University of Southern California Dissertations and Theses

Prospective associations of stress, compulsive internet use, and posttraumatic growth among emerging adults

(USC Thesis Other)

Prospective associations of stress, compulsive internet use, and posttraumatic growth among emerging adults

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content Copyright 2024 Sheila Yu
Prospective Associations of Stress, Compulsive Internet Use, and Posttraumatic Growth Among
Emerging Adults
By
Sheila Yu
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
(PREVENTIVE MEDICINE - HEALTH BEHAVIOR)
May 2024

ii
Dedication
To my parents. Thank you for your endless and unwavering love and support.

iii
Acknowledgements
This academic journey would not have been possible without the support of my
incredible mentors, colleagues, and friends. Thank you, Dr. Steve Sussman, for being my
guiding light since 2015. Despite my numerous shortcomings and moments of dropping the
ball—several at that—you never stopped believing in me. I truly believe that your support was
one of the most critical factors that resulted in the successful completion of my dissertation and
passing of the defense. Words cannot express the extent of my gratitude, appreciation, and
respect for you. Thank you, Dr. Jennifer Unger, for being a huge source of hope and support for
me throughout the years. Your timely words of encouragement served as a lifeline when things
seemed too daunting and motivated me to keep going. Thank you, Dr. Lawrence Palinkas, for
joining my committee at such short notice, providing valuable insights along the way, and being
so supportive of my endeavors. Thank you to my former committee members, for your guidance
and support as well. Thank you, Dr. Em Arpawong, for patiently guiding me through the theory
of posttraumatic growth and ensuring that I fully understood and grasped the intricacies of PTG.
Thank you, Dr. Jimi Huh, for patiently advising me on methodology and graciously accepting
my “statistically-challenged” self. I fondly look back at our MyQuit lab with Christian and the
others. Thank you, Dr. Kimberly Miller, for being sunshine and rainbows in human form.
Whenever I interacted with you, I felt so comforted by your warm kindness and encouragement.
For all mentors, your levels of patience and understanding were unrivaled—I truly appreciate
you all so much.
Thank you, Leah, for supporting me since I started the program. I think you were the
most familiar with all of the health doozies I went through—thank you for always reaching out to
see how I was doing, sympathizing with my pain, and encouraging me to push forward.
To my former and current colleagues, thank you for your streams of encouragement and
fully understanding the challenges that I experienced, having gone through similar experiences
yourselves. Thank you, Jen, for being by my side since day one. I am so grateful for our
friendship, for your incredible love and support through all the ups and downs, and for expertly
knowing exactly what kind of sweet treats I love. Thank you, Christine and Shirlene, for setting
up our last writing retreat, which was a much needed time of group productivity and fun. Also,
along with Trevor, thank you for giving me crucial feedback on my first practice run: I had major
doubts in the beginning but you three eased my mind and allowed me to broaden my perspective.
Christine, your bright personality and positive vibes always refresh and rejuvenate me, thank you
for being so incredibly amazing. Shirlene, bonding over shared experiences has made us grow
closer, and I am so thankful for you in my life. Your life updates give me life, haha, I will
forever be a fan. Thank you, Artur and Ellen, my longtime lab mates, for being patient with me
and helping me in so many ways—you are both so awesome. Thank you, Kellie, my strongest
ally, our camaraderie knows no bounds. Thank you for always understanding exactly what I’m
going through, saying the right things at the right time, and taking care of me. I appreciate you so
very much. Thank you, Angelica, for wholly supporting me despite the busyness of your
schedule and the distance between us. You have always had my back and rooted for me, and for
that I am so grateful. Thank you, Michelle, for lifting my spirits with your perfect memes, shared
love for kpop and kdramas and just being your cute and awesome self.
To my friends, near and far, thank you for being the best social support circle in the
world—sob—I love you all so dearly. Thank you, Edith, for loving and accepting me as I am and
for being my beautiful twin—but especially thank you for encouraging me with your dear words:

iv
“Congration, you done it.” You are the bestest. Thank you, Alice, for being my confidant and my
safe space, where I always feel so loved and never judged. You were my anchor during some
dark times and allowed me to get to where I am now, accepting of myself as best as I can. You
mean the world to me. Thank you, Xi Xi, for your immense love and support, for pushing me
forward and showering me with encouragement, and just being an absolute Queen. A sweet
friendship refreshes the soul (Proverbs 27:9). Thank you, Shanese, for our long comforting talks
and picking up right where we left off, no matter how much time has passed. Thank you, Jihyae
unni, for being there for me and for being so much fun to hang out with whenever I get to see
you. Thank you, Christina, for being such a lovely friend, always bringing a smile to my face and
making me feel so loved. I cherish our times together and appreciate you so much. Thank you,
Jules, for being a loving older sister to me. Thank you for grounding me in the truth, gently
redirecting my focus to Jesus each time, and genuinely caring for my physical and spiritual
wellbeing. Thank you, Thais, for being my hype woman, and a totally hot mama at that. I most
appreciate our instant reconnections, which make me feel seen and known. Thank you, Andrew,
for being a loyal brother and friend by my side all these years. Thank you for checking in on me,
even during the times I pushed you and everyone away. Especially during the times of my lowest
lows, you were there for me—for that I will be forever grateful. To my high school frans: Thank
you, Janet, for your constant love and support—having you by my side has been crucial for my
mental health. Thank you for recharging my social battery each and every time and nudging me
along with your warm love. Thank you, Knoelle, for being my biggest supporter. This has been a
long time coming, and your support throughout all these years have meant so much to me. Thank
you, Juri, for always checking in on me and also understanding the challenges of being in school
for so long—love that we are growing from these experiences as we navigate life together.
And of course, Mom and Dad, I love you so much. Sorry for being a whining brat
sometimes—sorry for being an insufferable ball of temper tantrums other times. How blessed I
am to have you as my parents… I will never take that for granted. Thank you for all of your
sacrifices, great and small, seen and unseen—thank you for reminding me that the light at the
end of the tunnel is indeed reachable. Thank you for everything, truly.

v
TABLE OF CONTENTS
Dedication.................................................................................................................................................ii
Acknowledgements..................................................................................................................................iii
List of Tables ........................................................................................................................................... vi
List of Figures.........................................................................................................................................vii
List of Abbreviations .............................................................................................................................viii
Abstract.................................................................................................................................................... ix
Introduction............................................................................................................................................... 1
Chapter 1: Background & Significance .................................................................................................... 6
Emerging Adulthood................................................................................................................ 6
Compulsive Internet Use ......................................................................................................... 7
Post Traumatic Growth.......................................................................................................... 11
Subgroup Differences............................................................................................................ 16
Dissertation Overview ........................................................................................................... 19
Chapter 2: Study 1................................................................................................................................... 22
Introduction............................................................................................................................ 22
Methods ................................................................................................................................. 25
Results.................................................................................................................................... 31
Discussion.............................................................................................................................. 39
Chapter 3: Study 2................................................................................................................................... 45
Introduction............................................................................................................................ 45
Methods ................................................................................................................................. 48
Results.................................................................................................................................... 54
Discussion.............................................................................................................................. 63
Chapter 4: Conclusion............................................................................................................................. 67
References............................................................................................................................................... 69
Appendices.............................................................................................................................................. 79
Appendix A: Tables............................................................................................................... 79
Appendix B: Figures.............................................................................................................. 85

vi
List of Tables
Table 1 Study 1 sample demographics (n=882). ................................................................. 32
Table 2 Sample means of relevant variables by data collection year. ............................. 32
Table 3 Frequencies of stressful life events (SLEs) reported in Years 2-5. .................... 33
Table 4 Frequencies of most life-altering events (MLAEs) reported in Years 2-5. ....... 33
Table 5 Unconditional linear growth curve model of PTG over time (n=882). ............. 34
Table 6 Unconditional linear growth curve model of CIU over time (n=882). .............. 35
Table 7 Conditional linear growth curve model of PTG with covariates and predictor
variables (n=882). ..................................................................................................... 38
Table 8 Study 2 sample demographics (n=410). ................................................................. 54
Table 9 Sample means of relevant variables by data collection year. ............................. 55
Table 10 Longitudinal growth curve model of CIU over time (n=410). ........................... 56
Table 11 Moderation model of PTG on SLEs and CIU (n=410). ....................................... 57
Table 12 Moderation model of PTG on SLEs and internet browsing (n=410). ............... 59
Table 13 Moderation model of PTG on SLEs and social networking (n=410). ............... 60
Table 14 Moderation model of PTG on SLEs and online shopping (n=410). .................. 62

vii
List of Figures
Figure 1 Study 1 sample exclusion criteria. ............................................................................ 30
Figure 2 CIU trajectory over time. .......................................................................................... 35
Figure 3 Study 2 sample exclusion criteria. ............................................................................ 52
Figure 4 Probing the significant interaction of time and between-school-SLEs revealed
diverging trajectories. ................................................................................................ 56
Figure 5 The moderating effect of PTG on the association of SLEs and CIU. ................. 58
Figure 6 The moderating effect of PTG on the association of SLEs and self-perceived
addiction to internet browsing. ................................................................................. 59
Figure 7 The moderating effect of PTG on the association of SLEs and self-perceived
addiction to social networking. ................................................................................ 61
Figure 8 The moderating effect of PTG on the association of SLEs and self-perceived
addiction to online shopping. ................................................................................... 62

viii
List of Abbreviations
CIU Compulsive internet use
CHS Continuation high schools
DSM Diagnostic and Statistical Manual of Mental Disorders
EA Emerging adult
IA Internet addiction
IRB Institutional Review Board
LCHD Life Course Health Development model
MI Motivational interviewing
PTG Posttraumatic growth
PTGI Posttraumatic growth inventory
PTSD Posttraumatic stress disorder
SLE Stressful life event
TND Project Towards No Drug Abuse

ix
Abstract
Background: Emerging adults (EAs) transitioning into adulthood may respond to stressful life
events (SLEs) by developing a harmful coping mechanism, an example being compulsive
internet use (CIU). CIU may serve as a means of escapism for EAs who are overwhelmed by
stress. A negative emotional response to an SLE may be trauma, ranging from acute (trauma
from one experience) to complex (trauma from repeated and/or prolonged experiences).
Increasing scientific literature supports the capability to experience positive psychological
change (e.g., posttraumatic growth, PTG) as a result of experiencing the SLEs. This dissertation
explores the potential moderating effect that PTG has on the association of SLEs and CIU.
Methods: Project Towards No Drug Abuse (TND) is a drug abuse prevention curriculum
targeted towards EAs in alternative/continuation high schools, focusing on developing skills,
motivation factors, and decision-making competence to avoid behaviors harmful to health. In
2008, at baseline, 1,676 participants (mean age 16 years, SD=0.93, range 14-21 years) completed
a pre-test survey. There were six waves of follow-up data collection from 2009 to 2015.
Results: PTG was found to be significantly negatively associated with CIU and positively
associated with identifying as female, identifying as Hispanic, and having participated in any
component of Project TND. Although not statistically significant, moderating effects of PTG on
SLEs and CIU were observed: with increasing SLEs, those who endorsed positive PTG reported
decreasing CIU, while those who endorsed negative PTG reported increasing CIU, suggesting a
protective factor of PTG against CIU engagement.
Conclusions: Improving EA health outcomes is critical, especially in identifying harmful health
behaviors that may continue into adulthood. CIU has been found to be prevalent among EAs and
harmful to their health and wellbeing. Literature is currently growing on the notion that PTG

x
could potentially have a buffering effect where higher PTG endorsement could be associated
with health behavior benefits, such as increased ability to not participate in harmful coping
behaviors. After experiencing stressful trauma, channeling this buffering effect of PTG against
harmful coping mechanisms may help in developing and supplementing stress management
skills that would enable EAs to reach a higher level of both life functioning and psychological
wellbeing for a healthier transition to adulthood.

1
Introduction
Compared to earlier generations at their age, 18- to 29-year-olds today face more societal
pressures partly stemming from social media: comparing themselves with others and to
unrealistic expectations, having low body image/self-esteem, being pressured to achieve success,
being competitive and unique from others, academically performing well while managing a
heavy educational workload, digesting information overload efficiently, approaching
perfectionism (Meyers, 2018), and undergoing the overall transition to adulthood (e.g., changing
identities, regulating emotions, being independent, maintaining relationships, pursuing and
achieving goals, and missing daily structure in life; Murray et al., 2020), all of which are
stressful and have the potential to impact health negatively (Hanna et al., 2018; Matud et al.,
2020; Murray et al., 2020). Researchers have termed this group as emerging adults (EAs) (Arnett,
2000; Frye & Liem, 2011; Rogers & Maytan, 2012; Arnett et al., 2014; Meyers, 2018; Murray et
al., 2020). The concept of emerging adulthood applies to areas where extended development
after adolescence exists, i.e., emerging adulthood is less relevant in countries where youth jump
directly into adult roles (Arnett, 2000).
EAs have different stressors than those who have more established identities (adults) or
those who are dependent upon others for resources and direction in life (children or adolescents)
(Frye & Liem, 2011; Rogers & Maytan, 2012; Meyers, 2018). Amidst a unique developmental
stage in life, EAs are questioning their place in the world, as well as their worldviews, goals, and
responsibilities (Arnett, 2000). EAs transitioning into adulthood may feel overwhelmed due to
the instability of this developmental period, which influences how they respond and adapt to
stressful life events (SLEs). The resulting psychological stress from SLEs negatively affects their
mental health and wellbeing (Matud et al., 2020; Murray et al., 2020). According to the

2
American Psychological Association (APA), the prevalence of American EAs who have
experienced mental health disorders has significantly increased over the last ten years (APA,
2019). Twenge et al. (2019) suggested a “generational shift” in the occurrence of mood disorders
among EAs in the U.S., according to data from the National Survey on Drug Use and Health
(NSDUH): Of the 400,000 EAs (aged 18 to 25) surveyed, nearly 13% reported major depressive
disorder symptoms in the last 12 months in 2017 compared to approximately 8% in 2009; also,
about 13% reported having experienced psychological distress in the last 30 days in 2017,
compared to about 8% in 2008. Cultural trends having a sizable effect on mood disorders may be
due in part to rising use of electronic social interaction and digital media, and “results suggest a
need for more research to understand how digital communication versus face-to-face social
interaction influences mood disorders” (Twenge et al., 2019).
Stress is a prominent part of our lives: not only is experiencing a highly stressful event
common for people of all ages and from all cultural contexts, but also most of the general
population have been or will be exposed to some sort of traumatic experience at some point in
their life (Powers et al., 2013; Berger, 2015). While stress is not harmful in some cases, trauma is
harmful in nearly all cases. An emotional response to an SLE may be trauma, ranging from acute
trauma to complex (National Network for Youth, 2022):
• Acute trauma: trauma from one stressful/traumatic experience
• Chronic/repetitive trauma: (e.g., domestic violence or abuse)
• Complex trauma: from several varying traumatic events, commonly of an intrusive
and interpersonal nature (e.g., family violence)
• Complex developmental trauma: repeated trauma experienced as a youth
• Vicarious trauma: from empathetic exposure to another’s trauma

3
• Historical/intergenerational trauma: from events that get passed on through
time/generations (e.g., war, slavery, colonization, loss of culture, racism, older
generation’s trauma affecting younger generation.)
Trauma and stress in turn creates the need for coping, and there are several ways of
coping among EAs that lead to different outcomes or processes (Berger, 2015). EAs who find it
more difficult to manage the stress are more likely to partake in harmful coping mechanisms to
escape or avoid the stressor (Edgerton, Keough, & Roberts, 2018), such as engaging in substance
use or addictive behaviors (e.g., gambling, Internet, shopping, love, sex, exercise, and work)
(Sussman, 2017). Such risky behaviors among EAs may be tolerated—or even promoted—by
society, as a rite of passage of sorts (Sussman & Arnett, 2014).
An example of a harmful coping mechanism is compulsive internet use (CIU). CIU may
serve as a means of escapism for EAs (Fernandes et al., 2020) who are overwhelmed by trauma
or stress, and the reality that the internet is readily accessible for this technologically savvy group
supports the possibility of high prevalence rates of CIU among EAs (Tang et al., 2017).
Development of CIU is especially relevant to EAs as they have been found to access the internet
more than any other age group, which leads to a higher risk of experiencing impaired daily
functioning, strained interpersonal relationships, and decreased emotional well-being (Anderson
et al., 2017).
Besides the development of harmful coping behaviors following adversity, another
common outcome from experiencing trauma is post-traumatic stress disorder (PTSD) in which
an individual faces intense and persistent recollections of the traumatic event, along with several
crippling side effects that interfere with daily life and can last indefinitely (Zoellner & Maercker,
2006; APA, 2019). Although negative stress outcomes are well-known, increasing scientific

4
literature supports the capability to experience positive psychological change as a result of
experiencing the trauma. This concept, known as posttraumatic growth (PTG), posits that
individuals do not only recover from trauma but also grow from the efforts of coping with the
trauma and are able to thrive afterwards (Tedeschi & Calhoun, 1996; Tedeschi & Calhoun, 2004).
While reporting positive changes does not negate the adverse impacts that individuals typically
experience following SLEs, those positive changes seem to be part of the overall experience of
coping with and adjusting to SLEs (Park & Fenster, 2004).
Tedeschi & Calhoun (1996) operationalized trauma as an extremely stressful or
challenging event that overturns individuals’ pre-trauma view of the world—this implies that
traumatic events do not necessarily have to be life-threatening or narrowly defined, as by
Diagnostic and Statistical Manual of Mental Disorders (DSM)-5 criteria, as long as it is selfperceived to be life-altering (Tedeschi & Calhoun, 1996; Tedeschi et al., 2018, p.4; Tedeschi &
Calhoun, 2004). Prior studies had rationale for measuring other SLEs besides those that were
solely life-threatening or traumatic because SLEs are still impactful and important to health
outcomes (Roth & Cohen, 1986; Park et al., 1996; Park, 1998) and have used those SLEs as a
proxy measure of stress. When developing a measure of stress-related growth, Park (1996; 1998)
established that the cumulative number of SLEs represent the overall stress being experienced by
an individual. In turn, the individual may experience positive changes in life resulting from
undergoing one or more SLEs, and the positive changes may include PTG where the individual
develops psychological functioning levels that exceed the level from before the SLEs. This
dissertation will focus on cumulative stress and refer to stress as the full range of stressors, from
low to traumatic levels.

5
Purpose
As emerging adults (EAs) face numerous stressful life events (SLEs) as they transition to
adulthood (Wood et al., 2018), they must deal with the stress, either in healthy or unhealthy ways
(Murray et al, 2020). This dissertation will focus on EAs who choose to cope with their stress by
using the internet to escape from the stressor. However, some may exceed normative internet use
to excessive amounts, to the point of being considered a behavioral addiction (criteria for
behavioral addictions will be expanded upon in a subsequent section, p.7-8). This is significant
because the physical, mental, and emotional risk factors associated with compulsive internet use
(CIU) negatively impact health and EAs are most likely to continue these risky behaviors onto
adulthood (Arnett et al., 2014; Anderson et al., 2017).
To address the public health threat that CIU poses, this dissertation proposes that
posttraumatic growth (PTG) may moderate the association between SLEs and CIU among EAs
who have undergone PTG-eliciting stress. If PTG helps EAs to better cope with stress, rather
than turn to harmful coping mechanism (in this case, excessive internet use) to cope, then
cultivating PTG could be effective in preventing and/or ameliorating CIU.
Emerging adulthood (transitioning from adolescence to adulthood), CIU (a harmful
coping mechanism to deal with the stress associated with the transition), and PTG (a radical shift
in view/sense of self after experiencing extreme stress) will be further defined and expanded
upon in the subsequent section to explore the proposed moderation relationship.

