ENABLING TECHNOLOGIES: EFFICACY OF
JUST-IN-TIME, ADAPTIVE INTERVENTIONS TO REDUCE RUMINATION
While extant interventions have explored cognitive behavioral therapy like mindfulness
to reduce depression related symptoms (Zoogman et al., 2015), none of them have looked at the
possibility of reducing the rumination to interpersonal impairment link. Indeed, cognitive
behavioral therapies may not be able to capture the moment within which negative rumination
occurs, not to mention providing the in-the-moment means for participants to reduce ruminative
responses (Nahum-Shani et al., 2015). Chapter three therefore presents a meta-analysis regarding
the potential efficacy of an emerging type of intervention: just-in-time, adaptive interventions to
reduce interpersonal rumination.
A Just-in-time, adaptive approach to health behavior change
3

Just-in-time adaptive intervention (JITAI) is an emerging design of intervention brought
about by new mobile technologies (Nahum-Shani et al., 2016; Nahum-Shani, Hekler, & Spruijt-
Metz, 2015a; Spruijt-Metz et al., 2015). “A JITAI is an intervention designed to address the
dynamically changing needs of individuals via the provision of the type/amount of support
needed, at the right time, and only when needed (p.3. Nahuam-Shani et al., 2016)”. “In various

3
In preparing of the dissertation projects, a more detailed reports of this chapter has already been published on Health Communication, under the
name: “Wang, L., & Miller, L. C. (2019). Just -in-the-Moment Adaptive Interventions (JITAI): A Meta-Analytical Review. Health
Communication, 0(0), 1–14. https://doi.org/10.1080/10410236.2019.1652388 ’. Therefore, to avoid self-plagiarism, the current paper marked all
parts that are the same as the Wang & Miller (2019) piece with page numbers
”

28

fields, different labels have been applied to a variety of JITAI-like intervention approaches such
as dynamic tailoring (Kennedy et al., 2012; Krebs et al., 2010), Ecological Momentary
Interventions (EMI) (Heron, 2011; Heron & Smyth, 2010) , and intelligent real time therapy
(Kelly et al., 2012): Because all of these approaches share similar components we use the JITAI
umbrella term to refer to all of them (Nahum-Shani et al., 2016). Advocates of just-in-time
treatments (Canel, Rosen, & Anderson, 2000; Nahum-Shani et al., 2016; van Merrienboer,
Merrienboer, Kirschner, & Kester, 2003) have a long tradition in seeking to “produce the right
item, at the right time, in the right quantities (Canel et al., 2000, p. 52).” In the realm of health
interventions, the idea of just-in-time is rooted in capturing the exact time needs of individual
clients/intervention participants arise to provide just the right type of support. Therefore, JITAI
could be provided when a smoker has experiencing day-long craving (e.g., Free et al., 2013;
Naughton, 2017), or at the exact moment a drinker at risk for driving while intoxicated walks
into a bar (e.g., Gustafson et al., 2011).
JITAI has been receiving increasing attention because of its potential to generate more
effective interventions through targeting at the exact moment of users’ needs and to provide just -
in-time messages or information (Fry & Neff, 2009; Nahum-Shani et al., 2016, 2016; Spruijt-
Metz et al., 2015). JITAI takes advantages of advancements in smartphones, mobile apps, and
various sensors to track the conditions of the targeted participants, such that through self -reports
and sensor technology, participants’ behaviors are monitored, and the intervention can be
adaptive. In JITAI, specific types of supports for behavior change, be they specific skills to cope
with current distress (e.g., Riordan, Conner, Flett, & Scarf, 2015), or emotional support to assist
users in adhering to their exercise routines (e.g., Brookie, Mainvil, Carr, Vissers, & Conner,

29

2017), are individualized based on ongoing situational changes: The goal is to maintain users’
engagement and minimize the burden of excessive self-reports. Indeed, JITAI is also gaining
attention from funding agencies: A number of JITAI interventions (Free et al., 2011; Gustafson
et al., 2014) have been found by the National Institutes of Health, and the National Science
Foundation.
The rapid unprecedented technological advances in today’s mobile devices afford suites
of tools for user assessments: Given the technical complexities involved in designing JITAI, it is
perhaps not surprising that the majority of research to date has been focusing on evaluating the
applicability and feasibility of JITAI (e.g. Achterkamp, Cabrita, Akker, Hermens, &
Vollenbroek-Hutten, 2013; Consolvo et al., 2008; Consolvo, Everitt, Smith, & Landay, 2006;
Shrier & Spalding, 2017). While those studies have shown us the promise of this technology for
use in JITAIs, we still have not addressed a fundamental question: How effective are JITAIs in
producing desired changes regarding health behaviors? Especially, can JITAI be more
efficacious in creating changes in behaviors that are nonetheless hard to change? Especially,
given the dearth of interventions to change rumination patterns, we ask can JITAI be an effective
means to deliver rumination-reduction interventions to improve mental health?
There is also an extensive literature on non-JITAI tailored interventions: These are
similar to JITAI in that tailored interventions have components regarding using individualized
information for targeted participants (Kreuter & Wray, 2003; Lustria et al., 2013). In the meta-
analysis on tailored interventions delivered via the internet, Lustria and colleagues (2013) found
stronger efficacy of tailored health interventions (1999 through 2009) compared to the non-
treatment control condition. Tailoring, in the non-JITAI literature, consists of two types: Static

30

and Dynamic tailoring. Static tailoring refers to using one-time baseline assessments as the basis
for tailoring rather than using one’s in -the-moment assessments (Krebs, Prochaska, & Rossi,
2010). Dynamic tailoring, being very similar to JITAI, is also rooted in the idea of providing
message interventions based on active assessment of one’s in -the-moment reports (Kennedy et
al., 2012; Krebs et al., 2010). Krebs and colleagues (2010), in their meta-analysis, further
compared interventions using dynamic tailoring and those that used static tailoring from 1988 to
2009. Results of their study demonstrated that long-term effects of dynamic tailoring
interventions can persist much longer than those statically tailored, therefore suggesting the
potentials of JITAI.
Evaluations on Ecological Momentary Interventions (EMI) can also give us indirect
information on the potential efficacy of JITAI. EMI is a type of intervention that is built on EMA
(Ecological Momentary Assessment). While EMA has a focus of using different instruments to
track individual behaviors (Shiffman, Stone, & Hufford, 2008; Shiffman & Waters, 2004; Stone
& Shiffman, 1994), it was not used for interventions to target changes in behaviors. EMI, similar
to JITAI, makes use of those real-time assessments to provide adequate feedback for intervention
purposes (Heron & Smyth, 2010). Indeed, a systematic review conducted by Heron and Smyth
(Heron & Smyth, 2010) examined EMI studies conducted before 2009, their results documented
the efficacy of EMI in various fields such as weight control, anxiety regulation, and diabetes
management.
However, JITAIs are not just delivered via the internet, as in the case of Lustria and
colleagues’ study (2013). Instead, JITAIs rely more on the portability of mobile devices to assess
one’s in-the-moment states and more rigorously tailored information. While dynamic tailoring

31

and EMI are analogous to the tenants behind JITAIs (Nahum-Shani et al., 2016; Sharmin et al.,
2015)given the advancements and efficacy of JITAIs over the past decade since the Heron and
Smyth (2010) and Kreb et al (2010) reviews, it is not clear whether newer interventions are also
similarly – or even more -- effective. Indeed, ten years of technological advancement in mobile
devices and theoretical development in creating effective interventions call for a systematic
examination regarding the efficacy of recent JITAIs. Therefore, this study is to focus on JITAIs
conducted in the past decade and to offer a meta-analytical evaluation regarding their efficacy in
producing desired health outcomes. In the following section, JITAI are effective compared to
controls, and there is significant between study variance to explain, we intent to examine a
number of potential moderators of this effect including: (a) study participant average age and (b)
health outcomes examined. Therefore, we ask,
RQ1: How effective are JITAIs in achieving health outcomes?
RQ2: Does JITAI demonstrate stronger efficacy on one age group than others?
RQ3: Does JITAI demonstrate stronger efficacy on achieving certain health outcomes
than others? (p. 1531, Wang & Miller, 2019)”
Methods
Search strategy and study selection
“In February 2018, we did a series of key word searches on a variety of databases
including the Web of Science, PubMed, Psych Info, and ProQuest. A combination of search
terms was used: JITAI, just-in-the-moment, adaptive, EMI, ecological momentary intervention,
intelligent real-time therapy, dynamic tailoring intervention, and health. We also used asterisks
on some search terms such as “health*” to expand the search scope such that key words that are

32

related to “healthy behaviors” or “health interventions” were also included. (Note that asterisk
automatically truncates the search terms such that instead of directly searching for the word
“healthy,” search terms like “health*” the search can find all other key terms that are related to
health). Excluding those studies conducted before 2008, a total number of 289 entries were
retrieved.
There were three major inclusion criteria for the meta-analysis. (a) All studies should be
empirical research studies with at least one specific health outcome. (b) Each study should have
at least one JITAI condition. Criteria on judging an intervention as an JITAI was based on
former reviews by Nahuam-Shani and colleagues (2016). For example, one inclusion criterion
was dynamic tailoring. If a study only entailed tailoring based on the baseline goals/personalities
of participants, as in the case of static tailoring, it was not included. (c) All studies should use
some form of technology (e.g., phone, tablet, wearables, sensors, pc, etc.). Additional criteria
included whether studies retrieved had appropriate quantitative information such as post-
intervention scores. Using the email addresses offered in potentially eligible studies, we
contacted relevant authors to inquire if they had or knew of unpublished research meeting these
criteria. We also emailed authors if further information was needed for computing study effect
sizes. Using the criteria above, Tables 6, 7, and 8 show the 35 studies found to be eligible and
that were therefore included in subsequent analyses. Among those studies, k=3 were unpublished
dissertations, k=33 were peer-reviewed journal articles. Figure 8 is the PRISMA chart for study
inclusion. (p. 1532, Wang & Miller, 2019)”
[Table 6 through 8]
[Figure 8]

33

Coding Strategies for Moderators
“In addition to coding basic information of each study such as publication date and
authorship, two major moderators were coded. First, we aimed to evaluate whether JITAI
interventions would be more effective for health outcomes in one domain (e.g., mental health)
versus other domains (e.g., controlling diet, improving level of exercise). Therefore, each study
was first coded regarding the outcomes towards which it was targeted. For example, a number of
studies were coded as “mental health”, focusing on managing problematic emotional reactions
(Moody, Tegge, Poe, Koffarnus, & Bickel, 2017; Proudfoot et al., 2013). Other studies have
interests in improving diet and increasing exercise levels, thus they were coded as involving the
domain of increasing healthy habits (Atienza, King, Oliveira, Ahn, & Gardner, 2008; Brookie et
al., 2017). In general, we found studies of four health outcomes: mental health, healthy habits,
addiction (including smoking, alcohol reduction, and substance use), and medication adherence
towards chronic diseases (e.g., pain management, diabetic self -management).
We aim to understand whether JITAI is particularly effective for some age groups versus
others. Specifically, we want to know how JITAI affects the elderly since JITAI relies on a
certain level of technology literacy. We also want to know whether JITAI has stronger effects
among those who are young compared to those who are older. As such, we coded studies in
terms of their participants’ mean age: 0 to 25 as 1, 26 to 49 as 2, 50 and above as 3.
Effect Size Extraction and Determination
Effect sizes were retrieved via Comprehensive Meta-analysis version 2 (Borenstein,
Hedges, Higgins, & Rothstein, 2005). We employ Hedges g as the effect size (ES) indicator.
Defined by Borestein and colleagues (2005), Hedges g is a derivation of the standardized mean

34

differences (Cohen’s d) effect sizes that is rigorous towards testing pooled variance component s.
Hedges g of 1 indicates that the effect is one standard deviation above the mean. Indeed,
interpretation of Hedge g is also similar to Cohen’s d, Hedges g =0.8 which indicates a large
effect size, 0.5 indicates a medium effect, and 0.2 indicates a small effect. It also uses a factor (J)
to correct the underestimation of the population standard deviation, as in eq.1,
𝑔 = 𝑑 ∗ 𝐽 (eq.1)
Where
(eq.2)

And
(eq.3)

Two sets of Hedges g were extracted, among the intervention and control group studies,
we extracted Hedges g using the immediate post-intervention differences between the
intervention and the control group. Among the one-arm intervention or assessment studies, we
extracted Hedges g by using the differences between the pre- and post-intervention scores. We
extracted the effect sizes using immediate post-intervention health outcome scores. Some studies
also include multiple conditions (e.g. JITAI treatment, non-JITAI treatment, and no treatment
control). To avoid comparing apples with oranges, we only used the differences between JITAI

35

treatment and the no treatment control condition to extract effect sizes. We also encountered
studies that have compares two JITAI treatments.
Each effect size was weighted regarding its inversive variance weight (Borenstein et al.,
2005). Following random effects procedures using weighted factorial models across
interventions and control groups, effects sizes were evaluated, which is recommended for smaller
sample sizes and for treatment-control studies (Borenstein et al., 2005; Lipsey & Wilson, 2000).
To detect heterogeneity, variance component was tested with the Q test for the existence of
additional variances that can be explained by the existence of moderators (Borenstein et al.,
2005; Lipsey & Wilson, 2000). We also conducted a number of tests to assess potential file
drawer bias in the sample of studies we included in this analysis. (p. 1536, Wang & Miller,
2019)””
Results: Can JITAI Be the Key to Improve Mental Health?
Main results regarding the efficacy in the meta-analysis is already reported elsewhere
(Wang & Miller, 2019). In short, JITAI was found to be a highly efficacious intervention design:
“there was moderate to large effect sizes of JITAI treatments over (1) waitlist -control conditions
(k=9), Hedges’s g=1.65 and (2) non -JITAI treatments (k=21), g=0.89. Also, in the current study,
this meta-analysis found participants of JITAI interventions showed significant changes (k=13)
in the positive direction (g=0.79).” This chapter also found that interventions focusing on mental
health (e.g. emotion intensity, depression, and PTSD) showed highest improvements on
participants (k=6), g=0.080, I
2
=67.35, which is significantly higher efficacy than interneurons
that targets at addiction (g=0.83), diet (g=0.74), and weight management (g=0.30). Therefore,

36

results of the meta-analytical analysis did show that JITAI could be a more efficacy intervention
design for mental health problems.
[Table 9]
However, because there were not sufficient JITAI studies focusing on mental health
related interventions, no further moderators were tested. A qualitative review reveals the
feasibility, efficacy, but also some drawbacks of JITAI to improve mental health: First, all six
studies are pilot tests that focused on whether JITAI is a feasible and efficacious means to
provide additional support for mental health related coping and skills to improve on mood,
catastrophizing thoughts, emotion intensity, etc. Three studies (Ahmedani et al., 2015;
Kristjánsdóttir et al., 2013; Rizvi et al., 2011) evaluated JITAI as a complement to human-
counselor based cognitive behavior therapies. All three studies suggested that JITAI can be a
more effective intervention than human-counselor-only therapies. Three other studies (Burns et
al., 2011; Miner et al., 2016; Possemato et al., 2016) focuses on whether mobile technologies
would be sufficient to provide necessary therapies without intensive participants of a human-
counselor. Two studies (Miner et al., 2016; Possemato et al., 2016), using two different samples,
studied the same mobile application: PTSD coach. As their results have shown, a JITAI based
mobile app can have the potential for PTSD patient to manage and improve upon their
symptoms. Another study by Burns and colleagues (2011) focused on creating machine learning
models that can predict in-the-moment emotional vulnerabilities of major depressive disorder
participants to provide tailored feedback. All six studies significantly improved their major
behaviors of interests. And all of them have also shown high acceptability among participants.

