Interpersonal risk factors, shame, and depression among Asian Americans: A longitudinal study of perfectionistic family discrepancy and selfcompassion as moderators Stephanie Carrera

Capstones, Theses and
Dissertations
2016
Interpersonal risk factors, shame, and depression
among Asian Americans: A longitudinal study of
perfectionistic family discrepancy and selfcompassion as moderators
Stephanie Carrera

TABLE OF CONTENTS
LIST OF FIGURES iii
LIST OF TABLES v
ABSTRACT vi
CHAPTER 1. INTRODUCTION 1
CHAPTER 2. LITERATURE REVIEW 16
CHAPTER 3. MATERIALS AND METHODS 48
CHAPTER 4: RESULTS 56
CHAPTER 5. DISCUSSION 79
REFERENCES 91
APPENDIX A: IRB APPROVAL 102
APPENDIX B: INVITATION EMAIL 103
APPENDIX C: UNIVERSITY RECRUITMENT EMAILS 104
APPENDIX D: INFORMED CONSENT FORMS 106
APPENDIX E: DEBRIEFING FORMS 110
ACKNOWLEDGMENTS 111
iii
LIST OF FIGURES
Figure 1. The Hypothesized Conceptual Model. 11
Figure 2. The Hypothesized Moderated Mediation Model. 12
Figure 3. Hypothesized Interaction Effects of Thwarted Belongingness 13
on Interpersonal Shame (Perfectionistic Family Discrepancy).
Figure 4. Hypothesized Interaction Effects of Perceived Burdensomeness on 13
Interpersonal Shame (Perfectionistic Family Discrepancy).
Figure 5. Hypothesized Interaction Effects of Interpersonal Shame 14
on Depression (Perfectionistic Family Discrepancy).
Figure 6. Hypothesized Interaction Effects of Thwarted Belongingness 14
on Interpersonal Shame (Self-Compassion).
Figure 7. Hypothesized Interaction Effects of Perceived Burdensomeness 15
on Interpersonal Shame (Self-Compassion).
Figure 8. Hypothesized Interaction Effects of Interpersonal Shame 15
on Depression (Self-Compassion).
Figure 9. Conceptual Diagram for Model “58” from Hayes’s (2013) 60
Command Guide in PROCESS.
Figure 10. Conceptual Diagram for Model “5” from Hayes’s (2013) 62
Command Guide in PROCESS.
Figure 11. The Conceptual Model for Post Hoc Analyses. 70
Figure 12. Perfectionistic Family Discrepancy Moderated the Direct Effect 72
from Thwarted Belongingness to Depression.
Figure 13. Perfectionistic Family Discrepancy Moderated the Direct Effect 72
from Thwarted Belongingness to Depression (scale range of 0-1).
Figure 14. Perfectionistic Family Discrepancy Moderated the Direct Effect 74
from Perceived Burdensomeness to Depression.
Figure 15. Perfectionistic Family Discrepancy Moderated the Direct Effect 74
from Perceived Burdensomeness to Depression through
Interpersonal Shame (scale range of 0-1).
iv
Figure 16. Self-Compassion Did Not Moderate the Direct Effect 76
from Thwarted Belongingness to Depression.
Figure 17. Self-Compassion Did Not Moderate the Direct Effect 76
from Thwarted Belongingness to Depression (scale range of 0-1).
Figure 18. Self-Compassion Did Not Moderate the Direct Effect 78
from Perceived Burdensomeness to Depression through
Interpersonal Shame.
Figure 19. Self-Compassion Did Not Moderate the Direct Effect 78
from Perceived Burdensomeness to Depression through
Interpersonal Shame (scale range of 0-1).
v
LIST OF TABLES
Table 1. Means, Standard Deviations, and Zero-Order Correlations. 58
Table 2. Path Coefficients Estimating Whether Perfectionistic Family 63
Discrepancy Moderated the Mediation Effect (Thwarted Belongingness).
Table 3. Path Coefficients Estimating Whether Perfectionistic Family 64
Discrepancy Moderated the Mediation Effect (Perceived Burdensomeness).
Table 4. Path Coefficients Estimating Whether Self-Compassion 65
Moderated the Mediation Effect (Thwarted Belongingness).
Table 5. Path Coefficients Estimating Whether Self-Compassion 66
Moderated the Mediation Effect (Perceived Burdensomeness).
Table 6. Path Coefficients Estimating Whether Perfectionistic Family 71
Discrepancy Moderated the Direct Effect (Thwarted Belongingness).
Table 7. Path Coefficients Estimating Whether Perfectionistic Family 73
Discrepancy Moderated the Direct Effect (Perceived Burdensomeness).
Table 8. Path Coefficients Estimating Whether Self-Compassion 75
Moderated the Direct Effect (Thwarted Belongingness).
Table 9. Path Coefficients Estimating Whether Self-Compassion 77
Moderated the Direct Effect (Perceived Burdensomeness).
vi
ABSTRACT
This longitudinal study (Time 1 = beginning of the spring semester [February-March];
Time 2 = end of the spring semester [April-June]; Time 3 = middle of the fall semester
[September-October]) applied Joiner’s (2005) Interpersonal-Psychological Theory of Suicide to
Asian Americans’ experiences with depression. A moderated mediation model was tested.
Perfectionistic family discrepancy and self-compassion were hypothesized to moderate the
mediation effect of interpersonal shame on the interpersonal risk factors (i.e., thwarted
belongingness and perceived burdensomeness) and depression relationships while controlling for
initial level of depression.
A total of 605 Asian Americans attending predominantly White, Midwestern universities
completed three online surveys. Conditional process modeling via Hayes’s (2013) PROCESS
was used to analyze the data. Neither perfectionistic family discrepancy nor self-compassion
moderated the hypothesized mediation effects. However, post hoc analyses supported the
mediation hypothesis, and perfectionistic family discrepancy moderated the direct effects of
thwarted belongingness on depression while controlling for initial level of depression.
Specifically, results from conditional direct effects indicated that the significantly positive
association between thwarted belongingness and depression was stronger for those with higher
than lower perfectionistic family discrepancy. Perfectionistic family discrepancy also moderated
the direct effect of perceived burdensomeness on depression while controlling for initial level of
depression. While the conditional direct effects were not significant, the significant interaction
demonstrated that the slopes at different levels of the moderator were significantly different from
each other. When perfectionistic family discrepancy was experienced at a high level, the slope
was positive; at a low level, the slope was negative. Therefore, Asian Americans with higher
vii
discrepancy in meeting their families’ standards for performance are more vulnerable to
depression when they believe they burden others compared to those with lower discrepancy.
Self-compassion did not significantly moderate the thwarted belongingness-depression or
perceived burdensomeness-depression relations.
In sum, this study extended Joiner’s (2005) theory to Asian Americans’ risk for future
depression. Our model also identified that Asian American college students with high family
expectations for perfectionism may be at higher risk for depression that stems from experiences
of interpersonal shame, thwarted belongingness, and perceived burdensomeness. Limitations,
future research directions, and counseling implications were discussed.
1
CHAPTER 1. INTRODUCTION
Asian Americans have been viewed as members of the “model minority” group who may
adjust well to college life upon leaving their families. That is, they may be held to high regard for
their ability and desire to excel academically (Kim, Atkinson, & Yang, 1999; Kwan, 2000). Yet,
literature on Asian American mental health has demonstrated that they may report concerns such
as anxiety (Sangalang & Gee, 2012), depression (Wang, 2010), shame (Wong, Kim, Nguyen,
Chen, & Saw, 2014), or thoughts of suicide (Cheng et al., 2010) resulting from interpersonal
concerns (e.g., lacking a sense of belongingness and fearing that one is burdening others). While
the effect of interpersonal stressors on Asian American mental health has been studied, few
studies have assessed mediating or moderating factors that protect or harm them while in college.
The purpose of this study was to apply Joiner’s (2005) Interpersonal-Psychological Theory of
Suicide to Asian American college students and to psychological outcomes other than suicide
(i.e., depression). Given that they are at risk for exhibiting higher depression (National Alliance
on Mental Illness, 2003), a better understanding of risk and protective factors related to
depression is imperative. For example, focusing on how moderators (e.g., perfectionistic family
discrepancy and self-compassion) alter the strength of the mediating role interpersonal shame
may play in the interpersonal risk factors-depression links over time could provide clinicians
knowledge on how to treat depression among their Asian American college students.
Applying Joiner’s (2005) Interpersonal-Psychological Theory to Depression
Joiner (2005) originally proposed an interpersonal theory in which the desire to die by
suicide results from the simultaneous presence of two negative interpersonal states: thwarted
belongingness and perceived burdensomeness. Both interpersonal states can be applied to
depression (T. Joiner, personal communication, September 23, 2014). Thwarted belongingness
2
describes individuals’ perception that they are isolated from others and do not form a vital part of
their valued social group. Perceived burdensomeness refers to individuals’ perception that they
are a burden to their friends, family, and society in general. According to Joiner, when
individuals experience both states as stable and unchanging, they are proximal and sufficient
causes of passive suicidal ideation (Joiner et al., 2009; Van Orden, Witte, Gordon, Bender, &
Joiner, 2008) and possibly depression (T. Joiner, personal communication, September 23, 2014).
More than 20 empirical tests have been conducted with Joiner’s Interpersonal-Psychological
Theory using Caucasian college samples (Ribeiro & Joiner, 2009) and non-college samples (e.g.,
Van Orden, Lynam, Hollar, & Joiner, 2006). Recently, some studies have applied Joiner’s theory
to Asian Americans due to its relevant emphasis on two negative interpersonal states that are
congruent with Asian cultural norms (i.e., the importance of maintaining relationships and group
cohesion; Wong, Brownson, & Schwing, 2011a; Wong, Koo, Tran, Chiu, & Mok, 2011b; Wong
& Poon, 2010). Cross-cultural validity of Joiner’s theory has also been demonstrated among
Chinese university students (Zhang, Lester, Zhao, & Zhou, 2013).
It is reasonable to apply Joiner’s interpersonal states to depression (T. Joiner, personal
communication, September 23, 2014). For instance, a sense of belonging is a basic human
motivation that, when absent, can affect the onset and course of depression (Baumeister & Leary,
1995; Hagerty & Williams, 1999; Maslow, 1954). Hagerty, Williams, Coyne, and Early (1996)
found that personal involvement in a system or environment and feeling valued by others are
closely related to social and psychological functioning among college students. Previous research
has also alluded to constructs related to perceived burdensomeness (e.g., mattering) that may
influence the onset of depression (Fisher-Beckfield & McFall, 1982). Specifically, mattering
describes the extent to which individuals believe they make a difference in the world around
3
them; individuals feel as though they matter if they perceive others attend to, invest in, or look to
them for resources. If individuals do not matter to others, they may feel irrelevant, without
purpose, or even harmful to those around them, which can contribute to low self-esteem, high
academic stress, and depression (Dixon & Robinson Kurpius, 2008; Taylor & Turner, 2001).
With specific reference to Asian Americans, both thwarted belongingness and perceived
burdensomeness may relate to the value they place on upholding their membership status in their
in-group(s) by attending to others and maintaining a sense of harmonious interdependence
(Markus & Kitayama, 1991; Triandis, Bontempo, & Villareal, 1988). First, experiencing a sense
of social integration within their community may contribute to a sense of belongingness which,
according to Joiner, is relevant in the development of depression; thwarted belongingness has
been operationally defined as low levels of support from others (Hill & Pettit, 2014). A lack of
connection with one’s friends, family, and community members may ultimately bring about
feelings of discomfort and insecurity, as well as increase Asian Americans’ susceptibility to
depression. Second, Asian Americans may be wary of when they burden others due to
collectivist culture advocating for interpersonal harmony, avoiding conflict, and avoiding
unnecessary use of community resources. If they do not believe they are serving an important
role in their group or deserve the attention of others, Asian Americans may believe they are
burdening others, which can contribute to their feeling depressed, embarrassed, and ashamed
(Taylor et al., 2004). Based on Joiner’s theory as applied to depression, the present study
examined a moderated mediation model for the associations between interpersonal risk factors
(i.e., thwarted belongingness and perceived burdensomeness) and depression among Asian
American college students across three time points.
4
Interpersonal Shame
While the literature on shame, its many operational and conceptual definitions, and its
distinctive characteristics as compared with guilt is extensive, literature exploring racial/ethnic
minority experiences with shame to future psychological outcomes is relevantly new. Several
scholars have begun to highlight the salience of shame among Asian Americans due to its
presence in several Asian languages, parenting techniques (e.g., discussing children’s
transgressions in front of others to induce shame and socialize children to behave properly;
Louie, 2014), and higher levels of shame experiences compared to Caucasians (as cited in Wong
et al., 2014). Conceptually, Markus and Kitayama (1991) argue that aspects of the self among
Asians rest upon a sense of interdependence on others, which influences their concern with
gaining approval from others. Those who believe they fail to fulfill socially prescribed duties
may experience shame (Lutwak, Razzino, & Ferrari, 1998).
Recently, Wong et al. (2014) reviewed the literature of shame experiences among Asian
Americans and conceptualized their experiences as predominantly interpersonal in nature. They
proposed the construct of interpersonal shame, or the experience of shame arising from
interpersonal concerns consisting of two dimensions: external shame (i.e., shame resulting from
concerns about others’ negative evaluations of the self) and family shame (i.e., shame resulting
from the perception that one has brought shame to one’s family). While the former reflects the
consequences individuals feel when they believe others see flaws within them, the latter reflects
perceived consequences of bringing shame upon multiple people (i.e., family members). Among
Asian American college students, external and family shame positively correlate with depression
and suicidal ideation and negatively correlate with self-esteem (Wong et al., 2014).
5
Proposed Moderated Mediation Model
This study proposed two moderating variables that may impact the mediation effect of
interpersonal shame on the interpersonal risk factors (i.e., thwarted belongingness and perceived
burdensomeness) and depression links (see Figure 1). First, one culturally salient variable that
may serve as risk factor for the interpersonal stressors-interpersonal shame and interpersonal
shame-depression links is perfectionistic family discrepancy (see below). Due to these students’
exposure to U.S. mainstream and Asian cultures, it is possible that their parents may demonstrate
high expectations for academic excellence and greater criticism when these expectations are not
fulfilled (Yoon & Lau, 2008). On the other hand, incorporating high expectations for perfection
may reflect a culturally appropriate form of parenting (DiBartolo & Rendón, 2012). This study
responds to a call from previous studies to assess the nature of perfectionistic family
discrepancy, given the importance Asian American families place on expectations for
achievement (Castro & Rice, 2003).
Similarly, another culturally salient variable that may influence the hypothesized
mediation relationships is self-compassion (see below). Neff (2003) argued that the relationship
between self-criticism and depression may be distinct for those who identify with East Asian
culture, or a culture that may advocate awareness of personal shortcomings to aid selfimprovement. In fact, Asian American students may demonstrate self-compassion and selfcriticism tendencies to a greater degree compared to American students, even though their
overall self-compassion scores do not differ significantly (Birkett, 2014). To effectively unravel
the role self-compassion may serve for Asian Americans, I included it as a possible moderating
variable for the proposed mediation hypotheses.
6
Perfectionistic family discrepancy
First, it was expected that positive associations between interpersonal risk factors (i.e.,
thwarted belongingness and perceived burdensomeness) and interpersonal shame would be
stronger for those with higher perfectionistic family discrepancy than for those with lower
perfectionistic family discrepancy (see Figures 3-4). Wang (2010) developed the construct of
perfectionistic family discrepancy to describe the degree to which individuals perceive
themselves as unable to meet their families’ standards. Asian culture advocates a strong
connection and obligation to the family, particularly with respect to meeting their expectations
(Kwan, 2000). Asian Americans with high perfectionistic family discrepancy may perceive that
they are not meeting their family’s standards for performance. They are more likely to believe
they have lost face, disgraced their family, or not honored their family via their achievements.
The interpersonal shame they perceive may increase if they do not believe they can meet their
families’ expectations; this perceived discrepancy may thus strengthen the negative effect of
these risk factors on their interpersonal shame.
Empirically, Wang, Wong, and Fu (2013) conducted a study with Asian international
students examining the moderating effects of perfectionistic personal discrepancy, perfectionistic
family discrepancy, and discrimination on the relation between interpersonal risk factors (i.e.,
thwarted belongingness and perceived burdensomeness) and suicidal ideation. Results
demonstrated that perfectionistic family discrepancy and discrimination intensified the
interpersonal risk factors-suicidal ideation links. Therefore, based on the above rationale and
relevant empirical support, it was hypothesized that perfectionistic family discrepancy would
strengthen the positive associations between interpersonal risk factors and interpersonal shame.
7
Second, it was expected that the positive association between interpersonal shame and
depression would be stronger for those with higher perfectionistic family discrepancy than for
those with lower perfectionistic family discrepancy (see Figure 5). As mentioned above, Asian
Americans with higher perfectionistic family discrepancy may perceive a gap between their
families’ standards for performance and their actual performance. They may be more likely to
believe that they are unable to fulfill their families’ expectations, which can make them
vulnerable to depression in the face of high interpersonal shame.
Although this specific hypothesis has not been tested empirically, two studies evidence
the possible moderating role of perfectionistic family discrepancy. Hewitt and Flett (1993) tested
their hypothesis that socially prescribed perfectionism might moderate the relations between
daily stress (i.e., achievement [e.g., workload] and interpersonal [e.g., social commitments]
hassles) and depression among adult inpatients. Those who were high in socially prescribed
perfectionism showed an increase in depression as the amount of daily stress increased. More
recently, Wei, Mallinckrodt, Russell, and Abraham (2004) explored the moderating role of
maladaptive perfectionism on the relation between adult attachment and depression. Their
findings suggested that depression was significantly positively associated with attachment
anxiety for undergraduate students with higher maladaptive perfectionism. Thus, conceptual and
related empirical findings support the hypothesis that perfectionistic family discrepancy may also
strengthen the positive association between interpersonal shame and depression.