6
Chapter 1: Background & Significance
Emerging Adulthood
According to the Life Course Health Development (LCHD) model (Halfon et al., 2018),
there are distinct stages in life: preconception, early childhood, middle childhood, adolescence,
emerging adulthood, and adulthood, which includes young, middle-aged, and older adulthood
(Steger et al., 2009). At each stage, numerous factors interact simultaneously at multiple levels to
influence an individual’s health and development trajectory. The emerging adulthood stage was
added because it could be neither categorized as adolescence (dependent on others) nor
adulthood (taking on responsibilities), as emerging adults (EAs) are not yet bound by the
expectations of typical adult social roles and norms (Frye & Liem, 2011; Wood et al., 2018).
Emerging adulthood may be the only stage in the LCHD model where an individual “experiences
such dynamic and complex changes on the personal, social, emotional, neuroanatomical, and
development levels,” (Wood et al., 2018, p.124), which implies that EAs undergo changes that
may only occur during this volatile stage and not in other stages. For example, EAs have the
freedom to explore a variety of possible life directions, which include relationships, careers, and
worldviews. Therefore, emerging adulthood could be considered a critical time point when,
generally by the end of their late twenties, EAs have made choices in life that have an enduring
effect on the direction of the rest of their lives (Arnett, 2000; Wood et al., 2018).
EAs do not follow the same developmental paths that are characteristically followed by
youth or older adults, as various factors, life transitions, and pivotal events may affect EAs
differently than those in another stage (Frye & Liem, 2011; Rogers & Maytan, 2012; Meyers,
2018). A life transition is a turning point in one’s life that could be biological (e.g., menarche,
menopause) or social (e.g., graduating from school, securing a job, retiring from a job), which

7
are in turn shaped by the individual’s interaction with one’s environment. Emerging adulthood is
also characterized by EAs exploring their identity as they transition to adulthood, which could be
an exciting yet daunting and confusing period to EAs.
These numerous life transitions and choices, along with the awareness of being in an inbetween state, have increased the risk of psychological distress, mental illness, and adoption of
harmful coping mechanisms (in this case, compulsive internet use) among EAs (Rogers &
Maytan, 2012). Although internet use and vulnerability to internet-related addictions may affect
individuals of all ages, EAs are particularly prone as they are constantly exposed to the internet
for work, school, and daily life (Tang et al., 2017).
Behavioral Addictions: Compulsive Internet Use
Recently, there has been general acknowledgement of the potential for certain behaviors
to develop into behavioral addictions, such as gaming, internet use, smartphone use, and
gambling (Sussman, 2017; van Rooij et al., 2010; Edgerton, Keough, & Roberts, 2018). For
example, pathological gambling disorder was first introduced in DSM-III (1980), updated in
DSM-IV (1994) as an “impulse control disorder,” and then officially considered an addictive
disorder in DSM-V (APA, 2013). Currently, gambling disorder is the only established behavioral
addiction in DSM-V, which included internet gaming disorder (IGD) in a section that
recommends further clinical research before its inclusion as a formal disorder. “At this time, the
criteria for this condition are limited to Internet gaming and do not include general use of the
Internet, online gambling or social media” (APA, 2013), the criteria being:
• Preoccupation with gaming

8
• Withdrawal symptoms when gaming is unavailable or taken away (e.g., sadness,
anxiety, irritability)
• Tolerance in that more time spent on gaming is needed to satisfy the urge to game
• Inability to reduce gaming/unsuccessful attempts to quit gaming
• Forgoing/loss of interest in activities outside of gaming
• Continuity of gaming despite issues
• Lying to friends/family/others about gaming time
• Use of gaming to relieve negative emotions (e.g., guilt, hopelessness)
• Loss of jobs/relationships due to gaming
Being diagnosed with IGD would entail experiencing five or more of these symptoms
within a year and includes gaming online alone or with others. Although the debate on whether
internet addiction (IA) should be categorized as a behavioral addiction is ongoing (Block, 2008;
Pies, 2009; Guillot et al., 2016), some studies support the notion that IA qualifies as a
compulsive-impulsive spectrum disorder with four classic symptoms that are characteristic of an
addiction disorder: 1) excessive/ compulsive use, e.g., having a distorted sense of time,
neglecting daily functions; 2) withdrawal, e.g., feeling angry, tense, and/or depressed when
unable to use internet; 3) tolerance, e.g., needing more time on the internet; and 4) negative
repercussions, e.g., being defensive or hostile when asked to stop using the internet, not telling
the truth about use, doing poorly in school/work, socially isolating self, and feeling fatigue
(Block, 2008; Guillot et al., 2016).
However, scholars who disagree that IA is an actual addiction disorder, based on the lack
of conclusive evidence, prefer the terms compulsive internet use (CIU), problematic internet use
(PIU), or internet dependence (van Rooij et al., 2010; Donald et al., 2019; Griffiths, 2020). The

9
present study acknowledges that there has yet to be standardized criteria for IA and that IA
research has not been clearly integrated so far, which is why CIU will be the main study variable
in this dissertation. Meerkerk et al. (2009) proposed that the criteria for CIU can be mirrored to
the DSM criteria for substance abuse and pathological gambling:
• Loss of control: Desire or unsuccessful efforts to control, cut down, or stop use (e.g.,
finding it difficult to stop using the internet when online; continuing to use despite
intentions to stop; not getting enough sleep because of use; having unsuccessfully
tried to spend less time on the internet).
• Conflict: Continued use despite knowing negative consequences; includes inter- and
intrapersonal conflict (e.g., others saying you should use internet less; you yourself
thinking you should use internet less; rushing through responsibilities to go on the
internet; neglecting daily obligations because prefer to go on the internet).
• Preoccupation: Much time spent in use (e.g., preferring internet use over spending
time with others; thinking about the internet, even when not online; looking forward
to next internet use sessions).
• Mood modification: Use becomes a coping mechanism (e.g., going on the internet
when feeling down; using the internet to escape from sorrows or get relief from
negative feelings).
• Withdrawal symptoms: Restless or irritable when use cut down or stopped (e.g.,
feeling restless, irritable, or frustrated when unable to use internet).
These criteria are similar to Griffiths' (2005) 6-component model of addiction:
salience/preoccupation, mood modification, tolerance, withdrawal symptoms, and conflict.
Although more research is needed for CIU to be considered a formal disorder, CIU has been

10
increasingly studied (Meerkerk et al., 2009; van Rooij et al., 2010; Donald et al., 2019; Griffiths,
2020), with research supporting the criteria of five or more symptoms outlined above being met
within the past year or even within the past 30 days.
Concerning CIU, the issue may either be with the medium itself (i.e., general overuse of
and dependency on the internet; Donald et al., 2019) or with the numerous activities made
available by the internet, (e.g., checking social media, watching Youtube, shopping online,
playing online games; van Rooij et al., 2010; Griffiths, 2020). Socializing seems to hold one type
of significant role in internet use being addictive (van Rooij et al., 2010; Donald et al., 2019), as
interpersonal exchanges are used to experience social contact. Reinforcement from these
exchanges has been found to serve as an alternative to, or even replace, face-to-face contact with
others (Donald et al., 2019).
Generally, studies have found that CIU is associated with negative health outcomes (e.g.,
mental health issues, stress, decreased well-being in general) as well as experiencing feelings of
loneliness, particularly among adolescents (van Rooij et al., 2010; Donald et al., 2019). The
debilitating condition of CIU has a high prevalence in the U.S.: According to the Pew Research
Center, 48% of Americans aged 18 to 29 indicated that they use the internet “almost constantly”
in 2021 (timing of the internet use was not specified, e.g., use in the past 30 days or past 12
months). This was in comparison to 42% of Americans aged 30 to 49 indicating constant use in
2021, 22% of those aged 50 to 64, and 8% of those aged 65 or older (Perrin & Atske, 2021).
Thus, it seems that excessive internet use is especially problematic among EAs aged 18 to 29
compared to other age groups. It is important to note that the 2021 rates of internet use for 30-49
and 50-64 aged groups increased from 28% and 22%, respectively, in 2015 (Perrin & Atske,
2021), which supports the notion that harmful/risky internet use may continue onto adulthood

11
(Arnett et al., 2014; Anderson et al., 2017). To address the public health threat of CIU
development among an impressionable group of EAs who have undergone intense stress, the
present dissertation proposes that fostering posttraumatic growth after experiencing the intense
stress from a significant life-altering event could potentially prevent or impede the development
of CIU among EAs by acting as a buffer to the stress.
Posttraumatic Growth
Tedeschi & Calhoun (2004) coined the term, posttraumatic growth (PTG), in the 1980s to
mean the generally “positive psychological change experienced as a result of the struggle with
highly challenging life circumstances” (p.1), which could include trauma, a crisis, or a highly
stressful event (terms used interchangeably). These circumstances significantly affect people’s
ability to adapt to and understand the world or situation that they are in, as well as those involved
(Tedeschi et al., 2018). While it is typical to see psychological distress and negative reactions to
SLEs, positive reactions to SLEs do occur as the SLEs act as a catalyst for the development of
PTG among different types of people facing a wide range of traumatic experiences (Tedeschi &
Calhoun, 2004; Arpawong et al., 2015; 2016). Evidence exists indicating that “PTG can develop
as a result of various stressors” (Arpawong et al., 2016, p.2) that the individual undergoes and
considers to be impactful enough to self-perceive as life-changing.
To date, PTG has been conceptualized in several ways (Tedeschi & Calhoun, 2004;
Joseph & Linley, 2005; Zoellner & Maercker, 2006; Jayawickreme & Blackie, 2014), including:
• Benefit finding: Trauma survivors perceiving at least some positive change that
resulted from their struggle with the aftermath of trauma.

12
• Meaning-making coping: A form of coping with trauma where the survivor tries to
make sense of what happened and perceive it as meaningful rather than harmful.
• Positive illusions: An adaptive function to psychologically adjust to the trauma;
creating a sense of hope to counterbalance psychological distress.
PTG has also been called transformational growth: Aldwin’s (1994) model of
“transformational coping” theorizes that coping allows an individual to experience a positive
change in life following adversity, which results in a higher level of functioning than the level
functioning than that of pre-adversity level (Aldwin, 1994; Tedeschi, 1999; Joseph & Linley,
2005; Zoellner & Maercker, 2006). The process of reaching that higher level of functioning (i.e.,
growth) involves self-transcendence, which is defined as the expanding of one’s conceptual
boundaries in three ways (Reed, 1991):
1. Intrapersonal expansion: performing introspective activities (e.g., reflecting back on
one’s life having received a diagnosis of terminal/chronic illness and coming to terms
with the diagnosis).
2. Interpersonal expansion: concerning self with others’ welfare (e.g., sympathizing
and/or empathizing with others going through adversity; putting self in others’ shoes).
3. Temporal expansion: integrating perceptions of one’s past and future to enhance
one’s present (e.g., past history of a life crisis allows one to be grateful for the present
rather than being traumatized by the past and letting that negatively affect the present).
A key characteristic of the positive change experienced after trauma is the transformation
of the individuals’ understanding of themselves, their priorities in life, and their place in the
context of self and others (Tedeschi et al., 1998). PTG is a perception, whether constructive (real)
or illusory (not real). The constructive side of PTG is considered to be self-transcending (defined

13
as expanding of one’s conceptual boundaries; Tedeschi et al., 1998; Tedeschi & Calhoun, 2004)
and manifested by functional adjustment or cognitive restructuring, while the illusory side
(Taylor et al., 2000) is considered to be self-deceptive (using cognitive avoidance strategies, e.g.,
intentionally not thinking about the trauma) and associated with self-consolidation (integration of
life goals and ambitions into a coherent narrative) or palliation (relief of symptoms and
suffering). For example, Taylor’s (1983) concept of positive illusion argues that when
experiencing a traumatic event, individuals muster up an adaptive form of cognition (e.g., search
for meaning in the event, attempt to regain a sense of control or mastery, and effort to restore a
positive sense of self) that allows them to return to or exceed their previous form of selfperception and outlook on life.
Zoellner and Maercker (2006) suggested that constructive PTG may be positively
associated with mental health functions in general while illusory PTG and overcoming traumatic
events may partly be due to processes of distortion (e.g., denial, avoidance, wishful thinking, and
distortion of meaning). Glad et al. (2013) noted, however, that there is a key difference between
(a) positive side effects of addressing trauma and (b) positive aspects of having experienced
trauma: Acknowledging that dealing with trauma may result in positive life changes is not the
same as denying the negative post-trauma outcomes. For example, among trauma survivors who
reported experiencing growth after a traumatic event, their experience of distress and
vulnerabilities remained (Chun & Lee, 2008). So, experiencing PTG may paradoxically co-occur
with psychological distress and actually result in increased vulnerability (Tedeschi & Calhoun,
2004; Chun & Lee, 2008; Glad et al., 2013). The illusory side of PTG, in contrast to the
constructive side, has been associated with cognitive avoidance strategies that may be perceived
as maladjustment to trauma (Taylor et al., 2000; Zoellner & Maercker, 2006; Sumalla et al.,

14
2009). But that is not necessarily always the case: the illusory perception of PTG may have a
helpful temporary coping effect to reduce high levels of distress and keep the sense of identity
intact when threatened by a traumatic event.
The current literature suggests that individuals report experiencing PTG following some
sort of SLE, ranging from 58% to 83% among survivors of SLEs, indicating that PTG is not
uncommon (Joseph & Linley, 2005; Marziliano, Tuman, & Moyer, 2019). This finding was
based on the use of the PTG inventory (PTGI; Tedeschi & Calhoun, 1996), which was developed
to assess outcomes following trauma, with the purpose of measuring an individual’s ability to
reconstruct their life after having experienced trauma. The original PTGI was a 21-item scale
with five key factors: 1) new possibilities, 2) relating to others, 3) personal strength, 4) spiritual
change, and 5) appreciation of life. Example items included “I developed new interests” and “My
priorities of what is important in life.” Participants had to indicate the degree to which this
change occurred in their lives as a result of their traumatic event, following a 6-point Likert scale:
1) I did not experience this change as a result of my crisis, 2) I experienced this change to a very
small degree, 3) a small degree, 4) a moderate degree, 5) a great degree, and 6) a very great
degree.
Though some studies have criticized the validity of the PTGI items (Joseph & Linley,
2005; Frazier et al., 2009), one study addressed this criticism by comparing the levels of positive
psychological change through corroboration of subjective reports by an observer (e.g., significant
other of a survivor), thereby providing convergent validity and supporting the use of the PTGI as
an appropriate PTG-measuring assessment (Shakespeare‐Finch & Enders, 2008). It is worth
noting, however, that due to the multidimensionality of PTG (Tedeschi & Calhoun, 1996),
individuals may report growth experiences in one area but not another (e.g., reported improved

15
relationships but not spiritual growth). These differences in PTG aspects may be influenced by
several factors, including differences in personality, in social support, in types of trauma, and in
processes of developing PTG (e.g., shifts in behaviors, goals, and identity affecting the routes to
PTG; Tedeschi, Park, & Calhoun, 1998). The following eight dimensions of PTG were measured
in this dissertation [adapted from the original Post-Traumatic Growth Inventory (PTGI) by
Tedeschi and Calhoun (1996)]: appreciation for the value of own life, direction for life, handling
difficulties in life, understanding of spiritual matters, sense of closeness with others, involvement
in things of interest, compassion for others, and own inner strength.
It is also possible that some individuals may experience PTG regardless of whether it
manifests as observable behaviors, which would make it difficult to measure in that case. Despite
the criticism directed at the validity of PTG reports and measuring it, PTG is worth exploring
because it serves as an integral part in the recovery process for individuals who have undergone
traumatic experiences (Zoellner & Maercker, 2006; Aspinwall & Tedeschi, 2010) by cultivating
hope in seemingly hopeless situations and shifting the focus away from negative thinking (i.e.,
associated with trauma) to positive thinking (Zoellner & Maercker, 2006). In some longitudinal
cohort studies examining the relationship between PTG and various health outcomes over time,
those who self-reported PTG exhibited better health outcomes compared to those who did not
report PTG, such as physically adapting to an illness/disease (Milam, 2006; Tomich et al., 2012),
experiencing an improved quality of life (Penedo et al., 2006; Tomich et al., 2012), and
experiencing fewer depressive symptoms (Tomich et al., 2012). There were also differences by
subgroups (e.g., gender, ethnicity) in PTG, which will be discussed in the following section.

16
Subgroup Differences in PTG
Gender differences in PTG. When it comes to managing SLEs in terms of gender, males
may handle stress differently than females do (Rogers & Maytan, 2012). In terms of gender
differences in PTG reports, females were generally more likely to report higher levels of PTG
following stress than males (Swickert & Hittner, 2009; Vishnevsky et al., 2010; Taku & Cann,
2014). However, there have also been studies reporting no gender differences, or cases where
males report higher PTG levels than females (Vishnevsky et al., 2010). Coping with stress
through social support has been found to facilitate PTG (Tedeschi & Calhoun, 2004), and social
support in turn has been found to influence the relationship between gender and PTG as females
are more prone to seeking social support (Swickert & Hittner, 2009).
Ethnic/racial differences in PTG. In studies comparing African American participants to
White, one study found that African American youth reported higher mean score of benefit
finding (M=39.7, SD=6.1) than their Caucasian counterparts (M=36.5, SD=8.1; p < .05) (Phipps,
Long, & Ogden, 2007), and two found higher mean endorsement of PTG among African
Americans compared to Whites (Bellizzi et al., 2010; Kent et al., 2013).
In studies focusing on Hispanic participants, several found higher mean endorsement of
PTG among Hispanics versus non-Hispanic Whites (Milam, Ritt-Olsen, & Unger, 2004; Milam,
2006; Smith et al., 2008; Tobin et al., 2018; Schneider et al., 2019). Tobin et al. (2018) discussed
that this difference may be due to cultural aspects (e.g., interdependency of Hispanic family
members being a significant source of social support for this group; stronger cultural and ethnic
identity, which has been associated with self-efficacy and positive coping) acting as protective
factors against adverse mental health outcomes and promoters of psychosocial adaptation and
development of PTG after hardship. Smith et al. (2008) reported a large effect of ethnicity on

17
PTG and suggested that identifying as being an ethnic minority, or being in the Hispanic culture,
possibly promotes growth. This may be due to protective factors of a cultural nature, such as
religion/spirituality and family being a significant source of encouragement, support, and
strength. For example, there have been studies reporting that Hispanic women with cancer
perceived spirituality and family to be key for quality of life (Ashing-Giwa et al., 2004; Smith et
al., 2008).
Conversely, Kent et al. (2013) noted a case where PTG (associated with confiding in
healthcare providers) was reported to be significantly higher among non-Hispanic White but not
among African American and Hispanic/Latina participants. The authors attributed this
inconsistent result to care experiences (actual versus perceived) among ethnic/racial groups,
mentioning that the non-significant association among African American and Hispanic
participants may have been due to lower perceived closeness/trust with healthcare providers or
lack of providers in these groups compared to the non-Hispanic White group (Kent et al., 2013).
PTG across time. Milam (2004) examined PTG among HIV/AIDS patients and found
that PTG endorsement was stable, with patients maintaining high PTG scores over a 1.6-year
period. Dekel et al. (2012) examined the bidirectional relation between PTG and PTSD and
found that those with PTSD had higher PTG endorsement across the two data collection time
points than those without PTSD, implying that PTG is maintained and promoted by the presence
rather than absence of PTSD, i.e., not orthogonal constructs. Likewise, Danhauer et al. (2012)
suggested that newly-diagnosed acute leukemia patients who developed PTG reported increased
PTG endorsement over the weeks following diagnosis and initial chemotherapy sessions; and
Tsai et al. (2016) prospectively examined PTG among U.S. veterans, finding that those who
reported moderate levels of PTG after a significant SLE maintained those levels two years later.