37

Therefore, it is possible to develop a saleable JITAI intervention that can target ruminative
responses towards interpersonal conflicts.
However, a closer look at the six studies on mental health also speaks to one major
concern regarding designing JITAI interventions: the excessive cognitive burdens of multiple
self-reports. All studies require participants to self-rate their mood, major depressive events, and
even to keep diaries from 3 to 5 times per day: this could be very taxing and boring, leading to
cognitive overload and early withdrawals from a JITAI (Sharmin et al., 2015). Indeed, the study
conducted by Burns and colleagues (2011), while taking advantages of non-invasive
technologies, must rely on one’s self -reports of their social and emotional contexts five times per
day. In the foregoing chapter, I propose a possible means to reduce such cognitive burdens.

38

CHAPTER FOUR: INTERATIVE NARRATIVE JITAI
A Call for Novel Mental Health Interventions During the Time of COVID-19
The need for mental health interventions is increasing: The global COVID-19 outbreak
has profoundly increased a range of psychological outcomes such as uncertainty stress, anxiety,
frustration, and depression (Luo et al., 2020; Serafini et al., 2020; Webb, n.d.). Psychiatrists, at
the onset of the outbreak, noted that there was alarming potential for long-term mental health
illness attributable from COVID-19 related trauma (fear of infection, anxiety during quarantine,
loss of family members or friends, unemployment, infected/risk of infections) (Ho et al., 2020;
Liu et al., 2020; Pfefferbaum & North, 2020; Zhai & Du, 2020). Dating back to April 2020, 30.8
% of Americans have reported psychological distress, with only 8.2% reporting similar
conditions at the same time in 2019. The number was even more staggering for those who are
aged from 18-29, who reported the highest increase (29%) of distress since 2019 (3% for this
group in 2019) (CDC, 2020; McGinty et al., 2020). As the outbreak progressed into June 2020,
we saw another increase to 40.9% who reported adverse mental or behavioral health conditions
(Czeisler, 2020). Past experiences have also shown a strong need for mental healthcare services
during an international epidemic outbreak. Dating back to the severe acute respiratory syndrome
epidemic in 2003, epicenters of outbreaks saw an 30% increase in suicide rates in those aged 65
and older, and the prevalence of stress among the general population, recovered patients, and
health-care workers (Nickell, 2004; Yip et al., 2010).
Cognitive Behavioral Therapy (CBT) has long been one of the chief face-to-face
interactive interventions that is promising in this domain. NICE, 2004, 2009) Patients entered
for CBT learn means to monitor and improve their psychological symptoms through a series of

39

tasks like diaries, emotion regulation, and mindfulness (Reavell et al., 2018). CBT offers a wide
range of strategies that help patients maintain, trigger, and mitigate their depressive symptoms
(Hofmann et al., 2012; Reavell et al., 2018). A number of meta-analytical reviews have shown
CBT is of strong efficacy in mitigating depression outcomes and recommend it as the “gold
standard” of the frontline treatment for depression (Hofmann et al., 2012, 2012; Horsch et al.,
2017; Karyotaki et al., 2017; Sijbrandij et al., 2016).
With new outbreaks projected well into 2021 and an economic crisis bound to persist,
elevated mental health risks are sure to continue, which can further complicate matters for the
mental healthcare system and the delivery CBTs (Haleem et al., 2020; Serafini et al., 2020).
First, direct encounters with patients for the sake of screening and monitoring have been reduced
to the minimum due to various social distancing and stay-at-home measures. As the virus ebbs
and flows, this can make face-to-face clinical treatment like CBT very difficult to consistently
provide and when immediate consultations are needed most. It can also delay immediate
consultations for those who are experiencing severe symptoms like suicidal ideation. Second, a
growing number of people has been fallen victims of the economic recession and large-scale
unemployment during the pandemic, they may be denied access to regular healthcare or ha ve a
hard time getting their referrals for formal evaluation and care. Lastly, with the significant
increase in those who need care during the time of the pandemic, available mental health services
could be highly strained at the time when they are the most needed. Therefore, to compensate for
the already-strained mental healthcare services at the time of the COVID-19 and future national
epidemics, there is a strong need for novel means that are accessible at scale for the prevention
and treatment of adverse mental and behavioral conditions.

40

One key challenge is to prevent the mental healthcare systems from being overwhelmed in
order to “flatten the curve” of the sudden peak s of need for mental health services during the
outbreak and new waves of the virus (e.g., as new, more dangerous variants take hold). Increases
in the need for mental health services have been associated with the prolonged COVID-19
outbreak and new virus waves. Constant uncertainties and elevated loads of stressors gradually
undermine the capacity to restore a healthy mental state, increasing the need for mental
healthcare. Indeed, psychological resilience (Chmitorz et al., 2018; Kalisch & Robins, 2006) -
outcomes indicating mental health is maintained or regained despite exposure to significant
stress and adversity- is especially vulnerable under extremely stressful circumstances like the
constant fear of infection and death. Another key challenge is the pervasive, acute need for
psychiatric services: It far exceeds our mental health services clinical capacity. Highly scalable
and efficacious mental health interventions are urgently needed.
JITAI as Means to Target Mental Health: a Pilot Randomized Control Trial
The meta-analytic review suggests that peoples’ everyday technologies (e.g., mobile
phones, wearable devices) not requiring human intervention, are not only scalable but provide
efficacious mental health treatment alternatives(Chandrashekar, 2018). Meta-analytical reviews
(Chmitorz et al., 2018; Joyce et al., 2018) suggest CBT is a “gold standard treatment” in
bolstering/maintaining psychological resilience and mobile phone-delivered CBTs (MCBT)
through apps are an efficacious scalable alternative to in-person CBT treatments(Chandrashekar,
2018). Specifically, none of the current MCBT programs take into account the just-in-time,
adaptive intervention (JITAI) features (i.e., algorithms), which can detect vulnerabilities in the
exact moment of need (Sharmin et al., 2015), a within-person approach affording intensive

41

measurement (i.e., 100+ measures per variable/person) is needed. To achieve this, these apps use
Ecological Momentary Assessments (EMA) where such moment-to-moment EMA assessments
in real time have been successfully applied in numerous health science domains (Heron &
Smyth, 2010). Researchers testing MCBT successfully, without human clinicians, have given
participants feedback based on their self-reported problems (L. Wang & Miller, 2019).Therefore,
the current project is designed to test the efficacy of a MCBT with JITAI features in reducing
ruminative thoughts- one major symptom of depression. And we hypothesize (H
1
) that a JITAI-
delivered intervention would be more effective than the wait-list control condition (WLC).
Engaging JITAIs
In designing and implementing JITAI, poor engagement and burden are always the major
hurdles to cross. Indeed, the degree to which participants are engaged in the intervention can be
reflected on their emotional, cognitive, and behavioral investment in the interventions (Sharmin
et al., 2015). JITAI sometimes may require a great amount of motivational and cognitive
investment. Participants may feel the burden in having to be engaged with the intervention
through diaries, self-reports, and other forms of assessment procedures that they have to go
through. As shown in previous chapters, even though there is relatively non-invasive sensor
technology and the highly efficient model buildings of machine learning, participants are still
prompted to report their current conditions five times per day (Burns et al., 2011). Multiple
suggestions have been made regarding keeping participants engaged (e.g. optimize motivational
towards receiving treatment, using different message prompts, use of video games, etc.). In this
study, we explore the potential of creating an interactive narrative JITAI to reduce user burden
and maximize engagement.

42

An interactive narrative is essential a form of storytelling that involves certain
interactive features such that audiences or readers of the story can have the feeling that they are
actively engaging in a certain form of interaction with the story characters (Green & Jenkins,
2014). An interactive narrative can be a simple linear story: at the crossroad of the forest, the
reader can turn to page 61 if he wants the hero to take the unpaved the road, or page 101 for a
flower embedded road (Green & Jenkins, 2014; Walter et al., 2017). An interactive narrative can
also be either entertaining or serious video games that allow players to choose-the-own-
adventures via their avatar (Peng et al., 2010). Recently, interactive narratives were also found to
take a trans-media form, that is, audiences can watch TV shows of their favorite characters,
meanwhile, they can play video games to interact with story characters through their own mobile
devices (Sangalang et al., 2013).
Interactive narratives can create a high level of engagement between the audiences and
the mediated platform. When designed for interventions or persuasive prompts, interactive
narratives can have the same potential to maintain participant engagement. Indeed, research have
shown that even simple linear story lines, with features of character characterization (e.g. choose
a piece of clothes for the character) can have the potential to allow audiences to view things from
the perspective of the character and therefore make more informed decisions in providing help
for people with whom they do not identify in the first place (Walter et al., 2017). Rooted in
theories of media richness (Sundar et al., 2013), interactive features ( e.g. character
customization, chat boxes, etc.) can increase immediate liking therefore allowing audiences to
fully explore the content offered by the mediated platform. Meanwhile, narratives, because of
their potential to bring the audiences to the story world, can have the potential to overcome

43

reactance therefore allowing audience to be more receptive towards different values and
opinions. It is proposed that while playing video games, navigating a likeable game world full of
interactive features, players can easily identify with their game character, assuming an other’s
identity, and act on behalf of the social expectations of their beloved avatars (Heffner et al.,
2015; Klimmt et al., 2010). Indeed, when embedded in serious games, interactive narratives have
been found to be more effective in eliciting desirable behaviors from participants (Peng et al.,
2010; Sangalang et al., 2013).
Therefore, in the current study, a more engaging JITAI design is proposed: an interactive
narrative JITAI, where the components of the intervention are derived from those used in
cognitive behavioral therapy that has been shown effective for depression (E. R. Watkins et al.,
2011). Appendix a shows the standard materials from a CBT that is used in the current study,
appendix b is the adapted narrative counterparts.

That is, participants of an interactive narrative JITAI are offered with storylines regarding
interpersonal encounters, and they get to take control of their characters to walk through different
settings, during which narrative stories were designed to help them reframe their relationships
and attribution of conflicts. Meanwhile, interactive features offered by JITAI could serve as non-
invasive means of assessment. It is hypothesized (H
2
) that a narrative JITAI-delivered
intervention would be more effective than the wait-list control condition (WLC).

44

Method
Study Design
The current study is a 35-day, randomized clinical control trial (RCT) of three conditions:
a WLC condition, a JITAI condition (using more conventional CBT materials), and a narrative
JITAI (NJITAI) condition (using CBT components). As the goal of the current project is to focus
on one major symptom of depression, which is excessive negative rumination, we adapted
materials from rumination-focused CBT (RFCBT), the efficacy of which has already been tested
in previous RCTs (E. Watkins, 2015; E. R. Watkins et al., 2011). RFCBT has activities, training
exercises, and diaries that are designed to reduce ruminative episodes for clinically depressed
patients (E. R. Watkins et al., 2011). Therapists can take advantage of diaries provided by
patients to offer tailored behavioral training (e.g., reframing emotions, attribution perceptions,
goal setting, and problem solving) for patients. The RFCBT program is ideal to be adapted to
MCBT with JITAI features: (1) The diary and rating systems with which therapists delivering
tailored training sessions could be conveniently designed into questionnaires for participants to
self-rate and document their conditions. (2) the JITAI system can have the capacity to deliver
automatic training sessions tailored to the results of individual self -reports. (3), participants can
have access to training materials and messages at the exact time of their requests. Therefore,
three training material sets (i.e., problem-solving, conflict attribution, and emotion regulation)
from RFCBT were adapted into the current JITAI condition. Further, an NJITAI condtion, which
is the narrative adaption of the JITAI, was also created based on the training materials in the
JITAI training sessions (for details of training materials, see Appendix A). Both JITAI and
NIJITAI condition ask participants to document their ruminative episode for seven days. After

45

that, they will receive intervention materials and training prompts for three weeks (21days). They
will also be asked to document their ruminative episodes for another week after accomplishing
the three-week training materials. A WLC doesn’t have any training materials from the RFCBT.
Participants are simply asked to document their ruminative episodes for 35 days.
Participants
Participants of the current study were recruited via a volunteering website managed by
National Institute of Health for clinical trials: Researchmatch.org (details of this website and the
screening procedures are described in chapter two). Participants were eligible if they were (1) 18
years of age or older, and self-reported 2) recently (twelve-month) given a diagnosis of clinical
depression, 3) and concurrently have no other diagnosis of any other mental health disorder.
Participants were also asked to indicate their medication and to specify their most recent visit to
a therapist. Those who had not had medication or had not visited a therapist in the past three
months were excluded. 450 participants were eligible and agreed to participate in the current
RCT.
Participants signed up and consented to participate before they were shown a link to the
intervention website. The design of the current RCT follows the necessary procedures that are
required in RFCBT. That is, to accomplish the RCT, participants are asked to self -report and
document their ruminative episodes every three hours (4-5 times a day). Each time, if a
participant documented their recent ruminative episode in the past three hours, they are prompted
to document the trigger, immediate environments, duration, and emotional experiences related to
the current ruminative episode. The intervention system was designed to tailor participant’'
ruminative patterns in three ways: (1) based on participants’ ruminative episodes (triggers,

46

immediate environments), participants would receive a set of training materials that best target
the trigger of their ruminative episode (e.g., problem-solving, interpersonal conflicts, etc.), (2)
participants can choose to review their training materials any time they need, as a means to
provide in-the-moment feedback, and (3) to skip and withdraw from answering some of their
documented prompts in they are, at the time, otherwise engaged in another activity. Therefore,
each participant can have up to 175 times within the 35-day of intervention to report their
ruminative episodes. Participants were encouraged, for the sake of the intervention outcome, to
document their ruminative episodes as many times as possible. They are also to be entered in a
lottery with a one in three odds of receiving a $50 gift card. A total of 150 participants enrolled
in the current RCT. However, only 18 of them (to date) have accomplished the current RCT
(i.e., finishing 80% of the questionnaires and training tasks). However, on average, those 18
participants have answered 145.76 out of 175 times of survey prompts, giving us sufficient
number of data points to conduct analysis on within-person changes. Table 10 shows the
demographic backgrounds and other information of those 18 participants.
[Table 10]
Measurements
In addition to a set of demographic questions, we also asked participants to document
their level of COVID-19 stress and total amount of standard drinks. To evaluate the ruminative
experiences, participants also indicate their responses on a battery of depression measurements.
Ruminative Response Scale (RRS). Rumination patterns for each individual (and
differences therein) towards their depressive feelings were measured with the Rumination
Response Scale (RRS; Nolen-Hoeksema, 1991). The RRS consists of 22 items for assessing

47

ruminative patterns, and dimensions of them (i.e., depression, brooding, and reflection): This is a
trait-like assessment of persistent rumination responses. In the current study, participants rate
each of these items on a 1 to 7 scale (1=not at all like me, 7=extremely like me) such as, “I think
about how alone I feel,” or “what am I doing to deserve this?” This measure exhibits excellent
reliability, Cronbach’s α =0.95.
Ruminative Episode and Duration. To evaluate participants’ experiences in
rumination, we asked two questions. First, in the past three hours, did you experience a
ruminative episode (by ruminative episode, we mean the experiencing a battery of negative
thoughts that distracted from your current situation)? Second, if the participant answered “yes”,
they were asked, how long did this ruminative episode last?
State Depression Scale (SDS). SDS is a 3-item scale for participants to rate their in-
the-moment level of depressed mood, anhedonia, and irritability on 7-point Likert scales ranging
from 1 (Not at all) to 7 (Very much) at the time of the alert. The SDS has already been tested and
used in previous EMA studies of rumination and depression (for a review, see Moberly &
Watkins, 2008). This scale was presented as a slider bar to fit with participants’ mobile -phone
screen size. Alphas on this scale has been very consistently in the range from 0.8 to 0.9 each time
it is used in research.
Emotional Experiences. Participants’ in-the-moment emotions at the time of sampling
were measured by a pictorial self-report by the name o Pick-A-Mood. This is essentially a 7-
point Likert Scales measured a range of feelings such as tense, annoyed, calm, and relaxed. Pick-
a-Mood was designed especially for mobile use (Desmet et al., 2016), which serves as a better
alternative than other emotion measurements. The eight emotional items were later combined

48

into a set of positive, neutral, and negative feelings. As in the case with SDS, alphas on this scale
have been very consistently from 0.8 to 0.9 each time it is used.