Self-compassion
With respect to self-compassion, it was expected that positive associations between
interpersonal risk factors (i.e., thwarted belongingness and perceived burdensomeness) and
interpersonal shame would be weaker for those with higher self-compassion than for those with
8
lower self-compassion (see Figures 6-7). Neff’s (2003) construct of self-compassion entails
treating oneself with kindness, recognizing one’s shared humanity with others, and holding one’s
emotional pain (e.g., shame) in mindful awareness. Asian Americans who display high selfcompassion may be better able to show themselves kindness and understand that they are not
alone for feeling a lack of belongingness. In the face of thwarted belongingness, they might
recognize that it may take time to feel connected with others and hold their loneliness in mindful
awareness. Moreover, greater self-kindness may help them to feel less perceived
burdensomeness; high self-compassion may make it less likely that individuals criticize
themselves and judge their self-worth as low (i.e., as a burden or harm to others). In this way,
self-compassion as a resource might decrease the interpersonal shame they experience because
they have difficulty connecting with others and believing in their worth and competence.
Empirically, I was not able to locate any studies that have specifically tested the moderating role
of self-compassion on the interpersonal risk factors-interpersonal shame links. One reason for
this may be that Joiner’s interpersonal states are relatively new to the literature; however, the
moderating role of self-compassion with respect to other psychological outcomes has been
shown elsewhere (see below). Nonetheless, it was hypothesized that self-compassion would
moderate the links of interpersonal risk factors and interpersonal shame.
Additionally, it was expected that the positive relation between interpersonal shame and
depression would be weaker for those with a higher level of self-compassion than for those with
a lower level of self-compassion (see Figure 8). The rationale for this hypothesis is the
following. Asian Americans who exhibit high self-compassion may be less critical of themselves
when feeling inadequate, negatively evaluated by others, and believing they are bringing shame
upon their family. They may also realize that there are others like themselves who struggle with
9
shame resulting from interpersonal stressors they encounter while at college. Highly selfcompassionate Asian Americans may be more aware of their shamed feeling and better able to
avoid over-identifying with their shame to the point of feeling depressed. Empirical studies have
showed evidence for the associations among self-compassion, shame-related concepts, and
negative outcomes (Hall, Row, Wuensch, & Godley, 2013; Van Dam, Sheppard, Forsyth, &
Earleywine, 2010; Woods & Proeve, 2014) as well as its moderator role. Wong and Mak (2012)
found that the associations between cognitive-personality vulnerability styles and depression
among Chinese adults in Hong Kong were weaker among those with high self-compassion.
Kyeong (2013) also showed that the effect of academic burden on Korean students’ depression
was weaker for those who exhibited high self-compassion. According to the above reasoning and
relevant empirical studies, it was hypothesized that self-compassion would lessen the positive
association between interpersonal shame and depression.
The Present Study
In sum, limited studies have applied Joiner’s (2005) Interpersonal-Psychological theory
to Asian American college student populations. To my knowledge, this was also the first study to
test this theory with the outcome variable of depression over time. Data were collected over three
academic semesters (Time 1[T1] = beginning of spring semester; Time 2 [T2] = end of spring
semester; and Time 3 [T3] = middle of fall semester) to test the hypothesized moderated
mediation model shown in Figure 2. Perfectionistic family discrepancy and self-compassion
were hypothesized to moderate the mediation effects from interpersonal risk factors (i.e.,
thwarted belongingness and perceived burdensomeness) through interpersonal shame to
depression. First, it was expected that the associations between interpersonal risk factors and
interpersonal shame would be stronger for those with higher than with lower perfectionistic
10
family discrepancy (see Figures 3-4). It was also hypothesized that the associations between
interpersonal shame and depression would be stronger for those with higher than with lower
perfectionistic family discrepancy (see Figure 5). Next, it was expected that the associations
between interpersonal risk factors and interpersonal shame would be weaker for those with
higher than with lower self-compassion (see Figures 6-7). It was also hypothesized that the
association between interpersonal shame and depression would be weaker for those who have
higher than lower self-compassion (see Figure 8). Finally, initial level of depression was used as
a covariate to account for its possible impact on future depression (see Figure 1). It is possible
that Asian Americans who exhibit high depression at T3 may also experience high depression at
T1. Therefore, to fully understand the hypothesized moderated mediation effects, it was
important to control for the confounding variable of depression level at T1.
11
Figure 1. The Hypothesized Conceptual Model. T = Time.
11
11
Figure 2. The Hypothesized Moderated Mediation Model. T = Time.
12
13
Figure 3. Hypothesized Interaction Effects of Thwarted Belongingness at T1 on Interpersonal
Shame at T2 for Those with Higher and Lower Levels of Perfectionistic Family Discrepancy at
T2. PFD = Perfectionistic Family Discrepancy. T = Time.
Figure 4. Hypothesized Interaction Effects of Perceived Burdensomeness at T1 on Interpersonal
Shame at T2 for Those with Higher and Lower Levels of Perfectionistic Family Discrepancy at
T2. PFD = Perfectionistic Family Discrepancy. T = Time.
14
Figure 5. Hypothesized Interaction Effects of Interpersonal Shame at T2 on Depression at T3 for
Those with Higher and Lower Levels of Perfectionistic Family Discrepancy at T2. PFD =
Perfectionistic Family Discrepancy. T = Time.
Figure 6. Hypothesized Interaction Effects of Thwarted Belongingness at T1 on Interpersonal
Shame at T2 for Those with Lower and Higher Levels of Self-Compassion at T2. SC = SelfCompassion. T = Time.
15
Figure 7. Hypothesized Interaction Effects of Perceived Burdensomeness at T1 on Interpersonal
Shame at T2 for Those with Lower and Higher Levels of Self-Compassion at T2. SC = SelfCompassion. T = Time.
Figure 8. Hypothesized Interaction Effects of Interpersonal Shame at T2 on Depression at T3 for
Those with Lower and Higher Levels of Self-Compassion at T2.
SC = Self-Compassion. T = Time.
16
CHAPTER 2. LITERATURE REVIEW
I will use this literature review to first explore Joiner’s (2005) InterpersonalPsychological Theory of Suicide, its background, concepts associated with this theoretical
framework, and the Interpersonal Needs Questionnaire Joiner and his colleagues created in
accordance with this framework. I will also review its applicability to the outcome variable of
depression. Second, I will provide background on the literature on shame and Wong et al.’s
(2014) development of the interpersonal shame construct. Third, I will include an allencompassing discussion of how interpersonal risk factors (i.e., thwarted belongingness and
perceived burdensomeness), interpersonal shame, and psychological outcomes have been linked
in previous literature and how they are linked theoretically in the present study. Fourth, I will
briefly review the literatures that pertain to the two moderating variables hypothesized to affect
the proposed mediation model (i.e., perfectionistic family discrepancy and self-compassion). I
will also explore the conceptual arguments behind the two moderating variables. Lastly, the
chapter will conclude with an explanation for why initial level of depression was accounted for
as a control variable in this short-term longitudinal study.
Joiner’s (2005) Interpersonal-Psychological Theory of Suicide
Foundations
Several studies have examined the numerous risk factors contributing to suicidal ideation,
attempts, and suicide in a variety of populations (see Van Orden et al., 2010). Factors ranging
from family conflict, genetic predispositions, psychological disorders, and previous history of
suicidal attempts contribute to why individuals may exhibit suicidal thoughts, desire, and
behavior. Researchers have also suggested that interpersonal difficulties are especially relevant
risk factors, particularly among non-clinical and clinical adult samples. Joiner with his
17
Interpersonal-Psychological Theory of Suicide proposed a proximal, problem-based framework
to explain individuals’ desire to die by suicide and later their capability to engage in suicidal
behavior. Joiner argued that the simultaneous presence of three factors highly contributes to
suicidal behavior: a thwarted sense of belongingness, a perceived sense of burdensomeness, and
an acquired capacity to attempt suicide. Individuals’ sense of belonging has been conceptualized
as a fundamentally human motive that, when unmet, can activate similar parts of the brain when
individuals undergo physical pain. Joiner (2005) argued that suicidal individuals may experience
interactions that do not satisfy their need to belong (e.g., relationships that are unpleasant,
unstable, infrequent, or without proximity) or may not feel connected to or cared about by others
(Van Orden et al., 2010). Thus, the need to belong is a need central to the development of
suicidal desire, particularly when a perceived lack of connection is prolonged for a significant
period of time.
When individuals perceive they are a burden to others, Joiner argues that they may view
themselves as ineffective and incompetent to the point where they believe others would benefit
more from their death. Individuals with high perceived burdensomeness tend to exhibit a
negative self-image, low internal locus of control, and negative emotions such as anger, sadness,
shame, and disappointment (Van Orden et al., 2010). Experiences with family conflict,
unemployment, mental illness, and physical illness can contribute to perceptions of
burdensomeness. Joiner (2005) explains that both cognitive affect states can fluctuate, but
individuals perceiving both states simultaneously over time without any prospect of change can
bring about suicidal ideation. Ideation may transform into lethal or non-lethal suicidal attempts
when individuals acquire the capability for suicide. That is, individuals may have had previous
suicidal attempts, no longer fear death, and/or can tolerate physical pain. Ultimately, Joiner’s
18
hypotheses for his Interpersonal-Psychological Theory can be summarized as the following: (a)
thwarted belongingness and perceived burdensomeness are proximal and sufficient causes of
passive suicidal ideation; (b) when thwarted belongingness and perceived burdensomeness are
simultaneously present and perceived as stable and unchanging, they are proximal and sufficient
causes for active suicidal desire; (c) suicidal intent forms when suicidal desire and lowered fear
of death are simultaneously present; and (d) serious suicidal behavior most likely occurs in the
context of thwarted belongingness, perceived burdensomeness, and acquired capability for
suicide (i.e., reduced fear of suicide and elevated tolerance for physical pain).
Measuring interpersonal risk factors
In accordance with Joiner’s model, the original 25-item Interpersonal Needs
Questionnaire (INQ) was compiled to assess thwarted belongingness and perceived
burdensomeness. While the number of items in the INQ has varied in the past eight years, Van
Orden, Cukrowicz, Witte, and Joiner (2012) recently revised the number of items via latent
variable modeling with five different samples. A series of exploratory and confirmatory factor
analyses yielded a 2-factor latent structure with 15 items: nine items loaded on the thwarted
belongingness factor and six on the perceived burdensomeness factor in young adult, clinical
outpatient adult, and elderly adult samples. The thwarted belongingness subscale assesses
individuals’ levels of perceived disconnection from others (e.g., whether or not they feel close to
other people or feel like an outsider in social gatherings). The perceived burdensomeness
subscale examines the degree to which individuals feel they are burdens to others. Items on this
subscale include individuals’ perception that others would like to be rid of them or would be
better off without them. High scores on both subscales indicate greater thwarted belongingness
and perceived burdensomeness.
19
Evidence for criterion validity demonstrates that higher levels of thwarted belongingness
and perceived burdensomeness are related to greater odds of reporting suicidal ideation at the
present time and one month later (Van Orden et al., 2012). With respect to convergent validity,
statistically significant negative relations were found between thwarted belongingness and
loneliness, social support, self-liking, and relatedness using undergraduate samples; negative
relations between perceived burdensomeness and autonomy, responsibility to family, and selfcompetence were also statistically significant. In reference to the present study, no studies to my
knowledge have used the INQ-15 with Asian Americans. However, other forms of the INQ have
demonstrated adequate reliability. Wong et al. (2011b) found internal consistencies for the
thwarted belongingness and perceived burdensomeness subscales in the INQ-12 of .88 and .74,
respectively, among Asian American college students. Both subscales were significantly (a)
positively correlated with suicidal ideation and depression; and (b) negatively correlated with
interdependent and independent self-construal. Zhang et al. (2013) tested Joiner’s theory in a
sample of Chinese university students. Results demonstrated strong internal consistency (α =
.87), and thwarted belongingness and perceived burdensomeness were (a) positively associated
with suicidal ideation, depression, and capability for self-harm; and (b) negatively associated
with social support and self-esteem (Zhang et al., 2013). Given that the earlier versions of the
INQ used a subset of items from the latest version of the INQ (T. Joiner, personal
communication, October 17, 2014), I used Van Orden et al.’s (2012) INQ-15, which is the most
rigorously tested version with respect to its psychometric properties.
Empirical testing of the theory is evident primarily in the work of Joiner, Van Orden, and
their colleagues. Van Orden et al. (2008) found that the combined presence of high thwarted
belongingness and perceived burdensomeness are dangerous for the development of suicidal
20
desire. An interaction effect of thwarted belongingness ˟ perceived burdensomeness on current
suicidal ideation was also found, such that the relationship between thwarted belongingness and
suicidal desire was significant at high levels of perceived burdensomeness. Furthermore, the
more past attempts and exposure participants had to pain and provocation experiences, the higher
individuals’ acquired capability for suicide. Lastly, the perceived burdensomeness ˟ acquired
capability interaction significantly predicted change in suicidal risk, even when the effects of
other risk factors (i.e., depression, gender, and age) were statistically controlled. These findings
are consistent with the theory’s hypothesis that suicidal desire and an acquired capability for
suicide are necessary for suicidal behavior to occur.
Based on Van Orden et al.’s (2008) results, the theory was tested further by Joiner and his
colleagues in 2009. In particular, a main effect of family social support was found, such that
those with low family social support tended to exhibit greater suicidal ideation. While no main
effect of mattering was found, the mattering ˟ family social support interaction predicted the
highest levels of suicidal ideation controlling for depression. In other words, participants who
were low in mattering and social support from their families were more likely to exhibit suicidal
thoughts. These finds are consistent with the theory’s hypotheses that suicidal ideation emerges
from high thwarted belongingness and perceived burdensomeness. Support for a 3-way
interaction of high thwarted belongingness, perceived burdensomeness, and acquired capability
to predict current suicide attempt was also found. Thus, in the presence of high acquired
capability, the combination of Joiner’s interpersonal states that predict suicidal desire can
transform into suicide attempts (Joiner et al., 2009).
While Joiner’s (2005) theory has been empirically tested, only recently have researchers
begun to tests its applicability to Asian Americans. Wong et al. (2011a) used Joiner’s theory to
21
examine risk and protective factors associated with Asian American college students’ suicidal
ideation. Their results demonstrated that variables such as mental health medication, identifying
as female, low GPA, and undergraduate status were significant risk factors for suicidal thoughts
in the past year. Among those who had seriously considered suicide in the past year, recent
family, academic, and financial concerns were the most frequent significant events occurring
before developing suicidal ideation (Wong et al., 2011a). Wong et al. (2011b) also conducted a
mixed-methods study guided by Joiner’s theory with Asian American college students. Results
suggested that perceived burdensomeness significantly positively predicted suicidal ideation
among participants with high thwarted belongingness. A more recent test of Joiner’s theory was
conducted by Zhang et al. (2013) with a sample of Chinese university students. Thwarted
belongingness, perceived burdensomeness, and acquired capability for self-harm were associated
with suicidal ideation above and beyond the effects of gender, social support, self-esteem,
depression, and age. These three studies may provide evidence that suicidal ideation and desire
results from the joint presence of thwarted belongingness and perceived burdensomeness in
Asian American populations. It is imperative that future research continue extending Joiner’s
theory with Asian Americans and via longitudinal studies.
Applying Joiner’s (2005) Interpersonal-Psychological Theory to Depression
To my knowledge, research using Joiner’s (2005) theory has examined factors related to
suicidal ideation or suicidal behavior as outcome variables. However, it is reasonable to apply
Joiner’s interpersonal states to depression (T. Joiner, personal communication, September 23,
2014) and was one of the objectives for this study. Joiner’s conceptualization of thwarted
belongingness stems from the notion that experiencing a sense of belonging is a basic human
need (Baumeister & Leary, 1995; Maslow, 1954). Without a sense of feeling valued, needed, and
22
similar to others, a host of negative psychological outcomes may ensue, including depression
(Hagerty & Williams, 1999), loneliness, anxiety, and suicidality (Hagerty et al., 1996). In
particular, college students who are not able to connect with others or cannot perceive that they
are part of a larger group may feel isolated, which can lead them to question their ability to
adequately relate with others. Doubting and ruminating over the effectiveness of their
interpersonal skills can lower their mood enough to feel depressed (Nolen-Hoeksema, 1991).
Empirical studies with college students support the finding that belonging is a robust precursor or
vulnerability factor for depression (Hagerty & Patusky, 1995; Hagerty & Williams, 1999;
Hagerty et al., 1996).
With respect to perceived burdensomeness, Joiner referenced the construct of mattering
in his original conceptualization of this interpersonal state. According to Rosenberg and
McCullough (1981), mattering is defined as individuals’ sense that, as far as others are
concerned, they are objects of attention, important to others (i.e., individuals’ actions are relevant
to others), wanted by others, dependable (i.e., competent enough to be reliable), and worthy of
emotional investment (i.e., others would miss them if they were gone). Individuals who feel that
they matter little or do not matter to others tend to have higher depression and lower self-esteem
(Rosenberg & McCullough, 1981). Among college students, feeling irrelevant, insignificant, and
inconsequential to their peers can contribute to depression (Taylor & Turner, 2001). Dixon and
Robinson Kurpius (2008) found that self-esteem and mattering accounted for nearly 40% of the
variance in depression. It seems as though mattering, a similar construct to perceived
burdensomeness, can be an important factor in college students’ psychological well-being. While
mattering was not the specific construct of interest, a lack of mattering (i.e., perception that
23
individuals harm or are a hindrance to others) corresponds with Joiner’s definition of perceived
burdensomeness (see above).
Specifically for Asian Americans, both thwarted belongingness and perceived
burdensomeness may relate to the value they place on upholding their membership status in their
in-group(s). Markus and Kitayama (1991) explained that Asian cultures may emphasize
individuals relating to each other, rather than standing out as unique. From a young age, Asian
American children are socialized to attend to others, concern themselves with fitting in, and
promote interpersonal harmony as a way of expressing and enhancing their “selves.”