18
Additionally, Stanton, Bower, and Low (2014) noted that of four studies finding a positive
relationship between time since onset of a stressor and PTG endorsement, one was a longitudinal
study (Manne et al., 2004) reporting a steady and significant increase in PTG scores over a 1.5-
year period among women with breast cancer. These results seem to suggest that PTG, especially
with the presence of stress that elicits PTG, is a long-term concept rather than a short-term one
that is depleted after a certain amount of time.
Significance
As a life stage that developed based on the changing times, emerging adulthood
emphasizes the critical transitions between the life stages of adolescence and adulthood and the
health consequences of those transitions. Referring back to the purpose of this dissertation (p.5),
studying how EAs deal with stress is important in that this research contributes to efforts to
improve health outcomes among EAs (a group that has not been studied as extensively as
adolescents and adults but has been gaining a lot of acknowledgement by researchers) by
identifying potential harmful coping mechanisms (e.g., turning to addictive behaviors) that may
continue into adulthood. While there are numerous coping mechanisms, many of which could
occur simultaneously, that EAs could use to deal with their stress, the current dissertation will
focus on CIU as the reported coping method.
In terms of previous research focusing on addictive behaviors and PTG among EAs,
Arpawong et al. (2015) examined the association between PTG and substance use and found that
higher PTG scores were associated with lower recurrences of alcohol and marijuana use and less
substance abuse at two-year follow up. To date, there has yet to be a study examining the
association between PTG and CIU. The hypothesized association (i.e., that higher PTG would be

19
associated with lower CIU) is plausible since those who develop PTG usually want to achieve
full functionality after SLEs, and risky addictive behaviors may not align with that goal
(Arpawong et al., 2015), so EAs who have experienced PTG may perceive CIU as detrimental
moving forward.
Dissertation Overview
Study 1: Prospective Associations of Stressful Life Events (SLEs), Compulsive Internet Use
(CIU), and Posttraumatic Growth (PTG) Among Emerging Adults (EAs)
Study 1 examined the trajectories of CIU and PTG among EAs over the course of four
years: PTG was first formally assessed in Year 2 and follow-up assessments were given in Years
3, 4, and 5. A prospective study on this topic aimed to add to the predominantly cross-sectional
literature by analyzing the direction of self-report CIU trends of over time. Examining if there
was an association between PTG and CIU was another goal of Study 1, as there were no studies
connecting the two concepts in the literature. CIU outcome was operationalized as the sum score
from the adapted CIU Scale (Meerkerk et al., 2009), with higher sum scores indicating greater
CIU levels. Gender and ethnic/racial differences in PTG were also examined.
Study 2: Examining the Moderating Effect of Posttraumatic Growth on the Association of
Cumulative Stress and Internet Use in Emerging Adults
Study 2 tested the hypothesis that PTG moderates the relationship of cumulative SLEs
and CIU levels: EAs who report higher PTG levels following stress would report lower CIU sum
scores. Based on the literature, while stress puts EAs at risk of developing harmful coping
mechanisms (in this case, CIU), we expected to find that PTG-equipped EAs were less likely to

20
turn to CIU when dealing with stress. Stress was operationalized as the number of SLEs reported
in the last two years, with the cumulative number of SLEs serving as an indicator of stress levels
experienced, adapted from the SLEs scale (Newcomb & Harlow, 1986; Wills et al., 1992). The
goal was to determine whether CIU scores increased or not over time among those who reported
experiencing PTG from stress, which would suggest that PTG weakens the association between
stress and CIU. Another goal was to test if PTG also had a moderating effect on the association
between stress and specific components of problematic internet use, such as internet browsing,
social networking, and online shopping. Reports of self-perceived addiction to each of these
three components were first assessed in Year 3 and follow-up assessments were given in Years 4
and 5.
Overall Design and Methods
The present dissertation used data collected from a larger study, Project Towards No
Drug Abuse (TND) (PI: Dr. Steve Sussman), a drug abuse prevention curriculum targeted
towards EAs in alternative or continuation high schools (CHS; detailed in Lisha et al., 2014).
EAs attending a CHS may experience more stress than their traditional high school counterparts,
and the reason for going to CHS may be due to excessive truancy, poor academic performance,
drug use, violence, or other illegal/disruptive behavior (Arpawong et al., 2015).
Data Collection Details. In the original procedure for TND implemented in the U.S.,
students aged 18 or older provided signed informed consent to participate in the study. Those
who were younger than 18 were required to submit a written assent as well as signed or verbal
permission from a parent. Twenty-four alternative high schools were randomly assigned to one
of the study conditions (intervention or control group) focusing on drug abuse prevention. There

21
were two subgroups for the intervention group, one with only the TND intervention and another
adding a motivational interviewing booster to the intervention. A pretest survey (close-ended,
self-report questionnaire) was administered at the school site immediately prior to
implementation of TND to assess program quality in 2008, and a posttest survey was
administered immediately after the completion of the program. Five Years of follow-up data
(Years 1-5) were gathered from 2009 to 2014.
A variable for program participation was dichotomously coded as TND or control
(1=intervention, 0=control) and included as a covariate to control for the study condition to
which students were assigned since program participation was not a primary research outcome.
Previous studies have shown no differences in substance use outcomes between the two
intervention conditions (see Sussman et al., 2012; Arpawong et al., 2016).

22
Chapter 2:
Study 1: Prospective Associations of Stressful Life Events, Compulsive Internet Use, and
Posttraumatic Growth Among Emerging Adults
Introduction
As individuals who are no longer children but not yet full adults, emerging adults (EAs)
are in a transitionary life stage where they are making enduring life choices (Arnett et al., 2014).
In the process of making these choices, EAs are bound to face numerous stressful life events
(SLEs) and adversity, including the death or loss of a loved one, loss of social support, medical
diagnoses, and financial hardships (Arpawong et al., 2016; Jankowski et al., 2021), which create
the need to cope with the stress and trauma (Roth & Cohen, 1986; Schroevers et al., 2007)
caused by the events.
Amidst the various ways to cope with stress and trauma, a common coping mechanism
among EAs is using the internet to escape from the stressor (Tang et al., 2017; Fernandes et al.,
2020). However, some may exceed normative internet use to excessive amounts, to the point of
being considered an addiction or compulsive/problematic use. The DSM-V criteria for internet
gaming disorder (APA, 2013) could be applied for assessing compulsive internet use (CIU):
preoccupation with internet use; withdrawal symptoms (e.g., sadness, anxiety, irritability) when
internet is unavailable or taken away; tolerance in that more time spent on the internet is needed
to satisfy the urge to use the internet; inability to reduce internet use; loss of interest in activities
outside of internet use, continuity of internet use despite issues; lying about internet use and the
amount of time on the internet; using the internet to relieve negative emotions (e.g., guilt,
hopelessness); and loss of jobs and/or relationships due to internet use. The physical, mental, and

23
emotional risk factors associated with CIU negatively impact health and EAs are likely to
continue these risky behaviors into adulthood (Arnett et al., 2014; Anderson et al., 2017).
While it is typical to see psychological distress causing negative reactions to SLEs,
positive reactions to SLEs have been also found to occur as the SLEs act as a catalyst for the
development of posttraumatic growth (PTG) among different types of people facing a wide range
of traumatic experiences (Tedeschi & Calhoun, 2004; Arpawong et al., 2015; 2016). Tedeschi &
Calhoun (2004) coined the term PTG in the 1980s to mean the generally “positive psychological
change experienced as a result of the struggle with highly challenging life circumstances” (p.1),
which could include trauma, a crisis, or a highly stressful event (terms used interchangeably).
These circumstances significantly affect people’s ability to adapt to and understand the world or
situation that they are in, as well as those involved (Tedeschi et al., 2018).
Previous studies have found that experiencing various stressors can result in the
endorsement of PTG (Tedeschi & Calhoun, 1996; Milam, Ritt-Olsen, & Unger, 2004; Arpawong
et al., 2016) as the SLEs are considered to be impactful enough to be perceived as life-changing.
The current study examined the association of PTG and SLEs over time, which would add to the
predominantly cross-sectional literature of PTG’s association with stress. Additionally, there is a
gap in the literature on the association between PTG and CIU: there are studies examining the
association of PTG and various conditions (such as depression [Milam, 2004; Milam, Ritt-Olsen,
& Unger, 2004], negative impacts of cancer [Tedeschi et al., 2018], and substance use behaviors
[Milam, Ritt-Olsen, & Unger, 2004; Arpawong et al., 2015]) but not many studies looking at
PTG’s association with internet use.
Gender and ethnicity were considered covariates to determine if there are demographic
differences in the reported mean values of PTG. In previous research, females were generally

24
more likely to report higher levels of PTG following stressful events than males (Swickert &
Hittner, 2009; Vishnevsky et al., 2010; Taku & Cann, 2014). However, there have been studies
reporting no gender differences or cases where males report higher PTG levels than females
(Vishnevsky et al., 2010).
In terms of the association between PTG and ethnicity, one study found that African
American youth reported a higher mean score of benefit finding (M=39.7, SD=6.1) than their
Caucasian counterparts (M=36.5, SD=8.1; p < .05) (Phipps, Long, & Ogden, 2007). Among
other studies with diverse populations, two found higher mean endorsement of PTG among
African Americans compared to Whites (Bellizzi et al., 2010; Kent et al., 2013). Among studies
focusing on Hispanic/Latinx populations, several studies found higher mean endorsement of
PTG among Hispanics versus non-Hispanic Whites (Milam, Ritt-Olsen, & Unger, 2004; Milam,
2006; Smith et al., 2008; Tobin et al., 2018; Schneider et al., 2019).
With these associations in mind, the purpose of the present study was to assess the
prevalence of CIU among EAs and determine whether there was an association between PTG
and the harmful behavior of CIU.
Proposed Aims and Hypotheses
Aim 1. To examine the trajectory of PTG and CIU to see general trends over time. PTG
was hypothesized to stay stable across time points based on previous studies reporting that
participants maintained PTG levels over time (Milam, 2004; Danhauer et al., 2015; Tsai et al.,
2016; Tedeschi et al., 2018). CIU was hypothesized to increase over time, based on results
indicating that CIU prevalence was high in the U.S. (Perrin & Atske, 2021) and on the notion

25
that harmful/risky internet use may continue onto adulthood (Arnett et al., 2014; Anderson et al.,
2017).
Aim 2. To examine the association between PTG and predictor variables and covariates
over time. Based on theory, PTG was hypothesized to be positively associated with SLEs
(Tedeshi et al., 2018).
Since those who endorse PTG usually want to achieve full functionality after their
traumatic experience, and risky addictive behaviors may not align with that goal (Arpawong et
al., 2015), EAs who have experienced PTG would theoretically perceive CIU as detrimental to
their post-SLE self. Therefore, PTG was hypothesized to be negatively associated with CIU.
As for covariate associations, based on literature, PTG was hypothesized to be positively
associated with identifying as female (Swickert & Hittner, 2009; Vishnevsky et al., 2010; Taku
& Cann, 2014) and identifying as Hispanic (Milam, Ritt-Olsen, & Unger, 2004; Milam, 2006;
Smith et al., 2008; Tobin et al., 2018; Schneider et al., 2019).
Methods
Participants were older teens and young adults (16-22 years old during 2010-2012) who
had attended an alternative high school in a southern California county and previously
participated in a larger longitudinal study testing the efficacy of a school-based substance abuse
prevention program, Project Towards No Drug Abuse (TND), consisting of 12 sessions of
curriculum focusing on developing motivation, skills, and decision-making abilities (Sussman et
al., 2012). There was also a motivational interviewing (MI) component that supplemented the
TND material. The main focus of MI was on the motivation to change by highlighting the

26
difference and incongruities between current lifestyle (that may have included behaviors harmful
to health) and future goals of the students.
A total of 24 schools participated, with each school having an average of 70 students
participate (range: 56-103 students). Participants’ informed consent were obtained prior to data
collection. Each student was asked their age, gender, and ethnicity in the initial baseline pretest,
which were designated as demographic covariates. Participants were 16-22 years old during Year
2 and were followed for 3 years. PTG items were added to Project TND beginning in Year 2 of
data collection, which is why the present study uses data starting from Year 2 (2010-11) to Year
5 (2013-15). Since assessing the program effects was not the main focus of this study, TND
participation was included as a control variable (refer to Measures section below).
Measures
Demographics. In 2008, baseline data collection included the demographic variables of
age, gender (1 = male, 0 = female), and race/ethnicity (1 = Asian or Asian American, 2 = Latino
or Hispanic, 3 = African American or Black, 4 = White, Caucasian; not Hispanic, 5 = American
Indian or Native American, 6 = Mixed, 7 = Other). Due to insufficient sample sizes for some of
the race/ethnicity categories, a dichotomous variable was created to recode race as Hispanic/nonHispanic (1 = Hispanic, 0 = non-Hispanic). At baseline, Project TND study condition was also
recorded (1 = any TND component (intervention or intervention plus motivational booster, 0 =
no TND component/control).
Stressful Life Events. SLEs were repeatedly measured in Years 2, 3, 4, and 5. Nine items
adapted from a shortened version of the Adolescent Negative Life Events Inventory (Newcomb
& Harlow, 1986; Wills et al., 1992) asked if participants experienced the following events in the

27
past two years: “I got disciplined or suspended from school or work," "Someone in my family
had a serious illness, accident, or injury," "I did not have enough money for basics (like food),"
"A new person joined the household (baby or young child, grandparent, stepbrother or sister,
stepparent, other)," "I was a victim of a violent or abusive crime," "Someone in my family or I
was arrested," "I broke up with my girlfriend/boyfriend/partner," "There were a lot of arguments
that happened at home," and “Other.” Of those nine options, participants were asked to indicate
which event, if any, affected them the most (i.e., was their most life-altering event, MLAE) as
this was relevant in assessing PTG. Some participants who reported experiencing SLEs did not
specify which event was their MLAE, and some reported that multiple SLEs were equally lifealtering.
The questionnaire did not assess the severity of the SLEs, so a weighted score of the
various life events could not be created because each person had their own way of interpreting
and processing SLEs. Instead, the number of SLEs (among those who indicated experiencing at
least one or more SLEs) was summed as a total—since cumulative stress is related to PTG—and
was entered as a continuous correlate. While there may be several explanations for relating
cumulative stress to PTG, one rationale was that more SLEs reported may be interrelated,
particularly to an event that was reported as most life altering (Arpawong et al., 2016). Another
was that more SLEs reported may affect how individuals adapt to adversity: if the stressors are
perceived to be challenges to overcome, one may use the adversity to reach an optimal
functionality/performance level and build resilience, e.g., emotional resilience, better tolerance to
pain, and increased cognitive function (Robertson, 2017).
Posttraumatic Growth. PTG was repeatedly measured in Years 2, 3, 4, and 5. Individuals
were asked to respond to the PTG items in relation to their reported most life-altering life

28
event(s). Eight items adapted from the PTGI (Tedeschi & Calhoun, 1996; Milam et al., 2004;
Milam, 2006; Arpawong et al., 2015) assessed changes (negative change, no change, or positive
change) in the participants’ lives that may have occurred since experiencing the life-altering
event(s) in the following: appreciation for the value of their own life, direction for their life,
handling their difficulties, their understanding of spiritual matters, their sense of closeness with
others, involvement in things that interest them, their compassion for others, and their own inner
strength. Responses were measured on the following 3-point Likert-type scale: 1 (negative
change/got worse), 2 (no change), and 3 (positive change/got better). PTG scores were averaged
across the eight items according to the PTG coding of 1, 2, or 3 to gauge whether there was an
overall negative, positive, or no change in a participants’ life. Cronbach’s alpha for this scale, on
average, was 0.83.
To clarify, the scale used in this study broadly measured PTG as negative, neutral, or
positive change, as compared to other studies that measured the degree of PTG using a Likert
scale (e.g., the original PTG Inventory measured on a 6-point Likert scale, with a score of 0
being “I did not experience this change as a result of my crisis” to a score of 5 being “I
experienced this change to a very great degree as a result of my crisis”; Tedeschi & Calhoun,
1996). Therefore, the results of this study solely indicated whether PTG was negatively, neutrally,
or positively present, rather than being able to specifically distinguish higher PTG levels from
lower PTG levels.
Compulsive internet use. CIU was repeatedly measured in Years 2, 3, 4, and 5. Four
items adapted from the Compulsive Internet Use (CIU) Scale (Guillot et al., 2016) asked how
often the participants: 1) stayed on the internet longer than planned, 2) used the internet more
than they ought, 3) could not cut down internet use despite wanting to, and 4) felt that their

29
internet use seemed beyond their control. Responses were measured on the following 5-point
Likert-type scale: 1 (never), 2 (rarely), 3 (sometimes), 4 (most of the time), and 5 (always). For
data analysis, this variable was re-coded as 0-4, with 0 representing no CIU issues versus 4
representing constant CIU issues. The CIU items were used as a sum score (0-16), with a higher
score indicating a greater level of CIU. Cronbach’s alpha for this scale, on average, was 0.87.
Sample exclusion criteria
Of the 1676 participants from baseline, 42 did not indicate their ethnicity. Of the
remaining 1634, 700 participants were excluded because they did not participate during Years 2-
5 of data collection, 35 were excluded for missing values for all relevant variables, and 17 were
excluded for indicating that they did not experience any SLE in the last two years (which could
possibly be due to that individual having experienced significant trauma prior to the two-year
mark) during Years 2-5. This resulted in a final sample size of 882 participants (refer to Figure
1).

30
Figure 1: Study 1 sample exclusion criteria.
Assessing attrition
To ensure that the attrition rate (47.4%) did not result in the working follow-up sample
being significantly different from the baseline sample, t-tests were performed to see if there were
differences in baseline CIU sum scores and demographics between the working follow-up
sample (n=882) and those who were not (n=794). CIU was compared since the variable was
assessed at baseline, while the SLE and PTG items were added to the questionnaire starting in
Year 2. There were no significant differences between the baseline sample lost to follow-up and
the remaining group in TND participation, Hispanic status, age, and internet use sum scores
(p=0.72, 0.07, 0.66, and 0.28, respectively). There was a significant difference in gender (p<.05),

31
however, in that there seems to have been slightly more boys lost to follow-up (60% males) than
the remaining group (55% males).
Analyses
All statistical analyses were conducted using SAS 9.4. Multilevel models (MLM) were
used to account for nesting (Bauer, D., J., & Curran, P. J., 2023). A total of 2,228 repeated
measures were nested within 882 students, who were nested within 24 schools. Of the 882
students, 203 participated once, 235 twice, 221 thrice, and 223 participated all four times during
the four years of data collection.
To address Aim 1, an unconditional linear growth curve model of PTG was run to map
the overall trajectory of PTG over time. A separate unconditional linear growth curve model of
CIU was run to determine the overall trajectory of CIU over time. Time-varying predictor
variables (CIU and SLEs) and time-invariant covariates (baseline demographic variables) were
added to the PTG model to test their associations with PTG. To address Aim 2, a multivariate
growth curve model was run with time-varying predictor variables (SLEs and transformed PTG)
and time-invariant covariates (demographics) to examine their associations with PTG.
Interactions with time were run to see the rate of change in CIU.
Results
Study 1 sample characteristics
Of the 882 eligible participants, slightly more than half (55%) identified as male, on
average was aged 19 years old (ranging from 16 to 22 years) at two year follow-up from baseline,
and nearly two-thirds (66%) identified as Latinx/Hispanic (refer to Table 1).