Statistical Analysis
As the goal of the current RCT involves evaluating changes in ruminative episodes
during the intervention, it follows a design for an EMA intervention. There is no concrete
standard to calculate the required power and sample size (Ruwaard et al., n.d.). However, within-
person differences require fewer participants but more EMA events within each participant. That
is, to track individual changes, each participant is required to document their changes in the
behavior of interests with sufficient time points (n=60~75). Moreover, it usually takes seven to10
participants within each comparison group to detect within-person differences (Beltz & Gates,
2017; Lane & Gates, 2017). Nine out of 18 participants who accomplish the study were in the
WLC, five JITAI condition, and four NJITAI condition. Therefore, due to the limited number of
participants, in the current study, we collapsed the JITAI and NJITAI conditions to evaluate the
efficacy of the JITAI delivered MCBT. Differences in ruminative outcomes are evaluated using
Group Interactive Multiple Model Estimation (GIMME). GIMME (Gates et al., 2017; Gates &
Molenaar, 2012) can reliably obtain the presence and direction of effects among variables
collected among multiple time points. That is, GIMME can both estimate associations among
variables measured simultaneously, and how variables measured at a particular time point predict
the effects and directions of relationships with variables collected in subsequent time points.
GIMME estimates those associations and directions through both Unified Structural Equation
Modelling (uSEM) and network analysis. That is, in estimating a GIMME model, the algorithm

49

first uses uSEM, a structural vector autoregressive model, to fit both contemporaneous and
lagged parameters. After fitting each individual model, GIMME also fits a group or sub-group
level model that looks at between group effect. In the meantime, GIMME analysis also produces
network connections for each participant and group (if there is a group-level effect). Network
graphs are produced by treating each variable of interest as a node, and the associations (both
contemporaneous and lagged) among each variables as edges (Beltz et al., 2018; Beltz & Gates,
2017).
Results
In answering our H
1
, which predicts stronger effects of JITAI conditions in reducing
depressive rumination, Table 10 shows each individual participant’s reported ruminative episode
(in the past three hours, have you experienced a negative thinking episode) and average time they
spent on each ruminative episode. The table shows the comparison between the differences in the
total reported ruminative episode and average rumination time during the first 2.5-week and the
second 2.5-week of the intervention. Because of the limited sample size (each group has an n of
9), we didn’t provide a test of differences at the between group level. However, it is worth
pointing out that participants in the JITAI condition uniformly showed a reduction in ruminative
episodes and average minutes spend ruminating during the second 2.5-week of the intervention.
However, because of limited number of participants in the current study, H
2
, which predicted
stronger effects of NJITAI over other conditions, was not tested.
[Table 10]
Because traditional t-tests could not provide rigorous estimation for RCT effects due to
sample size at the between group level, GIMME analysis was conducted to evaluate within

50

person changes, which provide robust results in explaining the effects of the current RCT. We
specified the analysis as a multiple group estimation that compares differences between the
JITAI condition and the control condition. Figures 10 and 11 provide the results of comparisons
both at the individual and at the group level. Table 11 is a set of fit indices that indicates the
fitness of each within individual changes using the uSEM approach. Our results have shown that
a great majority (15 out of 18, or 83%) of those models achieved excellent fit. The estimation of
group-level effects is set at 75%, meaning that group-level analysis will not be conducted unless
75% of participants have shown excellent fit. Therefore, in the current analysis, GIMME model
also estimates the association between each variable as beta-coefficients at each individual level
(if there is any), a sample of this set of beta-correlations could be found in table 12 and 13 (note
we didn’t provide the full estimation for all 18 participants here to avoid repetition).
[Figure 10, 11, Table 11, 12, and 13]
Discussion and Limitation of Current RCT
While the current analysis could not produce robust estimation between different
conditions due to limited sample size, across the entire sample, a simple comparison of the
reduction in ruminative episodes and average time one spends on those ruminative episodes
showed significant time1 to time2 reduction levels. This analysis limited us to finding evidence
of the effectiveness of a JITAI delivered in the form of narratives. Nonetheless, results of this
analysis suggest JITAI, regardless of narrative potential, as an essential means to deliver CBT.
Indeed, consistent with the previous meta-analysis that suggests the efficacy of internet-delivered
CBTs (Karyotaki et al., 2017), the current study provides initial evidence suggesting mobile
phones can also be an essential mode of delivery for CBTs. Especially, as no study so far has

51

tested mobile-phone delivered JITAI in reducing depressive rumination (Watkins et al., 2011),
this study fill in the gap.
Furthermore, we examined the data for control and JITAI condition participants with
GIMME modeling of group differences in figure 10. Indeed, participants within the WLC
condition demonstrated some time-lagged effects, such that experiencing ruminative episodes or
negative emotions at a time point were predictive with an increase of such experiences
subsequently. However, such lagged effects were not present in the JITAI condition. That is, no
dotted lines are indicating lagged effects. This difference at the group-level shows that those
participants in the JITAI condition didn’t bring their nega tive thoughts and previous ruminative
experience into their next ruminative episode. And it somehow stopped the “echo chamber” of
the previous negative experiences. Indeed, the main goal of the RFCBT was to train participants
to make different attributions and focus on problem-solving when they experience a particular
ruminative episode. This was the goal such that when the next ruminative episode occurs,
participants can immediately become aware of their ruminative thoughts and adopt one of the
strategies offered by RFCBT to reduce their ruminative thoughts. Patterns of the current GIMME
modeling speak to the possible intervention effects of such goals.
However, limitations of the current RCT are apparent. First, because of the small sample
of participants, the present analysis could offer neither a robust significant pre-/post-estimation
of the RCT, nor comparisons between the narrative and non-narrative versions of the JITAI
conditions. Second, because of delays in the IRB process of over one year, some exploratory
research questions on how COVID-19 and drinking patterns at the time of COVID-19 can affect
participants’ ruminative patterns didn’t show any significant results. Among participants who

52

have accomplished daily entries of their COVID-19 related stress, most of them indicate zero or
very low COVID-stress scores. Therefore, we didn’t detect any significant role of COVID -19
and related drinking behaviors (shown in the GIMME models). That is, the results of project
three could only be evaluated as a pilot for subsequent RCTs. Third, while the current RCT
showed some effects in reducing ruminative episodes among clinically depressed participants, it
did not examine the effect of the RCT on participants who are not clinically depressed.
Therefore, we can not generalize the efficacy of the RCT to a sample of participants who are not
clinically depressed. Lastly, in answering the results from project one, the current RCT materials
were specifically designed for participants to re-analysis their hurtful interpersonal feelings and
the supports they receive; both factors were designed as diaries for participants to self-report.
However, as we could not generate sufficient participants to code those variables, we could not
evaluate perceived hurtfulness and social support as mediators in statistical analysis. Therefore,
in subsequent analysis, there is a need to further designed and better measure those mediators.
Conclusion
The implications of the three projects could be profound in designing depression
interventions form both network science and interpersonal communication perspectives. The first
project shows the importance of interpersonal factors such as social support and perceived level
of hurtfulness in predicting subsequent behavioral outcomes of clinically depressed. Indeed,
situating clinically depressed individuals within a network of members who know and care about
the ego’s situation (i.e. size) could be an important component in creating interventions for
depressed participants. In the meantime, project one also calls for further attention towards
interventions that can address their rumination-related hurtful events. That is, project one was to

53

create a path-model indicating how communication could be an important part for depression
interventions.
The second project evaluated the development of social media platforms and mobile
technologies enabled JITAIs as a delivery means of reaching participants with an in-the-moment
intervention (Nahum-Shani et al., 2016). Results of this project have shown that in participating
in JITAI interventions, participants can receive in-the-moment feedback on their social network
platforms or mobile phones. And JITAI has already shown to be quite advantageous in
improving mental health outcomes (L. Wang & Miller, 2019). Adopting the JITAI framework,
participants can get in-the-moment feedback both on support seeking behaviors and maintaining
privacy. Specifically, results of project two have already shown JITAI is an efficient way to
reduce depressive symptoms.
Lastly, results of the third project were consistent with project two. Together they both
indicate that participants in JITAI conditions suggest promise for producing lower amounts of
ruminative episodes and lower average rumination times. These studies, therefore, indicate that
project 3 provides a pilot test suggesting the viability of a full-scale RCT for reducing rumination
in a population diagnosed with depression.
This line of projects, when put together, identified necessary procedures of generating
effective JITAIs (Sharmin et al., 2015). Indeed, as delineated by Sharmin and colleagues (2015),
an effective JITAI first relies on identifying treatment components for targeted behavior of
interests. The first project speaks to this requirement by pinpointed interpersonal components
(perceived level of hurtfulness and social support) in exacerbating depression outcomes.
Secondly, modalities of delivering effective JITAI were identified on targeting mental health

54

outcomes, especially when those JITAIs are designed to deliver CBT .. Lastly, the third project,
following suggestions by Sharmin and colleagues, pilot-tested the efficacy of such an
intervention. Therefore, the three projects serve an overarching goal of identifying means that
can deliver JITAI focusing on depression outcomes, and it can serve as a means for subsequent
clinical trials.

55

References
Afifi, T., Afifi, W., Merrill, A. F., Denes, A., & Davis, S. (2013). “You Need to Stop Talking
About This!”: Verbal Rumination and the Costs of Social Support. Human
Communication Research, 39(4), 395–421. https://doi.org/10.1111/hcre.12012
Ahmedani, B. K., Crotty, N., Abdulhak, M. M., & Ondersma, S. J. (2015). Pilot Feasibility Study
of a Brief, Tailored Mobile Health Intervention for Depression among Patients with
Chronic Pain. Behavioral Medicine, 41(1), 25–32.
https://doi.org/10.1080/08964289.2013.867827
Barnett, P. A., & Gotlib, I. H. (1988). Psychosocial functioning and depression: Distinguishing
among antecedents, concomitants, and consequences. Psychological Bulletin, 104(1), 97.
Beltz, A. M., & Gates, K. M. (2017). Network Mapping with GIMME. Multivariate Behavioral
Research, 52(6), 789–804. https://doi.org/10.1080/00273171.2017.1373014
Beltz, A. M., Moser, J. S., Zhu, D. C., Burt, S. A., & Klump, K. L. (2018). Using person-specific
neural networks to characterize heterogeneity in eating disorders: Illustrative links
between emotional eating and ovarian hormones. International Journal of Eating
Disorders, 51(7), 730–740.
Bergman, E. J., & Haley, W. E. (2009). Depressive Symptoms, Social Network, and
Bereavement Service Utilization and Preferences among Spouses of Former Hospice
Patients. Journal of Palliative Medicine, 12(2), 170–176.
https://doi.org/10.1089/jpm.2008.0237
Berkman, L. F., Glass, T., Brissette, I., & Seeman, T. E. (2000). From social integration to
health: Durkheim in the new millennium ☆☆This paper is adapted from Berkman, L.F.,

56

& Glass, T. Social integration, social networks, social support and health. In L. F.
Berkman & I. Kawachi, Social Epidemiology. New York: Oxford University Press; and
Brissette, I., Cohen S., Seeman, T. Measuring social integration and social networks. In
S. Cohen, L. Underwood & B. Gottlieb, Social Support Measurements and Intervention.
New York: Oxford University Press. Social Science & Medicine, 51(6), 843–857.
https://doi.org/10.1016/S0277-9536(00)00065-4
Bevan, J. L., & Sparks, L. (2014). The Relationship Between Accurate and Benevolently Biased
Serial Argument Perceptions and Individual Negative Health Perceptions
, The Relationship Between Accurate and
Benevolently Biased Serial Argument Perceptions and Individual Negative Health
Perceptions. Communication Research, 41(2), 257–281.
https://doi.org/10.1177/0093650212438391
Birnbaum, H. G., Kessler, R. C., Kelley, D., Ben‐Hamadi, R., Joish, V. N., & Greenberg, P. E.
(2010). Employer burden of mild, moderate, and severe major depressive disorder:
Mental health services utilization and costs, and work performance. Depression and
Anxiety, 27(1), 78–89. https://doi.org/10.1002/da.20580
Burns, M. N., Begale, M., Duffecy, J., Gergle, D., Karr, C. J., Giangrande, E., & Mohr, D. C.
(2011). Harnessing Context Sensing to Develop a Mobile Intervention for Depression.
Journal of Medical Internet Research, 13(3). https://doi.org/10.2196/jmir.1838
Camper, P. M., Jacobson, N. S., Holtzworth-Munroe, A., & Schmaling, K. B. (1988). Causal
attributions for interactional behaviors in married couples. Cognitive Therapy and
Research, 12(2), 195–209. https://doi.org/10.1007/BF01204931

57

CDC. (2020). Mental Health—Household Pulse Survey—COVID-19.
https://www.cdc.gov/nchs/covid19/pulse/mental-health.htm
Chandrashekar, P. (2018). Do mental health mobile apps work: Evidence and recommendations
for designing high-efficacy mental health mobile apps. MHealth, 4.
https://doi.org/10.21037/mhealth.2018.03.02
Child, S. T., & Lawton, L. E. (2020). Personal networks and associations with psychological
distress among young and older adults. Social Science & Medicine, 246, 112714.
https://doi.org/10.1016/j.socscimed.2019.112714
Chmitorz, A., Kunzler, A., Helmreich, I., Tüscher, O., Kalisch, R., Kubiak, T., Wessa, M., &
Lieb, K. (2018). Intervention studies to foster resilience – A systematic review and
proposal for a resilience framework in future intervention studies. Clinical Psychology
Review, 59, 78–100. https://doi.org/10.1016/j.cpr.2017.11.002
Cloven, D. H., & Roloff, M. E. (1991). Sense-Making Activities and Interpersonal Conflict:
Communicative Cures for the Mulling Blues. Western Journal of Speech
Communication: WJSC, 55(2), 134–158.
Cloven, D. H., & Roloff, M. E. (1993). Sense‐making activit ies and interpersonal conflict, II:
The effects of communicative intentions on internal dialogue. Western Journal of
Communication, 57(3), 309–329. https://doi.org/10.1080/10570319309374456
Cohen, S., & Wills, T. A. (1985). Stress, social support, and the buffering hypothesis.
Psychological Bulletin, 98(2), 310–357.
http://dx.doi.org.libproxy2.usc.edu/10.1037/0033-2909.98.2.310