Experiencing a sense of social integration within their community may contribute to a sense of
belongingness which, according to Joiner, is relevant in the development of depression. Asian
Americans who fail to engage and cooperate with others may risk a sense of belongingness with
their in-group. Thwarted belongingness has been operationally defined as low levels of support
from others (Hill & Pettit, 2014). Therefore, a lack of connection with one’s friends, family, and
community members may ultimately bring about feelings of discomfort and insecurity, as well as
increase Asian Americans’ susceptibility to depression.
Similarly, Asian Americans may be wary of when they burden others due to collectivist
culture advocating for interpersonal harmony, avoiding conflict, and avoiding unnecessary use of
community resources. Disrupting harmony or placing too much attention on themselves (e.g., via
asking for help, attempting to pursue individual goals) may come at the cost of burdening others,
creating conflict, and failing to preserve relationships. Thus, experiencing a sense of community
and being wary of when they overwhelm others may be especially important to avoid feeling
depressed and ashamed (Taylor et al., 2004). Based on Joiner’s theory as applied to depression,
the present short-term longitudinal study examined a hypothesized model for the associations
24
between interpersonal risk factors (i.e., thwarted belongingness and perceived burdensomeness)
and depression among Asian American college students.
Interpersonal Shame
The literature on shame, its many operational and conceptual definitions, and its
distinctive characteristics is extensive (see Blum, 2008 and Tangney, 1996 for a review);
therefore, I only provide a brief overview of the literature as it pertains to Asian Americans’
experience with shame. Several scholars have highlighted the salience of shame among Asian
Americans due to its presence in several Asian languages, parenting techniques (e.g., discussing
children’s transgressions in front of others to induce shame and socialize children to behave
properly; Louie, 2014), and higher levels of shame experiences compared to Caucasians (see
Wong et al., 2014). In particular, the internalization of shame among Asian Americans can be
deeply rooted in and influenced by interpersonal relationships with family, peers, colleagues, and
other in-group members. In fact, researchers argue that Asian Americans’ experience with shame
includes themselves, a “shamed other” (e.g., family members who feel shame as a result of other
individuals’ actions), and an audience who witnesses the shaming act. Asian Americans’ shame
can therefore be embedded in their social relationships, particularly with family members (Liem,
1997). Shame experienced as others’ negative evaluation of oneself and possible shame brought
upon others may result in a wish to hide, avoid, or escape from a defective, inadequate self.
In the past twenty years, there have been a few quantitative and qualitative studies
exploring Asian and Asian American experiences with shame. Lutwak et al. (1998) measured
subcultural differences in shame, guilt, fear of intimacy, self-deprecation, and perceptions of
inauthenticity among Asian American, Latina/o American, African American, and Caucasian
college students. With respect to Asian Americans, the study found that (a) Asian Americans
25
may be the most susceptible to shame as compared to other sampled groups; (b) there were
positive associations among shame-proneness, tendency to engage in self-deprecation, and
perceiving oneself as inauthentic; and (c) tendency to engage in self-deprecation was the best
predictor of shame-proneness. Results demonstrated that Asian Americans may be more likely to
disparage themselves, which may increase their probability of experiencing shame. More
recently, Bedford (2004) interviewed middle-aged women in Taiwan to better understand how
shame as an emotion can be described. Participants’ sharing of their experiences yielded four
emotions related to shame that involve a loss of one’s own face (i.e., diu lian, or to lose the
respect of others, to be humiliated or publicly disgraced), harming others via loss of one’s own
face (i.e., xiu kui), or both. It is clear that Asian Americans’ experiences with shame may involve
a heightened focus on the self and may be diffuse in its effects (i.e., one’s shameful act or
experience may also involve others close to them). Shame has been shown to be associated with
negative mental health outcomes such as depression (r = .43; Kim, Thibodeau, & Jorgensen,
2011) and suicide (Blum, 2008).
Recently, Wong and his colleagues (2014) proposed the construct of interpersonal shame
to better depict Asian Americans’ experiences of shame. Interpersonal shame is shame that arises
from individuals’ interpersonal difficulties. The construct can be conceptualized as having two
dimensions. First, shame resulting from individuals’ concerns that others negatively evaluate
them is known as external shame. This form of shame has a strong intrapersonal focus with
respect to the consequences individuals feel when they believe others view them as defective.
The second dimension is known as family shame, or the shame resulting from individuals
believing that they have brought shame to their families. Compared to external shame, family
shame has an interpersonal focus in which multiple people may be negatively affected by shame.
26
Asian Americans may vicariously experience shame due to their sense of self being deeply
embedded in their group memberships, especially their family group (Wong et al., 2014).
Measuring interpersonal shame
For their Interpersonal Shame Inventory (ISI), Wong et al. (2014) developed 34
preliminary items reflecting the construct of interpersonal shame based on their review of the
literature on Asian and Asian Americans’ shame experiences. The second author and two
doctoral students provided initial feedback on the wording of the items, the scale instructions,
and the Likert-type range of options; modifications were made, and 18 items (i.e., nine items
representing external shame and family shame, respectively) were sent to eight other research
psychologists for a second round of feedback. Further modifications were made to verify that the
items described the experience of shame and to develop a state-based measure of shame by
adding “These days” to the beginning of each item. In this way, participants focus on their recent
experiences of shame. Exploratory factor analyses conducted to test the scale’s factor structure
yielded a two-factor solution. Further revision of the scale based on the above analysis yielded a
10-item scale with item loadings above .70 on each subscale (i.e., five items each on the external
and family shame subscales). Confirmatory factor analyses demonstrated that a two-factor
oblique model fit the data well.
Therefore, the ISI is a 10-item self-report measure that assesses two dimensions of
interpersonal shame: external shame (i.e., ISI-E) and family shame (i.e., ISI-F). The ISI-E
assesses the degree to which individuals are concerned about others’ negative evaluations of
them. Items describe individuals’ experiences of wanting to hide from, escape, or shrink away
from others so they do not see their flaws. The ISI-F examines the degree to which participants
perceive that they have brought shame to their families. Items describe individuals’ experiences
27
of wanting to disappear or escape because their deficits may negatively impact their families.
Higher scores on the two subscales reflect higher external shame and family shame. Both
dimensions demonstrated strong internal consistency reliability (α = .94 for the ISI-E and α = .97
for the ISI-F) and high correlations between ISI-E and ISI-F scores at T1 and Time 2 (T2; two
weeks later; .72 and .69, respectively; Wong et al., 2014). Concurrent validity in a study with
Asian American college students revealed evidence of a positive link between the ISI-E and ISIF, generic state shame, thwarted belongingness, perceived burdensomeness, self-face concerns,
depression, and suicidal ideation. Both the ISI-E and ISI-F were also significantly negatively
related to self-esteem (Wong et al., 2014). Based on Wong et al.’s (2014) conceptualization of
shame, I used this scale to measure interpersonal shame as a possible mediator for the relation
between interpersonal risk factors (i.e., thwarted belongingness and perceived burdensomeness)
and depression (see below).
Given that the ISI was published within the last year, no studies have been conducted to
test interpersonal shame and its two dimensions, other than what was tested in Wong et al.’s
(2014) studies. Wong and his colleagues tested mediation hypotheses in which (a) external and
family shame mediate the effects of thwarted belongingness and perceived burdensomeness on
suicidal ideation, and (b) thwarted belongingness and perceived burdensomeness mediate the
effects of external shame and family shame on suicidal ideation. Results supported hypothesis
(a); among Asian American college students, thwarted belongingness (but not perceived
burdensomeness) was significantly and positively related to external and family shame. More
specifically, the mediation effect from thwarted belongingness to family shame to suicidal
ideation was significant, while the effect from thwarted belongingness to external shame to
suicidal ideation was not (Wong et al., 2014). Conceptually, this finding may signify that Asian
28
Americans who lack a sense of belonging with their family, friends, or community may be more
likely to believe that they have brought shame upon their family; this experience of family
shame, in turn, may anticipate suicidal thoughts. Nevertheless, more studies are needed to
confirm the mediating role interpersonal shame may play with predictors and outcomes other
than suicidal ideation (e.g., depression).
Proposed Moderated Mediation Model
This study proposed two moderating variables that may impact the mediation effect of
interpersonal shame on the interpersonal risk factors (i.e., thwarted belongingness and perceived
burdensomeness) and depression links. First, one culturally salient variable that may serve as a
risk factor for the interpersonal stressors-interpersonal shame and interpersonal shamedepression links is perfectionistic family discrepancy (see below). Due to these students’
exposure to U.S. mainstream and Asian cultures, it is possible that their parents may demonstrate
high expectations for academic excellence and greater criticism when these expectations are not
fulfilled (Yoon & Lau, 2008). On the other hand, incorporating high expectations for perfection
may reflect a culturally appropriate form of parenting (DiBartolo & Rendón, 2012). This study
responds to a call from previous studies to assess the nature of perfectionistic family
discrepancy, given the importance Asian American families place on expectations for
achievement (Castro & Rice, 2003).
Similarly, another culturally salient variable that may influence the hypothesized
mediation relationships is self-compassion (see below). Neff (2003) argued that the relationship
between self-criticism and depression may be distinct for those who identify with East Asian
culture, or a culture that may advocate awareness of personal shortcomings to aid selfimprovement. In fact, Asian American students may demonstrate self-compassion and self-
29
criticism tendencies to a greater degree compared to American students, even though their
overall self-compassion scores did not differ significantly (Birkett, 2014). To unravel the role
self-compassion may serve for Asian Americans, I included it as a moderating variable for the
proposed mediation hypotheses.
Perfectionistic family discrepancy
According to Adler (1956), striving to meet high personal standards, or perfectionism,
can be seen as a normal part of human growth and development (Alder, 1956; Flett & Hewitt,
2002). However, he argued that the pursuit of perfection becomes maladaptive when individuals
set unrealistic goals for themselves or perceive that others set unrealistic goals for them.
Unachievable goals and aspirations can set the stage for unrealistic or harmful standards, which
can contribute to negative psychological outcomes. Hamachek (1978) later defined “neurotic”
perfectionism, now commonly referred to as maladaptive perfectionism, as a type of
perfectionism in which individuals set inflexibly high standards, fear failure, and base their selfworth on their performance. This form of perfection may arise either from environments of
conditional approval (i.e., children earn their parents’ approval when they have completed their
tasks perfectly), conditions of non-approval, or conditions of inconsistent approval (i.e., children
do not learn how to please their parents). Ultimately, children may try to adopt standards high
enough that meeting them is more likely to please their parents, themselves, or others they hold
in high regard.
Since Hamachek (1978) presented his rationale for the origins of perfectionism, several
researchers have focused their attention on the nature of parent-child relationships and
interactions. This area of research on the development of perfectionism has continued to increase
in the past thirty years (Flett & Hewitt, 2002; see Shafran & Mansell, 2001 for a review).
30
Shafran and Mansell (2001) summarized primary findings of this research by explaining that
early family experiences influencing the development of perfectionistic thinking include (a)
overly critical and demanding parents; (b) excessively high parental expectations and standards
of performance; (c) parental approval that is absent, inconsistent, or conditional; and (d) parents
modeling perfectionistic attitudes and behaviors. Rice, Ashby, and Preusser (1996) found that
neurotic perfectionists’ perception of their parents as harsh and demanding (e.g., overprotective
mothers) significantly contributed to their self-esteem. Indeed, critical parents who expect
perfection from their children may not promote a sense of cohesion, emotional warmth, or
support among family members; rather, family relations may be wrought with conflict, extreme
family enmeshment, authoritarian parenting styles, and high psychological control (Craddock,
Church, & Sands, 2009; Flett & Hewitt, 2002). As a result, children may doubt their own
abilities and believe they are being held to higher parental expectations and criticism. Such
conditions may make these children susceptible to a host of negative psychological outcomes,
including poor adjustment, anxiety, depression, and suicidal ideation (see Castro & Rice, 2003).
Recently, researchers have begun to describe and operationally define family influences
on perfectionism with respect to racial/ethnic minority individuals’ experiences (e.g., DiBartolo
& Rendón, 2012; Ortega, Wang, Slaney, Hayes, & Morales, 2014; Wang, 2010). Preliminary
studies conducted with Asian Americans addressed the influence of parenting styles and parental
expectations on perfectionism and negative psychological outcomes. Kawamura, Frost, and
Harmatz (2002) sought to examine cultural differences in whether or not perfectionism develops
in families with overly critical parents. When they compared the relationship between perceived
parental characteristics and perfectionism among Asian Americans and Caucasians, they found
that Asian Americans rated both parents as more harsh and authoritarian than Caucasians. In fact,
31
Asian Americans seemed to show more concern over mistakes and doubt their actions, both of
which are components of maladaptive perfectionism. Castro and Rice (2003) later replicated and
extended these results when they assessed ethnic differences in perfectionism among Asian
American, African American, and Caucasian college students. Asian Americans scored
significantly higher on concerns over making mistakes, parental criticism, and doubts about
actions as compared to the other groups. With respect to psychological and academic outcomes,
higher scores on the aforementioned scales explained significant variance in depression and
cumulative GPA. This early research with Asian Americans already began to demonstrate that
they may be more susceptible to self-criticism, self-doubt, and concerns over not meeting
expectations. Such susceptibility may also place them at risk for depression.
In the past few years, researchers have attempted to specifically account for cultural
differences in their studies of Asian Americans’ perfectionistic tendencies. For example, Yoon
and Lau (2008) and Chang (2013) added in the construct of independent (i.e., emphasis on the
uniqueness of the self) and interdependent (i.e., emphasis on seeing oneself in relation to others)
self-construal in their study of maladaptive perfectionism among Asian American college
students. The former tested whether a sense of interdependence and parental support buffered the
association between perfectionism and depression. Analyses revealed that participants high on
interdependence were more vulnerable to depression in the face of perfectionistic tendencies.
Positively, perfectionism and depression were unrelated among those with parental support.
Based on their results, the authors speculated that Asian Americans are at greatest risk for
depression in the face of expectations from parents to perform well and when parents’ support is
contingent upon those expectations. The latter study compared Asian Americans and European
Americans on whether or not perfectionism, loneliness, and self-construal schemas (i.e.,
32
independent and interdependent self-construal) were unique predictors of depression and anxiety.
In line with previous studies, Asian Americans displayed greater concern over mistakes, parental
expectations, parental criticism, doubts about actions, loneliness, and interdependence compared
to European Americans. Furthermore, hierarchical regression analyses indicated perfectionism
and loneliness were additive predictors of depression and anxiety across both groups. Selfconstrual (i.e., lack of independent self-construal) was also found to significantly predict
depression, above and beyond perfectionism and loneliness. Interdependent self-construal did not
emerge as a significant predictor. In light of these few studies, future studies should continue to
develop a better understanding of perfectionism across ethnicities (e.g., Asian Americans).
Given the influence the family can have on how Asian Americans relate to others, it is
reasonable to focus on how it contributes to the development of perfectionistic tendencies. The
notion of filial piety, or one’s primary duty of respect, obedience, and care for one’s parents and
elders, may relate to the importance youth place on meeting their family’s expectations. Children
are disciplined and socialized at a young age to reciprocate what their parents and elders have
provided them (e.g., food, shelter, education); oftentimes they are expected to reciprocate via
scholastic achievement or perfectionism (Fuligni, Tseng, & Lam, 1999). In this way, they fulfill
their duty as children to honor their parents’ affection and devotion to them. Such strong
connections with family members should be considered in the operational definition of
perfectionism for Asian Americans. To my knowledge, Wang’s (2010) Family Almost Perfect
Scale (FAPS) captures Asian Americans’ experiences with perfectionism as it relates to their
connection with and obligation to the family.
33
Perfectionistic family discrepancy and its measurement
Wang’s (2010) review of the literature noted that family environment and parenting
styles are logical areas of study to better understand the effect of family environment on
perfectionistic tendencies. Currently, two commonly used perfectionism scales that peripherally
address the concept of family perfectionism are Frost’s Multidimensional Perfectionism Scale
(FMPS; Frost, Marten, Lahart, & Rosenblate, 1990) and Hewitt and Flett’s Multidimensional
Perfectionism scale (HFMPS; Hewitt & Flett, 1991). The FMPS includes subscales measuring
parental expectations and parental criticism, and the HFMPS includes a subscale measuring
socially prescribed perfectionism, or perceived expectations for perfectionism from others (e.g.,
family members). Nevertheless, a major critique of both scales is their broad assessment of the
features of perfectionism (Shafran & Mansell, 2001). Wang argued that creating a scale to
measure the distinct adaptive and maladaptive aspects of perfectionism from a family perspective
is essential when studying individuals from collectivistic backgrounds or who stress the
importance of the family. The FAPS was developed in accordance with the items from the
original Almost Perfect Scale-Revised (Slaney, Rice, Mobley, Trippi, & Ashby, 2001) to reflect
perceived perfectionistic expectations and evaluations from family.
In line with the three APS-R subscales (i.e., Discrepancy, Standards, and Order), three
subscales make up the FAPS: perfectionistic family discrepancy (i.e., seven items),
perfectionistic family standards (i.e., six items), and perfectionistic family order (i.e., four items).
For the purpose of this study and its hypotheses, only the perfectionistic family discrepancy
subscale was used. The perfectionistic family discrepancy scale measures the extent to which
participants perceive that they fail to meet their families’ standards for performance. Subscale
items describe individuals’ experiences with their efforts not being good enough for their
34
families, rarely living up to family standards, and imperfection being unacceptable for their
families. Higher scores indicate higher levels of perfectionistic family discrepancy. Results from
confirmatory factor analyses demonstrated an adequate fit and structural invariance for Asian
American and European American college student samples (Wang, 2010). With respect to
validity, Wang found (a) positive correlations between perfectionistic family discrepancy and
measures of parental expectations, parental criticism, anxiety and depression; and (b) negative
correlations with self-esteem among Asian American college students. Cronbach’s alpha values
ranged from .91 to .94 with Asian and European American college students (Wang, 2010). In
concordance with Wang’s conceptualization of family perfectionism and its sub-constructs, I
used this subscale to measure perfectionistic family discrepancy as a possible moderator of the
interpersonal risk factors-shame and interpersonal shame-depression links.