32
Table 1. Study 1 sample demographics (n=882).
Variable Freq (%) or Mean (SD)
Gender
Female 396 (44.90)
Male 486 (55.10)
Age (at Year 2 assessment) 18.92 (0.92)
Race/ethnicity
Asian or Asian American 22 (2.49)
Latinx or Hispanic 583 (66.10)
African American or Black 31 (3.51)
White, Caucasian, Anglo; not Hispanic 97 (11.00)
Other1 149 (16.89)
Assigned study condition
TND 595 (67.46)
Control 287 (32.54)
Notes. 1Other includes American Indian/Native American and mixed ethnicities.
The mean cumulative number of SLEs across time was 2.74 (SD=1.64), with values
ranging from one to nine. The mean sum score for CIU across time was 3.77 (SD=3.74), with
values ranging from zero to 16. The mean PTG score across time was 2.66 (SD=0.38), with
values ranging from one to three, indicating that most participants reported relatively positive
changes (i.e., some aspects of improvement in their life) after having experienced SLEs in the
past two years (refer to Table 2).
Table 2. Study 1 sample means of relevant variables by data collection year.
Year 2 Year 3 Year 4 Year 5 Overall
n 596 605 534 493 882
Variable Mean(SD) Mean(SD) Mean(SD) Mean(SD) Mean(SD)
SLEs1 3.09 (1.75) 2.84 (1.61) 2.48 (1.54) 2.47 (1.54) 2.74 (1.64)
CIU2 3.39 (3.51) 3.52 (3.57) 4.01 (3.92) 4.26 (3.97) 3.77 (3.74)
PTG3 2.64 (0.38) 2.67 (0.38) 2.66 (0.38) 2.66 (0.39) 2.66 (0.38)
Notes. 1SLEs: Stressful life events, cumulative, range: 0-9. 2CIU: Compulsive internet use sum score, range: 0-16.
3PTG: Posttraumatic growth average score, range: 1-3.
Examining the frequencies of SLEs among those who reported PTG from Years 2-5
(refer to Table 3) revealed that the most commonly reported SLE across time was Someone in my
family had a serious illness, accident, or injury (20%), which was also the most reported ‘most
life-altering event’ (MLAE) across time (25%). Next commonly reported SLEs across age

33
groups included Lots of arguments at home (15%), Broke up with girlfriend/boyfriend/partner
(15%), and New person joined household (15%). Taking a closer look at the “Other” category,
responses included death of a family member or friend, having a pregnancy or miscarriage,
experiencing mental health concerns, parents divorcing, and experiencing homelessness. Several
of the responses to “Other” indicated that two or more SLEs occurred simultaneously.
Table 3. Frequencies of stressful life events (SLEs) reported in Years 2-5.
SLEs Year 2 %1 Year 3 %1 Year 4 %1 Year 5 %1
Got disciplined/suspended from school/work 138 7.50 71 4.13 53 4.00 49 4.03
Family had serious illness, accident, or injury 370 20.10 335 19.47 259 19.53 249 20.48
Did not have enough money for basics 159 8.64 183 10.63 169 12.75 148 12.17
New person joined household 258 14.01 270 15.69 189 14.25 174 14.31
Was victim of violent or abusive crime 60 3.26 53 3.08 37 2.79 37 3.04
Family member or I was arrested 191 10.37 162 9.41 124 9.35 106 8.72
Broke up with girlfriend/boyfriend/partner 273 14.83 267 15.51 190 14.33 168 13.82
Lot of arguments at home 279 15.15 262 15.22 203 15.31 188 15.46
Other 113 6.14 118 6.86 102 7.69 97 7.98
Total SLEs reported 1841 100 1721 100 1326 100 1216 100
Notes. 1Percentage out of the total number of reported SLEs for that year.
Table 4. Frequencies of most life-altering events (MLAEs)
1
reported in Years 2-5.
SLEs Year 2 %2 Year 3 %2 Year 4 %2 Year 5 %2
Got disciplined/suspended from school/work 27 4.77 7 1.22 8 1.73 6 1.37
Family had serious illness, accident, or injury 150 26.50 147 25.70 108 23.33 104 23.74
Did not have enough money for basics 35 6.18 32 5.59 44 9.50 32 7.31
New person joined household 62 10.95 78 13.64 79 17.06 63 14.38
Was victim of violent or abusive crime 21 3.71 19 3.32 15 3.24 11 2.51
Family member or I was arrested 63 11.13 38 6.64 35 7.56 30 6.85
Broke up with girlfriend/boyfriend/partner 66 11.66 91 15.91 68 14.69 55 12.56
Lot of arguments at home 68 12.01 81 14.16 47 10.15 51 11.64
Other 74 13.07 79 13.81 59 12.74 86 19.63
Total MLAEs reported 566 100 572 100 463 100 438 100
Notes. 1Respondents were asked to designate one SLE as their most life-altering event (MLAE). While there were
multiple SLEs reported, individuals were asked to designate one MLAE on which to base their responses to the PTG
items. Non-respondents could have experienced an MLAE more than two years ago. 2Percentage out of the total
number of reported MLAEs for that year.
Aim 1 results
PTG linear growth curve model results (refer to Table 5): Testing for random effects in
the unconditional linear growth curve, the random intercepts-only model fit better than one
without (LRT p<.0001), but the random intercepts-random slopes model did not fit better than

34
the random intercepts-only model (LRT p=0.07). Moving forward with the random interceptsonly model, the fixed effect of PTG reported an overall mean estimate of 2.65 (SE=0.02) across
all schools and students. The average rate of change for PTG marginally increased over time by
0.004 points but was not statistically significant (p=0.53). According to the variance component
estimates, the intraclass correlation coefficients (ICCs) indicated that approximately 4% of the
individual differences in PTG during Year 2 data collection could be explained by differences
between schools. This implied that the vast majority of individual differences could be attributed
to differences within schools.
Table 5. Unconditional linear growth curve model of PTG over time (n=882).
Notes. 1
id=student ID, sid=school ID.
*p<0.05. **p<0.01. ***p<0.001.
CIU linear growth curve model results (refer to Table 6): Testing for random effects on
the unconditional model showed that the random intercepts-only model fit better than the one
without (LRT p<.0001), and the random intercepts-random slopes model fit better than the
random intercepts-only model (LRT p<.0001). The fixed effect of CIU significantly increased
over time by 0.31 points (p<.0001), with an overall mean estimate of 3.34 (SE=0.15). According
to the ICCs, approximately 2% of the individual differences in CIU during Year 2 data collection
could be explained by differences between schools, but 0% of the individual CIU change over
time could be attributed to between-school differences. This also implied that the vast, if not all,
majority of individual differences could be attributed to differences within schools.
Fixed Effects β SE
Intercept 2.6540*** 0.0157
Time 0.0041 0.0065
Random Effects β SE
UN(1,1) id(sid)1 0.0414*** 0.0043
UN(1,1) sid 0.0016 0.0012
Residual 0.1041*** 0.0039

35
Comparing school trajectories of CIU rates of change (refer to Figure 2), there seemed to
be slight variations in the slopes but were overall similar to each other in that all CIU score
averages were increasing over time.
Table 6. Unconditional linear growth curve model of CIU over time (n=882).
Fixed Effects β SE
Intercept 3.3416*** 0.1497
Time 0.3070*** 0.0613
Random Effects β SE
UN(1,1) id(sid)1 6.3333*** 0.6713
UN(2,1) id(sid) -0.1377 0.2606
UN(2,2) id(sid) 0.6475*** 0.1553
UN(1,1) sid 0.1600 0.1477
UN(2,1) sid -0.0382 0.0372
UN(2,2) sid 0 .
Residual 5.9393*** 0.2905
Notes. 1
id=student ID, sid=school ID.
*p<0.05. **p<0.01. ***p<0.001.
Figure 2: CIU trajectory over time. The light blue lines represent the mean compulsive internet use (CIU) scores
across students for each school, and the bolded black line represents the mean internet use score averaged across
students and across schools. Overall, CIU seems to increase over time.

36
Aim 2 results
Time-invariant covariates and time-varying predictors were added to the PTG linear
growth curve model from Aim 1 (refer to Table 7). The first model tested the association of timeinvariant covariates with PTG: gender, Hispanic status, and TND program participation. An LRT
indicated that the random effects (random intercepts and random slopes) model had a better fit
(p<0.05). The fixed effects of all three covariates were found to be statistically significantly
associated with PTG (p<0.05, p<0.01, p=0.05, respectively), signifying that those who identified
as female, identified as Hispanic, and participated in any component of TND were more likely to
report higher PTG than their respective counterparts.
The second model added the predictor of cumulative SLEs to test the within- and
between-school effects of SLEs, which were both school-mean centered and mean-centered prior
to being tested. In terms of fixed effects, the within-school effects of SLEs were statistically
significantly associated with PTG (p<0.0001): those who reported less cumulative SLEs were
more likely to have a higher PTG score within a school. Although the between-school effects of
SLEs were similar in value to the within-school effects, they were not significantly associated
(p=0.63). The fixed effects for TND participation, gender, and Hispanic status remained
statistically significant. Although small, the variance component estimate for within-school
differences was statistically significant (p<0.0001), which was similar to the one in the
unconditional model (p<0.0001).
The third model added CIU sum score to test the within- and between-school effects of
CIU, which were also school-mean centered and mean-centered prior to being tested. In terms of
fixed effects, the within-school effects of CIU were statistically significantly associated with
PTG (p<0.05): after controlling for stress and demographics, those who reported lower CIU sum

37
scores were more likely to have a higher PTG score within a school. The between-school effects
of CIU were similar in value to the within-school effects, but they were not significantly
associated (p=0.79). The fixed effects for TND participation, gender, and Hispanic status
remained statistically significant. The variance component estimate for within-school differences
remained small yet statistically significant (p<0.0001).

38
Table 7. Conditional linear growth curve model of PTG with covariates and predictor variables (n=882).
1. Time-invariant covariates only
Fixed Effects β SE
Intercept 2.5842*** 0.0288
Time 0.0037 0.0085
Gender -0.0453* 0.0200
Hispanic status 0.0693
** 0.0214
TND participation 0.0716** 0.0249
Random Effects β SE
UN(1,1) id(sid)1 0.0477*** 0.0079
UN(2,1) id(sid) -0.0056 0.0034
UN(2,2) id(sid) 0.0048** 0.0020
UN(1,1) sid 0.0001 0.0014
UN(2,1) sid -0.0001 0.0007
UN(2,2) sid 0.0006 0.0005
Residual 0.0962*** 0.0045
2. TICs2 + SLEs
Fixed Effects β SE
Intercept 2.5994*** 0.0290
Time -0.0025 0.0083
Gender -0.0470* 0.0197
Hispanic status 0.0614** 0.0215
TND participation 0.0700* 0.0257
SLEs: WSC3
-0.0278*** 0.0051
SLEs: BSC4
-0.0229 0.0477
Random Effects β SE
UN(1,1) id(sid) 0.0457*** 0.0079
UN(2,1) id(sid) -0.0055 0.0034
UN(2,2) id(sid) 0.0046* 0.0020
UN(1,1) sid 0.0003 0.0015
UN(2,1) sid -0.0001 0.0007
UN(2,2) sid 0.0005 0.0005
Residual 0.0961*** 0.0045
3. TICs + SLEs + CIU
Fixed Effects β SE
Intercept 2.5969*** 0.0295
Time -0.0006 0.0082
Gender -0.0457* 0.0197
Hispanic status 0.0577** 0.0216
TND participation 0.0727* 0.0271
SLEs: WSC -0.0273*** 0.0051
SLEs: BSC -0.0217 0.0486
CIU: WSC -0.0053* 0.0023
CIU: BSC 0.0053 0.0199
Random Effects β SE
UN(1,1) id(sid) 0.0459*** 0.0078
UN(2,1) id(sid) -0.0057 0.0034
UN(2,2) id(sid) 0.0048** 0.0020
UN(1,1) sid 0.0003 0.0015
UN(2,1) sid 0.00001 0.0007
UN(2,2) sid 0.0004 0.0005
Residual 0.0957*** 0.0045
Notes. 1
id: student ID, sid: school ID.
2TIC: time-invariant covariates. 3WSC: within-school-centered variable. 4BSC:
between-school-centered variable. *p<0.05. **p<0.01. ***p<0.001.

39
Discussion
General PTG and CIU trajectories
Consistent to theory (Tedeschi et al., 2018) and previous research (Milam, 2004;
Danhauer et al., 2015; Tsai et al., 2016), mean PTG stayed relatively constant across the four
data collection time points: there was a marginal increase over time without controlling for any
covariates and a marginal decrease over time after having controlled for covariates, albeit both
were not statistically significant. If this was a cross-sectional study, it would have been difficult
to address the temporality of PTG—but the results of this prospective study indicated that PTG
levels were maintained over time. In the case of this study, there seemed to have been a
consistently positive PTG endorsement over the course of four years, i.e., on average,
participants experienced overall positive changes in life following the stressful life events. Since
the relationship between PTG and time since the event was non-significant in this case, which
was similar to other findings examining the trajectories of PTG, it could be assumed that several
varying temporal courses of PTG could occur. This suggested that rather than a single
occurrence, PTG may be a stable and long-term perspective.
Overall, mean CIU consistently increased over time, supporting the view that
harmful/risky internet use during young adulthood may continue onto older adulthood (Arnett et
al., 2014; Anderson et al., 2017). In terms of CIU prevalence, there have been inconsistent
reports of compulsive/problematic internet use estimates, with estimates ranging widely across
studies, even among those using the same assessment scales (i.e., 8% to 47%) partly due to
sampling and methodological differences (Burkauskas et al., 2022). Guertler et al. (2014) used
the same scale as the one used in this study, the Compulsive Internet Use Scale (Meerkerk et al.,
2009), and suggested a cutoff of 21 out of 54 to estimate prevalence of problematic internet use.

40
Following that cutoff (i.e., 6 out of 16 for this study), the CIU prevalence estimates in this
sample were 25%, 22%, 29%, and 31% across the data collection years, respectively. These
estimates fell within the range of previously reported prevalence estimates (Guertler et al., 2014;
Perrin & Atske, 2021; Burkauskas et al., 2022) and indicated that CIU may pose a health risk to
some EAs.
PTG and relationship with predictors
SLEs. Contrary to the hypothesis that PTG would increase with increased cumulative
stress, SLEs were negatively associated with PTG, with the fixed effect association between
PTG and SLEs being statistically significant for within-school differences: Controlling for
internet use and all covariates, within a school, those who reported higher PTG scores were more
likely to report lower cumulative SLEs. The significant random effects suggested that the effect
of cumulative SLEs influencing PTG varied across students (i.e., the number of SLEs
experienced increased PTG in some students and decreased PTG in others) and could be
attributed to varying individual traits and to the varying factors of the SLEs experienced.
Compulsive internet use. CIU was negatively associated with PTG, with the withinschool differences being significant: Controlling for cumulative SLEs and all covariates, within a
school, those who reported higher PTG scores were more likely to report lower CIU sum scores.
Theoretically, those who endorse PTG usually desire to thrive after having gone through
hardship, so EAs who have experienced growth following adversity would generally perceive the
debilitating condition of CIU as detrimental to their post-SLE self since CIU has been associated
with negative health outcomes, such as mental health issues, stress, and decreased well-being in
general (van Rooij et al., 2010; Donald et al., 2019). Therefore, PTG endorsement among EAs
may discourage them from engaging in CIU.

41
PTG and relationship with covariates
Gender. Females were more likely to report higher PTG scores than males. Although
there have been conflicting findings in the literature, other studies that found that women
reported higher PTG than men attributed this to the ability/willingness to express emotions
(Jaarsma et al., 2006); to females naturally seeking social support during times of stress and that
social support, in turn, allowing for greater levels of stress-related growth (Swickert & Hittner,
2009); to women’s tendency to engage in more rumination (i.e., the deliberate act of thinking
about, processing, contemplating or meditating over, and/or reflecting on the issue) than men
and that leading to greater growth (Vishnevsky et al., 2010); and to cultural and individual
differences (Taku & Cann, 2014). However, further studies are needed to better understand the
causes of gender differences in PTG.
Race/ethnicity. Identifying as Hispanic was positively associated with PTG, which was
consistent with existing literature (Milam, Ritt-Olsen, & Unger, 2004; Milam, 2006; Smith et al.,
2008; Tobin et al., 2018; Schneider et al., 2019). Smith et al. (2008) reported a large effect of
ethnicity on PTG and suggested that identifying as being part of an ethnic minority, or being in
the Hispanic culture, possibly promotes growth due to protective cultural factors, such as
religion/spirituality and family being a significant source of encouragement, support, and
strength (e.g., there have been studies reporting that Hispanic women with cancer perceived
spirituality and family to be key for quality of life; Ashing-Giwa et al., 2004; Smith et al., 2008).
Tobin et al. (2018) discussed that this difference may be due to certain aspects of culture, such as
how Hispanic/Latinx family members are interdependent of each other, that the independency
provides a significant source of social support for this group, and that stronger ethnic/cultural
identity has been associated with positive coping—these cultural aspects then act as protective

42
factors against adverse mental health outcomes and promote the development of PTG after
hardship. These may serve as explanations as to why those who identified as Hispanic in this
sample were more likely to report higher PTG than their non-Hispanic counterparts. Future
research on racial/ethnic differences in PTG may need to bridge how these protective cultural
factors can be replicated for those without a cultural identity or those isolated from a supportive
community.
Program participation. Participation in Project TND was also positively associated with
PTG in this sample: those who participated in TND reported higher PTG than the control group
who did not participate in TND. This association was similar to the one of CIU and PTG, but
inversely: TND focused on factors predicting substance use and problem behaviors in order to
prevent them, including motivation (students’ attitudes, perceptions, expectations, and goals
avoiding problem behaviors), skills (efficacy in communication, self-control in social settings,
and coping methods), and decision-making (cultivating the ability to make decisions leading to
health-promoting behaviors; Sussman et al., 2012). Since those endorsing PTG usually desire to
thrive after adversity, the values taught in TND aligned with the desire to grow. Therefore,
cultivating health-promoting skills may in turn promote the cultivation of PTG among EAs.
Limitations and future directions
There were a few limitations to note. First, the results may have only been applicable to
CHS students rather than to all EAs in general, as CHS students were more likely to have
encountered a different high school experience than other high school/college students their age:
CHS students have had to typically take leave due to excessive school absences, poor scholastic
performance, disorderly behavior, drug/substance use, violent acts, or other illegal activity
(Rohrbach et al., 2005; Arpawong et al., 2016). Second, the data collected were self-reported.

43
Third, it is worth noting that there may have been special circumstances in certain internet use
cases, such as excessive use due to work-related reasons (e.g., professional gamer, social media
used as income platform, academic researcher), which would have affected findings. Fourth, the
data were collected several ago, which could be considered outdated data. Future studies could
address these limitations, respectively, by 1) expanding the participant pool to all EAs, whether
in college, working, or neither (this may allow for examination of age differences in PTG
reporting, i.e., younger EAs versus older EAs); 2) comparing the levels of positive psychological
change through corroboration of these subjective reports by an observer (e.g., a participant’s
family member, partner, colleague), thereby providing convergent validity and supporting the
use of the PTGI as an appropriate PTG-measuring assessment (Shakespeare‐Finch & Enders,
2008); 3) developing studies examining the health effects and consequences of careers requiring
heavy internet use; and 4) updating the stressful life event items to include current events that
may be considered stressful by EAs (e.g., how and to what extent the COVID-19 pandemic has
shaped health outcomes among EAs; the concern for future pandemics; shootings and deaths
related to shootings; political events, such as presidential, senatorial, and local elections;
inflation and its consequences, such as food instability and inability to purchase basic needs;
global warming and its consequences; racism and hate crimes; gender discrimination; wars and
war crimes, etc.) and incorporating an item to measure the self-perceived severity and threat of
the event to EAs’ health and wellbeing.
Despite the limitations, the present study observed that CIU was significantly negatively
associated with PTG in terms of within-school differences. EAs differ in life experiences, and
even if some have similar histories and backgrounds, their health outcomes may still vary
greatly (Frye & Liem, 2011; Halfon et al., 2018), which implies that individual factors may be

44
influencing those differences. Schmidt et al. (2019) suggested that there is a difference in PTG
experienced between those who have experienced trauma versus those who are experiencing
extreme shifts in worldviews—which also elicit PTG—and future studies are needed to
distinguish those differences in order to inform assessment and program designs to cultivate and
encourage the development of PTG. As CIU prevalence is particularly high among EAs, future
studies could explore whether PTG has a direct or indirect effect on preventing or ameliorating
the effects of CIU on EAs’ health and wellbeing.