58

Cornwell, B., Laumann, E. O., & Schumm, L. P. (2008). The Social Connectedness of Older
Adults: A National Profile. American Sociological Review, 73(2), 185–203.
https://doi.org/10.1177/000312240807300201
Czeisler, M. É. (2020). Mental Health, Substance Use, and Suicidal Ideation During the COVID-
19 Pandemic—United States, June 24–30, 2020. MMWR. Morbidity and Mortality
Weekly Report, 69. https://doi.org/10.15585/mmwr.mm6932a1
Desmet, P. M. A., Vastenburg, M. H., & Romero, N. (2016). Mood measurement with Pick-A-
Mood: Review of current methods and design of a pictorial self -report scale. Journal of
Design Research, 14(3), 241–279. https://doi.org/10.1504/JDR.2016.079751
Dimidjian, S., Hollon, S. D., Dobson, K. S., Schmaling, K. B., Kohlenberg, R. J., Addis, M. E.,
Gallop, R., McGlinchey, J. B., Markley, D. K., & Gollan, J. K. (2006). Randomized trial
of behavioral activation, cognitive therapy, and antidepressant medication in the acute
treatment of adults with major depression. Journal of Consulting and Clinical
Psychology, 74(4), 658.
Dong, M., Wang, S.-B., Li, Y., Xu, D.-D., Ungvari, G. S., Ng, C. H., Chow, I. H. I., & Xiang,
Y.-T. (2018). Prevalence of suicidal behaviors in patients with major depressive disorder
in China: A comprehensive meta-analysis. Journal of Affective Disorders, 225, 32–39.
https://doi.org/10.1016/j.jad.2017.07.043
Durkheim, E. (1951). Suicide: A study in sociology (JA Spaulding & G. Simpson, trans.).
Glencoe, IL: Free Press.(Original Work Published 1897).
Ferrari, A. J., Charlson, F. J., Norman, R. E., Patten, S. B., Freedman, G., Murray, C. J. L., Vos,
T., & Whiteford, H. A. (2013). Burden of Depressive Disorders by Country, Sex, Age,

59

and Year: Findings from the Global Burden of Disease Study 2010. PLOS Medicine,
10(11), e1001547. https://doi.org/10.1371/journal.pmed.1001547
Flynn, M., Kecmanovic, J., & Alloy, L. B. (2010). An Examination of Integrated Cognitive-
Interpersonal Vulnerability to Depression: The Role of Rumination, Perceived Social
Support, and Interpersonal Stress Generation. Cognitive Therapy and Research, 34(5),
456–466. https://doi.org/10.1007/s10608-010-9300-8
Franck, L., Molyneux, N., & Parkinson, L. (2016). Systematic review of interventions addressing
social isolation and depression in aged care clients. Quality of Life Research, 25(6),
1395–1407. https://doi.org/10.1007/s11136-015-1197-y
Free, C., Knight, R., Robertson, S., Whittaker, R., Edwards, P., Zhou, W., Rodgers, A., Cairns,
J., Kenward, M. G., & Roberts, I. (2011). Smoking cessation support delivered via
mobile phone text messaging (txt2stop): A single-blind, randomised trial. Lancet,
378(9785), 49–55.
Gariépy, G., Honkaniemi, H., & Quesnel-Vallée, A. (2016). Social support and protection from
depression: Systematic review of current findings in Western countries. The British
Journal of Psychiatry, 209(4), 284–293. https://doi.org/10.1192/bjp.bp.115.169094
Gates, K. M., Lane, S. T., Varangis, E., Giovanello, K., & Guiskewicz, K. (2017). Unsupervised
Classification During Time-Series Model Building. Multivariate Behavioral Research,
52(2), 129–148. https://doi.org/10.1080/00273171.2016.1256187
Gates, K. M., & Molenaar, P. C. M. (2012). Group search algorithm recovers effective
connectivity maps for individuals in homogeneous and heterogeneous samples.
NeuroImage, 63(1), 310–319. https://doi.org/10.1016/j.neuroimage.2012.06.026

60

Gerteis, A. K. S., & Schwerdtfeger, A. R. (2016). When rumination counts: Perceived social
support and heart rate variability in daily life. Psychophysiology, 53(7), 1034–1043.
https://doi.org/10.1111/psyp.12652
Green, M. C., & Jenkins, K. M. (2014). Interactive Narratives: Processes and Outcomes in User-
Directed Stories. Journal of Communication, 64(3), 479–500.
Greenberg, M. T., Domitrovich, C., & Bumbarger, B. (2001). The prevention of mental disorders
in school-aged children: Current state of the field. Prevention & Treatment, 4(1).
https://doi.org/10.1037/1522-3736.4.1.41a
Greenberg, P. E., Fournier, A.-A., Sisitsky, T., Pike, C. T., & Kessler, R. C. (2015). The
Economic Burden of Adults With Major Depressive Disorder in the United States (2005
and 2010). The Journal of Clinical Psychiatry, 76(02), 155–162.
https://doi.org/10.4088/JCP.14m09298
Gustafson, D. H., McTavish, F. M., Chih, M.-Y., Atwood, A. K., Johnson, R. A., Boyle, M. G.,
Levy, M. S., Driscoll, H., Chisholm, S. M., Dillenburg, L., Isham, A., & Shah, D. (2014).
A Smartphone Application to Support Recovery From Alcoholism: A Randomized
Clinical Trial. JAMA Psychiatry, 71(5), 566–572.
https://doi.org/10.1001/jamapsychiatry.2013.4642
Haleem, A., Javaid, M., Vaishya, R., & Deshmukh, S. G. (2020). Areas of academic research
with the impact of COVID-19. The American Journal of Emergency Medicine, 38(7),
1524–1526. https://doi.org/10.1016/j.ajem.2020.04.022

61

Hall, E. D., Meng, J., & Reynolds, R. M. (2020). Confidant network and interpersonal
communication associations with depression in older adulthood. Health Communication,
35(7).
Harris, P. A., Scott, K. W., Lebo, L., Hassan, N., Lighter, C., & Pulley, J. (2012).
ResearchMatch: A National Registry to Recruit Volunteers for Clinical Research.
Academic Medicine : Journal of the Association of American Medical Colleges, 87(1),
66–73. https://doi.org/10.1097/ACM.0b013e31823ab7d2
Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process analysis: A
regression-based approach. Guilford Press.
Heffner, J. L., Vilardaga, R., Mercer, L. D., Kientz, J. A., & Bricker, J. B. (2015). Feature-level
Analysis of a Novel Smartphone Application for Smoking Cessation. The American
Journal of Drug and Alcohol Abuse, 41(1), 68–73.
https://doi.org/10.3109/00952990.2014.977486
Heron, K. E. (2011). Ecological Momentary Intervention [EMI]: Incorporating mobile
technology into a disordered eating treatment program for college women [Ph.D.,
Syracuse University].
https://search.proquest.com/docview/883986008/abstract/230876F0AAC74F24PQ/1
Heron, K. E., & Smyth, J. M. (2010). Ecological momentary interventions: Incorporating mobile
technology into psychosocial and health behaviour treatments. British Journal of Health
Psychology, 15(1), 1–39. https://doi.org/10.1348/135910709X466063
Ho, C. S., Chee, C. Y., & Ho, R. C. (2020). Mental health strategies to combat the psychological
impact of COVID-19 beyond paranoia and panic. Ann Acad Med Singapore, 49(1), 1–3.

62

Hofmann, S. G., Asnaani, A., Vonk, I. J. J., Sawyer, A. T., & Fang, A. (2012). The Efficacy of
Cognitive Behavioral Therapy: A Review of Meta-analyses. Cognitive Therapy and
Research, 36(5), 427–440. https://doi.org/10.1007/s10608-012-9476-1
Horsch, C. H., Lancee, J., Griffioen-Both, F., Spruit, S., Fitrianie, S., Neerincx, M. A., Beun, R.
J., & Brinkman, W.-P. (2017). Mobile Phone-Delivered Cognitive Behavioral Therapy
for Insomnia: A Randomized Waitlist Controlled Trial. Journal of Medical Internet
Research, 19(4), e6524. https://doi.org/10.2196/jmir.6524
Johnson, D. P., & Whisman, M. A. (2013). Gender differences in rumination: A meta-analysis.
Personality and Individual Differences, 55(4), 367–374.
https://doi.org/10.1016/j.paid.2013.03.019
Joiner, T. E. (2000). Depression’s Vicious Scree: Self -Propagating and Erosive Processes in
Depression Chronicity. Clinical Psychology: Science and Practice, 7(2), 203–218.
https://doi.org/10.1093/clipsy.7.2.203
Joiner, T. E., & Lonigan, C. J. (2000). Tripartite Model of Depression and Anxiety in Youth
Psychiatric Inpatients: Relations With Diagnostic Status and Future Symptoms. Journal
of Clinical Child Psychology, 29(3), 372–382.
https://doi.org/10.1207/S15374424JCCP2903_8
Joiner, T. E., & Metalsky, G. I. (1995). A prospective test of an integrative interpersonal theory
of depression: A naturalistic study of college roommates. Journal of Personality and
Social Psychology, 69(4), 778–788. https://doi.org/10.1037/0022-3514.69.4.778

63

Jones, E. E., & Davis, K. E. (1965). From acts to dispositions the attribution process In person
perception1. In Advances in experimental social psychology (Vol. 2, pp. 219–266).
Elsevier.
Joyce, S., Shand, F., Tighe, J., Laurent, S. J., Bryant, R. A., & Harvey, S. B. (2018). Road to
resilience: A systematic review and meta-analysis of resilience training programmes and
interventions. BMJ Open, 8(6), e017858. https://doi.org/10.1136/bmjopen-2017-017858
Kalisch, Y., & Robins, G. (2006). Psychological predispositions and network structure: The
relationship between individual predispositions, structural holes and network closure.
Social Networks, 28(1), 56–84. https://doi.org/10.1016/j.socnet.2005.04.004
Kanter, J. W., Mulick, P. S., Busch, A. M., Berlin, K. S., & Martell, C. R. (2006). The
Behavioral Activation for Depression Scale (BADS): Psychometric Properties and Factor
Structure. Journal of Psychopathology and Behavioral Assessment, 29(3), 191.
https://doi.org/10.1007/s10862-006-9038-5
Kanter, J. W., Mulick, P. S., Busch, A. M., Berlin, K. S., & Martell, C. R. (2007). The
Behavioral Activation for Depression Scale (BADS): Psychometric Properties and Factor
Structure. Journal of Psychopathology and Behavioral Assessment, 3(29), 191–202.
Karyotaki, E., Riper, H., Twisk, J., Hoogendoorn, A., Kleiboer, A., Mira, A., Mackinnon, A.,
Meyer, B., Botella, C., Littlewood, E., Andersson, G., Christensen, H., Klein, J. P.,
Schröder, J., Bretón-López, J., Scheider, J., Griffiths, K., Farrer, L., Huibers, M. J. H., …
Cuijpers, P. (2017). Efficacy of Self-guided Internet-Based Cognitive Behavioral
Therapy in the Treatment of Depressive Symptoms: A Meta-analysis of Individual

64

Participant Data. JAMA Psychiatry, 74(4), 351–359.
https://doi.org/10.1001/jamapsychiatry.2017.0044
Kasch, K. L., Rottenberg, J., Arnow, B. A., & Gotlib, I. H. (2002). Behavioral activation and
inhibition systems and the severity and course of depression. Journal of Abnormal
Psychology, 111(4), 589–597.
Kawachi, I., & Berkman, L. F. (2001). Social ties and mental health. Journal of Urban Health,
78(3), 458–467. https://doi.org/10.1093/jurban/78.3.458
Kelly, J., Gooding, P., Pratt, D., Ainsworth, J., Welford, M., & Tarrier, N. (2012). Intelligent
real-time therapy: Harnessing the power of machine learning to optimise the delivery of
momentary cognitive-behavioural interventions. Journal of Mental Health, 21(4), 404–
414. http://dx.doi.org.libproxy1.usc.edu/10.3109/09638237.2011.638001
Kennedy, C. M., Powell, J., Payne, T. H., Ainsworth, J., Boyd, A., & Buchan, I. (2012). Active
assistance technology for health-related behavior change: An interdisciplinary review.
Journal of Medical Internet Research, 14(3).
Klimmt, C., Hefner, D., Vorderer, P., Roth, C., & Blake, C. (2010). Identification With Video
Game Characters as Automatic Shift of Self-Perceptions. Media Psychology, 13(4), 323–
338. https://doi.org/10.1080/15213269.2010.524911
Krebs, P., Prochaska, J. O., & Rossi, J. S. (2010). DEFINING WHAT WORKS IN
TAILORING: A META-ANALYSIS OF COMPUTERTAILORED INTERVENTIONS
FOR HEALTH BEHAVIOR CHANGE. Preventive Medicine, 51(3–4), 214–221.
https://doi.org/10.1016/j.ypmed.2010.06.004

65

Kristjánsdóttir, Ó. B., Fors, E. A., Eide, E., Finset, A., Stensrud, T. L., van Dulmen, S., Wigers,
S. H., & Eide, H. (2013). A Smartphone-Based Intervention With Diaries and Therapist-
Feedback to Reduce Catastrophizing and Increase Functioning in Women With Chronic
Widespread Pain. Journal of Medical Internet Research, 15(1).
https://doi.org/10.2196/jmir.2249
Lane, S. T., & Gates, K. M. (2017). Automated selection of robust individual-level structural
equation models for time series data. Structural Equation Modeling: A Multidisciplinary
Journal, 24(5), 768–782.
Lee, J.-S. (2019). Perceived social support functions as a resilience in buffering the impact of
trauma exposure on PTSD symptoms via intrusive rumination and entrapment in
firefighters. PLOS ONE, 14(8), e0220454. https://doi.org/10.1371/journal.pone.0220454
Lin, N., Dean, A., & Ensel, W. M. (1986). Social Support, Life Events, and Depression. Elsevier.
https://doi.org/10.1016/C2013-0-11093-8
Lin, N., Ye, X., & Ensel, W. M. (1999). Social Support and Depressed Mood: A Structural
Analysis. Journal of Health and Social Behavior, 40(4), 344–359. JSTOR.
https://doi.org/10.2307/2676330
Litwin, H., & Stoeckel, K. J. (2014). Confidant Network Types and Well-Being Among Older
Europeans. The Gerontologist, 54(5), 762–772. https://doi.org/10.1093/geront/gnt056
Liu, C. H., Zhang, E., Wong, G. T. F., Hyun, S., & Hahm, H. “Chris.” (2020). Factors associated
with depression, anxiety, and PTSD symptomatology during the COVID-19 pandemic:
Clinical implications for U.S. young adult mental health. Psychiatry Research, 290,
113172. https://doi.org/10.1016/j.psychres.2020.113172