The FAPS has been used with Asian American, Asian international, and, more recently,
Latina/o college students (Ortega et al., 2014). With respect to the population of interest for the
current study, two studies are described here. First, Wang (2012) examined whether Taiwanese
college students with perfectionistic tendencies have high perfectionistic family discrepancy and
family standards scores as compared to non-perfectionists. Results suggested that they tended to
exhibit higher perfectionistic family discrepancy but not higher family standards; these
participants also reported lower academic grades and low self-esteem. Second, Wang et al.
(2013) assessed possible moderating effects of perfectionistic personal discrepancy,
perfectionistic family discrepancy, and discrimination on the associations between thwarted
belongingness and perceived burdensomeness on suicidal ideation. Within a sample of Asian
international students studying in the United States (U.S.), perfectionistic family discrepancy
intensified the positive associations between interpersonal risk factors and suicide ideation.
35
Specifically, both thwarted belongingness and perceived burdensomeness more strongly
positively related to suicidal ideation when participants exhibited high perfectionistic family
discrepancy. Thus, previous studies have begun to explore the relations among the variables of
interest in this study (i.e., perfectionistic family discrepancy, thwarted belongingness, and
perceived burdensomeness). This study contributed to the growing number of studies by using
the FAPS to explore perfectionistic family discrepancy among Asian American college students.
Rationale for moderation hypotheses
First, it was expected that positive associations between interpersonal risk factors (i.e.,
thwarted belongingness and perceived burdensomeness) and interpersonal shame would be
stronger for those with higher perfectionistic family discrepancy than for those with lower
perfectionistic family discrepancy. Asian culture advocates for a strong connection and
obligation to the family, particularly with respect to meeting their expectations (Kwan, 2000).
Asian Americans with high perfectionistic family discrepancy may perceive that they are not
meeting their family’s standards for performance. They are more likely to believe they have lost
face, disgraced their family, or not honored their family via their achievements. The
interpersonal shame they perceive may increase if they do not believe they can meet their
families’ expectations; this perceived discrepancy may thus strengthen the negative effect of
these risk factors on their interpersonal shame. Empirically, Wang et al. (2013) found that
perfectionistic family discrepancy intensified the interpersonal risk factors-suicidal ideation links
among Asian international students. Therefore, based on the above rationale and relevant
empirical support, it was hypothesized that perfectionistic family discrepancy would strengthen
the positive association between interpersonal risk factors and interpersonal shame.
36
Second, it was expected that the positive association between interpersonal shame and
depression would be stronger for those with higher perfectionistic family discrepancy than for
those with lower perfectionistic family discrepancy. As mentioned above, Asian Americans with
higher perfectionistic family discrepancy may perceive a gap between their families’ standards
for performance and their actual performance. Those who perceive a sense of perfectionistic
family discrepancy may be more vulnerable to depression in the face of high interpersonal shame
due to their tendency to engage in frequent and intense self-evaluation. That is, they are strict
self-evaluators with a “broad” definition for failure (i.e., anything that is not done perfectly is a
failure; Tangney and Dearing, 2002) that makes it difficult to succeed. In the case of Asian
Americans, if they perceive that they are not meeting or have failed to meet their family’s
standards for performance, they may interpret this as their personal failure and feel depressed
and disappointed in themselves. They may believe they have lost face, disgraced their family, or
not honored their family via their achievements. Individuals may also be more prone to
depression as their shame increases because they hold themselves to standards their family
places on them; they may not have a say in their choice of tasks. Therefore, they could feel
increased depression when they cannot meet family expectations (Flett & Hewitt, 2002).
Although this specific hypothesis has not been tested empirically, two studies evidence
the possible moderating role of constructs related to perfectionistic family discrepancy. Hewitt
and Flett (1993) tested their hypothesis that perfectionism interacts with specific stressors (i.e.,
achievement [e.g., workload] and interpersonal [e.g., social commitments] hassles) to predict
depression in depressed patient and general psychiatric samples. Specifically, they examined
socially prescribed perfectionism, or individuals’ perception that others have unrealistically high
standards for them (Shafran & Mansell, 2001). Results suggested that depressed patients who
37
were high in socially prescribed perfectionism showed an increase in depression as the amount of
interpersonal hassles they experienced increased. Among participants in the general psychiatric
sample, those who were high in socially prescribed perfectionism showed an increase in
depression as the amount of achievement hassles increased. Moreover, Wei et al. (2004)
explored the moderating role of maladaptive perfectionism on the relation between adult
attachment (i.e., attachment anxiety and attachment avoidance) and depression. While no
evidence of a moderation effect was found with attachment avoidance, their findings suggested
that depression was significantly positively associated with attachment anxiety for undergraduate
students with higher maladaptive perfectionism. Maladaptive perfectionism aligns with Wang’s
definition of perfectionistic family discrepancy because they both capture the consequences of
perfectionistic tendencies (e.g., high or unrealistic expectations, greater probability for perceived
failure due to high standards for success). Thus, conceptual and related empirical findings
support the hypothesis that perfectionistic family discrepancy may also strengthen the positive
association between interpersonal shame and depression.
Self-compassion
One’s self-esteem, or the evaluation of one’s self-worth in comparison to others, has been
a common indicator of one’s mental health (as cited in Neff, 2003). Also embedded within this
definition are the perceived conditions individuals set on their self-esteem (i.e., “if others like
me, I can like myself too”). While enhancing one’s self esteem may be related to positive
psychological outcomes, there are also negative outcomes (e.g., depression, suicidal ideation,
narcissism; see Harter, 1999, for a review). Kristen Neff’s review of the self-esteem literature led
her to question how beneficial it is to raise individuals’ self-esteem when working through
mental disorders. Self-esteem can be difficult to change, may prevent true personal change or
38
growth, and may distort one’s awareness of one’s strengths and growth areas. Therefore, Neff
drew on the idea of self-compassion as an alternative means by which researchers and clinicians
could expand their understanding of healthy attitudes toward oneself.
Self-compassion is a concept common in Eastern philosophical thought but not as
familiar in Western philosophy or psychology. Neff (2003) argued that conducting research on
self-compassion would allow the field of psychology to better understand how mental well-being
can be promoted and maintained. Neff took on a positive psychological perspective to mental
disorders and psychopathology when she explained that those who exhibit self-compassion are
“touched by and open to one’s own suffering, not avoiding or disconnecting from it, generating
the desire to alleviate one’s suffering and to heal oneself with kindness” (p. 87). They may
understand their failures, inadequacies, and pain without judgment and as part of the larger
human experience. According to Neff’s (2003) Theory of Self-Compassion, individuals who
demonstrate self-compassion do so by (a) demonstrating self-kindness, (b) perceiving one’s
experiences as common among humans (i.e., common humanity), and (c) exhibiting mindfulness
of one’s painful thoughts and feelings. A brief description of each is given below.
First, individuals who demonstrate kindness to oneself extend kindness and understanding
to, rather than judging or criticizing, themselves. Next, perceiving one’s experiences as common
among humans (i.e., common humanity) allows individuals to feel connected to others who may
be undergoing experiences of pain or failure similar to theirs. In this way, individuals do not feel
isolated or alone in their suffering. Lastly, self-compassionate individuals may experience
painful feelings and thoughts associated with their suffering but, rather than over-identifying
with them, they are aware of them and hold them in “balanced awareness” (Neff, 2003, p. 89).
The act of holding their thoughts and feelings in such a manner is known as mindfulness. Neff
39
argued that these three aspects are distinct factors of self-compassion, but they can influence
each other. For instance, those who demonstrate self-kindness may be more mindful of negative
thoughts and feelings that could lead them to judge themselves. To prevent self-criticism,
individuals then practice mindfulness to avoid completely immersing themselves in their anger,
shame, or sadness. Those who are highly mindful of their negative emotions or thoughts may
also be aware of times when they focus solely on their plight; greater awareness of their tendency
to separate themselves may then encourage them to reconnect with others in spite of their
difficulties. Thus, rather than building up or solidifying the uniqueness of one’s identity as one
might do when enhancing self-esteem, self-compassion entails de-emphasizing how one’s sense
of self is different or “separate” from others (Neff, 2008).
Measuring self-compassion
Neff’s (2003) Self-Compassion Scale (SCS) measures the three components of selfcompassion on three separate subscales. Pilot testing was conducted among undergraduate
students in two separate phases. First, students met in small focus groups and answered openended questions about processes relevant to self-compassion and their reactions to experiences of
pain or failure. They then filled out a brief questionnaire containing a number of potential items
previously generated by the researchers. Items were modified based on students’ feedback, and a
larger pool was generated. Second, the modified items were administered to another group of
students in order to determine the revised items’ comprehensibility. Further modifications were
made to eliminate confusing or unclear items. Upon pilot testing the items, the scale was
administered to a larger group of undergraduate students to verify adequate psychometric
properties. Findings supported a six-factor oblique model to fit the data well, with each factor
representing one of the six hypothesized sub-constructs of self-compassion.
40
Thus, self-compassion can be seen as an overarching factor as a result of the moderate to
high inter-correlations among the sub-constructs. Items on the Self-Kindness subscale (i.e., five
items) include experiences with understanding, tolerating, and showing love towards oneself,
while items on the Self-Judgment subscale (i.e., five items) describe times when individuals are
tough on or disapproving of themselves. Experiences such as seeing one’s experiences as part of
the human condition make up the Common Humanity subscale (i.e., four items), and items
describing feeling alone or cut off from others in one’s suffering characterize the Isolation
subscale (i.e., four items). Finally, items on the Mindfulness subscale (i.e., four items) depict
individuals’ ability to view their emotions in a curious, open, and balanced way, as compared to
items on the Over-Identification subscale (i.e., four items) that describe obsession or fixation
with one’s feelings. Higher scores on the six dimensions reflect higher self-compassion.
In terms of reliability and validity, internal consistency for the 26-item SCS was .92 in a
sample of predominantly Caucasian college students (Neff, 2003). Construct validity was
evidenced by significant negative correlations with another measure of self-criticism and
significant positive correlations with a measure of social connectedness. Furthermore, the SCS
was found to significantly negatively predict mental health outcomes such as depression, anxiety,
and neurotic perfectionism. SCS scores also significantly positively predicted life satisfaction.
With respect to racial/ethnic minority samples, to my knowledge, the scale has not been used
specifically with Asian American samples. Related research, however, has been conducted with
Asian samples. For example, Neff, Pisitsungkagarn, and Hsieh (2008) found internal reliability
for the total score used among a sample of Taiwanese students was .86, and the reliabilities for
the individual subscales ranged from .65 to .74. Concurrent validity in a study with Taiwanese
college students revealed evidence of a positive link between self-compassion and self-esteem.
41
Moreover, with respect to its subscales, interdependent and independent self-construal were
associated with increased self-kindness, mindfulness, and a sense of common humanity.
Interdependent self-construal was associated with increased self-judgment and overidentification, whereas independent self-construal was negatively associated with isolation (Neff
et al., 2008). In concordance with Neff’s (2003) Theory of Self-Compassion, I used this scale to
measure self-compassion as a possible moderator for the relations between (a) interpersonal risk
factors and shame and (b) interpersonal shame and depression (see below).
With respect to the college-aged population, numerous studies have been conducted
examining the effect of self-compassion on positive (e.g., positive relations with life satisfaction,
emotional intelligence, social connectedness, well-being; Neely, Schallert, Mohammed, Roberts,
& Chen, 2009) and negative psychological outcomes (e.g., negative relations with self-criticism,
depression, anxiety, rumination, shame, and perfectionism; see Neff, 2008 for a review; Woods
& Proeve, 2014). Self-compassion has also been shown to significantly positively associate with
personality traits such as extraversion, conscientiousness, and agreeableness and negatively
associate with neuroticism (Neff, Kirkpatrick, & Rude, 2007). Currently, research has expanded
from testing direct associations or correlations to examining self-compassion in relation to future
psychological outcomes and as a moderating variable (see below). For example, Neff, Rude, &
Kirkpatrick (2007) had undergraduate students fill out the SCS on two separate occasions that
were one month apart; students’ levels of self-compassion were then rated by therapists before
and after they led them through a therapeutic exercise. Results demonstrated that selfcompassion was associated with increased psychological well-being. Raes (2011) had university
students complete the SCS and the Beck Depression Inventory to assess whether self-compassion
prospectively predicts depression five months later. Similarly, results demonstrated that self-
42
compassion significantly predicted change in depression symptoms; high levels of selfcompassion at T1 were associated with greater reductions in depression from T1 to T2.
Unfortunately, I could not locate any studies that examine self-compassion specifically
among Asian American college students. The study of self-compassion among Asians, however,
has also suggested its relation to psychological outcomes across cultures. Neff et al. (2008)
sought to answer the questions of whether or not (a) Asian individuals tended to have higher selfcompassion compared to Western individuals; (b) self-compassion would be associated with
psychological well-being; and (c) self-compassion would be linked to independent and
interdependent self-construal. Data was collected from participants in Thailand, Taiwan, and the
U.S. The authors found that level of self-compassion and its six dimensions was highest in
Thailand and lowest in Taiwan, with the U.S. falling in between (Neff et al., 2008); crosscultural differences in self-compassion remained even when controlling for type of selfconstrual. They hypothesized that low self-compassion in Taiwanese participants may be due to
high levels of negative self-relevant emotions (e.g., shame). Nevertheless, self-compassion was
related to well-being (i.e., less depression and greater life satisfaction) in all participants from all
three cultures. Finally, interdependent self-construal was only related to self-compassion among
Thai participants, while independent self-construal was linked to self-compassion in Taiwan and
the U.S. While results of this cross-cultural study provide some insight into cultural differences
in self-compassion, future research should disentangle how cultural values and practices within
collectivistic and individualistic cultures promote or discourage self-compassion.
Rationale for moderation hypothesis
With respect to self-compassion, it was expected that positive associations between
interpersonal risk factors (i.e., thwarted belongingness and perceived burdensomeness) and
43
interpersonal shame would be weaker for those with higher self-compassion than for those with
lower self-compassion. Asian Americans who display high self-compassion may be better able to
show themselves kindness and understand that they are not alone for feeling a lack of
belongingness. In the face of thwarted belongingness, they might recognize that it may take time
to feel connected with others and hold their loneliness in mindful awareness. Moreover, greater
self-kindness may help them to feel less perceived burdensomeness; high self-compassion may
make it less likely that individuals criticize themselves and judge their self-worth as low (i.e., as
a burden or harm to others). In this way, self-compassion as a resource might decrease the
interpersonal shame they experience because they have difficulty connecting with others and
believing in their worth and competence. Empirically, I was not able to locate any studies that
have specifically tested the moderating role of self-compassion on the interpersonal risk factorsinterpersonal shame links. One reason for this may be that Joiner’s interpersonal states are
relatively new to the literature; however, the moderating role of self-compassion with respect to
other psychological outcomes has been shown elsewhere (see below). Nonetheless, it was
hypothesized that self-compassion would moderate the links of interpersonal risk factors and
interpersonal shame.
Additionally, it was expected that the positive relation between interpersonal shame and
depression would be weaker for those with a higher level of self-compassion than for those with
a lower level of self-compassion. The rationales for these hypotheses are the following. First,
Asian Americans who exhibit high self-compassion may be less critical of themselves when
feeling inadequate, negatively evaluated by others, and believing they are bringing shame upon
their family. Instead, they may wish to promote their well-being and heal their pain. Next, they
may also realize that there are others like themselves who struggle with shame resulting from a
44
failure to meet their family’s expectations for perfection. Rather than feeling isolated or separate
from other Asian Americans who face interpersonal shame, they may realize that others struggle
the way they do. It may be easier for Asian Americans high on self-compassion to see that their
experience is part of a greater human condition (Neff, 2004). Lastly, highly self-compassionate
Asian Americans may be more aware of their shamed feeling. They do not ignore or repress their
pain, and they also do not over-identify with their shame to the point of feeling depressed.
Awareness of the extent of their shame in the present moment may help prevent them from
drawing harsh conclusions about their value and worth to others (e.g., if I am inadequate and
bringing shame to my family). Ultimately, when Asian Americans experience interpersonal
shame, kindness and understanding towards themselves and acceptance of their limitations or
imperfections may protect them from depression. High self-compassion may buffer the shamedepression link because it is a personal resource to regulate their interpersonal shame feelings in
order to prevent future depression.
While no studies have specifically tested for self-compassion as a moderator in the
interpersonal shame-depression link, recent empirical studies have showed some evidence for the
moderating role of self-compassion among college students. Terry, Leary, and Mehta (2012)
analyzed how self-compassion may buffer the effect of social and academic difficulties first-year
students experience in their transition to college on their depression, homesickness, and
dissatisfaction with their decision to attend university. The relations between (a) social
difficulties and homesickness; (b) social difficulties and dissatisfaction with decision to attend
university; and (c) academic difficulties and homesickness were weaker for students who scored
higher on self-compassion. Terry et al.’s (2012) results suggest that highly self-compassionate
students may be better able to overcome difficulties characteristic of the college transition by
45
quieting their urge to self-blame; separate themselves from others while they endure their
difficulties; or over-identifying with their negative emotions.