45
Chapter 3:
Study 2: Exploring Associations between Stress and Compulsive Internet Use Moderated by
Posttraumatic Growth
Introduction
Emerging adulthood is a developmental period where emerging adults (EAs) undergo
multiple stressors in life as they transition to becoming adults (Wood et al., 2018), and they must
deal with the stress, either in healthy or unhealthy ways (Murray et al, 2020). Particularly among
EAs, a common coping mechanism to deal with stress is by using the internet to escape from the
stressors (Tang et al., 2017; Fernandes et al., 2020). Internet use can also be broken down into
specific components, such as internet browsing, using social networking applications (e.g.,
Facebook, Myspace, Twitter, etc.), and online shopping. However, some individuals may exceed
normative internet use to excessive amounts—to the point of being considered a problematic
behavior or addiction—displaying loss of control, conflict, preoccupation, coping or mood
modification, and withdrawal symptoms (based on DSM-IV CIU criteria; APA, 1994). While
numerous coping mechanisms exist, the current study will focus on the harmful coping method
of compulsive internet use (CIU). The physical, mental, and emotional risk factors associated
with CIU, such as mental health issues, stress, and decreased well-being in general (van Rooij et
al., 2010; Donald et al., 2019) negatively impact health. To compound that, EAs may be likely to
continue risky behaviors, in this case CIU, into adulthood (Arnett et al., 2014; Anderson et al.,
2017).
Even though the prevalence of CIU among EAs is relatively high (Tang et al., 2017;
Perrin & Atske, 2021), there is currently a lack of theory-based intervention programming

46
targeting the prevention of CIU development among young adults (Chamberlain et al., 2018;
Pettorruso et al., 2020; Romero et al., 2021). This may be due to a lack of standardized criteria
for CIU (also known as problematic internet use or internet dependence; van Rooij et al., 2010;
Donald et al., 2019; Griffiths, 2020) as well as a lack of integrated CIU research. As a result,
there is also the issue of inconsistent assessments of CIU, with prevalence estimates ranging
widely, partly due to methodological differences (Burkauskas et al., 2022). Regardless of this,
CIU still presents health-comprising risk factors and needs to be addressed, particularly among
EAs.
If starting with the premise that extreme stress among EAs leads to CIU to cope with that
stress, then a plausible direction would be to examine the effects of external factors influencing
the strength of that association, i.e., a moderator. A potential moderator of stress and CIU among
EAs could be posttraumatic growth (PTG), which is the “positive psychological change
experienced as a result of the struggle with highly challenging life circumstances” (Tedeschi &
Calhoun, 2004). While psychological distress and negative reactions to stressful life events
(SLEs) are common, there are also positive reactions to SLEs, which promote the development
of PTG among those facing an extensive range of stressful and traumatic experiences (Tedeschi
& Calhoun, 2004; Arpawong et al., 2015; 2016). It has been established that PTG can be
cultivated as a response to various stressors (Arpawong et al., 2016) that the undergoing
individual perceives to be impactful enough to result in pivotal life changes (Tedeschi et al.,
2018). A previous study examined the association between PTG and substance use in students
participating in a drug abuse prevention program and found that higher PTG scores were
associated with lower recurrences of alcohol and marijuana use and less substance abuse at twoyear follow up (Arpawong et al., 2015). Another study found that CIU was significantly

47
negatively associated with PTG: those with higher PTG tended to report lower CIU (Yu, 2024).
Building upon those studies, PTG could serve as a moderator on the association of cumulative
SLEs and CIU (as well as to the components of CIU, e.g., internet browsing, social networking,
and online shopping), decreasing the strength of the stress-internet use association.
Proposed Aims and Hypotheses
Aim 1. To examine whether PTG or SLEs affected the rate of change in CIU. SLEs were
hypothesized to result in steeper increases in CIU than PTG. To also examine the moderating
effect of PTG on the association of cumulative stress and general CIU. In terms of main effects,
stress was hypothesized to be positively associated with CIU (i.e., as cumulative stress increases,
CIU would also increase), and PTG was hypothesized to be negatively associated with CIU (i.e.,
as PTG increases, CIU would decrease). In terms of moderating effects, the interaction of PTG
and stress would be negatively associated with CIU: the stress-CIU association would be
moderated by PTG in that those with higher PTG would report lower CIU with increasing stress
while those with lower PTG would report higher CIU with increasing stress.
Aim 2. To examine if PTG also had a moderating effect on the association between
cumulative SLEs and the specific internet use behaviors of internet browsing, social networking,
and online shopping. The same moderating effects as the ones in Aim 1 were hypothesized for
Aim 2, that the interaction of PTG and SLEs would be negatively associated with self-perceived
accounts of internet browsing, social network usage, and online shopping: PTG would moderate
the stress-CIU association by decreasing the strength of the association. That is, with increasing
stress, those who endorsed positive PTG would report having less self-perceived addictions to
internet browsing, social network usage, and online shopping.

48
Methods
Participants were emerging adults (average age of 20 year, ranging from 17 to 23) who
were former alternative high school attendees in a southern California county and had
participated in a larger longitudinal study testing the efficacy of a school-based substance abuse
prevention program, Project Towards No Drug Abuse (TND), consisting of curriculum focusing
on developing motivation, skills, and decision-making abilities to prevent risky health behaviors
(Sussman et al., 2012).
A total of 24 schools participated, with each school having an average of 70 students
participate (range: 56-103 students). Participants’ informed consent were obtained prior to data
collection. Each student was asked their age, gender, and ethnicity in the initial baseline pretest,
which were designated as demographic covariates. Starting in Year 3 (2011-2012) of data
collection, items assessing self-perceived addiction to various substances and activities were
included and followed-up for two years, which is why the present study uses data starting from
Year 3 to Year 5 (2013-15). Since assessing the program effects was not the main focus of this
study, TND participation was included as a control variable (refer to Measures section below).
Measures
Demographics. In 2008, baseline data collection included the demographic variables of
age, gender (1 = male, 0 = female), and race/ethnicity (1 = Asian or Asian American, 2 = Latino
or Hispanic, 3 = African American or Black, 4 = White, Caucasian; not Hispanic, 5 = American
Indian or Native American, 6 = Mixed, 7 = Other). Due to insufficient sample sizes for some of
the race/ethnicity categories, a dichotomous variable was created to recode race as Hispanic/non-

49
Hispanic (1 = Hispanic, 0 = non-Hispanic). At baseline, Project TND study condition was also
recorded (1 = any TND component (intervention or intervention plus motivational booster, 0 =
no TND component/control).
Stressful Life Events. SLEs were repeatedly measured in Years 3, 4, and 5. Nine items
adapted from a shortened version of the Adolescent Negative Life Events Inventory (Newcomb
& Harlow, 1986; Wills et al., 1992) asked if participants experienced the following events in the
past two years: “I got disciplined or suspended from school or work," "Someone in my family
had a serious illness, accident, or injury," "I did not have enough money for basics (like food),"
"A new person joined the household (baby or young child, grandparent, stepbrother or sister,
stepparent, other)," "I was a victim of a violent or abusive crime," "Someone in my family or I
was arrested," "I broke up with my girlfriend/boyfriend/partner," "There were a lot of arguments
that happened at home," and “Other.” Of those nine options, participants were asked to indicate
which event, if any, affected them the most (i.e., was their most life-altering event, MLAE) as
this was relevant in assessing PTG. Some participants who reported experiencing SLEs did not
specify which event was their MLAE, and some reported that multiple SLEs were equally lifealtering.
The questionnaire did not assess the severity of the SLEs, so a weighted score of the
various life events could not be created because each person had their own way of interpreting
and processing SLEs. Instead, the number of SLEs (among those who indicated experiencing at
least one or more SLEs) was summed as a total—since cumulative stress has been found to be
related to PTG—and was entered as a continuous correlate. While there may be several
explanations for relating cumulative stress to PTG, one rationale is that more SLEs reported may
be interrelated, particularly to an event that was reported as most life altering (Arpawong et al.,

50
2016). Another is that more SLEs reported may affect how individuals adapt to adversity: if the
stressors are perceived to be challenges to overcome, one may use the adversity to reach an
optimal functionality/performance level and build resilience, e.g., emotional resilience, better
tolerance to pain, and increased cognitive function (Robertson, 2017).
Posttraumatic Growth. PTG was repeatedly measured in Years 3, 4, and 5. Individuals
were asked to respond to the PTG items in relation to their reported most life-altering life
event(s). Eight items adapted from the Posttraumatic Growth Inventory (PTGI) (Tedeschi &
Calhoun, 1996; Milam et al., 2004; Milam, 2006; Arpawong et al., 2015) assessed changes
(negative change, no change, or positive change) in the participants’ lives that may have
occurred since experiencing the life-altering event(s) in the following: appreciation for the value
of their own life, direction for their life, handling their difficulties, their understanding of
spiritual matters, their sense of closeness with others, involvement in things that interest them,
their compassion for others, and their own inner strength. Responses were measured on the
following 3-point Likert-type scale: 1 (negative change/got worse), 2 (no change), and 3
(positive change/got better). Items were recoded as negative change (-1), no change (0), and
positive change (1). PTG scores were averaged across the eight items according to the PTG
coding of -1, 0, and 1 to gauge whether there was an overall negative, positive, or no change in a
participants’ life. Cronbach’s alpha for this scale, on average across the three years, was 0.84.
To clarify, the scale used in this study broadly measured PTG as negative, neutral, or
positive change, as compared to other studies that measured the degree of PTG using a Likert
scale (e.g., the original PTG Inventory measured on a 6-point Likert scale, with a score of 0
being “I did not experience this change as a result of my crisis” to a score of 5 being “I
experienced this change to a very great degree as a result of my crisis”; Tedeschi & Calhoun,

51
1996). Therefore, the results of this study solely indicated whether PTG was negatively, neutrally,
or positively present, rather than being able to specifically distinguish higher PTG levels from
lower PTG levels.
Compulsive internet use. CIU was repeatedly measured in Years 3, 4, and 5. Four items
adapted from the Compulsive Internet Use (CIU) Scale (Meerkerk et al., 2009) asked how often
the participants: 1) stayed on the internet longer than planned, 2) used the internet more than they
ought, 3) could not cut down internet use despite wanting to, and 4) felt that their internet use
seemed beyond their control. Responses were measured on the following 5-point Likert-type
scale: 1 (never), 2 (rarely), 3 (sometimes), 4 (most of the time), and 5 (always). For data analysis,
this variable was re-coded as 0-4, with 0 representing no CIU issues versus 4 representing
constant CIU issues. The CIU items were used as a sum score (0-16), with a higher score
indicating a greater level of CIU. Cronbach’s alpha for this scale, on average across the three
years, was 0.88.
Self-perceived addictions. These items were repeatedly measured in Years 3, 4, and 5.
Starting in Year 3, participants were asked if they perceived themselves to be “addicted” to
certain substances or activities in the last 30 days, with an explanation of “addiction” as an
occurrence that 1) is repeated over and over again to try to feel good, for excitement, or to stop
feeling bad; 2) cannot be stopped even if they wanted to stop; and 3) causes bad things to happen
to them or to the people they care about because of what they are doing. Self-perception has been
widely used in studies as an indicator of a condition or disease (Ballestar-Tarín et al., 2020), and
in this case, items related to CIU included internet browsing (surfing the web), Facebook or other
online social networking, and online shopping. Responses to the self-perceived addictions were
measured as Yes (1) or No (0).

52
Sample exclusion criteria
Of the 1676 participants from baseline, 42 did not indicate their ethnicity. Of the
remaining 1634, 585 were eligible to participate from Year 3, 518 from Year 4, and 483 from
Year 5. There were 760 eligible from all Years, and 350 were excluded for only having one data
point—this was because two or more data points were believed to give a better estimate of rates
of change over time. The final sample for the current study was 410 participants (refer to Figure
3).
Figure 3: Study 2 sample exclusion criteria. (Note in Year 5 exclusions: SPA = self-perceived addictions)

53
Assessing attrition
To ensure that the attrition rate (76%) did not result in the working follow-up sample
being significantly different from the baseline sample, t-tests were performed to see if there were
differences in baseline CIU sum scores and demographics between the working follow-up
sample (n=410) and those who were not (n=1266). CIU was compared since the variable was
assessed at baseline. There were no significant differences between the baseline sample lost to
follow-up and the remaining group in Hispanic status and age. (p=0.23 and 0.26, respectively).
However, there were differences in TND participation, gender, and internet use sum scores
(p=0.05, p<.0001, p<0.001), in that more of the lost-to-follow-up group had participated in TND
(68% vs 63%), identified as males (61% vs 47%), and had lower baseline CIU (mean sum score
9.38 vs 10.12) than the working sample. Because the present study’s focus was on the
moderation effects of PTG, rather than CIU prevalence, the CIU differences may not be so
relevant; however, future studies could use a propensity score to account for attrition.
Analyses
All statistical analyses were conducted using SAS 9.4. Multilevel models (MLM) were
used to account for nesting (Bauer, D., J., & Curran, P. J., 2023). A total of 1,001 repeated
measures were nested within 410 students, who were nested within 24 schools. Of the 410
students, 229 participated once and 181 participated all three times during the three years of data
collection.
To address Aim 1, a three-level longitudinal model of CIU was run to determine whether
PTG and cumulative SLEs affected rate of change in CIU over time by testing the slope
interactions with time. A separate model examined the moderating effects of PTG on the

54
association between cumulative stress and CIU by testing the interaction of PTG and SLEs on
CIU. Time-invariant demographic covariates were added to the model as control variables. To
address Aim 2, moderation models similar to that of Aim 1 were run, testing the interaction of
PTG and SLEs on self-perceived reports of internet browsing, social networking, and online
shopping.
Results
Study 2 sample characteristics
Of the 410 eligible participants, slightly less than half (47%) identified as male, on
average was aged 20 years old (ranging from 17 to 23 years) at three year follow-up from
baseline, and nearly two-thirds (67%) identified as Latinx/Hispanic (refer to Table 8).
Table 8. Study 2 sample demographics (n=410).
Variable Freq (%) or Mean (SD)
Gender
Female 216 (52.68)
Male 194 (47.32)
Age (at Year 3 assessment) 19.94 (0.89)
Race/ethnicity
Asian or Asian American 12 (2.93)
Latinx or Hispanic 273 (66.59)
African American or Black 12 (2.93)
White, Caucasian, Anglo; not Hispanic 46 (11.22)
Other1 67 (16.34)
Assigned study condition
TND 259 (63.17)
Control 151 (36.83)
Notes. 1Other includes American Indian/Native American and mixed ethnicities.
The mean cumulative number of SLEs across time was 2.67 (SD=1.56), with values
ranging from zero to eight. The mean sum score for CIU across time was 4.90 (SD=3.59), with
values ranging from one to 16. The mean PTG score across time was 0.65 (SD=0.38), with
values ranging from negative one to one, indicating that most participants reported relatively

55
positive changes (i.e., some aspects of improvement in their life) after having experienced SLEs
in the past two years (refer to Table 9).
Table 9. Sample means of relevant variables by data collection year.
Year 3 Year 4 Year 5 Overall
n 325 349 327 410
Variable Mean(SD) Mean(SD) Mean(SD) Mean(SD)
SLEs1 2.93 (1.54) 2.59 (1.52) 2.49 (1.60) 2.67 (1.56)
CIU2 4.54 (3.40) 5.08 (3.74) 5.06 (3.61) 4.90 (3.59)
PTG3 0.66 (0.39) 0.66 (0.37) 0.64 (0.39) 0.65 (0.38)
Notes. 1SLEs: Stressful life events, cumulative, range: 0-8.
2CIU: Compulsive internet use sum score, range: 1-16.
3PTG: Posttraumatic growth average score, range: -1 to 1.
Aim 1 results
CIU three-level model results (refer to Table 10): Testing for random effects in the
unconditional linear growth curve, the random intercepts-only model fit better than one without
(LRT p<.0001), but the random intercepts-random slopes model did not fit better than the
random intercepts-only model (LRT p=0.14). Moving forward with the random intercepts-only
model, the fixed effect of CIU reported an overall mean estimate of 5.16 (SE=0.38) across all
schools and students. There were no significant demographic differences in CIU by gender,
Hispanic status, or TND participation. The average rate of change for CIU slightly increased
over time by 0.24 points (p<0.05). According to the variance component estimates, the intraclass
correlation coefficient (ICC) indicated that all of the individual differences in CIU could be
attributed to differences within schools. However, there was a between-school effect on the rate
of change of CIU over time: on average, schools with higher mean SLEs had students with
higher increases of CIU over the course of three years (refer to Figure 4).