66

Luo, M., Guo, L., Yu, M., Jiang, W., & Wang, H. (2020). The psychological and mental impact
of coronavirus disease 2019 (COVID-19) on medical staff and general public – A
systematic review and meta-analysis. Psychiatry Research, 291, 113190.
https://doi.org/10.1016/j.psychres.2020.113190
Lyubomirsky, S., Caldwell, N. D., & Nolen-Hoeksema, S. (1998). Effects of ruminative and
distracting responses to depressed mood on retrieval of autobiographical memories.
Journal of Personality and Social Psychology, 75(1), 166–177.
http://dx.doi.org.libproxy2.usc.edu/10.1037/0022-3514.75.1.166
Lyubomirsky, S., & Nolen-Hoeksema, S. (1995). Effects of self-focused rumination on negative
thinking and interpersonal problem solving. Journal of Personality and Social
Psychology, 69(1), 176–190. http://dx.doi.org.libproxy2.usc.edu/10.1037/0022-
3514.69.1.176
Manos, R. C., Kanter, J. W., & Luo, W. (2011). The Behavioral Activation for Depression
Scale–Short Form: Development and Validation. Behavior Therapy, 42(4), 726–739.
https://doi.org/10.1016/j.beth.2011.04.004
Marsden, P. V. (1987). Core discussion networks of Americans. American Sociological Review,
122–131.
Marsden, P. V. (2011). Survey methods for network data. The SAGE Handbook of Social
Network Analysis, 25, 370–388.
McGinty, E. E., Presskreischer, R., Han, H., & Barry, C. L. (2020). Psychological Distress and
Loneliness Reported by US Adults in 2018 and April 2020. JAMA, 324(1), 93–94.
https://doi.org/10.1001/jama.2020.9740

67

McPherson, M., Smith-Lovin, L., & Brashears, M. E. (2006). Social isolation in America:
Changes in core discussion networks over two decades. American Sociological Review,
71(3), 353–375.
Michell, L. (2000). Smoke rings: Social network analysis of friendship groups, smoking and
drug-taking. Drugs: Education, Prevention and Policy, 7(1), 21–37.
Michl, L. C., McLaughlin, K. A., Shepherd, K., & Nolen-Hoeksema, S. (2013). Rumination as a
Mechanism Linking Stressful Life Events to Symptoms of Depression and Anxiety:
Longitudinal Evidence in Early Adolescents and Adults. Journal of Abnormal
Psychology, 122(2), 339–352. https://doi.org/10.1037/a0031994
Miner, A., Kuhn, E., Hoffman, J. E., Owen, J. E., Ruzek, J. I., & Taylor, C. B. (2016).
Feasibility, acceptability, and potential efficacy of the PTSD Coach app: A pilot
randomized controlled trial with community trauma survivors. Psychological Trauma:
Theory, Research, Practice, and Policy, 8(3), 384–392.
http://dx.doi.org.libproxy2.usc.edu/10.1037/tra0000092
Nahum-Shani, I., Hekler, E. B., & Spruijt-Metz, D. (2015). Building health behavior models to
guide the development of just-in-time adaptive interventions: A pragmatic framework.
Health Psychology, 34(Suppl), 1209–1219.
http://dx.doi.org.libproxy2.usc.edu/10.1037/hea0000306
Nahum-Shani, I., Smith, S. N., Spring, B. J., Collins, L. M., Witkiewitz, K., Tewari, A., &
Murphy, S. A. (2016). Just-in-Time Adaptive Interventions (JITAIs) in Mobile Health:
Key Components and Design Principles for Ongoing Health Behavior Support. Annals of
Behavioral Medicine, 1–17. https://doi.org/10.1007/s12160-016-9830-8

68

Newman, H. M., & Langer, E. J. (1981). Post-Divorce Adaptation and the Attribution of
Responsibility. Sex Roles; New York, N.Y., 7(3), 223–232.
Nickell, L. A. (2004). Psychosocial effects of SARS on hospital staff: Survey of a large tertiary
care institution. Canadian Medical Association Journal (CMAJ), 170(5), 793–798.
https://doi.org/10.1503/cmaj.1031077
Nolen-Hoeksema, S. (1991). Responses to depression and their effects on the duration of
depressive episodes. Journal of Abnormal Psychology, 100(4), 569–582.
http://dx.doi.org/10.1037//0021-843x.100.4.569
Nolen-Hoeksema, S., & Davis, C. G. (1999). “Thanks for sharing that”: Ruminators and their
social support networks. Journal of Personality and Social Psychology, 77(4), 801–814.
https://doi.org/10.1037/0022-3514.77.4.801
Nolen-Hoeksema, S., Wisco, B. E., & Lyubomirsky, S. (2008). Rethinking Rumination.
Perspectives on Psychological Science, 3(5), 400–424. https://doi.org/10.1111/j.1745-
6924.2008.00088.x
Olatunji, B. O., Naragon‐Gainey, K., & Wolitzky‐Taylor, K . B. (2013). Specificity of
Rumination in Anxiety and Depression: A Multimodal Meta-Analysis. Clinical
Psychology: Science and Practice, 20(3), 225–257. https://doi.org/10.1111/cpsp.12037
Peng, W., Lee, M., & Heeter, C. (2010). Lu. Journal of Communication, 60(4), 723–742.
https://doi.org/10.1111/j.1460-2466.2010.01511.x
Pettit, J. W., & Joiner, T. E. (2006). Interpersonal Conflict Avoidance. In Chronic depression:
Interpersonal sources, therapeutic solutions (pp. 73–84). American Psychological
Association. https://doi.org/10.1037/11291-006

69

Pfefferbaum, B., & North, C. S. (2020). Mental Health and the Covid-19 Pandemic. New
England Journal of Medicine, 0(0), null. https://doi.org/10.1056/NEJMp2008017
Possemato, K., Kuhn, E., Johnson, E., Hoffman, J. E., Owen, J. E., Kanuri, N., De Stefano, L., &
Brooks, E. (2016). Using PTSD Coach in primary care with and without clinician
support: A pilot randomized controlled trial. General Hospital Psychiatry, 38, 94–98.
https://doi.org/10.1016/j.genhosppsych.2015.09.005
Prince, M., Patel, V., Saxena, S., Maj, M., Maselko, J., Phillips, M. R., & Rahman, A. (2007).
No health without mental health. The Lancet, 370(9590), 859–877.
https://doi.org/10.1016/S0140-6736(07)61238-0
Pulley, J. M., Jerome, R. N., Bernard, G. R., Olson, E. J., Tan, J., Wilkins, C. H., & Harris, P. A.
(2018). Connecting the public with clinical trial options: The ResearchMatch Trials
Today tool. Journal of Clinical and Translational Science, 2(4), 253–257.
https://doi.org/10.1017/cts.2018.327
Puterman, E., DeLongis, A., & Pomaki, G. (2010). Protecting Us from Ourselves: Social Support
as a Buffer of Trait and State Rumination. Journal of Social and Clinical Psychology,
29(7), 797–820. https://doi.org/10.1521/jscp.2010.29.7.797
Rachman, S. (1998). A Cognitive Theory of Obsessions. In Behavior and Cognitive Therapy
Today (pp. 209–222). Elsevier. https://doi.org/10.1016/B978-008043437-7/50017-X
Reavell, J., Hopkinson, M., Clarkesmith, D., & Lane, D. A. (2018). Effectiveness of Cognitive
Behavioral Therapy for Depression and Anxiety in Patients With Cardiovascular Disease:
A Systematic Review and Meta-Analysis. Psychosomatic Medicine, 80(8), 742–753.
https://doi.org/10.1097/PSY.0000000000000626

70

Ribeiro, J., D., Franklin, J. C., Fox, K. R., Bentley, K. H., Kleiman, E. M., Chang, B. P., &
Nock, M. K. (2016). Self-injurious thoughts and behaviors as risk factors for future
suicide ideation, attempts, and death: A meta-analysis of longitudinal studies.
Psychological Medicine, 46(2), 225–236. https://doi.org/10.1017/S0033291715001804
Ribeiro, J., D., Huang, X., Fox, K. R., & Franklin, J. C. (2018). Depression and hopelessness as
risk factors for suicide ideation, attempts and death: Meta-analysis of longitudinal
studies. The British Journal of Psychiatry, 212(5), 279–286.
Rizvi, S. L., Dimeff, L. A., Skutch, J., Carroll, D., & Linehan, M. M. (2011). A pilot study of the
DBT coach: An interactive mobile phone application for individuals with borderline
personality disorder and substance use disorder. Behavior Therapy, 42(4), 589–600.
Roloff, M. E., Reznik, R. M., Miller, C. W., & Johnson, K. L. (2015). “I Thought We Settled
This⁈” Antecedents and Consequences of Resolution of an Initial Episode in a Serial
Argument. Argumentation and Advocacy, 52(1), 8–31.
https://doi.org/10.1080/00028533.2015.11821858
Rood, L., Roelofs, J., Bögels, S. M., Nolen-Hoeksema, S., & Schouten, E. (2009). The influence
of emotion-focused rumination and distraction on depressive symptoms in non-clinical
youth: A meta-analytic review. Clinical Psychology Review, 29(7), 607–616.
https://doi.org/10.1016/j.cpr.2009.07.001
Rueger, S. Y., Malecki, C. K., Pyun, Y., Aycock, C., & Coyle, S. (2016). A meta-analytic review
of the association between perceived social support and depression in childhood and
adolescence. Psychological Bulletin, 142(10), 1017–1067.
https://doi.org/10.1037/bul0000058

71

Ruwaard, J., Kooistra, L., & Thong, M. (n.d.). Ecological Momentary Assessment in Mental
Health Research. Retrieved August 27, 2020, from
https://jruwaard.github.io/aph_ema_handbook/
SAMHSA. (2018). Mental Health Services Administration (SAMHSA).(2017). Key substance use
and mental health indicators in the United States: Results from the 2016 National Survey
on Drug Use and Health (HHS Publication No. SMA 17-5044, NSDUH Series H-52).
Center for Behavioral Health Statistics and Quality.
https://www.samhsa.gov/data/sites/default/files/cbhsq-
reports/NSDUHDetailedTabs2017/NSDUHDetailedTabs2017.pdf
Sangalang, A., Johnson, J. M. Q., & Ciancio, K. E. (2013). Exploring audience involvement with
an interactive narrative: Implications for incorporating transmedia storytelling into
entertainment-education campaigns. Critical Arts, 27(1), 127–146.
https://doi.org/10.1080/02560046.2013.766977
Santini, Z. I., Koyanagi, A., Tyrovolas, S., Mason, C., & Haro, J. M. (2015). The association
between social relationships and depression: A systematic review. Journal of Affective
Disorders, 175, 53–65.
Serafini, G., Parmigiani, B., Amerio, A., Aguglia, A., Sher, L., & Amore, M. (2020). The
psychological impact of COVID-19 on the mental health in the general population. QJM:
An International Journal of Medicine, 113(8), 531–537.
https://doi.org/10.1093/qjmed/hcaa201
Sharmin, M., Raij, A., Epstien, D., Nahum-Shani, I., Beck, J. G., Vhaduri, S., Preston, K., &
Kumar, S. (2015). Visualization of Time-Series Sensor Data to Inform the Design of Just-

72

In-Time Adaptive Stress Interventions. Proceedings of the ... ACM International
Conference on Ubiquitous Computing . UbiComp (Conference), 2015, 505–516.
Showers, C. (1988). The effects of how and why thinking on perceptions of future negative
events. Cognitive Therapy and Research, 12(3), 225–240.
Sijbrandij, M., Kunovski, I., & Cuijpers, P. (2016). Effectiveness of Internet-Delivered Cognitive
Behavioral Therapy for Posttraumatic Stress Disorder: A Systematic Review and Meta-
Analysis. Depression and Anxiety, 33(9), 783–791. https://doi.org/10.1002/da.22533
Sundar, S. S., Oh, J., Kang, H., Sreenivasan, A., Dillard, J. P., & Shen, L. (2013). How does
technology persuade? Theoretical mechanisms for persuasive technologies. The Sage
Handbook of Persuasion: Developments in Theory and Practice, 388–404.
Uchino, B. N., Bowen, K., Kent de Grey, R., Mikel, J., & Fisher, E. B. (2018). Social Support
and Physical Health: Models, Mechanisms, and Opportunities. In E. B. Fisher, L. D.
Cameron, A. J. Christensen, U. Ehlert, Y. Guo, B. Oldenburg, & F. J. Snoek (Eds.),
Principles and Concepts of Behavioral Medicine: A Global Handbook (pp. 341–372).
Springer. https://doi.org/10.1007/978-0-387-93826-4_12
Wahl, K., Schönfeld, S., Hissbach, J., Küsel, S., Zurowski, B., Moritz, S., Hohagen, F., &
Kordon, A. (2011). Differences and similarities between obsessive and ruminative
thoughts in obsessive-compulsive and depressed patients: A comparative study. Journal
of Behavior Therapy and Experimental Psychiatry, 42(4), 454–461.
https://doi.org/10.1016/j.jbtep.2011.03.002
Walter, N., Murphy, S. T., & Gillig, T. K. (2017). To Walk a Mile in Someone Else’s Shoes:
How Narratives Can Change Causal Attribution Through Story Exploration and

73

Character Customization. Human Communication Research, n/a-n/a.
https://doi.org/10.1111/hcre.12112
Wang, L., & Miller, L. C. (2019). Just-in-the-Moment Adaptive Interventions (JITAI): A Meta-
Analytical Review. Health Communication, 0(0), 1–14.
https://doi.org/10.1080/10410236.2019.1652388
Wang, Y., Wang, Y., Chen, S., Zhu, C., & Wang, K. (2008). Theory of mind disability in major
depression with or without psychotic symptoms: A componential view. Psychiatry
Research, 161(2), 153–161. https://doi.org/10.1016/j.psychres.2007.07.018
Watkins, E. (2015). Psychological treatment of depressive rumination. Current Opinion in
Psychology, 4, 32–36.
Watkins, E. R., Mullan, E., Wingrove, J., Rimes, K., Steiner, H., Bathurst, N., Eastman, R., &
Scott, J. (2011). Rumination-focused cognitive–behavioural therapy for residual
depression: Phase II randomised controlled trial. The British Journal of Psychiatry,
199(4), 317–322.
Webb, L. (n.d.). COVID-19 lockdown: A perfect storm for older people’s mental health. Journal
of Psychiatric and Mental Health Nursing, n/a(n/a). https://doi.org/10.1111/jpm.12644
Wellman, B., & Wortley, S. (1989). Brothers’ keepers: Situating kinship relations in broader
networks of social support. Sociological Perspectives, 32(3), 273–306.
Yip, P. S. F., Cheung, Y. T., Chau, P. H., & Law, Y. W. (2010). The impact of epidemic
outbreak: The case of severe acute respiratory syndrome (SARS) and suicide among
older adults in Hong Kong. Crisis: The Journal of Crisis Intervention and Suicide
Prevention, 31(2), 86–92. https://doi.org/10.1027/0227-5910/a000015