Furthermore, two cross-cultural studies suggested that self-compassion may serve as a
moderator for psychological outcomes. For example, Wong and Mak (2012) examined the three
specific components of self-compassion (i.e., self-kindness, common humanity, and
mindfulness) and their possible buffering effect on the relation between cognitive-personality
vulnerability styles and depression. The authors defined three cognitive-personality vulnerability
styles as (a) sociotropy (i.e., the fear of abandonment and possessing excessive dependency
needs); (b) autonomy (i.e., the tendency to reject help from others); and (c) self-criticism. In a
sample of Chinese adults in Hong Kong, results suggested that the association between
autonomy and depression was weaker among individuals with high self-kindness and high
mindfulness; the association between self-criticism and depression was also weaker among
individuals with high self-kindness, high common humanity, and high mindfulness. Additionally,
Kyeong (2013) sampled undergraduate students from Korea to assess the possible moderating
effect of self-compassion on psychological well-being in the face of academic burden. Multiple
regression analyses showed that the effect of academic burden on students’ depression was
stronger for those who did not exhibit self-compassion. Given the above research, there is
evidence to show that self-compassion and its three components may play moderating roles
between various predictor variables and negative psychological outcomes such as depression.
Control Variable: Initial Level of Depression
Controlling for initial levels of depression was warranted to rule out this possible
confounding variable. Since I collected data at three time points, it was possible that participants
who exhibit higher depression at T3 may already exhibit high depression at T1. Likewise, if
46
participants do not exhibit high depression at T3, then perhaps they do not show symptoms of
depression at T1.
The depression subscale of the Depression, Anxiety, and Stress Scales-Short form
(DASS; S.H. Lovibond & P.F. Lovibond, 1995) was administered to measure the index of
depression at T1 and T3 of the study. The DASS is a self-report survey that covers a full range
of core symptoms of anxiety and depression. In comparison to the Beck Depression Inventory
(BDI) and the Beck Anxiety Inventory (BAI), the DASS better discriminates between depression
and other affective states (P.F. Lovibond & S.H. Lovibond, 1995) and possesses high reliability
(i.e., .88 for Depression; Henry & Crawford, 2005; S.H. Lovibond & P.F. Lovibond, 1995).
With respect to racial/ethnic minority samples, Wei et al. (2004) found that internal consistency
for the depression subscale was .88 among Asian American college students. Depression was
found to be positively correlated with attachment anxiety, attachment avoidance, and anxiety
(Wei et al., 2004). More recently, Norton (2007) demonstrated structural invariance of the
DASS-21 for each of the three subscales across Caucasian, African American, Hispanic, and
Asian college students. Cronbach’s alpha for the depression subscale among Asian college
students was .84.
Conclusion
The brief literature review suggests that perfectionistic family discrepancy and selfcompassion might moderate the mediating relationships (i.e., the relations between interpersonal
risk factors and depression through interpersonal shame). Therefore, the goals of the present
study were to demonstrate the following: (1) whether or not interpersonal shame mediates the
relationships between interpersonal risk factors and depression, and (2) whether or not the above
47
two variables moderate both paths of the mediating relationships across three time points. In both
sets of hypotheses, initial level of depression was accounted for in the analyses.
48
CHAPTER 3. MATERIALS AND METHODS
Power Analysis
Mediation
Researchers have suggested at least 200 to 300 participants for conducting a model
(Comrey & Lee, 1992). Given that I collected data across three time points, I also accounted for
a 30% to 50% attrition rate that may result from collecting data at multiple time points.
Considering Comrey and Lee’s (1992) suggestion of 200 participants, I expected to have
approximately 800, 400, and 200 participants at Times 1, 2, and 3, respectively. With reference
to their suggestion of 300 participants, I expected to have approximately 1,200, 600, and 300
participants at Times 1, 2, and 3, respectively. Therefore, I sought to obtain a range between 800
and 1,200 participants at T1, 400 to 600 participants at T2, and 200 to 300 participants at T3.
Moderation
Power in a quantitative study is a function of effect size, sample size, and alpha level.
Effect size can be expressed by R2
, or a standardized regression coefficient used to convey the
correlation among or between variables. A power analysis was conducted using G*Power
3.1.9.2 (Erdfelder, Faul, & Buchner, 1996) to estimate the number of participants needed. To
determine sample size requirements, Cohen (1992) recommended that each predictor variable
(i.e., interpersonal risk factors [i.e., thwarted belongingness or perceived burdensomeness] and
the moderating variable [i.e., perfectionistic family discrepancy or self-compassion]) be assigned
an effect size of f
2 =.02, .15, or .35 for small, medium, or large effect sizes for the two-way
interaction, respectively, in relation to the criterion variable (i.e., depression). Subsequent
calculations indicated that sample sizes of 602, 85, or 40, respectively, were needed for a power
49
of .80 or higher at p < .05. Based on these calculations, a sample size between 602 and 40
participants was chosen for a small to medium effect size for the two-way interaction.
Participants
The sample of this 3-time point study included 605 Asian American undergraduate
students (T1 [beginning of spring semester; N = 605]; T2 [end of spring semester; N = 280]; and
T3 [middle of fall semester; N = 172]) from large, public or private, predominantly White
universities in the Midwest. In addition to Iowa State University, universities that were sampled
included the University of Illinois at Chicago, Northern Illinois University, Northwestern
University, the University of Michigan at Ann-Arbor, Michigan State University, Wayne State
University, Wichita State University, the University of Wisconsin-Milwaukee, Purdue
University, Indiana University at Bloomington, Garden Valley State University, and the
University of Minnesota-Twin Cities.
Approximately 54 percent of the participants identified as female (i.e., 328; 54.2%), 150
participants identified as male (24.8%), three participants identified as transgender (0.5%), and
six participants preferred not to answer this question (1.0%).1 Participants’ ages ranged from 18
to 64 (M = 21.65, SD = 2.74). Of the 605 participants, 110 (18.2%) were first-year students, 104
(17.2%) were second-year students, 119 (19.7%) were third-year students, and 153 (25.3%) were
fourth-year students and above. With respect to sexual orientation, about 70% percent identified
as heterosexual/straight (i.e., 425; 70.2%), 5 participants identified as gay (0.8%), 3 participants
identified as lesbian (0.5%), 23 participants identified as bisexual (3.8%), 9 participants
identified as questioning (1.5%), and 21 participants preferred not to answer (3.5%).
Approximately 46 percent of the participants identified as 2nd generation Asian American (i.e.,

1 One reason for the high number of missing demographic data may be because the demographic
questionnaire was placed at the end of the survey.
50
277; 45.8%), while 122 (20.2%) identified as 1.5 generation, 37 (6.1%) as 1st generation, 18
(3.0%) as 4th generation, 13 (2.1%) as 5th generation, and 2 (0.3%) as 3rd generation. Lastly, with
respect to ethnic sub-group identification, participants identified as Chinese-American (92;
15.2%), Vietnamese-American (83; 13.7%), Asian Indian American (74; 12.2%), KoreanAmerican (41; 6.8%), Hmong-American (41; 6.8%), Filipino/a-American (34; 5.6%), JapaneseAmerican (21; 3.5%), Laotian-American (15; 2.5%), Taiwanese-American (13; 2.1%),
Malaysian-American (6; 1.0%), Cambodian-American (5; 0.8%), Thai-American (4; 0.7%),
Indonesian-American (4; 0.7%), Pan-Asian American (2; 0.3%), and Other (43; 7.1%; e.g.,
Bengali, Nepali-American).
Materials
Interpersonal risk factors
Interpersonal Needs Questionnaire (INQ-15; Van Orden et al., 2012). The INQ scale is a
15-item self-report measure that assesses individuals’ levels of perceived disconnection from
others (i.e., thwarted belongingness; nine items) and the degree to which they feel they are
burdens to others (i.e., perceived burdensomeness; six items). Sample items for the thwarted
belongingness construct include, “These days, I feel disconnected from other people” and “These
days, I have at least one satisfying interaction every day.” Sample items for the perceived
burdensomeness construct include, “These days, I think I make things worse for people in my
life” and “These days, I think I am a burden on society.” Participants rate items on a 7-point
scale from 1 (Not at all true for me) to 7 (Very true of me). High scores reflect higher levels of
thwarted belongingness and perceive burdensomeness. Although no studies to the author’s
knowledge have used the INQ-15 with Asian Americans, other forms of the INQ have
demonstrated adequate reliability for those with Asian heritages. For example, Wong et al.
51
(2011b) found internal consistencies for the thwarted belongingness and perceived
burdensomeness subscales in the INQ-12 of .88 and .74, respectively, among Asian American
college students. Both subscales were significantly (a) positively correlated with suicidal ideation
and depression and (b) negatively correlated with interdependent and independent self-construal.
Zhang et al. (2013) tested Joiner’s theory in a sample of Chinese university students. The INQ18 demonstrated strong internal consistency reliability (α = .87), and thwarted belongingness and
perceived burdensomeness were (a) positively associated with suicidal ideation, depression, and
capability for self-harm; and (b) negatively associated with social support and self-esteem
(Zhang et al., 2013). In this study, Cronbach’s alpha was .88 for thwarted belongingness at T1
and .93 for perceived burdensomeness at T1.
Interpersonal shame
Interpersonal Shame Inventory (ISI; Wong et al., 2014). The ISI is a 10-item self-report
measure that assesses two dimensions of interpersonal shame: external shame (i.e., ISI-E; five
items) and family shame (i.e., ISI-F; five items). The ISI-E assesses the degree to which
participants are concerned about others’ negative evaluations of them. Sample items include,
“These days, I feel like hiding because people might view me as flawed” and “These days, I wish
I could shrink away because others might perceive me as incompetent.” The ISI-F examines the
degree to which participants perceive that they have brought shame to their families. Sample
items include, “These days, I wish I could disappear because my deficits might cause my family
to lose face” and “These days, I feel like escaping because my defects might disgrace my
family.” Participants rate items on a 6-point Likert scale from 1 (Strongly Disagree) to 6
(Strongly Agree). Higher scores on the two subscales reflect higher interpersonal shame. Both
dimensions demonstrated strong internal consistency reliability (α = .94 for the ISI-E and α = .97
52
for the ISI-F) among Asian and Asian American college students (Wong et al., 2014). In the
present study, Cronbach’s alpha was .96 for the total score at T2. Concurrent validity in a study
with Asian American college students revealed evidence of a positive link between the ISI-E and
ISI-F, generic state shame, thwarted belongingness, perceived burdensomeness, self-face
concerns, depression, and suicidal ideation. Furthermore, both the ISI-E and ISI-F were
significantly negatively related to self-esteem among Asian and Asian American college students
(Wong et al., 2014).
Perfectionistic family discrepancy
Family Almost Perfect Scale-Perfectionistic Family Discrepancy subscale (FAPS-D;
Wang, 2010). The perfectionistic family discrepancy subscale is a 7-item self-report measure,
which assesses the extent to which participants perceive that they fail to meet their families’
standards for performance. Sample items include, “Doing my best never seems to be enough for
my family” and “Nothing short of perfect is acceptable in my family.” Participants rate items on
a 7-point Likert scale from 1 (Strongly Disagree) to 7 (Strongly Agree). Higher scores indicate
higher levels of perfectionistic family discrepancy. Wang (2010) found (a) positive correlations
of perfectionistic family discrepancy with parental expectations, parental criticism, anxiety and
depression; and (b) negative correlations with self-esteem. Cronbach’s alpha values ranged from
.91 to .94 among Asian and European American college students (Wang, 2010). Cronbach’s
alpha in this study was .95 at T2.
Self-compassion
Self-Compassion Scale (SCS; Neff, 2003). The SCS is a 26-item self-report measure that
assesses six dimensions of self-compassion: self-kindness, self-judgment, common humanity,
isolation, mindfulness, and over-identification. Participants rate items on a 5-point scale from 1
53
(Almost Never) to 5 (Almost Always). Higher scores on the six dimensions reflect higher selfcompassion. To the author’s knowledge, the scale has not been used specifically with Asian
American samples. Related research, however, has been conducted with Asian samples. Internal
reliability for the total score used among a sample of Taiwanese students was .86, and the
reliabilities for the individual subscales ranged from .65 to .74 (Neff et al., 2008). In this study,
Cronbach’s alpha for the total score was .74 at T2. Concurrent validity in a study with Taiwanese
college students revealed evidence of a positive link between self-compassion and self-esteem.
Moreover, with respect to its subscales, interdependent and independent self-construal were
associated with increased self-kindness, mindfulness, and a sense of common humanity.
Interdependent self-construal was associated with increased self-judgment and overidentification, whereas independent self-construal was negatively associated with isolation (Neff
et al., 2008).
Depression
Depression and Anxiety Stress Scales-Short Form, Depression subscale (DASS; S.H.
Lovibond & P.F. Lovibond, 1995). The depression subscale of the DASS-21 is a 7-item
subscale that measures depressive symptoms (e.g., “I felt downhearted and blue”). Individuals
rate each symptom on a 4-point scale ranging from 0 (Did not apply to me at all) to 3 (Applied to
me very much or most of the time) using “Over the past week, including today” as a reference
point. Higher scores indicate more depression. With respect to racial/ethnic minority samples,
Wei et al. (2004) found that internal consistency for the depression subscale was .88 among
Asian American college students. Depression was found to be positively correlated with
attachment anxiety, attachment avoidance, and anxiety (Wei et al., 2004). More recently, Norton
(2007) demonstrated structural invariance of the DASS-21 depression subscale across Caucasian,
54
African American, Hispanic, and Asian college students. Cronbach’s alpha for the depression
subscale among Asian college students was .84 (Norton, 2007). Cronbach’s alphas in this study
were .89 at T1 (i.e., control variable) and .92 at T3 (i.e., outcome variable).
Demographic information
Participants completed a 6-item questionnaire for information on their age, gender, sexual
orientation, education level, generation status, and ethnicity.
Procedure
Prior to administering questionnaires via an online survey site, I obtained approval from
the Institutional Review Board (IRB) at Iowa State University (IRB ID: 14-568, approved
12/10/14; see Appendix A) to conduct my study with Asian American college students. I
obtained a list of names and email addresses from the Office of the Registrar to invite selfidentified Asian American students to participate in the study (see Appendix B). Participants
signed up via an email. The study description invited them to participate in a study examining
factors related to interpersonal stressors and psychological outcomes. In addition, I contacted the
Institutional Review Board (IRB) at the other universities to find out whether or not IRB
approval was needed. I then contacted the Office of the Registrar, Student Records, or
Multicultural offices at each university for a list of names and emails for students that selfidentified as Asian American or emails for Asian American student organizations (see Appendix
C). I emailed students on the list given by each office using the same email template I used to
recruit participants at Iowa State University.
When participants signed up for the study at the beginning of the spring semester, they
were guided to the survey via a website hosted through qualtrics.com. After participants clicked
on the link to the survey, they first read through an informed consent page that explained the
55
nature of the study, the risks and benefits, and their rights as participants (see Appendix D). Once
participants confirmed that they were over the age of 18 and consented to continue with the
study, they were forwarded to the actual survey. Participants completed the INQ-15, ISI, FAPSD, SCS, and the DASS-Depression scales (see above). At the end of the survey at T1, the
participants were thanked for their participation, read through a debriefing form (see Appendix
E), and were given my contact information for their reference. Participants had the opportunity to
submit their contact information (i.e., name, phone number, and email address) to be entered into
a random drawing for a $25 VISA gift card.
I kept a record of participants’ contact information to contact participants again at the end
of the spring semester (T2) and in the middle of the fall semester (T3). At T2, participants were
contacted via email to complete the survey on the same website. All students were able to enter
into a random drawing for a $25 VISA gift card. At T3, participants were contacted via email to
complete the same survey they completed at T1 and T2. At the end of the T3 survey, all
participants had a third opportunity to enter into a random drawing for a $25 VISA gift card.
Once participants finished the surveys at T2 and T3, they were thanked for their participation,
debriefed on the nature of the study, and given the researcher’s contact information. The surveys
at each time point included three validity check items (e.g., “A week has 7 days”) to help filter
responses that may be submitted randomly. I maintained participant confidentiality by storing
participants’ contact information separate from the survey data after they completed the surveys
at all time points. Contact information was only reviewed at the end of the project for those who
entered the drawing for a $25 VISA gift card.
56
CHAPTER 4. RESULTS
Preliminary Analyses
Independent samples t-test
Preliminary analyses were conducted before analyzing the main hypotheses. First, an
independent samples t-test was conducted to examine whether my variables of interest (i.e.,
thwarted belongingness, perceived burdensomeness, and depression at T1) varied between Asian
American students who dropped out (i.e., those who only participated at T1) and those who did
not drop out (i.e., those who participated at all three time points or at least two time points).
Results indicated that there were no significant differences among participants who dropped out
and those who did not, t (488) = 1.28, p = .20, Cohen’s d = .12, for depression at T1 to t (603) = –
.54, p = .59, Cohen’s d =.04 for perceived burdensomeness at T1.There did not seem to be a
systematic reason for why students dropped out after the first wave of data collection.
Missing data analysis
Second, I examined the missing data. Data from 677, 280, and 172 Asian American
college students were collected at T1, T2, and T3, respectively. Eight participants were removed
because they answered all nine of the validity questions (e.g., “One week has seven days”)
incorrectly (i.e., N = 669). Furthermore, 45 participants were removed due to not having
completed any of the survey (i.e., N = 645), and 19 participants were removed because they had
completed only T2 or only T3 (i.e., N = 605). The result from Little’s Missing Completing at
Random (MCAR) test was not found to be significant, χ
2
(34, N = 605) = 38.94, p = .26.
Therefore, this finding suggests that the missing data were missing completely at random. Based
on Schlomer, Bauman, and Card’s (2010) recommendation, the expectation maximization
57
estimation method, a maximum likelihood method superior to other methods such as listwise
deletion or mean substitution (Schafer & Graham, 2002), was used to impute the missing data.
Means, standard deviations, and zero-order correlations (see Table 1)
With respect to the predictor variables, significantly moderate or large positive
correlations existed among thwarted belongingness at T1 and (a) perceived burdensomeness at
T1; (b) depression at T1; (c) interpersonal shame at T2; (d) perfectionistic family discrepancy at
T2; and (e) depression at T3. Thwarted belongingness at T1 was moderately negatively
correlated with self-compassion at T2. Significantly large positive correlations also existed
among perceived burdensomeness at T1 and (a) depression at T1; (b) interpersonal shame at T2;
(c) perfectionistic family discrepancy at T2; and (d) depression at T3. Perceived burdensomeness
at T1 was moderately negatively correlated with self-compassion at T2.