56
Table 10. Longitudinal growth curve model of CIU over time (n=410).
Notes. 1
id=student ID, sid=school ID.
2WSC: within-school-centered variable. 3BSC: between-school-centered
variable. *p<0.05. **p<0.01. ***p<0.001.
Figure 4: Probing the significant interaction of time and between-school-SLEs revealed diverging trajectories. The
increase of CIU is steeper for students from schools with higher cumulative stress while CIU is relatively steady for
students from schools with lower cumulative stress reported.
Fixed Effects β SE
Intercept 5.1632*** 0.3774
Time 0.2416* 0.1123
Gender 0.0973 0.2915
Ethnicity -0.3681 0.3101
TND participation -0.4965 0.3116
SLEs_WSC2
-0.0584 0.1060
SLEs_BSC3
-0.7921 0.7069
Time*SLEs_WSC 0.0537 0.0773
Time*SLEs_BSC 0.8629* 0.4249
PTG_WSC -2.3383 1.5036
PTG_BSC 2.5555 9.4645
Time*PTG_WSC 1.3776 1.1034
Time*PTG_BSC 10.7694 5.6108
Random Effects β SE
UN(1,1) id(sid)1 5.4326*** 0.6317
UN(1,1) sid 0 .
Residual 7.4300*** 0.4350

57
Moderation model of PTG on the association of SLEs and CIU (refer to Table 11). Since
ICC indicated that all of the individual differences in CIU could be attributed to differences
within schools, the within-school time-varying variables of SLEs and PTG were kept, and the
interaction of the two was also added. Keeping everything constant, PTG was negatively
associated with CIU: higher PTG was associated with lower CIU. And although the moderation
effect was not significant, probing the interaction estimates revealed that PTG did have a
moderating effect on the association of stress and CIU (refer to Figure 5): with increasing
cumulative stress, CIU increased when negative change was perceived (PTG closer to a score of
-1) while CIU decreased when positive change was perceived (PTG closer to a score of 1).
Table 11. Moderation model of PTG on SLEs and CIU (n=410).
Notes. *p<0.05. **p<0.01. ***p<0.001.
Fixed Effects β SE
Intercept 5.0261*** 0.3714
Time 0.2381* 0.1126
Gender 0.1310 0.2917
Ethnicity -0.3380 0.3090
TND participation -0.3511 0.3007
SLEs_WSC -0.0056 0.0732
PTG_WSC -1.0243 1.0222
SLEs_WSC*
PTG_WSC
-0.2415 0.6033
Random Effects β SE
UN(1,1) id(sid) 5.4571*** 0.6330
UN(1,1) sid 0 .
Residual 7.4843*** 0.4350

58
Figure 5: The moderating effect of PTG on the association of SLEs and CIU.
Aim 2 results
Moderation model of PTG on the association of SLEs and self-perceived addiction to
internet browsing (refer to Table 12). Similar to the moderation model in Aim 1, the withinschool time-varying variables of SLEs and PTG were kept, and the interaction of the two was
added. Keeping everything constant, PTG was negatively associated with internet browsing:
those with higher PTG were less likely to report internet browsing addiction. The moderation
effect was not significant and probing the interaction estimates revealed that PTG had a positive
moderating effect on the association of stress and self-perceived internet browsing addiction
(refer to Figure 6): with increasing cumulative stress, those who reported positive change (PTG
closer to a score of 1) were more likely to perceive themselves to be addicted to internet

59
browsing, while those who reported negative change (PTG closer to a score of -1) were less
likely to perceive themselves to be addicted to internet browsing.
Table 12. Moderation model of PTG on SLEs and internet browsing (n=410).
Notes. *p<0.05. **p<0.01. ***p<0.001.
Figure 6: The moderating effect of PTG on the association of SLEs and self-perceived addiction to internet
browsing.
Fixed Effects β SE
Intercept 0.1417*** 0.3714
Time 0.0273* 0.0121
Gender 0.0268 0.0292
Ethnicity 0.0002 0.0310
TND participation -0.0271 0.0301
SLEs_WSC 0.0021 0.0077
PTG_WSC -0.3001** 0.1074
SLEs_WSC*
PTG_WSC
-0.0356 0.0637
Random Effects β SE
UN(1,1) id(sid) 0.0499
*** 0.0063
UN(1,1) sid 0 .
Residual 0.0870*** 0.0050

60
Moderation model of PTG on the association of SLEs and self-perceived addiction to
social networking (refer to Table 13). Keeping everything constant, PTG was negatively
associated with social networking: those with higher PTG were less likely to report social
networking addiction. The moderation effect was reaching significance (p=0.085) and probing
the interaction estimates revealed that PTG had a moderating effect on the association of stress
and self-perceived social networking addiction (refer to Figure 7): with increasing cumulative
stress, while all who perceived either negative or positive change were likely to perceive
themselves to be addicted to social networking, those who reported negative change had a
steeper increase in reporting social networking addiction than those who reported positive
change.
Table 13. Moderation model of PTG on SLEs and social networking (n=410).
Notes. *p<0.05. **p<0.01. ***p<0.001.
Fixed Effects β SE
Intercept 0.2131*** 0.0393
Time -0.0058 0.0139
Gender -0.0489 0.0303
Ethnicity 0.0427 0.0321
TND participation -0.0163 0.0312
SLEs_WSC 0.0194* 0.0085
PTG_WSC -0.1822 0.1193
SLEs_WSC*
PTG_WSC
-0.1227 0.0713
Random Effects β SE
UN(1,1) id(sid) 0.0435*** 0.0069
UN(1,1) sid 0 .
Residual 0.1175*** 0.0068

61
Figure 7: The moderating effect of PTG on the association of SLEs and self-perceived addiction to social
networking.
Moderation model of PTG on the association of SLEs and self-perceived addiction to
online shopping (refer to Table 14). Keeping everything constant, PTG was negatively associated
with online shopping: those with higher PTG were less likely to report online shopping addiction.
The moderation effect was not significant and probing the interaction estimates revealed that
PTG had a moderating effect on the association of stress and self-perceived online shopping
addiction (refer to Figure 8): with increasing cumulative stress, those who perceived negative
change were constant in perceiving themselves to be addicted to online shopping, while those
who reported positive change had a steeper decrease in reporting online shopping addiction than
those who reported no change (PTG closer to 0).

62
Table 14. Moderation model of PTG on SLEs and online shopping (n=410).
Notes. *p<0.05. **p<0.01. ***p<0.001.
Figure 8: The moderating effect of PTG on the association of SLEs and self-perceived addiction to online shopping.
Fixed Effects β SE
Intercept 0.0436 0.0265
Time 0.0148 0.0082
Gender -0.0045 0.0189
Ethnicity 0.0000 0.0205
TND participation 0.0105 0.0233
SLEs_WSC -0.0036 0.0051
PTG_WSC -0.0225 0.0715
SLEs_WSC*
PTG_WSC
-0.0316 0.0427
Random Effects β SE
UN(1,1) id(sid) 0.0184*** 0.0026
UN(1,1) sid 0.0009 0.0009
Residual 0.0405
*** 0.0023

63
Discussion
Between-school stress effects on rate of change in CIU
While PTG did not influence the rate of change in CIU over time, cumulative SLEs
between schools did have a significant effect: schools with students who reported higher mean
SLEs had faster rates of increasing CIU over time than students who reported lower mean SLEs.
This could inform alternative high school administrations that the school environment may play
a role in encouraging CIU, e.g., perhaps the social norm in a school of high-stress enrollees is to
turn to excessive internet use to escape, find relief, or check out from the stress. Alternative high
schools could consider incorporating accessible stress management methods into school
curriculum, such as mental health counseling, mindfulness classes, pet therapy sessions, etc.
Main effects of PTG on internet use
Consistently, PTG was negatively associated with CIU, internet browsing, social
networking, and online shopping. This aligned with the perspective that those who endorse PTG
generally want to move forward in life and engaging in activities that would deter from growing
(in this case, CIU and internet-related addictions) would be seen as detrimental to their health
and keep them from progressing, so therefore they would try to avoid those activities. There
were also no significant differences by gender, ethnicity, or program participation in CIU,
suggesting that CIU may be a health concern to all regardless of one’s demographic background.
Moderation effects of PTG on stress and internet use
Although moderation effects were not significant, visualizing the interaction estimates
indicated that PTG still played a moderating role on the association of cumulative SLEs and CIU.
With increasing stress, those who endorsed positive PTG reported decreasing CIU, while those
who endorsed negative PTG reported increasing CIU. So, it seems that PTG was a protective

64
factor in pathways where rising stress levels and increasing stressful events lead to CIU as a
coping mechanism.
Evidence has suggested that some may feel compelled to maintain their engagement in
online social networks, which could lead to the excessive use of social networking sites
(Griffiths et al., 2014). The moderation effects of PTG on SLEs and social networking were
reaching significance and indicated that with increasing stress SLEs, social networking increased
regardless of PTG direction. This informs CIU research that compared to other internet use
behaviors, engaging in social networking may be more common and widely used. With
increasing stress, those who endorsed positive PTG perceived themselves to be addicted to social
networking increased at a slower rate than those who endorsed negative PTG. So, in this case, it
seems PTG served as a buffer to rising stress levels against social networking. Similarly, PTG’s
moderating effect on the association of stress and online shopping addiction indicated that with
increasing stress, those who endorsed positive PTG had a steeper decrease in online shopping
than those who reported neutral or negative PTG.
PTG had a moderating effect on the association of stress and internet browsing, but it was
contrary to the present study’s hypothesis: with increasing SLEs, those who endorsed positive
PTG were more likely to perceive themselves to be addicted to internet browsing than those who
endorsed negative PTG. One possible explanation could be that EAs with high PTG who
perceived themselves to be addicted to internet browsing were investing a lot of time browsing
the internet for growth-related activities. For example, job searching could be a time-intensive
task, such as sifting through numerous positions, looking up application requirements, and
frequently checking application sites for status updates. Another example could be browsing the

65
internet to look for self-development opportunities, such as webinars or online training to set the
foundation for work during the transition to adulthood (Wood et al., 2018).
Limitations and future directions
Some limitations of the present study included the results being pertinent to older
alternative high school students—who may have encountered a different high school experience
compared to other students their age—rather than to all EAs in general. In terms of data
limitations, the survey responses from the participants may have self-report and social
desirability bias, especially for the CIU and self-perceived internet-use-related addiction items.
Additionally, the high attrition rate and lower sample size for the present study needs to be noted.
These limitations should be addressed in future studies by sampling a broader range of EAs, e.g.,
those in school, working, or neither. Due to the preexisting inconsistencies of CIU assessment,
with widely-ranging estimates due to methodological differences (Burkauskas et al., 2022), more
efforts are needed in standardizing the term used to describe compulsive/excessive/problematic
internet use behavior, the diagnosing criteria, and assessment items.
Despite these limitations, the present study has found that the hypothesized moderation
effect of PTG on the stress-internet use association was supported, suggesting that PTG may play
a protective role against CIU as well as a buffer against self-perceived internet-related addictions.
While PTG may be realized naturally as a response to trauma, certain methods of self-help
techniques may aid in fostering PTG, such as going through a PTG workbook that guides an
individual’s thought process and encourages self-reflection (Tedeschi et al., 2018). This
workbook approach may be effective among EAs as it is accessible and could be a stress
management method that alternative high schools could incorporate into their curriculum (see
section Between-school stress effects on rate of change in CIU, p.62). For EAs who prefer

66
companionship when reflecting rather than be alone may find use in completing the workbook
with a professional expert counselor or another form of social support therapy. Previous studies
(Tedeschi et al., 2018) have indicated that foundational treatment approaches for trauma are
standard, and trauma therapists could integrate PTG into those approaches. Tailoring trauma
therapy to cultivate PTG and focus on the needs of the client would advance the research field of
PTG. More research is needed to determine the moderating and potential buffering effects of
PTG on weakening the association of SLEs and CIU. Future studies could also explore the
moderation effect of PTG on the association of life-altering stressful/traumatic experiences and
other harmful coping behaviors.

67
Chapter 4: Conclusion
Studying areas of research that may improve EA health outcomes is critical, especially in
identifying harmful health behaviors that may continue into adulthood. Problematic internet use
behaviors and CIU have been found to be prevalent among EAs and harmful to their health and
wellbeing. Literature is currently growing on the notion that PTG could potentially have a
buffering effect where higher PTG endorsement could be associated with health behavior
benefits, such as increased ability to not participate in harmful coping behaviors (Arpawong et al.,
2015; Schmidt et al., 2019). Channeling this buffering effect against harmful coping mechanisms
after experiencing some sort of stressful trauma may help in developing and supplementing
stress management skills that enable EAs to reach a higher level of both life functioning and
psychological wellbeing for a healthier transition to adulthood.
An example of a life-altering stressful/traumatic event affecting EAs is the COVID-19
pandemic. The concerning mental health trends among EAs in the past few years have not only
persisted but have become worse over the years and also exacerbated by the pandemic (e.g.,
social isolation, loneliness, restlessness from lockdown enforcements). According to the Stress in
AmericaTM 2020 report (APA, 2020), the pandemic has brought about a national mental health
crisis with lasting health consequences, including extreme stress and trauma from the devastating
loss of the lives of friends and family, harrowing recovery processes for the infected, and chaotic
upheavals in life (e.g., job loss, financial emergencies, and uncertain futures). Americans of all
ages have been negatively affected by the pandemic, but younger individuals are especially
experiencing high stress levels and reporting depression symptoms due to the unprecedented
uncertainty of the pandemic. In the last 30 days in the Stress in AmericaTM 2020 report, young
adults (ages 18-23) reported significantly higher mean stress levels (6.1 out of 10) compared to

68
older adults (ages 24-41, 42-55, 56-74, and 75+; 5.6, 5.2, 4.0, and 3.3 out of 10, respectively). In
the last two years, young adults have reported increasing stress levels: 5.6 out of 10 in 2018, 5.8
in 2019, and 6.1 in 2020. Several young adults reported experiencing common depression
symptoms in the past two weeks, such as feeling so tired they sat around and did nothing (75%),
felt very restless (74%), found it hard to think properly or concentrate (73%), felt lonely (73%)
or felt miserable or unhappy (71%). Additionally, 76% of 18- to 23-year-olds reported negative
health consequences (e.g., disrupted sleep patterns, 31%; eating more unhealthy foods than usual,
28%; weight changes, 28%; feeling very lonely, 63%; and feeling the need of more emotional
support than they received over the past year, 82%) due to the pandemic, compared to 71% of
24- to 41-year-olds, 59% of 42- to 55-year-olds, 53% of 56- to 74-year-olds, and 28% of 75-
year-old adults and older (APA, 2020).
Regarding stress stemming from the pandemic and issues related to it, this has resulted in
steeply increased problematic online gaming (King et al., 2020) and addictive internet use in
terms of longer use and more dependence on internet-related activities among EAs and adults
(mean age: 28 years, SD = 9.2; Sun et al., 2020). This is just one example of a traumatic SLE
that affects EAs—today, the sources of stress among EAs are increasing (the concern for future
pandemics; the effect of hearing about shootings and deaths related to shootings; political events,
such as presidential, senatorial, and local elections; inflation and its consequences, such as food
instability and inability to purchase basic needs; global warming and its consequences; racism
and hate crimes; gender discrimination; wars and war crimes, etc.), and some EAs may cope with
the stress in unhealthier ways than others. In light of this, future studies may inform how PTG
can alleviate negative health outcomes stemming from harmful coping behaviors caused by
SLEs–and the compounding of multiple events–and improve quality of life for EAs.

69
References
Aldwin. (1994). Stress, coping, and development : an integrative perspective. Guilford Press.
American Psychological Association (APA). (1980). Diagnostic and statistical manual of mental
disorders (3rd ed.).
APA. (1994). Diagnostic and statistical manual of mental disorders (4th ed.).
APA. (2013). Diagnostic and statistical manual of mental disorders (5th ed.).
https://doi.org/10.1176/appi.books.9780890425596
APA. (2019, March 13). Mental health issues increased significantly in young adults over last
decade [Press release]. https://www.apa.org/news/press/releases/2019/03/mental-healthadults
APA. (2019, October 30). How to cope with traumatic stress.
https://www.apa.org/topics/traumatic-stress
APA. (2020, October). Stress in America™ 2020: A National Mental Health Crisis.
https://www.apa.org/news/press/releases/stress/2020/sia-mental-health-crisis.pdf
Anderson, E. L., Steen, E., & Stavropoulos, V. (2017). internet use and Problematic internet Use:
a systematic review of longitudinal research trends in adolescence and emergent
adulthood. International Journal of Adolescence and Youth, 22(4), 430–454.
https://doi.org/10.1080/02673843.2016.1227716
Arpawong, T. E., Sussman, S., Milam, J. E., Unger, J. B., Land, H., Sun, P., & Rohrbach, L. A.
(2015) Post-traumatic growth, stressful life events, and relationships with substance use
behaviors among alternative high school students: A prospective study. Psychology &
Health, 30(4), 475-494. https://doi.org/10.1080/08870446.2014.979171
Arpawong, T. E., Rohrbach, L. A., Milam, J. E., Unger, J. B., Land, H., Sun, P., Spruijt-Metz, D.,
& Sussman, S. (2016). Stressful Life Events and Predictors of Post-traumatic Growth
among High-Risk Early Emerging Adults. The journal of positive psychology, 11(1), 1–
14. https://doi.org/10.1080/17439760.2014.994223
Arnett, J. J. (2000). Emerging adulthood: A theory of development from the late teens through
the twenties. American Psychologist, 55(5), 469-480. https://doi.org/10.1037//0003-
066X.55.5.469
Arnett, J., Žukauskienė, R., & Sugimura, K. (2014). The new life stage of emerging adulthood at
ages 18–29 years: implications for mental health. The Lancet. Psychiatry, 1(7), 569–576.
https://doi.org/10.1016/S2215-0366(14)00080-7

70
Ashing-Giwa, K. T., Padilla, G., Tejero, J., Kraemer, J., Wright, K., Coscarelli, A., Clayton, S.,
Williams, I., & Hills, D. (2004). Understanding the breast cancer experience of women:
A qualitative study of African American, Asian American, Latina and Caucasian cancer
survivors. Psycho-Oncology, 13(6), 408–428. https://doi.org/10.1002/pon.750
Aspinwall, L. G., & Tedeschi, R. G. (2010). Of babies and bathwater: A reply to Coyne and
Tennen’s views on positive psychology and health. Annals of Behavioral Medicine, 39(1),
27-34. https://doi.org/10.1007/s12160-010-9155-y
Ballestar-Tarín, M. L., Simó-Sanz, C., Chover-Sierra, E., Saus-Ortega, C., Casal-Angulo, C., &
Martínez-Sabater, A. (2020). Self-Perception of Dependence as an Indicator of
Smartphone Addiction—Establishment of a Cutoff Point in the SPAI–Spain Inventory.
International Journal of Environmental Research and Public Health, 17(11), 3838.
https://doi.org/10.3390/ijerph17113838
Bauer, D. J., & Curran, P. J. (2023). Multilevel Modeling: SAS Demonstration Notes. CurranBauer Analytics.
Berger. (2015). Stress, Trauma, and Posttraumatic Growth : Social Context, Environment, and
Identities. Routledge.
Block, J. J. (2008). Issues for DSM-V: internet addiction. The American Journal of Psychiatry,
165, 306-307. http://dx.doi.org/10.1176/appi.ajp.2007.07101556
Burkauskas, J., Gecaite-Stonciene, J., Demetrovics, Z., Griffiths, M. D., & Király, O. (2022).
Prevalence of problematic Internet use during the coronavirus disease 2019 pandemic.
Current opinion in behavioral sciences, 46, 101179.
https://doi.org/10.1016/j.cobeha.2022.101179
Chamberlain, S. R., Ioannidis, K., & Grant, J. E. (2018). The impact of comorbid
impulsive/compulsive disorders in problematic Internet use. Journal of Behavioral
Addictions, 7(2), 269–275. https://doi.org/10.1556/2006.7.2018.30
Chun, & Lee, Y. (2008). The Experience of Posttraumatic Growth for People With Spinal Cord
Injury. Qualitative Health Research, 18(7), 877–890.
https://doi.org/10.1177/1049732308318028
Danhauer, S. C., Russell, G. B., Tedeschi, R. G., Jesse, M. T., Vishnevsky, T., Daley, K., Carroll,
S., Triplett, K. N., Calhoun, L. G., Cann, A., & Powell, B. L. (2012). A Longitudinal
Investigation of Posttraumatic Growth in Adult Patients Undergoing Treatment for Acute
Leukemia. Journal of Clinical Psychology in Medical Settings, 20(1), 13–24.
https://doi.org/10.1007/s10880-012-9304-5
Danhauer, S. C., Russell, G., Case, L. D., Sohl, S. J., Tedeschi, R. G., Addington, E. L., Triplett,
K., Van Zee, K. J., Naftalis, E. Z., Levine, B., & Avis, N. E. (2015). Trajectories of
Posttraumatic Growth and Associated Characteristics in Women with Breast Cancer.