74

Zhai, Y., & Du, X. (2020). Loss and grief amidst COVID-19: A path to adaptation and
resilience. Brain, Behavior, and Immunity, 87, 80–81.
https://doi.org/10.1016/j.bbi.2020.04.053
Zoogman, S., Goldberg, S. B., Hoyt, W. T., & Miller, L. (2015). Mindfulness Interventions with
Youth: A Meta-Analysis. Mindfulness, 6(2), 290–302. https://doi.org/10.1007/s12671-
013-0260-4

75

Appendices A: Tables
Table 1. Chapter Two: Characteristics of Participant
Clinically Depressed Non-depressed
Age 39.05 13.59 37.85 11.92
Education
(in years)

12.31 5.12

13.32 4.5

n Percentage

n Percentage
Gender Male 69 17.7

144 21

Female 321 82.3

542 79
Income Less than 30,000 106 27.2

180 26.2

30,000 – 39,999 45 11.5

80 11.7

40,000 – 49,999 47 12

89 13

50,000 – 59,999 46 11.8

89 13

60,000 – 69,999 27 7

48 7

70,000 – 79,999 28 7.3

49 7.1

80,000 – 89,999 23 5.9

37 5.4

90,000 – 99,999 15 3.9

24 3.5

100,000 or more 51 13.2

77 11.2

Ethnicity White/Non-Hispanic 281 72

515 75

Hispanic 38 9.8

56 8.1

Black 39 10.1

55 8
76
Pacific Islander 4 1.1 8 1.1
Native American 2 0.6 4 0.6
Middle Eastern 2 0.6 4 0.6
Asian American 22 5.6 41 6
Depression
diagnosis
Major Depressive
Disorder
230 59
Persistent Depressive
Disorder
59 15
Bipolar Disorder 35 9
Seasonal Affective
Disorder
8 2
Psychotic Depression 1 0.2
Peripartum (Postpartum)
Depression
5 1.2
'Situational' Depression 15 3.9
Atypical
Depression
9 2.4
Mild depressive
diagnosis
46 11.7
77
Table 2. Chapter Two: Differences in depression outcomes by clinical diagnosis
Clinically Depressed Non-depressed
M SD M SD t Cohen’s d
Depression Exacerbation
(n=1054)
3.77 1.27 3.83 1.44 0.59 *
Depression Mitigation (n=1054) 4.36 1.34 4.26 1.38 -1.22
Social Support (n=1039) 4.20 1.63 4.16 1.72 -0.33
Trait Rumination( n=1063) 4.51 1.12 3.72 1.47 -9.11*
Hurtful Experience (n=1046) 9.10 1.96 2.63 2.15 -48.33*
* p<.05
*p<.01
*note: the differences in df was due to the fact that network measures could not be computed with a network size lower than 2 .
78
Table3. Chapter Two: Correlation Tables of Variables of Interests (Spearman’s rho)
1 2 3 4 5 6
1. Diagnosis 1.00
2. Trait Rumination (M = 4.01, SD = 1.40) .26
**
1.00
3. Perceived Hurtfulness (M = 5.00, SD = 1.40) .87
**
.31
**
1.00
4. Perceived Social Support (M = 4.17, SD =
1.47
.01 -.01 -.01 1.00
5. Depression Exacerbation -.01 .28
**
.12
**
.04 1.00
6. Depression Mitigation .04 -.01 -.01 .10
**
-.26
**
1.00
* p<.05
*p<.01
Note, all variables are recoded such that large value means high scores of a particular construct, lower mean low scores mean low
level
79
Table 4. Chapter Two: Direct and Indirect Effects of Parallel Mediation Analysis on Depression Mitigation
X:Trait Rumination X:Trait Rumination
M
i:
Perceived Hurtfulness M
i:
Perceived Hurtfulness
M
ii
: Social Support M
ii
: Social Support
W: Diagnosis W: Diagnosis
Y: Depression Mitigation
В SE 95% CI В SE 95% CI
a
i1: X predicting to Mi
.42** 0.06 0.3 0.54 b
i
: Mi to Y -0.02 0.02 -0.05 0.02
a
i2: W predicting to Mi
9.53** 0.5 8.54 10.52 b
ii: Mii to Y
.07* 0.02 0.02 0.11
a
i3: Interaction X*W to Mi
-.76** 0.11 -0.98 -0.54 c
1
’: X to Y -.28* 0.04 -0.36 -0.21
a
ii1: X predicting to Mii
-.10* 0.05 -0.2 0 c
2
’: W to Y -3.37** 0.37 -4.09 -2.66
a
ii2: W predicting to Mii
-0.72 0.41 -1.53 0.09 c
3
’: W*X to
Y
.82** 0.07 0.71 1
a
ii3: Interaction X*W to Mii
.19* 0.09 0 0.37

R
2
0.16
Note: legends of path notations are shown in figure 2b.
80
Table 5. Chapter Two: Direct and Indirect Effects of Parallel Mediation Analysis on Depression Exacerbation
X:Trait Rumination X:Trait Rumination
M
i:
Perceived Hurtfulness M
i:
Perceived Hurtfulness
M
ii
: Social Support M
ii
: Social Support
W: Diagnosis W: Diagnosis
Y: Depression Exacerbation
В SE 95% CI В SE 95% CI
a
i1: X predicting to Mi
.42** 0.06 0.3 0.54 b
i
: Mi to Y
0.06 0.02 0.48 0.62
a
i2: W predicting to Mi
9.53** 0.5 8.54 10.52 b
ii: Mii to Y

.06* 0.04 0.02 0.11
a
i3: Interaction X*W to Mi
-.76** 0.11 -0.98 -0.54 c
1
’: X to Y
.29** 0.04 0.36 0.21
a
ii1: X predicting to Mii
-.10* 0.05 -0.2 0 c
2
’: W to Y
3.38** 0.36 4.1 2.67
a
ii2: W predicting to Mii
-0.72 0.41 -1.53 0.09 c
3
’: W*X to
Y
-.76** 0.07 0.43 0.22
a
ii3: Interaction X*W to Mii
.19* 0.09 0 0.37

R
2
0.16
Note: legends of path notations are shown in figure 2b.
81
Table 6. Chapter Three: JITIAI and WLC-control Studies Included (k=9)
Authors, Year of publication Behaviors of interests Country Age: Mean (SD) JITAI n
WLC-control
n
Brookie, Mainvil, Carr, Vissers, &
Conner, 2017 Healthy Diet
New Zealand
19.4(1.45)
57 60
Heron, 2011 Healthy Diet US 19.50 (0.70) 44 43
Miner et al., 2016 Mental Health US 45.7 (13.9) 25 24
Proudfoot et al., 2013 Mental Health Australia 39.23(10.96) 126 198
Rodgers et al., 2005 Addiction New Zealand 22(3.5) 806 818
Suffoletto, Callaway, Kristan,
Kraemer, & Clark, 2012
Addiction US 21(1.8) 14 25
Svetkey et al., 2008 Weight Loss US 55.6 328 324
Witkiewitz et al., 2014 Addiction US 19.55 (1.45) 30 26
Total n 1430 1518
82
Table7. Chapter Three: JITIAI and non-JITAI Studies Included (k=21)
Authors, Year of publication Behaviors of interests Country Age: Mean (SD) JITAI n non-JITAI n
Atienza, King, Oliveira, Ahn, & Gardner,
2008
Healthy Diet US 60.12(7.64) 16 11
Brendryen & Kraft, 2008 Addiction Norway 36.15 (10.24) 197 199
Brendryen, Drozd, & Kraft, 2008 Addiction Norway 39.60(10.88) 144 146
Burnett, Taylor, & Agras, 1985 Weight Loss US 41.50 (7.24) 6 6
Brookie, Mainvil, Carr, Vissers, &
Conner, 2017
Healthy Diet US 19.4(1.45) 55 59
Depp et al., 2015 Mental Health (bipolar) US 47.5 (12.8) 41 41
Franklin, et al., 2006 Diabetic Management US 13.42 (2.33) 27 64
Gustafson et al., 2014 Addiction US 38(10) 170 179
Heron, 2011 Healthy Diet US 19.50 (0.70) 44 44
Hutcheson, 2013 Healthy Diet US 11.29(1.82) 16 15
Kim & Kim, 2008 Diabetic Management Korean 46.91(8.61) 18 16
83
Kristjánsdóttir et al., 2013 Mental Health Norway 44.20(11.03) 37 40
Moody, Tegge, Poe, Koffarnus, &
Bickel, 2017
Addiction US 35(12.08) 18 17
Newman, Przeworski, Consoli, & Taylor,
2014
Mental Health (anxiety) US 42.16(12.24) 11 23
Patrick et al., 2009 Weight Loss US 44.9(7.7) 26 26
Proudfoot et al., 2013 Mental Health Australia 39.23(10.96) 126 195
Quinn et al., 2011 Diabetic Management US 52.45(8.04) 77 51
Suffoletto, Callaway, Kristan, Kraemer,
& Clark, 2012
Addiction US 21(1.8) 14 25
Svetkey et al., 2008 Weight Loss US 55.6 328 324
Vidrine, Arduino, Lazev, & Gritz, 2006 Addiction US 42.85(8.1) 38 39
Witkiewitz et al., 2014 Addiction US 19.55 (1.45) 30 29
Total n 1345 1478
84
Table 8. Chapter Three: Studies eligible to extract pre/post changes (k=13)
Authors, Year of publication Behaviors of interests Country Age n
Ahmedani, Crotty, Abdulhak, & Ondersma, 2015 Mental Health US (n/a) 64
Brookie, Mainvil, Carr, Vissers, & Conner, 2017 Healthy Diet New Zealand 19.4 (1.45) 57
Burnett, Taylor, & Agras, 1985 Weight Loss US 41.50 (7.24) 6
Burns et al., 2011 Mental Health US 37.4 (12.2) 7
Gonzalez & Dulin, 2015 Addiction US 33.5 (6.54) 25
Hutcheson, 2013 Healthy Diet US 11.29(1.82) 16
King et al., 2013 Physical activity US 59.1 (9.2) 68
Kristjánsdóttir et al., 2013 Mental Health Norway 44.20 (11.3) 47
Miner et al., 2016 Mental Health US 45.7 (13.5) 25
Possemato et al., 2016 Mental Health US 42 (12) 10
Petersen, Sill, Lu, Mental Health US n/a 1602
Riley et al., 2008 Smoking(abstinence) US 20 (n/a) 31
Rizvi, Dimeff, Skutch, Carroll, & Linehan, 2011 Mental Health (depression) US 33.86 (10.27) 22
85
Total n 1908
86
Table 9. Chapter Three: Forest Plots of JITAI studies by outcomes of interests
Group by
outcomes
Study name Subgroup within study Hedges's g and 95% CI
addiction Gonzalez & Dulin, 2015 n/a
addiction Riley et al., 2008 n/a
addiction
diet Brookie, Mainvil, Carr, Vissers, & Conner, 2017 n/a
diet Hutcheson, 2013 n/a
diet King et al., 2013 n/a
diet
mental health Ahmedani, Crotty, Abdulhak, & Ondersma, 2015 n/a
mental health Burns et al., 2011 n/a
mental health Kristjánsdóttir et al., 2013 n/a
mental health Miner et al., 2016 n/a
mental health Possemato et al., 2016 counselor JTIAI
mental health Possemato et al., 2016 No counselor JITAI
mental health Rizvi, Dimeff, Skutch, Carroll, & Linehan, 2011 n/a
mental health
weight Burnett, Taylor, & Agras, 1985 n/a
weight Petersen, Sill, Lu, Young, & Edington, 2008 n/a
weight
Overall
-4.00 -2.00 0.00 2.00 4.00
Pre-Intervention Post-intervention
87
Table 10. Chapter 4. Participants’ Information
n percentage
Gender Male 6 33.33
Female 12 82.30
Income Less than 30,000 3 16.67
30,000 – 59,999 6 33.33
60,000 – 89,999 7 38.89
90,000 or more 2 11.11
18
Ethnicity White/Non-Hispanic 11 61.11
Hispanic 3 16.67
Black 2 11.11
Pacific Islander 0 0.00
Native American 1 5.56
Middle Eastern 1 5.56
Asian American 0 0.00
18
Depression
diagnosis
Major Depressive Disorder 12 66.67
Persistent Depressive
Disorder
1 5.56
Psychotic Depression 2 11.11
'Situational' Depression 1 5.56
Mild depressive diagnosis 2 11.11
88
Table 11. Chapter 4. Participants’ Information The count of rumination episodes and
average duration (in minutes) of each ruminative episode in week 1-2.5 (time 1), and week
2.5- (time2) of the intervention
episode (counts)
duration (in
minutes)
time1 time 2 time1 time 2
JITAI Conditions individual_1 20 13 18.75 14.23077
individual_2 5 0 15 0
individual_3 7 5 113.5714 76
individual_4 98 70 55.55 50
individual_5 45 32 110 79
individual_6 15 5 20.45 10.06
individual_7 30 21 55.9 30.7
individual_8 45 40 69 60
individual_9 78 70 79.99 72.3
Control Condition individual_10 98 88 75 52
individual_11 30 25 55 53.5
individual_12 45 40 56.5 53.5
individual_13 42 39 62.5 54.4
individual_14 19 15 89 79
individual_15 23 20 43 40.47
individual_16 32 12 31.23 28
individual_17 56 23 69.68 70
individual_18 25 20 65.3 60

89

Table 12. Within-individual fit indices in GIMME (WLC and JITAI comparison)

Χ
2
rmsea srmr nnfi cfi
value df p
JITAI
Conditions
individual_1 109.9389 78 0.01 0.07 0.07 0.96 0.97
individual_2 101.2896 77 0.03 0.06 0.07 0.96 0.97
individual_3 97.9413 79 0.07 0.07 0.09 0.97 0.98
individual_4 94.4551 79 0.11 0.05 0.06 0.98 0.99
individual_5 123.9945 77 0.00 0.07 0.05 0.95 0.97
individual_6 111.9069 80 0.02 0.06 0.07 0.95 0.97
individual_7 114.0346 79 0.36 0.02 0.06 1.00 1.00
individual_8 116.1622 76 0.02 0.07 0.06 0.96 0.97
individual_9 118.2899 74 0.00 0.09 0.05 0.94 0.96
Control
Condition
individual_10 120.4176 79 0.00 0.40 0.05 0.95 0.97
individual_11 122.5452 76 0.04 0.30 0.05 0.94 0.97
individual_12 124.6729 76 0.08 0.35 0.05 0.94 0.96
individual_13 126.8006 76 0.00 0.45 0.05 0.94 0.96
individual_14 128.9283 75 0.02 0.52 0.05 0.94 0.96
individual_15 131.0559 75 0.02 0.59 0.04 0.94 0.96
individual_16 95.5461 77 0.01 0.05 0.06 0.98 0.99
individual_17 85.0285 76 0.22 0.05 0.08 0.96 0.98
individual_18 146.7335 77 0.00 0.11 0.05 0.94 0.96

90

Note, fit indices provided by GIMME subject to the same rule of thumbs in SEM (Beltz & Gates,
2017)