Correlations among all other variables
Depression at T1 was positively correlated with (a) interpersonal shame at T2; (b)
perfectionistic family discrepancy at T2; and (c) depression at T3. Depression at T1 was
moderately negatively correlated with self-compassion at T2. With respect to interpersonal
shame at T2, results demonstrated large positive correlations with perfectionistic family
discrepancy at T2 and depression at T3 and a moderate negative correlation with selfcompassion at T2. Finally, perfectionistic family discrepancy at T2 was negatively correlated
with self-compassion at T2 and positively correlated with depression at T3. Self-compassion at
T2 was moderately negatively correlated with depression at T3.
11
Table 1
Means, Standard Deviations, and Zero-Order Correlations
Note. N = 605. T = Time. ** p < .01.
1 2 3 4 5 6 7
1. Thwarted Belongingness at T1 —–
2. Perceived Burdensomeness at T1 .54** —–
3. Depression at T1 .62** .64** —–
4. Interpersonal Shame at T2 .48** .67** .57** —–
5. Perfectionistic Family Discrepancy at T2 .35** .42** .35** .51** —–
6. Self-compassion at T2 -.36** -.47** -.47** -.52** -.35** —–
7. Depression at T3 .44** .52** .55** .56** .35** -.45** —–
Possible Range 1-7 1-7 0-3 1-6 1-7 1-5 0-3
Mean 2.82 1.73 .74 2.06 3.28 3.00 .60
SD 1.26 1.12 .69 1.23 1.71 .71 .66
58
59
Plan for Data Analysis
Conditional process modeling via Hayes’s (2013) PROCESS in SPSS was used to
examine the hypothesized moderated mediation effects using bias-corrected bootstrapping (e.g.,
MacKinnon, Lockwood, & Williams, 2004; Mallinckrodt, Abraham, Wei, & Russell, 2006;
Shrout & Bolger, 2002). A total of 1,000 bootstrap samples were requested to compute the
means of 1,000 estimated indirect effects. MacKinnon et al. (2004) indicated that the bootstrap
confidence interval (CI) adjusted for bias showed the highest levels of statistical power. Thus, 95
% bias-corrected bootstrap CIs for the indirect effects were reported. I identified significant
mediation effects at the .05 level if the 95% CI did not include zero.
In the command guide for PROCESS, the hypothesized moderated mediation model for
this study (see Figure 9) was first graphically depicted in conceptual and statistical terms as
“Model 58,” where “X” (i.e., predictors) represented thwarted belongingness at T1 or perceived
burdensomeness at T1, “Y” (i.e., outcome) represented depression at T3, “Mi” (i.e., mediator)
represented interpersonal shame at T2, and “W” (i.e., moderator) represented perfectionistic
family discrepancy at T2 or self-compassion at T2. This moderated mediation model allowed us
to test multiple moderator effects simultaneously (i.e., X ˟ W, Mi ˟ W). Second, the index of
moderated mediation was used to test for the presence of moderated mediation (Hayes, 2013;
2015). Moderation of the indirect effect exists if the index of moderated mediation is statistically
different from zero. Third, PROCESS generated conditional indirect (mediation) effects at one
standard deviation above and below the mean of the moderating variables to represent the nature
of the moderated mediation effects at high and low levels. The predictors (i.e., thwarted
belongingness at T1 and perceived burdensomeness at T1) and moderators (i.e., perfectionistic
60
family discrepancy at T2 and self-compassion at T2) were mean-centered to better interpret the
moderation effects.
Figure 9. Conceptual Diagram for Model “58” from Hayes’s (2013) Command Guide in
PROCESS.
Test for Moderated Mediation: Conditional Process Modeling
The original hypotheses were that the mediation effects from interpersonal risk factors
(i.e., thwarted belongingness at T1 and perceived burdensomeness at T1; X1 and X2 as two
predictors) on depression at T3 (Y as the dependent variable) through interpersonal shame at T2
(Mi as the mediator) would be moderated by perfectionistic family discrepancy at T2 or selfcompassion at T2 (W1 and W2 as two moderators; see Figure 2). I ran two moderated mediation
models (i.e., one including each moderator of interest). Therefore, an adjusted p value of .025
(i.e., .05/2 for each of the two moderators of interest) was used for the significant level. In
addition, it is important to note that PROCESS only allows one predictor to be entered into the
model for analysis. When one predictor (e.g., thwarted belongingness at T1) was entered in the
model, the other predictor (e.g., perceived burdensomeness at T1) served as a covariate.
Therefore, for each moderator, there were two sets of results (i.e., one for each predictor).
61
Perfectionistic family discrepancy at T2 as moderator
Results for thwarted belongingness at T1as a predictor indicated that the index of
moderated mediation (B = -.003, 95% CI [-.011, .002]) was not significant. This implies that
perfectionistic family discrepancy at T2 did not significantly moderate the indirect (mediation)
effect from thwarted belongingness at T1 to depression at T3 through interpersonal shame at T2.
As shown in Table 2, results consistently indicated that perfectionistic family discrepancy at T2
did not moderate the paths from thwarted belongingness at T1 (X1) to interpersonal shame at T2
(Mi; see path a3 in Table 2), as well as from interpersonal shame at T2 (Mi) to depression at T3
(Y; see path b3 in Table 2).
Results for perceived burdensomeness at T1 as a predictor indicated that the index of
moderated mediation (B = -.003, 95% CI [-.013, .001]) was not significant. That is,
perfectionistic family discrepancy at T2 did not significantly moderate the indirect (mediation)
effect from perceived burdensomeness at T1 to depression at T3 through interpersonal shame at
T2. As shown on Table 3, results indicated that perfectionistic family discrepancy at T2 did not
moderate the paths from perceived burdensomeness at T1 (X2) to interpersonal shame at T2 (Mi;
see path a3 in Table 3), as well as from interpersonal shame at T2 (Mi) to depression at T3 (Y; see
path b3 in Table 3).
Self-compassion at T2 as moderator
First, results indicated that the index of moderated mediation (B = .007, 95% CI [-.012,
.030]) was not significant for thwarted belongingness at T1 as a predictor. Therefore, selfcompassion at T2 did not significantly moderate the indirect (mediation) effect from thwarted
belongingness at T1 to depression at T3 through interpersonal shame at T2. Table 4 demonstrates
that self-compassion at T2 did not moderate the paths from thwarted belongingness at T1 (X1) to
62
interpersonal shame at T2 (Mi; see path a3 in Table 4), as well as from interpersonal shame at T2
(Mi) to depression at T3 (Y; see path b3 in Table 4).
Lastly, the index of moderated mediation, B = .016, 95% CI (-.002, .046), was not
significant for perceived burdensomeness at T1 as a predictor. Self-compassion at T2 did not
significantly moderate the indirect (mediation) effect from perceived burdensomeness at T1 to
depression at T3 through interpersonal shame at T2. Table 5 shows that self-compassion at T2
did moderate the path from perceived burdensomeness at T1 (X2) to interpersonal shame at T2
(Mi; see path a3 in Table 5). However, using my adjusted p value of .025 (see above), selfcompassion at T2 did not significantly moderate the path from perceived burdensomeness at T1
(X2) to interpersonal shame at T2 (Mi). Furthermore, self-compassion at T2 did not moderate the
path from interpersonal shame at T2 (Mi) to depression at T3 (Y; see path b3 in Table 5).
Post Hoc Analyses
Unexpectedly, my original hypotheses that the indirect (mediation) effects would be
moderated by my moderators of interest (i.e., perfectionistic family discrepancy at T2 and selfcompassion at T2) were not supported by this data. Therefore, post hoc analyses were conducted
to examine an alternative moderated mediation model. This alternative moderated mediation
model refers to Model 5 in PROCESS (see Figure 10 below; Hayes, 2013).
Figure 10. Conceptual Diagram for Model “5” from Hayes’s (2013) Command Guide in
PROCESS.
11
Table 2
Unstandardized Path Coefficients with Confidence Intervals Estimating Whether Perfectionistic Family Discrepancy at T2 Moderated
the Indirect (Mediation) Effect from Thwarted Belongingness at T1 to Depression at T3 through Interpersonal Shame at T2
Note. T = Time. CI = Confidence Interval. * p < .05.** p < .01. *** p < .001. P values are based on the Sobel test.
Interpersonal Shame at T2 (Mi) Depression at T3 (Y)
Path B SE 95% CI Path B SE 95% CI
Constant a0 -1.02*** .05 -1.12, -.916 b0 .407*** .03 .346, .469
Interpersonal Shame at T2 (Mi) b1 .165*** .02 .125, .204
Thwarted Belongingness at T1 (X1) a1 .052* .02 .004, .100 c’ .056*** .01 .034, .079
Perfectionistic Family Discrepancy at T2 (W1) a2 .176*** .02 .133, .220 b2 .028** .01 .006, .050
X1 ˟ W1 a3 -.022 .01 -.050, .005
Mi ˟ W1 b3 .015 .01 -.001, .031
Perceived Burdensomeness at T1 (covariate) a4 .458*** .03 .400, .516 b4 .008 .02 -.025, .040
Depression at T1 (covariate) a5 .328*** .05 .228, .427 b5 .256*** .02 .208, .305
R
2 = .713
F(5,599) = 297.108, p < .001
R
2 = .710
F(6,598) = 244.052, p < .001
63
11
Table 3
Unstandardized Path Coefficients with Confidence Intervals Estimating Whether Perfectionistic Family Discrepancy at T2 Moderated
the Indirect (Mediation) Effect from Perceived Burdensomeness at T1 to Depression at T3 through Interpersonal Shame at T2
Note. T = Time. CI = Confidence Interval. ** p < .01. *** p < .001. P values are based on the Sobel test.
Interpersonal Shame at T2 (Mi) Depression at T3 (Y)
Path B SE 95% CI Path B SE 95% CI
Constant a0 -.347*** .07 -.484, -.210 b0 .261*** .03 .194, .328
Interpersonal Shame at T2 (Mi) b1 .165*** .02 .125, .204
Perceived Burdensomeness at T1 (X2) a1 .481*** .04 .410, .551 c’ .008 .02 -.025, .040
Perfectionistic Family Discrepancy at T2 (W1) a2 .175*** .02 .131, .218 b2 .028** .01 .006, .050
X2 ˟ W1 a3 -.025 .02 -.055, .006
Mi ˟ W1 b3 .015 .01 -.001, .031
Thwarted Belongingness at T1 (covariate) a4 .045 .02 -.004, .093 b4 .056*** .01 .034, .079
Depression at T1 (covariate) a5 .330*** .05 .231, .429 b5 .256*** .02 .208, .305
R
2 = .713
F(5,599) = 297.083, p < .001
R
2 = .710
F(6,598) = 244.052, p < .001
64
11
Table 4
Unstandardized Path Coefficients with Confidence Intervals Estimating Whether Self-Compassion at T2 Moderated the Indirect
(Mediation) Effect from Thwarted Belongingness at T1 to Depression at T3 through Interpersonal Shame at T2
Note. T = Time. CI = Confidence Interval. ** p < .01. *** p < .001. P values are based on the Sobel test.
Interpersonal Shame at T2 (Mi) Depression at T3 (Y)
Path B SE 95% CI Path B SE 95% CI
Constant a0 -.990*** .06 -1.10, -.880 b0 .438*** .03 .376, .501
Interpersonal Shame at T2 (Mi) b1 .162*** .02 .125, .200
Thwarted Belongingness at T1 (X1) a1 .072** .02 .025, .120 c’ .056*** .01 .034, .078
Self-compassion at T2 (W2) a2 -.385*** .06 -.497, -.274 b2 -.136*** .03 -.189, -.082
X1 ˟ W2 a3 .032 .03 -.032, .096
Mi ˟ W2 b3 -.016 .02 -.050, .017
Perceived Burdensomeness at T1 (covariate) a4 .465*** .03 .407, .523 b4 .007 .02 -.026, .039
Depression at T1 (covariate) a5 .263*** .05 .157, .368 b5 .224*** .03 .175, .273
R
2 = .705
F(5,599) = 285.981, p < .001
R
2 = .718
F(6,598) = 253.289, p < .001
65
11
Table 5
Unstandardized Path Coefficients with Confidence Intervals Estimating Whether Self-Compassion at T2 Moderated the Indirect
(Mediation) Effect from Perceived Burdensomeness at T1 to Depression at T3 through Interpersonal Shame at T2
Note. T = Time. CI = Confidence Interval. * p < .05.** p < .01. *** p < .001. P values are based on the Sobel test.
Interpersonal Shame at T2 (Mi) Depression at T3 (Y)
Path B SE 95% CI Path B SE 95% CI
Constant a0 -.343*** .07 -.482, -.205 b0 .291*** .03 .226, .356
Interpersonal Shame at T2 (Mi) b1 .162*** .02 .125, .200
Perceived Burdensomeness at T1 (X2) a1 .506*** .04 .436, .576 c’ .007 .02 -.026, .039
Self-compassion at T2 (W2) a2 -.378*** .06 -.490, -.266 b2 -.136*** .03 -.189, -.082
X2 ˟ W2 a3 .074* .03 .008, .140
Mi ˟ W2 b3 -.016 .02 -.050, .017
Thwarted Belongingness at T1 (covariate) a4 .063** .02 .015, .111 b4 .056*** .01 .034, .078
Depression at T1 (covariate) a5 .262*** .05 .157, .367 b5 .224*** .03 .175, .273
R
2 = .707
F(5,599) = 288.627, p < .001
R
2 = .718
F(6,598) = 253.289, p < .001
66
67
The alternative moderated mediation model for this study was defined by two parts (see
Figure 11). The first part hypothesized the indirect (mediation) effects from the two interpersonal
risk factors at T1 to depression at T3 through interpersonal shame at T2. The second part
hypothesized that (a) perfectionistic family discrepancy at T2 and (b) self-compassion at T2
would moderate the direct effects from the two interpersonal risk factors at T1 to depression at
T3. Similar to the main analyses above, when one predictor (e.g., thwarted belongingness at T1)
was entered in the model, the other predictor (e.g., perceived burdensomeness at T1) served as a
covariate. Thus, for each moderator, there were two sets of results (i.e., one for each predictor).
Perfectionistic family discrepancy as a moderator
The first set of results for thwarted belongingness at T1 as a predictor demonstrated that
interpersonal shame at T2 significantly mediated the association between thwarted belongingness
at T1 and depression at T3, b = .015, SE = .01, 95% CI (.006, .030). Furthermore, the results also
indicated that perfectionistic family discrepancy at T2 significantly moderated the direct path
from thwarted belongingness at T1 to depression at T3, b = .017, SE = .01, 95% CI (.004, .030;
see path c3 in Table 6). To better understand the nature of the interaction, results from
conditional direct effects reported in PROCESS indicated that the simple slope was significant at
low (i.e., 1 SD below the mean; b =.031, SE = .01, 95% CI [.003, .059]) and high levels (i.e., 1
SD above the mean; b =.072, SE = .01, 95% CI [.046, .099]) of perfectionistic family
discrepancy at T2. Figures 12 and 13 demonstrate that the slope at high perfectionistic family
discrepancy at T2 was significantly deeper than that at low perfectionistic family discrepancy at
T2. That is, the direct effect of thwarted belongingness at T1 on depression at T3 was stronger
for those with higher than lower perfectionistic family discrepancy at T2.
68
The second set of results for perceived burdensomeness at T1 also demonstrated that
interpersonal shame at T2 significantly mediated the association between perceived
burdensomeness at T1 and depression at T3, b = .094, SE = .02, 95% CI (.062, .132). Moreover,
the results indicated that perfectionistic family discrepancy at T2 significantly moderated the
direct path from perceived burdensomeness at T1 to depression at T3, b = .018, SE = .01, 95% CI
(.004, .032; see path c3 in Table 7). Results from conditional direct effects reported in PROCESS
indicated that the simple slopes were not significant at low (i.e., 1 SD below the mean; b = -.033,
SE = .03, 95% CI [-.083, .017]) and high levels (i.e., 1 SD above the mean; b = .011, SE = .02,
95% CI [-.020, .042]; see Figures 14 and 15) of perfectionistic family discrepancy at T2. While
the simple slopes were not significant, the significant interaction demonstrated that the slopes at
different levels of the moderator were significantly different from each other. Figures 14 and 15
reflect an interesting interaction pattern. When perfectionistic family discrepancy at T2 was
experienced at a high level, the slope was positive. That is, Asian Americans with higher
discrepancy in meeting their families’ standards for performance might be more vulnerable to
depression when they believe they are burdens to others. Conversely, when perfectionistic family
discrepancy at T2 was experienced at a low level, the slope was negative. This implies that Asian
Americans with lower discrepancy in meeting their families’ standards for performance might
not be vulnerable to depression when they believe they are burdens to others.
Self-compassion as a moderator
The first set of results for thwarted belongingness at T1 as a predictor demonstrated that
interpersonal shame at T2 significantly mediated the association between thwarted belongingness
at T1 and depression at T3, b = .014, SE = .01, 95% CI (.005, .026). With respect to moderation,
the results did not indicate that self-compassion at T2 significantly moderated the direct path
69
from thwarted belongingness at T1 to depression at T3, b = -.007, SE = .01, 95% CI (-.036, .022;
see path c3 in Table 8). In general, there is no need to plot the simple slopes for non-significant
moderation results. Nevertheless, I plotted the simple slopes for this interaction to view the trend
of the interaction pattern. As indicated in Figures 18 and 19, the simple slopes for higher (b =
.051, SE = .01, 95% CI [.024, .078]) and lower self-compassion (b = .058, SE = .01, 95% CI
[.032, .084]) were significantly positive but the two slopes were relatively parallel. This also
illustrated that self-compassion at T2 did not significantly moderate the direct effect from
thwarted belongingness at T1 to depression at T3.