71
Annals of Behavioral Medicine, 49(5), 650–659. https://doi.org/10.1007/s12160-015-
9696-1
Dekel, S., Ein-Dor, T., & Solomon, Z. (2012). Posttraumatic Growth and Posttraumatic Distress:
A Longitudinal Study. Psychological Trauma, 4(1), 94–101.
https://doi.org/10.1037/a0021865
Donald, J. N., Ciarrochi, J, Parker, P. D., & Sahdra, B. K. (2019). Compulsive internet use and
the development of self‐esteem and hope: A four‐year longitudinal study. Journal of
Personality, 87(5), 981–995. https://doi.org/10.1111/jopy.12450
Edgerton, J. D., Keough, M. T., & Roberts, L. W. (2018). Co-development of Problem Gambling
and Depression Symptoms in Emerging Adults: A Parallel-Process Latent Class Growth
Model. Journal of Gambling Studies, 34(3), 949–968. https://doi.org/10.1007/s10899-
018-9760-4
Fernandes, B., Biswas, U. N., Tan-Mansukhani, R., Vallejo, A., & Essau, C. A. (2020). The
impact of COVID-19 lockdown on internet use and escapism in adolescents. Revista de
Psicología Clínica Con Niños y Adolescentes, 7(3), 59–65.
https://www.revistapcna.com/sites/default/files/010_0.pdf
Frazier, T., Tennen, H., Gavian, M., Park, C., Tomich, P., & Tashiro, T. (2009). Does SelfReported Posttraumatic Growth Reflect Genuine Positive Change? Psychological Science,
20(7), 912–919. https://doi.org/10.1111/j.1467-9280.2009.02381.x
Frye, A. A., & Liem, J. H. (2011). Diverse Patterns in the Development of Depressive Symptoms
Among Emerging Adults. Journal of Adolescent Research, 26(5), 570–590.
https://doi.org/10.1177/0743558411402336
Glad, K. A., Jensen, T. K., Holt, T., & Ormhaug, S. M. (2013). Exploring self-perceived growth
in a clinical sample of severely traumatized youth. Child Abuse & Neglect, 37(5), 331–
342. https://doi.org/10.1016/j.chiabu.2013.02.007
Griffiths, M. (2005). A 'components' model of addiction within a biopsychosocial framework.
Journal of Substance Use, 10(4), 191–197. https://doi.org/10.1080/14659890500114359
Griffiths, M. D., Kuss, D. J., & Demetrovics, Z. (2014). Social networking addiction: An
overview of preliminary findings. In K. P. Rosenberg & L. Curtiss Feder (Eds.),
Behavioral addictions: Criteria, evidence, and treatment (pp. 119–141). Elsevier
Academic Press. https://doi.org/10.1016/B978-0-12-407724-9.00006-9
Griffiths, M. D. (2020). Internet use disorders: What’s new and what’s not?: Commentary on:
How to overcome taxonomical problems in the study of Internet use disorders and what
to do with “smartphone addiction”? (Montag et al., 2019). Journal of Behavioral
Addictions, 9(4), 934–937. https://doi.org/10.1556/2006.2020.00072

72
Guertler, D., Rumpf, H. J., Bischof, A., Kastirke, N., Petersen, K. U., John, U., & Meyer, C.
(2014). Assessment of problematic internet use by the Compulsive Internet Use Scale and
the Internet Addiction Test: a sample of problematic and pathological gamblers.
European addiction research, 20(2), 75–81. https://doi.org/10.1159/000355076
Guillot, C. R., Bello, M. S., Tsai, J. Y., Huh, J., Leventhal, A. M., & Sussman, S. (2016)
Longitudinal associations between anhedonia and internet-related addictive behaviors in
emerging adults. Computers in Human Behavior, 62, 475-479.
https://doi.org/10.1016/j.chb.2016.04.019
Halfon, N., Forrest, C. B., Lerner, R. M., & Faustman, E. M. (2018). Handbook of Life Course
Health Development (1st ed.). Springer International Publishing.
https://doi.org/10.1007/978-3-319-47143-3
Hanna, K. M., Kaiser, K. L., Brown, S. G., Campbell-Grossman, C., Fial, A., Ford, A., Hudson,
D. B., Keating-Lefler, R., Keeler, H., Moore, T. A., Nelson, A. E., Pelish, P., & Wilhelm,
S. (2018). A Scoping Review of Transitions, Stress, and Adaptation Among Emerging
Adults. Advances in Nursing Science, 41(3), 203–215.
https://doi.org/10.1097/ANS.0000000000000214
Kent, E. E., Alfano, C. M., Smith, A. W., Bernstein, L., McTiernan, A., Baumgartner, K. B., &
Ballard-Barbash, R. (2013). The roles of support seeking and race/ethnicity in
posttraumatic growth among breast cancer survivors. Journal of Psychosocial Oncology,
31(4), 393–412. https://doi.org/10.1080/07347332.2013.798759
King, D. L., Delfabbro, P. H., Billieux, J., & Potenza, M. N. (2020). Problematic online gaming
and the COVID-19 pandemic. Journal of Behavioral Addictions, 9(2), 184–186.
https://doi.org/10.1556/2006.2020.00016
Lisha, N. E., Grana, R., Sun, P., Rohrbach, L., Spruijt-Metz, D., Reifman, A., & Sussman, S.
(2014). Evaluation of the Psychometric Properties of the Revised Inventory of the
Dimensions of Emerging Adulthood (IDEA-R) in a Sample of Continuation High School
Students. Eval Health Prof, 37(2), 156-77. https://doi.org/10.1177/0163278712452664
Jaarsma, T. A., Pool, G., Sanderman, R., & Ranchor, A. V. (2006). Psychometric properties of
the Dutch version of the Posttraumatic Growth Inventory among cancer patients. PsychoOncology, 15, 911–920. https://doi.org/10.1002/pon.1026
Jankowski, P. J., Hall, E., Crabtree, S. A., Sandage, S. J., Bronstein, M., & Sandage, D. (2021).
Risk, Symptoms, and Well‐Being: Emerging Adult Latent Profiles During Treatment.
Journal of Counseling and Development, 99(4), 440–451.
https://doi.org/10.1002/jcad.12396
Jayawickreme, E., & Blackie, L. E. R. (2014). Post-traumatic Growth as Positive Personality
Change: Evidence, Controversies and Future Directions. European Journal of
Personality, 28(4), 312-331. https://doi.org/10.1002/per.1963

73
Joseph, S., & Linley, P. A. (2005). Positive Adjustment to Threatening Events: An Organismic
Valuing Theory of Growth Through Adversity. Review of General Psychology, 9(3),
262–280. https://doi.org/10.1037/1089-2680.9.3.262
Manne, S., Ostroff, J., Winkel, G., Goldstein, L., Fox, K., & Grana, G. (2004). Posttraumatic
growth after breast cancer: Patient, partner, and couple perspectives. Psychosomatic
Medicine, 66(3), 442–454. https://doi.org/10.1097/01.psy.0000127689.38525.7d
Marziliano, A., Tuman, M., & Moyer, A. (2020). The relationship between post-traumatic stress
and post-traumatic growth in cancer patients and survivors: A systematic review and
meta-analysis. Psycho-oncology, 29(4), 604–616. https://doi.org/10.1002/pon.5314
Matud, M., Díaz, A., Bethencourt, J. M., & Ibáñez, I. (2020). Stress and Psychological Distress
in Emerging Adulthood: A Gender Analysis. Journal of Clinical Medicine, 9(9), 2859–.
https://doi.org/10.3390/jcm9092859
Meerkerk, G.-J., Van Den Eijnden, R. J. J. M., Vermulst, A. A., & Garretsen, H. F. L. (2009).
The Compulsive Internet Use Scale (CIUS): Some Psychometric Properties.
Cyberpsychology & Behavior, 12(1), 1–6. https://doi.org/10.1089/cpb.2008.0181
Meyers, L. (2018). The ‘storm and stress’ of adolescence and young adulthood. Counseling
Today, a publication of the American Counseling Association.
https://ct.counseling.org/2018/10/the-storm-and-stress-of-adolescence-and-youngadulthood/
Milam. (2004). Posttraumatic Growth Among HIV/AIDS Patients. Journal of Applied Social
Psychology, 34(11), 2353–2376. https://doi.org/10.1111/j.1559-1816.2004.tb01981.x
Milam, J., Ritt-Olsen, A., & Unger, J. B. (2004). Posttraumatic Growth Among Adolescents.
Journal of Adolescent Research, 19(2), 192-204.
https://doi.org/10.1177/0743558403258273
Milam, J. (2006). Posttraumatic Growth and HIV Disease Progression. Journal of Consulting
and Clinical Psychology, 74(5), 817–827. https://doi.org/10.1037/0022-006X.74.5.817
Murray, K., Crisp, D. A., Burns, R. A., & Byrne, D. (2020). Psychometric Validation of the
Emerging Adult Stress Inventory (EASI). Psychological Assessment, 32(12), 1133–1144.
https://doi.org/10.1037/pas0000952
National Network for Youth. (2022). What Is Trauma? Stress vs. Trauma.
https://nn4youth.org/learn/trauma-informed-care-toolkit/stress-vs-trauma/
Newcomb, M. D., & Harlow, L. L. (1986). Life events and substance use among adolescents:
mediating effects of perceived loss of control and meaninglessness in life. Journal of

74
personality and social psychology, 51(3), 564–577. https://doi.org/10.1037//0022-
3514.51.3.564
Park, C. L., Cohen, L. H., & Murch, R. L. (1996). Assessment and Prediction of Stress-Related
Growth. Journal of Personality, 64(1), 71–105. https://doi.org/10.1111/j.1467-
6494.1996.tb00815.x
Park, C. L. (1998). Stress-related growth and thriving through coping: The roles of personality
and cognitive processes. Journal of Social Issues, 54(2), 267–277.
https://doi.org/10.1111/0022-4537.651998065
Park, C. L., & Fenster, J. R. (2004). Stress-related growth: Predictors of occurrence and
correlates with psychological adjustment. Journal of Social and Clinical Psychology,
23(2), 195–215. https://doi.org/10.1521/jscp.23.2.195.31019
Penedo, F., Molton, I., Dahn, J., Shen, B., Kinsinger, D., Traeger, L., Siegel, S., Schneiderman,
N., & Antoni, M. (2006). A randomized clinical trial of group-based cognitive-behavioral
stress management in localized prostate cancer: Development of stress management skills
improves quality of life and benefit finding. Annals of Behavioral Medicine, 31(3), 261–
270. https://doi.org/10.1207/s15324796abm3103_8
Perrin, A., & Atske, S. (2021, March 26). About three-in-ten U.S. adults say they are ‘almost
constantly’ online. Pew Research Center. https://www.pewresearch.org/shortreads/2021/03/26/about-three-in-ten-u-s-adults-say-they-are-almost-constantly-online/
Pettorruso, M., Valle, S., Cavic, E., Martinotti, G., di Giannantonio, M., & Grant, J. E. (2020).
Problematic Internet use (PIU), personality profiles and emotion dysregulation in a cohort
of young adults: trajectories from risky behaviors to addiction. Psychiatry Research, 289,
1-6. https://doi.org/10.1016/j.psychres.2020.113036
Phipps, S., Long, A. M., & Ogden, J. (2007). Benefit Finding Scale for Children: Preliminary
Findings from a Childhood Cancer Population. Journal of Pediatric Psychology, 32(10),
1264–1271. https://doi.org/10.1093/jpepsy/jsl052
Pies, R. (2009). Should DSM-V Designate "Internet Addiction" a Mental Disorder?. Psychiatry
(Edgmont), 6(2), 31–37. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2719452/
Powers, M. B., Halpern, J. M., Ferenschak, M. P., Gillihan, S. J., & Foa, E. B. (2013). A MetaAnalytic Review of Prolonged Exposure for Posttraumatic Stress Disorder. Focus
(American Psychiatric Publishing), 11(3), 428–436.
https://doi.org/10.1176/appi.focus.11.3.428
Reed, P. G. (1991). Self-Transcendence and Mental Health in Oldest-Old Adults. Nursing
Research, 40(1), 5–11. https://doi.org/10.1097/00006199-199101000-00002

75
Robertson, I. H. (2017). The stress test: Can stress ever be beneficial? Journal of the British
Academy, 5, 163–176. https://doi.org/10.5871/jba/005.163
Rogers, H., & Maytan, M. (2012). Mindfulness for the next generation : helping emerging adults
manage stress and lead healthier lives. Oxford University Press.
https://psycnet.apa.org/record/2012-18602-000
Rohrbach, L. A., Sussman, S., Dent, C. W., & Sun, P. (2005). Tobacco, alcohol, and other drug
use among high-risk young people: A five-year longitudinal study from adolescence to
emerging adulthood. Journal of Drug Issues, 35(2), 333–355.
https://doi.org/10.1177/002204260503500206
Romero Saletti, S. M., Van den Broucke, S., & Chau, C. (2021). The effectiveness of prevention
programs for problematic Internet use in adolescents and youths: A systematic review
and meta-analysis. Cyberpsychology: Journal of Psychosocial Research on Cyberspace,
15(2), Article 10. https://doi.org/10.5817/CP2021-2-10
van Rooij, A. J., Schoenmakers, T. M., van de Eijnden, R. J. J. M., & van de Mheen, D. (2010).
Compulsive internet Use: The Role of Online Gaming and Other Internet Applications.
Journal of Adolescent Health, 47(1), 51-57.
https://doi.org/10.1016/j.jadohealth.2009.12.021
Roth, S., & Cohen, L. J. (1986). Approach, avoidance, and coping with stress. American
Psychologist, 41(7), 813–819. https://doi.org/10.1037/0003-066X.41.7.813
Schmidt, S. D., Blank, T. O., Bellizzi, K. M., & Park, C. L. (2019). Posttraumatic Growth
Reported by Emerging Adults: a Multigroup Analysis of the Roles of Attachment,
Support, and Coping. Current Psychology, 38, 1225–1234.
https://doi.org/10.1007/s12144-017-9670-0
Schneider, S., Rasul, R., Liu, B., Corry, D., Lieberman-Cribbin, W., Watson, A., Kerath, S. M.,
Taioli, E., & Schwartz, R. M. (2019). Examining posttraumatic growth and mental health
difficulties in the aftermath of Hurricane Sandy. Psychological trauma: theory, research,
practice and policy, 11(2), 127–136. https://doi.org/10.1037/tra0000400
Schroevers, M., Kraaij, V., & Garnefski, N. (2007). Goal disturbance, cognitive coping strategies,
and psychological adjustment to different types of stressful life event. Personality and
Individual Differences, 43(2), 413–423. https://doi.org/10.1016/j.paid.2006.12.009
Shakespeare‐Finch, J., & Enders, T. (2008). Corroborating evidence of posttraumatic growth.
Journal of Traumatic Stress: Official Publication of The International Society for
Traumatic Stress Studies, 21(4), 421-424. https://doi.org/10.1002/jts.20347
Smith, Dalen, J., Bernard, J. F., & Baumgartner, K. B. (2008). Posttraumatic Growth in NonHispanic White and Hispanic Women with Cervical Cancer. Journal of Psychosocial
Oncology, 26(4), 91–109. https://doi.org/10.1080/07347330802359768

76
Stanton, A. L., Bower, J. E., & Low, C. A. (2014). Posttraumatic growth after cancer. In
Handbook of Posttraumatic Growth: Research and Practice (pp. 138–175).
Steger, M. F., Oishi, S., & Kashdan, T. B. (2009). Meaning in life across the life span: Levels
and correlates of meaning in life from emerging adulthood to older adulthood. The
Journal of Positive Psychology, 4(1), 43–52.
https://doi.org/10.1080/17439760802303127
Sumalla, E. C., Ochoa, C., & Blanco, I. (2009). Posttraumatic growth in cancer: Reality or
illusion? Clinical Psychology Review, 29(1), 24–33.
https://doi.org/10.1016/j.cpr.2008.09.006
Sussman, S. (2017). Substance and Behavioral Addictions: Concepts, Causes, and Cures.
Cambridge University Press.
Sussman, S., & Arnett, J. J. (2014). Emerging Adulthood: Developmental Period Facilitative of
the Addictions. Evaluation & the Health Professions, 37(2), 147–155.
https://doi.org/10.1177/0163278714521812
Sussman, S., Sun, P., Rohrbach, L. A., & Spruijt-Metz, D. (2012). One-Year Outcomes of a
Drug Abuse Prevention Program for Older Teens and Emerging Adults: Evaluating a
Motivational Interviewing Booster Component. Health Psychology, 31(4), 476–485.
https://doi.org/10.1037/a0025756
Swickert, R., & Hittner, J. (2009). Social Support Coping Mediates the Relationship between
Gender and Posttraumatic Growth. Journal of Health Psychology, 14(3), 387–393.
https://doi.org/10.1177/1359105308101677
Taku, K., & Cann, A. (2014). Cross-National and Religious Relationships With Posttraumatic
Growth: The Role of Individual Differences and Perceptions of the Triggering Event.
Journal of Cross-Cultural Psychology, 45(4), 601–617.
https://doi.org/10.1177/0022022113520074
Tang, C. S.-K., Koh, Y. W., & Gan, Y. (2017). Addiction to Internet Use, Online Gaming, and
Online Social Networking Among Young Adults in China, Singapore, and the United
States. Asia-Pacific Journal of Public Health, 29(8), 673–682.
https://doi.org/10.1177/1010539517739558
Taylor, S. E. (1983). Adjustment to threatening events: A theory of cognitive adaptation.
American Psychologist, 38(11), 1161–1173. https://doi.org/10.1037/0003-
066X.38.11.1161
Taylor, S. E., Kemeny, M. E., Reed, G. M., Bower, J. E., & Gruenewald, T. L. (2000).
Psychological resources, positive illusions, and health. American psychologist, 55(1), 99-
109. https://doi.org/10.1037/0003-066X.55.1.99

77
Tedeschi, R., & Calhoun, L. (1996). The Posttraumatic Growth Inventory: Measuring the
positive legacy of trauma. Journal of Traumatic Stress, 9(3), 455–471.
https://doi.org/10.1007/BF02103658
Tedeschi, R. G., Park, C. L., & Calhoun, L. G. (1998). Posttraumatic growth : positive changes
in the aftermath of crisis. Erlbaum.
Tedeschi, R. G. (1999). Violence transformed: Posttraumatic growth in survivors and their
societies. Aggression and Violent Behavior, 4(3), 319-341.
https://doi.org/10.1016/S1359-1789(98)00005-6
Tedeschi, R. G., & Calhoun, L. G. (2004). Posttraumatic Growth: Conceptual Foundations and
Empirical Evidence. Psychological Inquiry, 15(1), 1-18.
https://doi.org/10.1207/s15327965pli1501_01
Tedeschi, R. G., Shakespeare-Finch, J., Taku, K., & Calhoun, L.G. (2018). Posttraumatic
Growth: Theory, Research, and Applications. Taylor & Francis Group.
https://doi.org/10.4324/9781315527451
Tobin, J. L., Allem, J.-P., Slaughter, R., Unger, J. B., Hamilton, A. S., & Milam, J. E. (2018).
Posttraumatic growth among childhood cancer survivors: Associations with ethnicity,
acculturation, and religious service attendance. Journal of Psychosocial Oncology, 36(2),
175–188. https://doi.org/10.1080/07347332.2017.1365799
Tomich, P. L., & Helgeson, V. S. (2012). Posttraumatic Growth Following Cancer: Links to
Quality of Life: Posttraumatic Growth. Journal of Traumatic Stress, 25(5), 567–573.
https://doi.org/10.1002/jts.21738
Tsai, J. Sippel, L. M., Mota, N., Southwick, S. M., & Pietrzak, R. H. (2016). Longitudinal
Course of Posttraumatic Growth Among U.S. Military Veterans: Results from the
National Health and Resilience in Veterans Study: Longitudinal Course of Posttraumatic
Growth. Depression and Anxiety, 33(1), 9–18. https://doi.org/10.1002/da.22371
Twenge, J. M., Cooper, A. B., Joiner, T. E., Duffy, M. E., & Binau, S. G. (2019). Age, period,
and cohort trends in mood disorder indicators and suicide-related outcomes in a
nationally representative dataset, 2005-2017. Journal of abnormal psychology, 128(3),
185–199. https://doi.org/10.1037/abn0000410
Vishnevsky, T., Cann, A., Calhoun, L. G., Tedeschi, R. G., & Demakis, G. J. (2010). Gender
Differences in Self-Reported Posttraumatic Growth: A Meta-Analysis. Psychology of
Women Quarterly, 34(1), 110–120. https://doi.org/10.1111/j.1471-6402.2009.01546.x
Wills, T. A., Vaccaro, D., & McNamara, G. (1992). The role of life events, family support, and
competence in adolescent substance use: a test of vulnerability and protective factors.