91

Table 13 Sample Path Analysis within Individual 1 (WLC condition)
predictor DV beta se z p
episode episodelag 0.06 0.06 1.02 0.31
durationmins durationminslag 0.04 0.09 0.48 0.63
stateRM stateRMlag 0.06 0.06 0.98 0.33
neutral neutrallag -0.01 0.11 -0.11 0.91
negative negativelag 0.06 0.05 1.25 0.21
positive positivelag 0.13 0.08 1.76 0.08
covidst covidstlag 0.80 0.04 20.36 0.00
drinks drinkslag 0.94 0.01 70.61 0.00
stateRM episode 0.83 0.03 24.80 0.00
durationmins episode 0.51 0.08 6.44 0.00
episode neutral -0.84 0.03 -26.84 0.00
negative stateRM 0.14 0.09 1.58 0.11
positive negative -0.71 0.05 -13.21 0.00
negative episode 0.77 0.08 9.61 0.00
episode episodelag 0.22 0.09 2.43 0.02
durationmins durationminslag 0.02 0.06 0.37 0.71
stateRM stateRMlag -0.06 0.04 -1.41 0.16
neutral neutrallag -0.11 0.10 -1.11 0.27
negative negativelag 0.24 0.09 2.52 0.01
positive positivelag 0.12 0.08 1.40 0.16
covidst covidstlag 0.91 0.02 52.88 0.00
drinks drinkslag 0.86 0.03 33.58 0.00
stateRM episode 1.17 0.06 20.82 0.00
durationmins episode 0.78 0.04 19.50 0.00
episode neutral -0.40 0.08 -4.73 0.00
negative stateRM 0.30 0.09 3.23 0.00
stateRM durationmins -0.35 0.07 -4.85 0.00
positive negative -0.44 0.07 -5.87 0.00
Note: episode is the occurrence of ruminative episode at each time, state RM is the state
depressive rumination level, negative, positive, and neutral indicate participants’ emotional
state.
The affix “ -lag” like episodelage, or positivelage indicate the subsequent scores/occurrences of
the variable

92

Table 14 Sample Path Analysis within Individual 9 (JITAI condition)

predictor DV beta se z p
episode episodelag -0.03 0.09 -0.32 0.75
durationmins durationminslag 0.00 0.09 0.00 1.00
stateRM stateRMlag 0.01 0.04 0.16 0.87
neutral neutrallag -0.08 0.09 -0.94 0.35
negative negativelag 0.18 0.08 2.23 0.03
positive positivelag 0.03 0.11 0.23 0.82
covidst covidstlag 0.96 0.01 98.47 0.00
drinks drinkslag 0.99 0.00 625.20 0.00
stateRM episode 0.73 0.05 14.79 0.00
durationmins episode 0.59 0.07 7.99 0.00
episode neutral -0.64 0.07 -9.78 0.00
negative stateRM 0.03 0.12 0.26 0.79
neutral positive 0.60 0.07 8.55 0.00
stateRM neutral -0.26 0.06 -4.61 0.00
negative neutral -0.64 0.11 -5.77 0.00
Note: episode is the occurrence of ruminative episode at each time, state RM is the state
depressive rumination level, negative, positive, and neutral indicate participants’ emotional
state.
The affix “ -lag” like episodelage, or positivelage indicate the subsequent scores/occurrences of
the variable

93

Appendices B: Figures
Figure 1. Chapter Two: Interpersonal Communication Mediates Trait Rumination and
Depression Outcomes

Trait Ruminatio n
Perceived Hurtfulness
Perceived Support
Avoidance (depression
exacerbation)
Activation(depression
mitigation)
Depression Diagnosis
Depression Outcomes
Intrapersonal Process Interpersonal Process

94

Figure 2. Chapter Two: Statistical Models for Analysis in Chapter Two
Figure 2a. Theoretical Model for Parallel Mediator Analysis (Hayes, 2006, p. 13)

95

Figure 2b. Analytical Models of the Current Study, depression mitigation as DV

Trait Rumination
Perceived Hurtfulness (M
i
)
Perceived Support (M
ii
)
depression mitigation
Diagnosis
Trait
Rumination*
Diagnosis
a
i1
a
i2
a
i3
b
i
C
1
C
2
C
3
a
ii1
a
ii2
a
ii3
b
ii

96

Figure 2c. Analytical Models of the Current Study, depression exacerbation as DV

Trait Rumination
Perceived Hurtfulness
Perceived Support
depression exacerbation
Diagnosis
Trait
Rumination*
Diagnosis

97

Figure 3. Chapter Two: Results from model testing

Trait Rumination
Perceived Hurtfulness
Perceived Support
Activation(depression
mitigation)
Depression Diagnosis
/**
/**
/**
/**
/**
/**
/**
/*

Trait Rumination
Perceived Hurtfulness
Perceived Support
Avoidance (depression
exacerbation)
Depression Diagnosis
/**
/**
/**
/**
/** /**
/**

/*

98

Figure 4. Chapter Two: Interaction Effects of depression diagnosis and trait rumination on
perceiving hurtfulness

For depressed participants, trait rumination positively associated with perceived hurtfulness,
В=.41, SE=.06, 95% CI (.30,.53)
For non-depressed participants, trait rumination negatively perceived hurtfulness, В=-.35,
SE=.10, 95% CI (-.53,-.16)

99

Figure 5. Chapter Two: Interaction Effects of depression diagnosis and trait rumination on
social support

For depressed participants, trait rumination negatively associated with perceived social support,
В=.10, SE=.08, 95% CI (-.06, .26)
For non-depressed participants, , trait rumination negatively perceived social support, В=-.09,
SE=.05, 95% CI (-.18,.01)

100

Figure 6. Chapter Two: Interaction Effects of depression diagnosis and trait rumination on
depression exacerbation

For depressed participants, trait rumination positively associated with depression exacerbation,
В=.55, SE=.04, 95% CI (.48,.62)
For non-depressed participants, trait rumination negatively predicts depression exacerbation,
В=-.34, SE=.06, 95% CI (-.48,-.24)

101

Figure 7. Chapter Two: Interaction Effects of depression diagnosis and trait rumination on
depression mitigation

For depressed participants, trait rumination negatively associated with depression mitigation,
В=-.27, SE=.04, 95% CI (-.35,-.19)
For non-depressed participants, trait rumination negatively predicts depression exacerbation,
В=-.57, SE=.06, 95% CI (.45,.69)

102

Figure 8. Chapter Three: PRISMA chart for paper inclusion

103

Figure 10. between group effects (WLC vs. JITAI)

Note. the graph on the left shows the associations of different variables within the WLC,
which the one on the right shows the overall path models. “Solid lines indicate a
contemporaneous effect; dashed lines indicate a lagged effect. Black lines indicate a group-level
effect; green paths indicate an effect at the subgroup level; gray paths indicate an effect unique to
an individual. Width of line corresponds to the count of
individuals for whom the path was estimated (Lane & Gates, 2017, p. 9). ”

104

Figure 11. contemporaneous individual effects within the control condition

Note. the graph on the left shows the associations of different variables within the WLC,
which the one on the right shows the overall path models. “Solid lines indicate a
contemporaneous effect; dashed lines indicate a lagged effect. Black lines indicate a group-level
effect; green paths indicate an effect at the subgroup level; gray paths indicate an effect unique to
an individual. Width of line corresponds to the count of
individuals for whom the path was estimated (Lane & Gates, 2017, p. 9). ”

105

Figure 12. contemporaneous individual effects within JITAI conditions

106

Appendices C: JITAI Materials for Project Three
Week 1
Sometimes, it is the way we think that lead us to positive outcomes. After monitoring
your rumination habits for a while, you can probably already tell that it is the “dwelling in the
moment” way of thinking that hindered your actions. Now let’s try to change the focus of those
thoughts to something more constructive.
Sometimes we are deep in our thoughts because we want to make sure when similar things
happen the next time, we can change things to the better. Those thoughts help us solve problems.
Sometimes we are deep in our thoughts simply because we were stuck. We don’t know
why unpleasant things keep happening to us. We focus on those negative feelings, not the
problems. Those are the thoughts that debilitate us. If we are too deep into those thoughts, we are
losing our ability to learn from those unpleasant experiences.
We need to learn from our own experiences to learn what works and direct our thoughts to
the helpful end of the spectrum.
Now let’s think of an example that can happen to any of us. Think of a Tuesday afternoon,
you suddenly remember that you need to make an important phone call for work ( report the
numbers of sales, costs of travel, or customer dispute) . And you were supposed to do this for
some time. But it is just very difficult to start this call. It is a simple trigger that can affect us. We
can easily fall into the question of why we could not get it done. And we start to ask the “why”
questions, like “why couldn’t I do this? ”, “Why is it me that has to make this phone call? ”
Things could be very different if we think differently. Now let’s think of a time that
making such phone calls was easier for us. In fact, we can probably remember several times that

107

making phone calls was much easier. It could be that when we were doing this, no one else was
in the room to make us self-conscious. And we kept detailed notes on what has happened.
Comparing our previous “success” with our current situation, we may find that now the room is a
little bit distracting because we have too many colleagues around. And we were just too
concerned that we may look bad if we could not get the number straight.
It seems like to make things easier, we should make this phone call in an undisturbed,
quiet place. And we should focus on details of the numbers we need to report for the phone call
rather than over-concerning for our performance.
Indeed, how we are thinking can determine how things turn out. The difference is to try to
focus our thoughts on “how to solve the problem (problem solving)” rather than “why I am
bothered by this (negative rumination).”
Now let's try this exercise, you can write your answer down on a paper and take a picture
of it to upload if typing on a mobile phone seems difficult. Please think of a situation that you
used your ruminative thinking as a means of problem-solving. Please try to describe as detailed
as possible by writing down what you were supposed to do that seemed difficult, what you were
thinking, where you were, and how you approach the problem. And how you were feeling after
you got this done.___________________________
Now, let’s imagine a similar situation happens in the future. Please think about a specific
ruminative episode within which you SUCCESSFULLY engaged in problem-solving thinking in
the future. Please think specifically, what is the trigger that will bring you into deep thoughts,
when will this happen, at what time of the day, where you are going to be, and what are the
surrounding environments? Please also think about the specific problem-solving strategies you

108

will be using? you can also write your answer down on a paper and take a picture of it to upload
if typing on a mobile phone seems difficult.__________________
Good, now that you have accomplished the session for today, let’s try something next
time.

JITAI PROMPT
It is a good sign that you noticed your intrusive thoughts. This week, we are practicing applying
your existing skills in managing those thoughts. we must practice so that when a difficult
situation arises. Now please think of a similar situation that you have successfully got out of
those thoughts. Did you think of a way that can solve the problem? If so, could you think of a
way that can solve your current problem?
Please write this down___
Now let’s practice this for the future, please think of a similar situation in the future.
Please be specific,
When will a similar situation happen?
Where will you be when this situation happens?
Who will you be with at that time?
What will be the ways to solve this problem?
Week 2
It is important that we can be aware of the ruminative thoughts at the time when it occurs.
For example, if we try, we can easily remember when the last time when we started to worry,
could be a few days ago, several hours ago, or even just before this session. If we can start to

109

analyze things as they were in slow motion. It could be about our need to talk to the boss to take
some time off work, or simply got bothered by our chores.
And now let’s trace back, what happened before the ruminative episode? We maybe
bothered by what would happen if I were to ask the boss for a leave? What would happen if the
boss says no? That would be embarrassing, and shameful.
Now please zoom in how your body was feeling, was it a little bit agitated and tense? As
the random thoughts came into mind, did you notice something more, like the tense across the
shoulders, or unease over the stomach?
Those moments could be accompanied by lots of abstract thoughts on ourselves. “It is
just pathetic. I am such a failure. Why couldn’t I just grow up and ask the question? I was always
like this. Why are things so difficult for me?
We maybe fell victims of our thoughts for hours. And realizing how this happens could
be the key for us to do things that can help us get rid of those thoughts.
“This is just getting worse. And I am not making any progress. I should just focus on
asking the question, or what I would say. I really can’t put it off. I have to act now. “
And things get done this way.
Taking an action rather than thinking about it usually worked.
Now let try this exercise, let’s try to be very aware of the physiological (e.g. tense on the
shoulder) and psychological (shameful, feeling pathetic) reactions during our ruminative
episodes. Did you have experienced any similar episode like the example?
Now please take some time and fill in the following form:
What was this episode about _______ ( unfinished chores, conflict with a friend, etc.)

110

To your best of knowledge, when did you had such an episode ( put in a date)
____
What time was it when it happens? ____ (put in time)
How did you feel before? ____ (e.g. anxious, peaceful, agitated, etc..)
How long did this ruminative episode last? ___ (xx minutes, xxx hours)
What were you thinking about_______ ? ( e.g. why do I feel this way? Why couldn’t I
sleep? Etc)
What were the consequences of this episode______? ( e.g. couldn’t fell asleep, felt
worse)
In our story, the key is to be decided in taking actions (talk to the boss) rather than
thinking over things he could not resolve (fearing of rejected by the boss, self -shame). Did you
do similar things as that can help you get out of this ruminative episode of yours? ( e.g. taking a
sleeping pill, figuring out ways that solve the problem).
If Yes _________
If no, please think of a way now that could have effectively end this negative
episode of yours.
______
Good, remember it is always important to practice our existing strategies for the future. Please
think of a similar ruminative episode that may happen shortly ( please do this in slow-motion,
detail-by-detail).
What will this episode about _______ ( unfinished chores, conflict with a friend, etc.)
How will your intrusive thoughts start? ____ (e.g. anxious, peaceful, agitated, etc..)

111

How will you try to become aware of those thoughts? ___
How will you apply the strategy you just mentioned in this scenario? ___
JITAI prompt
It is a good sign that you noticed your intrusive thoughts. This week, we are practicing
both (1) applying existing skills in managing those thoughts and (2) focus on solving the
problem rather than mulling over the potential negative results that may not even happen . It is
important that we practice every time when difficult situation arise. Now let think about what
would do to keep you off your recent ruminative episode.
Please write this down___
Now let’s practice this for the future, please think of a similar situation for the future.
Please be specific,
When will similar situation happen?
Where will the person be when this situation happens?
Who will the person be with at that time?
How will the person solve this problem? Please be very specific in terms of the strategies
that she will use.
Week 3
It is very important to understand the trigger of our ruminative thoughts. Interpersonal
problems could very easily become a trigger. We ran into small conflicts all the time.
Sometimes, it could be offensive jokes from a co-worker who teases us: “Hey, you know, you
are making us look really bad. You work too hard for unnecessary things; I don’t like you.” It
could be a disturbing scene of an ex, asking us to clear up the apartment. Even though we kept

112

telling ourselves to let it go, details of the altercation may keep coming back to us when we were
alone. We maybe keep thinking over what happened, wishing he didn’t say what he said or did.
Common thoughts from those interpersonal conflicts maybe: “I am not popular.”
“People feel that I can't take a joke.” “I don’t want to be at work.” “Why am I making myself
feel so bad?” “Why it is so hard for me with problems with people.”
We start to feel more isolated, ended up taking sick leaves to avoid the stressor.
Usually, it is important to realize that “It is not that I could hide from this forever. “And it
really shouldn’t be about us not popular. It should be about how our colleagues communicate.
And we may want the person to know that me working hard is not to make others look bad.
Thank you for coming to the third session. This week, we are focusing on how we could
be caught up by our own ruminative thinking style. That is, we kept focusing on abstract,
extreme, and over-general ideas. This style makes it harder for us to solve problems. Lots of
time, we dwelled on those out of context thoughts (Why I am not popular. Why couldn’t I do
this). Things can become easier for us when we use a specific and concrete style of thinking (our
co-worker was out of the line and should be more considerate, and we should just let the person
know about this).
Interpersonal problems (small conflicts, disagreements) can be triggers of ruminative
thinking. If you are to be in such a situation, what would you do?
___ (fill in answers)
We must visualize things in detail for the future. Now let’s think of a scenario in the
future that you can apply this strategy you planned.
Who is/are the person that will have a small interpersonal conflict with you? ___

113

What time will it happen? ____ (put in time)
What will be the conflict about?
How will you feel about it? ____ (e.g. anxious, peaceful, agitated, etc..)
How will you use the strategy ( according to this specific situation you just envisioned)?
JITAI prompt
Good that you are monitoring your intrusive thoughts. Let’s keep practice: (1) being
aware of major rumination triggers ( like our recent conflict), (2) applying existing skills in
managing those thoughts (and (3) focus on solving the problem rather than mulling over the
potential negative results that may not even happen. It is important that we practice every time
when difficult situation arise. Now let think about what you can do to change your current
ruminative episode? What will you do in the future if similar situation happens
Please write this down___
Now let’s practice this for the future, please think of a similar situation in the future.
Please be specific,
When will similar situation happen?
Where will the person be when this situation happens?
Who will the person be with at that time?
How will the person solve this problem? Please be very specific in terms of the strategies that
you will use.