For perceived burdensomeness at T1 as a predictor, interpersonal shame at T2
significantly mediated the association between perceived burdensomeness at T1 and depression
at T3, b = .087, SE = .02, 95% CI (.055, .127). With respect to moderation, the results did not
indicate that self-compassion at T2 significantly moderated the direct path from perceived
burdensomeness at T1 to depression at T3, b = -.008, SE = .02, 95% CI (-.038, .022; see path c3
in Table 9). Similar to my analysis with thwarted belongingness at T1 (see above), I plotted the
simple slopes for this interaction to view the trend of the interaction pattern. Figures 20 and 21
demonstrate that the simple slopes for higher (b = .001, SE = .02, 95% CI [-.047, .049]) and
lower self-compassion (b = .010, SE = .02, 95% CI [-.021, .041]) were not significant and were
relatively parallel. Thus, this indicated that self-compassion at T2 did not significantly moderate
the direct effect from perceived burdensomeness at T1 to depression at T3.
11
Figure 11. The Conceptual Model for Post Hoc Analyses. T = Time.
70
11
Table 6
Unstandardized Path Coefficients with Confidence Intervals Estimating Whether Perfectionistic Family Discrepancy at T2
Moderated the Direct Effect from Thwarted Belongingness at T1 to Depression at T3 through Interpersonal Shame at T2
Note. T = Time. CI = Confidence Interval. * p < .05.** p < .01. *** p < .001. P values are based on the Sobel test.
Interpersonal Shame at T2 (Mi) Depression at T3 (Y)
Path B SE 95% CI Path B SE 95% CI
Constant a0 .880*** .05 .779, .981 b0 .044 .03 -.018, .105
Interpersonal Shame at T2 (Mi) b1 .180*** .02 .142, .218
Thwarted Belongingness at T1 (X1) a1 .085*** .03 .036, .134 c’ .052 .01 .029, .074
Perceived Burdensomeness at T1 (covariate) a2 .520*** .03 .465, .576 b2 .005 .02 -.027, .037
Depression at T1 (covariate) a3 .365*** .05 .261, .470 b3 .254*** .02 .206, .302
Perfectionistic Family Discrepancy at T2 (W1)
X1 ˟ W1
c2
c3
.025*
.017**
.01
.01
.003, .046
.004, .030
R
2 = .681 R
2 = .712
F(3,601) = 427.933, p < .001 F(6,598) = 245.780, p < .001
71
72
Figure 12. Perfectionistic Family Discrepancy Moderated the Direct Effect from Thwarted
Belongingness at T1 to Depression at T3 (scale range from 0-3). Note. PFD = Perfectionistic
Family Discrepancy. T = Time.
Figure 13. Perfectionistic Family Discrepancy Moderated the Direct Effect from Thwarted
Belongingness at T1 to Depression at T3. Note. PFD = Perfectionistic Family Discrepancy. T =
Time. The original scale range for Depression at T3 is 0-3. The figure above presents a scale
range of 0-1 to view the differences in simple slopes.
11
Table 7
Unstandardized Path Coefficients with Confidence Intervals Estimating Whether Perfectionistic Family Discrepancy at T2
Moderated the Direct Effect from Perceived Burdensomeness at T1 to Depression at T3 through Interpersonal Shame at T2
Note. T = Time. CI = Confidence Interval. * p < .05. *** p < .001. P values based on the Sobel test.
Interpersonal Shame at T2 (Mi) Depression at T3 (Y)
Path B SE 95% CI Path B SE 95% CI
Constant a0 1.54*** .07 1.41, 1.68 b0 -.109* .05 -.200, -.018
Interpersonal Shame at T2 (Mi) b1 .180*** .02 .142, .217
Perceived Burdensomeness at T1 (X2) a1 .520*** .03 .465, .576 c’ -.011 .02 -.049, .027
Thwarted Belongingness at T1 (covariate) a2 .085*** .03 .036, .134 b2 .057*** .01 .035, .080
Depression at T1 (covariate) a3 .365*** .05 .261, .470 b3 .252*** .02 .204, .301
Perfectionistic Family Discrepancy at T2 (W1)
X2 ˟ W1
c2
c3
.026*
.018*
.01
.01
.005, .048
.004, .032
R
2 = .681 R
2 = .711
F(3,601) = 427.933, p < .001 F(6,598) = 245.464, p < .001
73
74
Figure 14. Perfectionistic Family Discrepancy Moderated the Direct Effect from Perceived
Burdensomeness at T1 to Depression at T3 (scale range from 0-3). Note. PFD = Perfectionistic
Family Discrepancy. T = Time.
Figure 15. Perfectionistic Family Discrepancy Moderated the Direct Effect from Perceived
Burdensomeness at T1 to Depression at T3 through Interpersonal Shame at T2. Note. PFD =
Perfectionistic Family Discrepancy. T = Time. The original scale range for Depression at T3 is
0-3. The figure above presents a scale range of 0-1 to view the differences in simple slopes.
11
Table 8
Unstandardized Path Coefficients with Confidence Intervals Estimating Whether Self-Compassion at T2 Moderated the
Direct Effect from Thwarted Belongingness at T1 to Depression at T3 through Interpersonal Shame at T2
Note. T = Time. CI = Confidence Interval. ** p < .01. *** p < .001. P values are based on the Sobel test.
Interpersonal Shame at T2 (Mi) Depression at T3 (Y)
Path B SE 95% CI Path B SE 95% CI
Constant a0 .880*** .05 .779, .981 b0 .096** .03 .032, .160
Interpersonal Shame at T2 (Mi) b1 .166*** .02 .129, .203
Thwarted Belongingness at T1 (X1) a1 .085*** .03 .036, .134 c’ .055*** .01 .033, .077
Perceived Burdensomeness at T1 (covariate) a2 .520*** .03 .465, .576 b2 .010 .02 -.022, .041
Depression at T1 (covariate) a3 .365*** .05 .261, .470 b3 .225*** .03 .176, .274
Self-compassion at T2 (W2)
X1 ˟ W2
c2
c3
-.133***
-.007
.03
.01
-.186, -.080
-.036, .022
R
2 = .681 R
2 = .717
F(3,601) = 427.933, p < .001 F(6,598) = 252.866, p < .001
75
76
Figure 16. While the Simple Effects were Significant at the .001 Level, Self-Compassion Did
Not Moderate the Direct Effect from Thwarted Belongingness at T1 to Depression at T3 (scale
range from 0-3). Note. SC = Self-Compassion. T = Time.
Figure 17. While the Simple Effects were Significant at the .001 Level, Self-Compassion Did
Not Moderate the Direct Effect from Thwarted Belongingness at T1 to Depression at T3. Note.
SC = Self-Compassion. T = Time. The original scale range for Depression at T3 is 0-3. The
figure above presents a scale range of 0-1 to view the differences in simple slopes.
11
Table 9
Unstandardized Path Coefficients with Confidence Intervals Estimating Whether Self-Compassion at T2 Moderated the
Direct Effect from Perceived Burdensomeness at T1 to Depression at T3 through Interpersonal Shame at T2
Note. T = Time. CI = Confidence Interval. *** p < .001. P values are based on the Sobel test.
Interpersonal Shame at T2 (Mi) Depression at T3 (Y)
Path B (SE) 95% CI Path B (SE) 95% CI
Constant a0 1.54*** .07 1.41, 1.69 b0 -.047 .05 -.136, .043
Interpersonal Shame at T2 (Mi) b1 .166*** .02 .130, .203
Perceived Burdensomeness at T1 (X2) a1 .520*** .03 .465, .576 c’ .006 .02 -.032, .043
Thwarted Belongingness at T1 (covariate) a2 .085*** .03 .036, .134 b2 .056*** .01 .034, .078
Depression at T1 (covariate) a3 .365*** .05 .261, .470 b3 .225*** .02 .176, .274
Self-compassion at T2 (W2)
X2 ˟W2
c2
c3
-.133***
-.008
.03
.02
-.186, -.080
-.038, .022
R
2 = .681 R
2 = .717
F(3,601) = 427.933, p < .001 F(6,598) = 252.893, p < .001
77
78
Figure 18. Self-Compassion Did Not Moderate the Direct Effect from Perceived
Burdensomeness at T1 to Depression at T3 through Interpersonal Shame at T2. Note. SC = SelfCompassion. T = Time.
Figure 19. Self-Compassion Did Not Moderate the Direct Effect from Perceived
Burdensomeness at T1 to Depression at T3 through Interpersonal Shame at T2. Note. SC = SelfCompassion. T = Time. The original scale range for Depression at T3 is 0-3. The figure above
presents a scale range of 0-1 to view the differences in simple slopes.
79
CHAPTER 5. DISCUSSION
Asian Americans have been viewed as members of the “model minority” group who may
be held to high regard for their ability to excel academically and adjust well to college life (Kim
et al., 1999; Kwan, 2000). Yet, the Asian American mental health literature has demonstrated
that college students may report anxiety (Sangalang & Gee, 2012), depression (Wang, 2010),
shame (Wong et al., 2014), or suicidal ideation (Cheng et al., 2010) resulting from interpersonal
concerns (e.g., thwarted belongingness and perceived burdensomeness). The consequences of
interpersonal stressors on Asian American mental health have been studied, but few studies have
assessed mediating or moderating factors that protect or harm students while at college. The
findings of the current empirical study support and expand on past literature (see below) by (a)
applying Joiner’s (2005) Interpersonal-Psychological Theory of Suicide to Asian American
college students with the psychological outcome of depression; (b) examining interpersonal
shame as a potential mediator in the links between interpersonal risk factors (i.e., thwarted
belongingness and perceived burdensomeness) and depression; and (c) assessing whether
moderators (e.g., perfectionistic family discrepancy and self-compassion) alter the strength of the
proposed mediated effects in a sample of Asian American college students attending
predominantly White, Midwestern universities.
Original Hypotheses for Moderation of the Indirect (Mediation) Effects
With respect to moderated mediation, this study proposed for whom (i.e., Asian
American college students who had high or low perfectionistic family discrepancy at T2 or high
or low self-compassion at T2) the associations between interpersonal risk factors at T1 and
depression at T3 through interpersonal shame at T2 (i.e., mediation effects) would change.
80
Contrary to my hypotheses, the current results did not support that perfectionistic family
discrepancy at T2 or self-compassion at T2 moderated the associations between interpersonal
risk factors (i.e., thwarted belongingness at T1 and perceived burdensomeness at T1) and
depression at T3 through interpersonal shame at T2. That is, interpersonal risk factors related to
depression at T3 through interpersonal shame at T2, no matter the level of each moderator (i.e.,
there were no significant differences for those with high or low perfectionistic family
discrepancy at T2 or self-compassion at T2). These results were found after statistically
controlling for initial levels of depression.
Post Hoc Analyses: Mediation Effects and Moderated Mediation of the Direct Effects
Mediation effects
While the original moderated mediation hypotheses were not supported, post hoc
analyses demonstrated significant mediation effects (i.e., interpersonal shame at T2 mediated the
links of interpersonal risk factors at T1 and depression at T3). These results may imply that the
data support significant mediation effects and do not support the hypothesis that the mediation
effects depend on my moderators of interest (i.e., perfectionistic family discrepancy at T2 and
self-compassion at T2). In other words, when Asian American college students experience higher
thwarted belongingness and perceived burdensomeness, these students will have a higher sense
of interpersonal shame, which in turn can be related to higher levels of depression in the future.
This makes sense because relationships are highly valued in Asian American culture.
Disconnection from others or the perception that they burden others likely bring about
interpersonal shame among Asian American college students (Wong et al., 2014). Therefore,
interpersonal shame is likely to make Asian Americans vulnerable to future depression. These
results are consistent with a previous study that examined interpersonal shame as a mediator
81
between interpersonal risk factors (i.e., thwarted belongingness and perceived burdensomeness)
and suicidal ideation (Wong et al., 2014). It is important to highlight that these mediation effects
occurred after controlling for initial level of depression as a possible confounding variable.
Perfectionistic family discrepancy at T2 as a moderator
With respect to moderated direct effects, I found that perfectionistic family discrepancy
at T2 significantly moderated the direct effect of thwarted belongingness at T1 to depression at
T3. Closer examination of the conditional direct effects with thwarted belongingness at T1 to
depression at T3 revealed that the simple slopes were significant at low and high levels of
perfectionistic family discrepancy at T2 (see Figures 12 and 13). Furthermore, the direct effect of
thwarted belongingness at T1 on depression at T3 was stronger for those with higher than lower
perfectionistic family discrepancy at T2. Those who face high discrepancies between how they
believe their families wish them to perform and their actual performance are more at risk for
depression while experiencing higher disconnection from others. My results confirm that
students’ lack of close connections with others and high perfectionistic family discrepancy may
strengthen students’ susceptibility to depression compared to those who experience low
perfectionistic family discrepancy. These results are consistent with the literature regarding the
moderating role of perfectionism in relations between interpersonal stressors (e.g., social
commitments) and depression (Hewitt & Flett, 1993) and between attachment anxiety and
depression (Wei et al., 2004).
Moreover, perfectionistic family discrepancy at T2 also significantly moderated the direct
effect of perceived burdensomeness at T1 to depression at T3. That is, the slopes at different
levels of perfectionistic family discrepancy at T2 were significantly different from each other.
However, results for the conditional direct effects with perceived burdensomeness at T1 to
82
depression at T3 revealed that the simple slopes were not significant at low and high levels of
perfectionistic family discrepancy at T2 (i.e., one standard deviation below and above the mean;
see Figures 14 and 15). Specifically, Asian Americans with high discrepancy in meeting their
families’ standards for performance might be more vulnerable to depression when they perceive
that they burden others. Perceiving discrepancy between their performance and families’
standards can be an added stressor that strengthens the effect of perceived burdensomeness on
future depression (Taylor et al., 2004). Nonetheless, those who perceive that they burden others
may not be susceptible to depression when their perfectionistic family discrepancy is low.
Self-compassion at T2 as a moderator
Finally, results of post hoc analyses revealed that self-compassion at T2 did not
significantly moderate the direct effect from thwarted belongingness at T1 to depression at T3.
The simple slopes for higher and lower self-compassion were relatively parallel (see Figures 16
and 17), which indicated that the positive relationship between thwarted belongingness at T1 and
depression at T3 seemed to be relatively constant across high and low levels of self-compassion
among Asian American students. It is noteworthy, however, that the simple slopes for higher and
lower self-compassion were significantly positive across different levels of self-compassion. One
reason that self-compassion was not a significant moderator may be related to within-group
differences. Neff et al. (2008) found that Thai and Taiwanese students differ in exposure to and
attitudes about self-compassion. Future researchers can further explore the moderation role selfcompassion plays in the interpersonal risk factors and depression links for specific Asian
American groups.
In addition, self-compassion at T2 also did not significantly moderate the direct effect
from perceived burdensomeness at T1 to depression at T3. In comparison to thwarted
83
belongingness at T1 (see above), the simple slopes for higher and lower levels of selfcompassion were not significant, and they were relatively parallel (i.e., the perceived
burdensomeness-depression link was constant across all levels of self-compassion; see Figures
18 and 19). These results seemingly contradict Wong and Mak’s (2012) and Kyeong’s (2013)
studies demonstrating that self-compassion serves as a buffer in the relationships between (a)
cognitive-personality styles (i.e., autonomy and self-criticism) and (b) academic burden with
depression. However, it is important to note that (a) cognitive-personality styles and academic
burden are conceptually different from Joiner’s (2005) interpersonal risk factors and (b) both
studies sampled Asian students studying in Hong Kong and Korea, respectively. Selfcompassion is a relatively new concept that has been studied with Asian college students but, to
my knowledge, has not been studied with Asian American college student samples (e.g., Neff et
al., 2008). Believing that one burdens others in a relational context is different from cognitivepersonality styles or academic burden. In Asian American culture, individuals are expected to
put others’ needs before their own needs. When Asian American college students believe that
they are others’ burden, it might be difficult to show themselves compassion and decrease their
depression. Future studies should confirm or disconfirm this hypothesis.
Contribution to the Asian American Literature
Current research in the Asian American mental health literature has begun moving
beyond assessing the nature of interpersonal stressors to explore possible mediators and
moderators that protect or put Asian American students at risk for negative psychological
outcomes (e.g., depression). Thus, the present study made at least three important contributions
to the Asian American mental health literature.
84
First, the current study identified specific interpersonal stressors (i.e., thwarted
belongingness and perceived burdensomeness) based on Joiner’s (2005) InterpersonalPsychological Theory of Suicide that can affect Asian American students’ risk for future
depression. It is only recently that multicultural researchers have applied Joiner’s theory to Asian
and Asian American students’ experiences (e.g., Wong et al., 2011a; Wong et al., 2011b; Zhang
et al., 2013). These studies have primary examined risk factors to specifically predict the
psychological outcome of suicidal ideation. Nevertheless, this study’s findings confirm T.
Joiner’s personal communication (September 23, 2014) that his interpersonal states of thwarted
belongingness and perceived burdensomeness may help researchers and clinicians to better
understand Asian American students’ risk factors for future depression. By applying Joiner’s
framework to this population of interest, the current study has introduced specific contextual
factors (i.e., thwarted belongingness and perceived burdensomeness) that may help future
researchers examine social situations in which Asian American students lack a sense of
belonging or perceive themselves as burdens to their peers at predominantly White institutions.
Next, my results advance the literature by highlighting the relevance of combining
Joiner’s (2005) interpersonal stressors with Wong et al.’s (2014) construct of interpersonal
shame. The present study found that, above and beyond initial level of depression, interpersonal
shame serves as a mediator between both interpersonal risk factors (i.e., thwarted belongingness
and perceived burdensomeness) and depression among Asian American college students. The
fact that interpersonal shame serves as a mediator between both of Joiner’s constructs and
depression adds complexity to the Asian American mental health literature. This study’s focus on
both Joiner’s (2005) framework and Wong et al.’s (2014) culturally-applicable construct of
interpersonal shame expanded the mediating role of interpersonal shame beyond suicidal
85
ideation (e.g., Wong et al., 2014) to depression in a growing population (i.e., Asian Americans).