78
American journal of community psychology, 20(3), 349–374.
https://doi.org/10.1007/BF00937914
Wood, D., Crapnell, T., Lau, L., Bennett, A., Lotstein, D., Ferris, M., & Kuo, A. (2018)
Emerging Adulthood as a Critical Stage in the Life Course. In: Halfon N., Forrest C.,
Lerner R., Faustman E. (eds) Handbook of Life Course Health Development (pp.123-143).
Springer, Cham. https://doi.org/10.1007/978-3-319-47143-3_7
Yu, S. (2024). Prospective Associations of Stress, Compulsive Internet Use, and Posttraumatic
Growth Among Emerging Adults. [Unpublished manuscript].
Zoellner, T., & Maercker, A. (2006). Posttraumatic growth in clinical psychology — A critical
review and introduction of a two component model. Clinical Psychology Review, 26(5),
626–653. https://doi.org/10.1016/j.cpr.2006.01.008

79
Appendix A: Tables
Table 1. Sample demographics (n=882).
Variable Freq (%) or Mean (SD)
Gender
Female 396 (44.90)
Male 486 (55.10)
Age (at Year 2 assessment) 18.92 (0.92)
Race/ethnicity
Asian or Asian American 22 (2.49)
Latinx or Hispanic 583 (66.10)
African American or Black 31 (3.51)
White, Caucasian, Anglo; not Hispanic 97 (11.00)
Other1 149 (16.89)
Assigned study condition
TND 595 (67.46)
Control 287 (32.54)
Notes. 1Other includes American Indian/Native American and mixed ethnicities.
Table 2. Study 1 sample means of relevant variables by data collection year.
Year 2 Year 3 Year 4 Year 5 Overall
n 596 605 534 493 882
Variable Mean(SD) Mean(SD) Mean(SD) Mean(SD) Mean(SD)
SLEs1 3.09 (1.75) 2.84 (1.61) 2.48 (1.54) 2.47 (1.54) 2.74 (1.64)
CIU2 3.39 (3.51) 3.52 (3.57) 4.01 (3.92) 4.26 (3.97) 3.77 (3.74)
PTG3 2.64 (0.38) 2.67 (0.38) 2.66 (0.38) 2.66 (0.39) 2.66 (0.38)
Notes. 1SLEs: Stressful life events, cumulative, range: 0-9. 2CIU: Compulsive internet use sum score, range: 0-16.
3PTG: Posttraumatic growth average score, range: 1-3.
Table 3. Frequencies of stressful life events (SLEs) reported in Years 2-5.
SLEs Year 2 %1 Year 3 %1 Year 4 %1 Year 5 %1
Got disciplined/suspended from school/work 138 7.50 71 4.13 53 4.00 49 4.03
Family had serious illness, accident, or injury 370 20.10 335 19.47 259 19.53 249 20.48
Did not have enough money for basics 159 8.64 183 10.63 169 12.75 148 12.17
New person joined household 258 14.01 270 15.69 189 14.25 174 14.31
Was victim of violent or abusive crime 60 3.26 53 3.08 37 2.79 37 3.04
Family member or I was arrested 191 10.37 162 9.41 124 9.35 106 8.72
Broke up with girlfriend/boyfriend/partner 273 14.83 267 15.51 190 14.33 168 13.82
Lot of arguments at home 279 15.15 262 15.22 203 15.31 188 15.46
Other 113 6.14 118 6.86 102 7.69 97 7.98
Total SLEs reported 1841 100 1721 100 1326 100 1216 100
Notes. 1Percentage out of the total number of reported SLEs for that year.

80
Table 4. Frequencies of most life-altering events (MLAEs)
1
reported in Years 2-5.
SLEs Year 2 %2 Year 3 %2 Year 4 %2 Year 5 %2
Got disciplined/suspended from school/work 27 4.77 7 1.22 8 1.73 6 1.37
Family had serious illness, accident, or injury 150 26.50 147 25.70 108 23.33 104 23.74
Did not have enough money for basics 35 6.18 32 5.59 44 9.50 32 7.31
New person joined household 62 10.95 78 13.64 79 17.06 63 14.38
Was victim of violent or abusive crime 21 3.71 19 3.32 15 3.24 11 2.51
Family member or I was arrested 63 11.13 38 6.64 35 7.56 30 6.85
Broke up with girlfriend/boyfriend/partner 66 11.66 91 15.91 68 14.69 55 12.56
Lot of arguments at home 68 12.01 81 14.16 47 10.15 51 11.64
Other 74 13.07 79 13.81 59 12.74 86 19.63
Total MLAEs reported 566 100 572 100 463 100 438 100
Notes. 1Respondents were asked to designate one SLE as their most life-altering event (MLAE). While there were
multiple SLEs reported, individuals were asked to designate one MLAE on which to base their responses to the PTG
items. Non-respondents could have experienced an MLAE more than two years ago. 2Percentage out of the total
number of reported MLAEs for that year.
Table 5. Unconditional linear growth curve model of PTG over time (n=882).
Notes. 1
id=student ID, sid=school ID.
*p<0.05. **p<0.01. ***p<0.001.
Table 6. Unconditional linear growth curve model of CIU over time (n=882).
Fixed Effects β SE
Intercept 3.3416*** 0.1497
Time 0.3070*** 0.0613
Random Effects β SE
UN(1,1) id(sid)1 6.3333*** 0.6713
UN(2,1) id(sid) -0.1377 0.2606
UN(2,2) id(sid) 0.6475*** 0.1553
UN(1,1) sid 0.1600 0.1477
UN(2,1) sid -0.0382 0.0372
UN(2,2) sid 0 .
Residual 5.9393*** 0.2905
Notes. 1
id=student ID, sid=school ID.
*p<0.05. **p<0.01. ***p<0.001.
Fixed Effects β SE
Intercept 2.6540*** 0.0157
Time 0.0041 0.0065
Random Effects β SE
UN(1,1) id(sid)1 0.0414*** 0.0043
UN(1,1) sid 0.0016 0.0012
Residual 0.1041*** 0.0039

81
Table 7. Conditional linear growth curve model of PTG with covariates and predictor variables (n=882).
1. Time-invariant covariates only
Fixed Effects β SE
Intercept 2.5842*** 0.0288
Time 0.0037 0.0085
Gender -0.0453* 0.0200
Hispanic status 0.0693
** 0.0214
TND participation 0.0716** 0.0249
Random Effects β SE
UN(1,1) id(sid)1 0.0477*** 0.0079
UN(2,1) id(sid) -0.0056 0.0034
UN(2,2) id(sid) 0.0048** 0.0020
UN(1,1) sid 0.0001 0.0014
UN(2,1) sid -0.0001 0.0007
UN(2,2) sid 0.0006 0.0005
Residual 0.0962*** 0.0045
2. TICs2 + SLEs
Fixed Effects β SE
Intercept 2.5994*** 0.0290
Time -0.0025 0.0083
Gender -0.0470* 0.0197
Hispanic status 0.0614** 0.0215
TND participation 0.0700* 0.0257
SLEs: WSC3
-0.0278*** 0.0051
SLEs: BSC4
-0.0229 0.0477
Random Effects β SE
UN(1,1) id(sid) 0.0457*** 0.0079
UN(2,1) id(sid) -0.0055 0.0034
UN(2,2) id(sid) 0.0046* 0.0020
UN(1,1) sid 0.0003 0.0015
UN(2,1) sid -0.0001 0.0007
UN(2,2) sid 0.0005 0.0005
Residual 0.0961*** 0.0045
3. TICs + SLEs + CIU
Fixed Effects β SE
Intercept 2.5969*** 0.0295
Time -0.0006 0.0082
Gender -0.0457* 0.0197
Hispanic status 0.0577** 0.0216
TND participation 0.0727* 0.0271
SLEs: WSC -0.0273*** 0.0051
SLEs: BSC -0.0217 0.0486
CIU: WSC -0.0053* 0.0023
CIU: BSC 0.0053 0.0199
Random Effects β SE
UN(1,1) id(sid) 0.0459*** 0.0078
UN(2,1) id(sid) -0.0057 0.0034
UN(2,2) id(sid) 0.0048** 0.0020
UN(1,1) sid 0.0003 0.0015
UN(2,1) sid 0.00001 0.0007
UN(2,2) sid 0.0004 0.0005
Residual 0.0957*** 0.0045
Notes. 1
id: student ID, sid: school ID.
2TIC: time-invariant covariates. 3WSC: within-school-centered variable. 4BSC:
between-school-centered variable. *p<0.05. **p<0.01. ***p<0.001.

82
Table 8. Study 2 sample demographics (n=410).
Variable Freq (%) or Mean (SD)
Gender
Female 216 (52.68)
Male 194 (47.32)
Age (at Year 3 assessment) 19.94 (0.89)
Race/ethnicity
Asian or Asian American 12 (2.93)
Latinx or Hispanic 273 (66.59)
African American or Black 12 (2.93)
White, Caucasian, Anglo; not Hispanic 46 (11.22)
Other1 67 (16.34)
Assigned study condition
TND 259 (63.17)
Control 151 (36.83)
Notes. 1Other includes American Indian/Native American and mixed ethnicities.
Table 9. Sample means of relevant variables by data collection year.
Year 3 Year 4 Year 5 Overall
n 325 349 327 410
Variable Mean(SD) Mean(SD) Mean(SD) Mean(SD)
SLEs1 2.93 (1.54) 2.59 (1.52) 2.49 (1.60) 2.67 (1.56)
CIU2 4.54 (3.40) 5.08 (3.74) 5.06 (3.61) 4.90 (3.59)
PTG3 0.66 (0.39) 0.66 (0.37) 0.64 (0.39) 0.65 (0.38)
Notes. 1SLEs: Stressful life events, cumulative, range: 0-8. 2CIU: Compulsive internet use sum score, range: 1-16.
3PTG: Posttraumatic growth average score, range: -1 to 1.
Table 10. Longitudinal growth curve model of CIU over time (n=410).
Notes. 1
id=student ID, sid=school ID.
2WSC: within-school-centered variable. 3BSC: between-school-centered
variable. *p<0.05. **p<0.01. ***p<0.001.
Fixed Effects β SE
Intercept 5.1632*** 0.3774
Time 0.2416* 0.1123
Gender 0.0973 0.2915
Ethnicity -0.3681 0.3101
TND participation -0.4965 0.3116
SLEs_WSC2
-0.0584 0.1060
SLEs_BSC3
-0.7921 0.7069
Time*SLEs_WSC 0.0537 0.0773
Time*SLEs_BSC 0.8629* 0.4249
PTG_WSC -2.3383 1.5036
PTG_BSC 2.5555 9.4645
Time*PTG_WSC 1.3776 1.1034
Time*PTG_BSC 10.7694 5.6108
Random Effects β SE
UN(1,1) id(sid)1 5.4326*** 0.6317
UN(1,1) sid 0 .
Residual 7.4300*** 0.4350

83
Table 11. Moderation model of PTG on SLEs and CIU (n=410).
Notes. *p<0.05. **p<0.01. ***p<0.001.
Table 12. Moderation model of PTG on SLEs and internet browsing (n=410).
Notes. *p<0.05. **p<0.01. ***p<0.001.
Table 13. Moderation model of PTG on SLEs and social networking (n=410).
Notes. *p<0.05. **p<0.01. ***p<0.001.
Fixed Effects β SE
Intercept 5.0261*** 0.3714
Time 0.2381* 0.1126
Gender 0.1310 0.2917
Ethnicity -0.3380 0.3090
TND participation -0.3511 0.3007
SLEs_WSC -0.0056 0.0732
PTG_WSC -1.0243 1.0222
SLEs_WSC*
PTG_WSC
-0.2415 0.6033
Random Effects β SE
UN(1,1) id(sid) 5.4571*** 0.6330
UN(1,1) sid 0 .
Residual 7.4843*** 0.4350
Fixed Effects β SE
Intercept 0.1417*** 0.3714
Time 0.0273* 0.0121
Gender 0.0268 0.0292
Ethnicity 0.0002 0.0310
TND participation -0.0271 0.0301
SLEs_WSC 0.0021 0.0077
PTG_WSC -0.3001** 0.1074
SLEs_WSC*
PTG_WSC
-0.0356 0.0637
Random Effects β SE
UN(1,1) id(sid) 0.0499
*** 0.0063
UN(1,1) sid 0 .
Residual 0.0870*** 0.0050
Fixed Effects β SE
Intercept 0.2131*** 0.0393
Time -0.0058 0.0139
Gender -0.0489 0.0303
Ethnicity 0.0427 0.0321
TND participation -0.0163 0.0312
SLEs_WSC 0.0194* 0.0085
PTG_WSC -0.1822 0.1193
SLEs_WSC*
PTG_WSC
-0.1227 0.0713
Random Effects β SE
UN(1,1) id(sid) 0.0435*** 0.0069
UN(1,1) sid 0 .
Residual 0.1175*** 0.0068

84
Table 14. Moderation model of PTG on SLEs and online shopping (n=410).
Notes. *p<0.05. **p<0.01. ***p<0.001.
Fixed Effects β SE
Intercept 0.0436 0.0265
Time 0.0148 0.0082
Gender -0.0045 0.0189
Ethnicity 0.0000 0.0205
TND participation 0.0105 0.0233
SLEs_WSC -0.0036 0.0051
PTG_WSC -0.0225 0.0715
SLEs_WSC*
PTG_WSC
-0.0316 0.0427
Random Effects β SE
UN(1,1) id(sid) 0.0184*** 0.0026
UN(1,1) sid 0.0009 0.0009
Residual 0.0405
*** 0.0023

85
Appendix B: Figures
Figure 1: Study 1 sample exclusion criteria.

86
Figure 2: CIU trajectory over time. The light blue lines represent the mean compulsive internet use (CIU) scores
across students for each school, and the bolded black line represents the mean internet use score averaged across
students and across schools. Overall, CIU seems to increase over time.

87
Figure 3: Study 2 sample exclusion criteria. (Note in Year 5 exclusions: SPA = self-perceived addictions)

88
Figure 4: Probing the significant interaction of time and between-school-SLEs revealed diverging trajectories. The
increase of CIU is steeper for students from schools with higher cumulative stress while CIU is relatively steady for
students from schools with lower cumulative stress reported.

89
Figure 5: The moderating effect of PTG on the association of SLEs and CIU.

90
Figure 6: The moderating effect of PTG on the association of SLEs and self-perceived addiction to internet
browsing.

91
Figure 7: The moderating effect of PTG on the association of SLEs and self-perceived addiction to social
networking.

92
Figure 8: The moderating effect of PTG on the association of SLEs and self-perceived addiction to online shopping.

Abstract (if available)

Abstract Background: Emerging adults (EAs) transitioning into adulthood may respond to stressful life events (SLEs) by developing a harmful coping mechanism, an example being compulsive internet use (CIU). CIU may serve as a means of escapism for EAs who are overwhelmed by stress. A negative emotional response to an SLE may be trauma, ranging from acute (trauma from one experience) to complex (trauma from repeated and/or prolonged experiences). Increasing scientific literature supports the capability to experience positive psychological change (e.g., posttraumatic growth, PTG) as a result of experiencing the SLEs. This dissertation explores the potential moderating effect that PTG has on the association of SLEs and CIU.

Methods: Project Towards No Drug Abuse (TND) is a drug abuse prevention curriculum targeted towards EAs in alternative/continuation high schools, focusing on developing skills, motivation factors, and decision-making competence to avoid behaviors harmful to health. In 2008, at baseline, 1,676 participants (mean age 16 years, SD=0.93, range 14-21 years) completed a pre-test survey. There were six waves of follow-up data collection from 2009 to 2015.

Results: PTG was found to be significantly negatively associated with CIU and positively associated with identifying as female, identifying as Hispanic, and having participated in any component of Project TND. Although not statistically significant, moderating effects of PTG on SLEs and CIU were observed: with increasing SLEs, those who endorsed positive PTG reported decreasing CIU, while those who endorsed negative PTG reported increasing CIU, suggesting a protective factor of PTG against CIU engagement.

Conclusions: Improving EA health outcomes is critical, especially in identifying harmful health behaviors that may continue into adulthood. CIU has been found to be prevalent among EAs and harmful to their health and wellbeing. Literature is currently growing on the notion that PTG could potentially have a buffering effect where higher PTG endorsement could be associated with health behavior benefits, such as increased ability to not participate in harmful coping behaviors. After experiencing stressful trauma, channeling this buffering effect of PTG against harmful coping mechanisms may help in developing and supplementing stress management skills that would enable EAs to reach a higher level of both life functioning and psychological wellbeing for a healthier transition to adulthood.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

The Internet activities, gratifications, and health consequences related to compulsive Internet use

PDF

Post-traumatic growth among high-risk youth: predictors, impact of stressful life events, and relationship with changes in substance use behaviors

PDF

Sociocultural stress, coping and substance use among Hispanic/Latino adolescents

PDF

A penta-dimensional longitudinal analysis of the predictors of compulsive internet use among adolescents using linear mixed model (LMM)

PDF

The role of social support in the relationship between adverse childhood experiences and addictive behaviors across adolescence and young adulthood

PDF

Role transitions, past life events, and their associations with multiple categories of substance use among emerging adults

PDF

The dynamic relationship of emerging adulthood and substance use

PDF

Cultural risk and protective factors for tobacco use behaviors and depressive symptoms among American Indian adolescents in California

PDF

Social self-control and adolescent substance use

PDF

A sociocultural and developmental approach to intimate partner violence among a sample of Hispanic emerging adults

PDF

Smoke-free housing policies and secondhand smoke exposure in low income multiunit housing in Los Angeles County

PDF

The influence of contextual factors on the processes of adoption and implementation of evidence-based substance use prevention and tobacco cessation programs in schools

PDF

Pathways of drug use among people who inject drugs

PDF

A multi-system evaluation of medication for opioid use disorder for people who use opioids

PDF

Emotion regulation and heart rate variability among smokers

PDF

Problematic alcohol use in Hispanic emerging adults: the role of perceived discrimination, cultural identity, and salient cultural values

PDF

Childhood cancer survivorship: parental factors associated with survivor's follow-up care behavior and mental health

PDF

Applications of mindfulness toward promoting psychosocial health in diverse communities

PDF

Multilevel sociodemographic correlates of the health and healthcare utilization of childhood cancer survivors

PDF

Effects of flavorings in electronic cigarettes on the use and appeal of e-cigarettes among youth and adults

Asset Metadata

Creator Yu, Sheila (author)

Core Title Prospective associations of stress, compulsive internet use, and posttraumatic growth among emerging adults

School Keck School of Medicine

Degree Doctor of Philosophy

Degree Program Preventive Medicine (Health Behavior)

Degree Conferral Date 2024-05

Publication Date 03/20/2024

Defense Date 03/04/2024

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

Tag compulsive internet use,emerging adulthood,OAI-PMH Harvest,posttraumatic growth,stressful life events

Format theses (aat)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Sussman, Steve (committee chair), Palinkas, Lawrence (committee member), Unger, Jennifer (committee member)

Creator Email sheilayu@usc.edu,sheilayu@utexas.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC113857447

Unique identifier UC113857447

Identifier etd-YuSheila-12704.pdf (filename)

Legacy Identifier etd-YuSheila-12704

Document Type Dissertation

Format theses (aat)

Rights Yu, Sheila

Internet Media Type application/pdf

Type texts

Source 20240325-usctheses-batch-1130 (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.

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