114

Appendices D: JITAI Materials for Project Three
Week 1
It is just a Tuesday afternoon, like any Tuesday afternoon these days. Rudy was still
sitting in her home office. Suddenly, she remembered something. Crap, she forgot. She was
supposed to call the manager of her sales division a week ago. Damn, why is this happening to
me again? Last week, her division received a very angry complaint from a customer, yelling at
their “big lies” for not returning an open -box air mattress. The operator on the line was very
courteous and explained very carefully the return policy and the required restocking fees for
items like an air mattress. But things apparently didn’t go so well. The phone call became a
nightmare of a yelling match, with the customer demanding to speak to the “highest person in
charge.” And yes, as the head of the division, Rudy was the person who had to brief this to the
“boss”.
It is not that her supervisor, Tina, is someone who is not understanding. But why she has
to be very detailed in the process of the briefing. Who answered the phone call? What happened
at first, then what happened? The whole thing was a mess. Why does this have to happen to my
division again, the third time in two weeks? What would Tina say this time? She probably still
remembers last time; I was almost out of words when I had to brief her. And this thing happened
a week ago, she probably would wonder what went wrong with me that I had to wait for a week.
Sitting there for a while and dwelling in her own mind of concerns, Rudy still couldn’t
make that phone call. She knew she was deeply in her thoughts simply because she was stuck.
She could not figure out why unpleasant things keep happening to her. And how would she
become so upset over a procedural phone call? Those things stressed her out and debilitated her.

115

God knows for how long, Rudy suddenly saw a picture of her and her colleagues during
last year’s Christmas party. Yes, it was Christmas time, and she had to report an even greater
amount of complaints to Tina. How did I do that? That wasn’t very hard. She was so focused
back then because Christmas was such a busy time. And she has to keep track of all things. And
yes, she was in her office with those who were involved in the customer complaint. Her
colleagues were so supportive. They are of course very supportive right now, except they are just
not in the room. And they kept detailed notes on what has happened. That wasn’t so difficult.
Rudy smiled, why not bring in those who are involved in this situation into this phone she was
about to make. And yes, she needs to have her “notes” ready. She always kept notes anyway, it
can definitely make her feel better though. She just needs to dig it out of her nightstand in her
bedroom. And now, Rudy feels ready to make that call.
Now let try this exercise, you can write your answer down on a paper and take a picture
of it to upload if typing on a mobile phone seems difficult. Please think of another situation that
Rudy can use her ruminative thinking as a means of problem-solving. Please try to describe as
detailed as possible by writing down what she was supposed to do that seemed difficult, what she
was thinking, where she was, and how she approaches the problem. And how she was feeling
after she got this done.
Now, let’s imagine a similar situation happen in Rudy’s future. Please think about a
specific ruminative episode within which Rudy SUCCESSFULLY engaged in problem-solving
thinking in the future. Please think specifically, what is the trigger that will bring her into deep
thoughts, when will this happen, at what time of the day, where she was going to be, and what

116

are the surrounding environments? Please also think about the specific problem-solving
strategies she will be using? You can also write your answer down on a paper and take a picture
of it to upload if typing on a mobile phone seems difficult.__________
Good, now that you have accomplished the session for today, let’s try something next
time._______
JITAI prompt
It is a good sign that you noticed your intrusive thoughts. This week, we are practicing
applying existing skills in managing those thoughts, like Rudy did in our story. It is important
that we practice every time when difficult situation arise. Now let think about what Rudy would
do if she is in your situation. In what way can she solve the problem? If so, could you think of a
way that she can solve a similar problem?
Please write this down___
Now let’s practice this for the future, please think of a similar situation for Rudy in the
future. Please be specific,
When will similar situation happen?
Where will Rudy be when this situation happens?
Who will Rudy be with at that time?
How will she solve this problem? Please be very specific in terms of the strategies that
she will use.
Week 2
The last time when Jimmy started to worry was a few days ago. He can still remember it
as things were in slow motion. He needed to talk to his boss to take some time off work, to take a

117

small trip with his younger brother. He had been working very hard during the final sales season
and needed to relax. He saw his boss passing by and casually nodded at him. But he didn’t get to
ask him for leave. Instead, he started to worry. What would happen if he asked?
“Was the boss looking at me as if I was stupid? What would happen if I am to ask him for
a leave? He probably would say no. That would be embarrassing, and shameful. “
Slowly Jimmy started to feel a little bit agitated and tense. As the random thoughts came
into his mind, he noticed that he was tense across the shoulders. Those thoughts were not very
helpful and even debilitating.
Jimmy started to feel even more stupid. “It is just pathetic. I am such a failure. Why
couldn’t I just grow up and ask the question? I was always like this. Why things are so difficult
for me? ” He started feeling worse. And it got harder for him to get started. He feels down and
lost motivation.
An hour has passed. The boss passed by several times.
“This is just getting worse. And I am not making any progress. I should just focus on
asking the question, or what I would say. I really can’t put it off. I have to ask now. My brother
is waiting for me to book the trip. “
And that’s what Jimmy did. He asked the question, and of course, the boss said yes.
It seems that taking an action rather than thinking about it usually worked. Jimmy started
to feel a huge relief. The tension was lifted.
Now let try this exercise, in our story, Jimmy was very aware of the physiological (e.g.
tense on the shoulder) and psychological (shameful, feeling pathetic) reactions during his

118

ruminative episodes. Did you have experienced any similar episode as Jimmy had in the past
week?
Now please take some time and fill in the following form:
What was this episode about _______ ( unfinished chores, conflict with a friend, etc.)
To your best of knowledge, when did you had such an episode ( put in a date)
____
What time was it when it happens? ____ (put in time)
How did you feel before? ____ (e.g. anxious, peaceful, agitated, etc..)
How long did this ruminative episode last? ___ (xx minutes, xxx hours)
What were you thinking about_______ ? ( e.g. why do I feel this way? Why couldn’t I
sleep? Etc)
What were the consequences of this episode______? ( e.g. couldn’t fell asleep, felt worse)
In our story, Jimmy decided to take action (talk to the boss) rather than thinking over
things he could not resolve (fearing of rejected by the boss, self -shame). Did you do
similar things as Jimmy that can help you get out of this ruminative episode of yours? (
e.g. taking a sleeping pill, figuring out ways that solve the problem).
If Yes _________
If no, please think of a way now that could have effectively end this negative
episode of yours.
______

119

Good, remember it is always important to practice our existing strategies for the future.
Please think of a similar ruminative episode that may happen soon ( please do this in slow-
motion, detail-by-detail).

What will this episode about _______ ( unfinished chores, conflict with a friend, etc.)
How will your intrusive thoughts start? ____ (e.g. anxious, peaceful, agitated, etc..)
How will you try to become aware of those thoughts? ___
How will you apply the strategy you just mentioned in this scenario? ___
JITAI prompt
It is a good sign that you noticed your intrusive thoughts. This week, we are practicing
both (1) applying existing skills in managing those thoughts (like Rudy did in our story) and (2)
focus on solving the problem rather than mulling over the potential negative results that may not
even happen ( like Jimmy did in our story series) . It is important that we practice every time
when difficult situation arise. Now let think about what our story characters would do if they are
in your situation. In what way can one of them solve the problem? If so, could you think of a
way that she can solve a similar problem?
Please write this down___
Now let’s practice this for the future, please think of a similar situation for Rudy or Jimmy
in the future. Please be specific,
When will similar situation happen?
Where will the person be when this situation happens?
Who will the person be with at that time?

120

How will the person solve this problem? Please be very specific in terms of the strategies
that the character will use.
Week 3
Eddie was always the quiet one. He always thinks himself not so quick-minded, especially
when this guy, Chris, always give gibes at work. He usually couldn’t come up words when Chris
half-mindedly laughed at his quietness or mocked his meticulousness. Yes, he likes keeping
things in order at work. He’d rather stay late to clean things like and put them back where they
belong. Last week, Chris really crossed the line, he laughed: “Hey, you know, you are making us
look really bad. You work too hard for unnecessary things, I don’t like you.” Eddie, this time,
very loudly responded: “Well, I didn’t ask you to like me.” Chris was still shameless: “Hey, why
can’t you just take a joke? ”After this, Eddie walked out of the office with a couple of co-
workers looked at him in disbelief.
This was one month ago. But Eddie finds himself stewing over it. Even though he kept
telling himself to let it go, details of the altercation kept coming back to him when he was alone.
He kept thinking over what happened, wishing he didn’t say what he said or did.
Eddie can’t help. “I am not popular.” “People feel that I can't take a joke.” “I don’t want
to be at work.” “Why am I making myself feel so bad?” “Why it is so hard for me with problems
with people.”
Things went even worse for Eddie. He felt more isolated, ended up taking a week off sick.
“It is not that I could hide from this forever. “ Eddie knew that he still needed this job.
And it really shouldn’t be about him not popular. It should be about how Chris communicates his

121

discomfort with him. Chris should know that I am not making him look bad. It is just my habit of
keeping things in good order. If things come up next. I’ll just let Chris know that.
Back-to-work Monday ended with a full day of sales ops. Eddie started to feel like
himself. He has always been very good at “real” problems, rather than “people” problems. Now
everyone needs him for his detailed spreadsheets of product serial numbers and matching
customer notes. No one seems to remember his thing with Chris. Eddie somehow managed to
smile while handing out a copy of the customer information to Chris: “You know, I just can’t
stomach messy spreadsheets.”
Thank you for coming to the third session. In our story, Eddie was caught up by his own
ruminative thinking style. That is, he kept focusing on abstract, extreme, and over-general ideas.
This style makes it harder for Eddie to solve problems. Lots of time, he dwelled on those out of
context thoughts (Why I am not popular. Why couldn’t I do this). Things become easier for him
when he uses a specific and concrete style of thinking (Chris was out of the line and should be
more considerate, and he should just let Chris know about this).
Interpersonal problems (small conflicts, disagreements) can be triggers of ruminative
thinking. If you are to be in Eddie's situation, what would you do?
___ (fill in answers)
It is important that we visualize things in detail for the future. Now let’s think of a
scenario in the future that you can apply this strategy you planned.
Who is/are the person that will have a small interpersonal conflict with you? ___
What time will it happen? ____ (put in time)
What will be the conflict about?

122

How will you feel about it? ____ (e.g. anxious, peaceful, agitated, etc..)
How will you use the strategy (according to this specific situation you just envisioned)?
JITAI prompt
Good that you are monitoring your intrusive thoughts. Let’s keep practice: (1) being
aware of major rumination triggers ( like Eddie’s recent conflict), (2) applying existing skills in
managing those thoughts (like Rudy did in our story) and (3) focus on solving the problem rather
than mulling over the potential negative results that may not even happen ( like Jimmy did in our
story series) . It is important that we practice every time when difficult situation arise. Now let
think about what our story characters would do if they are in your current situation. In what way
can one of them solve the problem? If so, could you think of a way that she can solve a similar
problem?
Please write this down___
Now let’s practice this for the future, please think of a similar situation for Rudy, Jimmy,
or Eddie in the future. Please be specific,
When will similar situation happen?
Where will the person be when this situation happens?
Who will the person be with at that time?
How will the person solve this problem? Please be very specific in terms of the strategies
the story character will use.

Abstract (if available)

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

The power of social media narratives in raising mental health awareness for anti-stigma campaigns

PDF

Using observed peer discussions to understand adolescent depressive symptoms and interpersonal interactions

PDF

Braver together: exploring the communicative and psychological experiences of LGBTQ youth and families

PDF

School-based Health Centers 2.0: Using social media to increase utilization of UMMA Wellness Center Services

PDF

Service to civilian: life after the Second Gulf War

PDF

Building personal wellness communities: meaningful play in the everyday life of a network society

PDF

Social norms intervention to promote help-seeking with depressed Asian and European Americans: A pilot study

PDF

Using virtual environments to unobtrusively measure real-life risk-taking: findings and implications for health communication interventions

PDF

Applications of mindfulness toward promoting psychosocial health in diverse communities

PDF

Gamification + HCI + CMC: effects of persuasive video games on consumers’ mental and physical health

PDF

Addressing federal pain research priorities: drug policy, pain mechanisms, and integrative treatment

PDF

Personalized normative feedback applied to undergraduates with problem drinking: a comparison with psychoeducation and an examination of cognitive-affective change mechanisms via the articulated ...

PDF

Voice, participation & technology in India: communication technology for development in the mobile era

PDF

Predictors of mHealth engagement in a mindful eating intervention for Type 2 diabetes

PDF

A mobile app for anxiety: an examination of efficacy and user perceptions

PDF

The influence of a humorous intervention on parent-child sexual communication

PDF

Mindfulness connection (MBSR) program integration into high school curriculum

PDF

Internet-delivered eating disorder prevention: a randomized controlled trial of dissonance-based and cognitive-behavioral treatments

PDF

The link between maternal depression and adolescent daughters' risk behavior: the mediating and moderating role of family

PDF

Climate change, air pollution and health: how policy can induce adaptation

Asset Metadata

Creator Wang, Liyuan (author)

Core Title Towards designing mental health interventions: integrating interpersonal communication and technology

School Annenberg School for Communication

Degree Doctor of Philosophy

Degree Program Communication

Degree Conferral Date 2021-12

Publication Date 12/15/2021

Defense Date 08/12/2021

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

Tag CBT,cognitive behavioral therapy,depressive rumination,JITAI,just-in-time adaptive intervention,mental health,mobile phones,OAI-PMH Harvest

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Miller, Lynn C. (committee chair), Leventhal, Adam (committee member), Valente, Thomas (committee member), Young, Lindsay E. (committee member)

Creator Email liyuan.wang1115@gmail.com,liyuanwa@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC18652766

Unique identifier UC18652766

Legacy Identifier etd-WangLiyuan-10302

Document Type Dissertation

Format application/pdf (imt)

Rights Wang, Liyuan

Type texts

Source 20211221-wayne-usctheses-batch-905-nissen (batch), University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Repository Email cisadmin@lib.usc.edu