Wong et al. (2014) argued that a relevant next step to take in the study of interpersonal shame is
to conduct longitudinal studies. The present study answers the authors’ call for complex research
methods (i.e., longitudinal studies) by providing short-term longitudinal data to support the
notion that interpersonal shame is an apparent causal mechanism in the interpersonal risk factorsdepression links.
Finally, the current study adds further depth to the literature on Asian Americans’
heightened risk for depression by demonstrating that perfectionistic family discrepancy
moderated the direct effects of (a) thwarted belongingness and (b) perceived burdensomeness on
depression. Specifically, the positive relationship between lacking a sense of belonging and
increased depression was stronger for Asian American students who strongly believe they do not
live up to their families’ standards. Thus, these results show for whom (i.e., Asian American
students with high perfectionistic family discrepancy [i.e., a culturally salient variable for
students navigating family expectations for excellence]) depression occurs in the face of
thwarted belongingness and perceived burdensomeness.
Limitations
The contributions of this study to the Asian American mental health literature should be
viewed in light of its limitations. First, the majority of participants in the sample identified as
East Asian American and second generation Asian American. Furthermore, participants attended
predominantly White Midwestern universities, which limits the extent of the study’s
generalizability to other Asian American subgroups. Future studies may benefit from sampling
specific ethnic groups (e.g., Chinese-Americans, South Asian Americans) in other areas of the
86
U.S., as well as those who identify as Asian American of various generational statuses (e.g., 1st
and 3rd generation) to confirm or disconfirm the findings of the present study.
Second, the study may be subject to a self-selection bias. Participants who were (a)
interested in the topic of interpersonal risk factors and psychological outcomes and (b) who were
comfortable disclosing their experiences with thwarted belongingness, perceived
burdensomeness, interpersonal shame, and depression via an online survey may have been more
likely to participate. Future researchers may advertise their studies in such a way that increases
interest among a larger population of Asian American college students (e.g., providing psychoeducation on the study’s topic before sampling and providing multiple means of participation
[e.g., paper format, online format, in-person interview, etc.]). To ensure credible and accurate
responses from participants at all three time points, researchers may also consider administering
a measure of social desirability.
Finally, the use of an online, self-report questionnaire to gather short-term, longitudinal
data poses as a limitation. This means of data collection may assume that participants will
accurately report their amount of thwarted belongingness, perceived burdensomeness,
perfectionistic family discrepancy, and self-compassion. Moreover, the study’s results rely on the
truthfulness and the ability of participants to reflect on their level of interpersonal shame and
current psychological state (i.e., depression) at three time points. Future studies may benefit from
obtaining (a) peers and/or parents’ reports of the degree to which participants are subjected to
interpersonal risk factors and interpersonal shame; and (b) live observation data collected by
researchers or third-party collaborators on each of the study’s variables. Data obtained in either
manner may be analyzed in combination with or separate from self-report data.
87
Future Research Directions
Despite the aforementioned limitations, the current study presents at least three directions
for future studies. First, given that self-compassion was not a significant moderator in this study,
future researchers could explore the possible moderating role of self-compassion with positive
psychological outcomes. For instance, recent literature supporting the moderating role of selfcompassion has operationally defined positive psychological outcomes as life satisfaction (Allen,
Goldwasser, & Leary, 2012), subjective well-being (Allen et al., 2012), and self-esteem
(Pisitsungkagarn, Taephant, & Attasaranya, 2014). It is imperative that future researchers expand
their study of self-compassion among Asian Americans.
Second, while this study found that perfectionistic family discrepancy moderated the
direct effects of thwarted belongingness and perceived burdensomeness on depression, the
results found were gathered two academic semesters apart from each other. Future studies may
benefit from collecting longitudinal data across a longer time period or at multiple time points to
draw causal inferences about the relationship among interpersonal risk factors, interpersonal
shame, perfectionistic family discrepancy, and depression. For instance, data may be collected
when participants are first-year students and when participants are in their last year of college to
capture the possible long-term increase or decrease in interpersonal shame and perfectionistic
family discrepancy that may influence their depression. Longitudinal designs can also be
combined with experimental designs or intervention studies that test for causal effects (i.e.,
decreasing participants’ level of interpersonal shame and perfectionistic family discrepancy).
Lastly, my primary variables of interest were interpersonal shame, perfectionistic family
discrepancy, and self-compassion and their possible influences on the relationships between
interpersonal risk factors (i.e., thwarted belongingness and perceived burdensomeness) and
88
depression in Asian American college students. However, Perera and Chang’s (2015) recent
study comparing Asian and European Americans’ experiences with socially prescribed
perfectionism (i.e., perceived expectations to meet high standards prescribed by others)
demonstrated that Asian American students may also perceive that their peers (above and beyond
their parents) set unrealistic expectations for their performance. Thus, it is important that future
researchers explore the possible influence of this construct as a moderating variable for the
interpersonal risk factors-depression links. Specifically, the relationship between interpersonal
risk factors and depression may be stronger for those with high socially prescribed perfectionism
from their peers. Furthermore, since self-compassion was not a significant moderator in this
study for a negative psychological outcome (i.e., depression), a protective factor relevant to
Asian Americans’ enhanced well-being such as family support (e.g., Wyatt, Ung, Park, Kwon, &
Trinh-Shevrin, 2015) may moderate the above associations. That is, the relationships between
interpersonal risk factors and depression may be weaker for those with high family support.
Implications for Counseling
Given that the Asian American population is at risk for exhibiting higher depression
(National Alliance on Mental Illness, 2003), the results obtained suggest at least three
implications applicable to mental health professionals working with Asian American college
students studying at predominantly White, Midwestern universities. First, counselors can
increase their awareness that Asian American college students may experience a lack of
belonging and perceive that they burden others, which can contribute to a sense of interpersonal
shame and contribute to depressive symptoms. Furthermore, these concerns may worsen in
severity due to family of origin concerns (i.e., high perfectionistic family discrepancy).
89
Counselors should obtain more information to understand the concerns related to students’
disconnection and burdensomeness to others in their social networks.
Next, once counselors have learned about their clients’ experiences with interpersonal
stressors, counselors can find ways to decrease their clients’ interpersonal shame to decrease
their risk for future depression. Clinical interventions based in an emotion-focused perspective
(e.g., Dearing & Tangney, 2011) may be specifically tailored for Asian American students who
experience concerns about negative evaluations of their “selves” and perceiving that they bring
shame to their families. For instance, students may greatly benefit from counselors
psychoeducating them about the construct of interpersonal shame and building rapport in their
therapeutic alliance. Once safety, empathy, and cultural understanding have been established,
they may increase their awareness of shame they experience from interpersonal struggles (e.g.,
feeling alone or disconnected; Wong et al., 2014). Increasing awareness may mean providing
Asian American clients corrective emotional experiences in therapy to expose them to
unconditional positive regard from another person. Guiding students to work through their
shame, increase their shame resilience, and seek out support from their social support networks
may ultimately decrease their depression (Brown, 2007).
Finally, culturally-sensitive interventions to decrease high levels of perfectionistic family
discrepancy may be used in conjunction with psycho-educational workshops or orientations that
includes both Asian American college students and their parents. Counselors may use the
knowledge of this study to inform and psycho-educate parents and their children on possible
sources of their children’s depression (i.e., interpersonal risk factors, interpersonal shame, and
perfectionistic family discrepancy). For example, parents and children may realize that
frequently perceiving that one cannot meet families’ standards for performance may heighten the
90
depression they experience due to lacking a sense of belonging and believing they are burdens to
others. Counselors and educators can help to facilitate discussion on the effects interpersonal risk
factors and expectations for perfectionism may have on parents and youth while the latter are at
college. Granting the opportunity for students to decrease their sense of perfectionistic family
discrepancy may be critical to keep current depression low and prevent future depression.
In conclusion, the current empirical study tested a moderated mediation model, in which
perfectionistic family discrepancy and self-compassion were hypothesized to moderate the
mediation effect of interpersonal shame on the interpersonal risk factors (i.e., thwarted
belongingness and perceived burdensomeness) and future depression relationships while
controlling for initial level of depression. Although perfectionistic family discrepancy and selfcompassion did not moderate the hypothesized indirect (mediation) effects, post hoc analyses
supported my mediation hypothesis and found that perfectionistic family discrepancy moderated
the direct effects of thwarted belongingness to future depression while controlling for initial level
of depression. The direct effect of thwarted belongingness on future depression was stronger for
those with higher than lower perfectionistic family discrepancy. Perfectionistic family
discrepancy also moderated the direct effect of perceived burdensomeness on future depression
while controlling for initial level of depression. This study contributes to the growing yet limited
literature on interpersonal stressors and its connections with shame, depression, and family
expectations for perfectionism specifically among Asian American college students. Implications
of this study suggest that students who lack a sense of belonging and perceive that they burden
others may keep depression low by learning how to decrease their interpersonal shame and work
through family expectations for perfectionism.
91
REFERENCES
Adler, A. (1956). The individual psychology of Alfred Adler. New York, NY: Basic Books, Inc.
Allen, A.B., Goldwasser, E.R., Leary, M.R. (2012). Self-compassion and well-being among
older adults. Self and Identity, 11, 428-453. doi: 10.1080/15298868.2011.595082
Baumeister, R.F., & Leary, M.R. (1995). The need to belong: Desire for interpersonal
attachments as a fundamental human motivation. Psychological Bulletin, 117, 497-529.
doi: 10.1037/0033-2909.117.3.497
Bedford, O.A. (2004). The individual experience of guilt and shame in Chinese culture. Culture
& Psychology, 10, 29-52. doi: 10.1177/1354067X04040929
Birkett, A. (2014). Self-compassion and empathy across cultures: Comparison of young adults in
China and the United States. International Journal of Research Studies in Psychology, 3,
25-34. doi: 10.5861/ijrsp.2013.551
Blum, A. (2008). Shame and guilt, misconceptions and controversies: A critical review of the
literature. Traumatology, 14, 91-102. doi: 10.1177/1534765608321070
Brown, B. (2007). I thought it was just me: Women reclaiming power and courage in a culture of
shame. New York, NY: Gotham Books.
Castro, J.R., & Rice, K.G. (2003). Perfectionism and ethnicity: Implications for depressive
symptoms and self-reported academic achievement. Cultural Diversity and Ethnic
Minority Psychology, 9, 64-78. doi: 10.1037/1099-9809.9.1.64
Chang, E.C. (2013). Perfectionism and loneliness as predictors of depressive and anxious
symptoms in Asian and European Americans: Do self-construal schemas also matter?
Cognitive Therapy and Research, 37, 1179-1188. doi: 10.1007/s10608-013-9549-9
Cheng, J.K.Y., Fancher, T.L., Ratanasen, M., Conner, K.R., Duberstein, P.R, Sue, S.,
…Takeuchi, D. (2010). Lifetime suicidal ideation and suicidal attempts in Asian
Americans. Asian American Journal of Psychology, 1, 18-30. doi:10.1037/a0018799
92
Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 155-159. doi: 10.1037/0033-
2909.112.1.155
Comrey, A.L., & Lee, H.B. (1992). A first course in factor analysis (2
nd ed.). Hillsdale, NJ:
Lawrence Erlbaum Associates, Inc.
Craddock, A.E., Church, W., & Sands, A. (2009). Family of origin characteristics as predictors
of perfectionism. Australian Journal of Psychology, 61, 136-144.
doi:10.1080/00049530802239326
Dearing, R.L., & Tangney, J.P. (2011). Shame in the therapy hour. Washington, DC: American
Psychological Association.
DiBartolo, P.M., & Rendón, M.J. (2012). A critical examination of the construct of
perfectionism and its relationship to mental health in Asian and African Americans using
a cross-cultural framework. Clinical Psychology Review, 32, 139-152. doi:
10.1016/j.cpr.2011.09.007
Dixon, S.K., & Robinson Kurpius, S.E. (2008). Depression and college stress among university
undergraduates: Do mattering and self-esteem make a difference? Journal of College
Student Development, 49, 412-424. doi: 10.1353/csd.0.0024
Erdfelder, E., Faul, F., & Buchner, A. (1996). GPOWER: A general power analysis program.
Behavior Research Methods, Instruments, & Computers, 28, 1-11. doi:
10.3758/BF03203630
Fisher-Beckfield, D., & McFall, R.M. (1982). Development of a competence inventory for
college men and evaluation of relationships between competence and depression.
Journal of Consulting and Clinical Psychology, 50, 697-705. doi: 0022-006X/82/5005-
0697
Frost, R.O., Marten, P., Lahart, C., & Rosenblate, R. (1990). The dimensions of perfectionism.
Cognitive Therapy and Research, 14, 449-468. doi: 0147-5916/90/1000-0449
Fuligni, A.J., Tseng, V., & Lam, M. (1999). Attitudes toward family obligations among
American adolescents with Asian, Latin American, and European backgrounds. Child
Development, 70, 1030-1044. doi: 10.1111/1467-8624.00075
93
Hagerty, B.M., & Patusky, K. (1995). Developing a measure of sense of belonging. Nursing
Research, 44, 9-13. doi: 10.1097/00006199-199501000-00003
Hagerty, B.M., & Williams, R.A. (1999). The effects of sense of belonging, social support,
conflict, and loneliness on depression. Nursing Research, 48, 215-219. doi:
10.1097/00006199-199907000-00004
Hagerty, B.M., Williams, R.A., Coyne, J.C., & Early, M.R. (1996). Sense of belonging and
indicators of social and psychological functioning. Archives of Psychiatric Nursing, 4,
235-244. doi: 10.1016/S0883-9417(96)80029-X
Hall, C.W., Row, K.A., Wuensch, K.L., & Godley, K.R. (2013). The role of self-compassion in
physical and psychological well-being. The Journal of Psychology, 147, 311-323. doi:
10.1080/00223980.2012.693138
Hamachek, D.E. (1978). Psychodynamics of normal and neurotic perfectionism. Psychology: A
Journal of Human Behavior, 15, 27-33. Retrieved from
http://search.proquest.com.proxy.lib.iastate.edu/docview/616292295?accountid=10906
Harter, S. (1999). The construction of the self: A developmental perspective. New York, NY:
Guilford Press.
Hayes, A.F. (2013). An introduction to mediation, moderation, and conditional process analysis:
A regression-based approach. New York, NY: Guilford Press.
Hayes, A.F. (2015). An index and test of linear moderated mediation. Multivariate Behavioral
Research, 50, 1-22. doi: 10.1080/00273171.2014.962683
Henry, J.D., & Crawford, J.R. (2005). The short-form version of the depression anxiety stress
scales (DASS-21): Construct validity and normative data in a large non-clinical sample.
British Journal of Clinical Psychology, 44, 227-239. doi: 10.1348/014466505X29657
Hewitt, P.L., & Flett, G.L. (1991). Perfectionism in the self and social contexts:
Conceptualization, assessment, and association with psychopathology. Journal of
Personality and Social Psychology, 60, 456-470. doi: 10.1037/0022-3514.60.3.456
94
Hewitt, P.L., & Flett, G.L. (1993). Dimensions of perfectionism, daily stress, and depression: A
test of the specific vulnerability hypothesis. Journal of Abnormal Psychology, 102, 58-
65. doi: 10.1037/0021-843X.102.1.58
Hill, R.M., & Pettit, J.W. (2014). Perceived burdensomeness and suicide‐related behaviors in
clinical samples: Current evidence and future directions. Journal of Clinical Psychology,
70, 631-643. doi: 10.1002/jclp.22071
Joiner, T. (2005). Why people die by suicide. Cambridge, MA: Harvard University Press.
Joiner, T.E., Van Orden, K.A., Witte, T., Selby, E.A., Ribeiro, J.D., Lewis, R., & Rudd, M.D.
(2009). Main predictions of the interpersonal-psychological theory of suicidal behavior:
Empirical tests in two samples of young adults. Journal of Abnormal Psychology, 118,
634-646. doi: 10.1037/a0016500
Kawamura, K.Y., Frost, R.O., & Harmatz, M.G. (2002). The relationship of perceived parenting
styles to perfectionism. Personality and Individual Differences, 32, 317-327. doi:
10.1016/S0191-8869(01)00026-5
Kim, B.S.K., Atkinson, D.R., & Yang, P.H. (1999). The Asian values scale: Development, factor
analysis, validation, and reliability. Journal of Counseling Psychology, 46, 342-352. doi:
10.1037/0022-0167.46.3.342
Kim, S., Thibodeau, R., & Jorgensen, R.S. (2011). Shame, guilt, and depressive symptoms: A
meta-analytic review. Psychological Bulletin, 137, 68-96. doi: 10.1037/a0021466
Kwan, K-L.K. (2000). Counseling Chinese people: Perspectives of filial piety. Asian Journal of
Counselling, 7, 23-41. Retrieved from
http://search.proquest.com.proxy.lib.iastate.edu/docview/619739505?accountid=10906
Kyeong, L.W. (2013). Self-compassion as a moderator of the relationship between academic
burn-out and psychological health in Korean cyber university students. Personality and
Individual Differences, 54, 899-902. doi: 10.1016/j.paid.2013.01.001
Liem, R. (1997). Shame and guilt among first- and second-generation Asian Americans and
European Americans. Journal of Cross-Cultural Psychology, 28, 365-392. doi:
10.1177/0022022197284001
95
Louie, S. (2014). Asian shame and honor: A cultural conundrum and case study. Psychology
Today. Retrieved from http://www.psychologytoday.com/blog/minorityreport/201406/asian-shame-and-honor
Lovibond, P.F., & Lovibond, S.H. (1995). The structure of negative emotional states:
Comparison of the depression anxiety stress scales (DASS) with the beck depression and
anxiety inventories. Behavior Research and Therapy, 33, 335-343. doi: 10.1016/0005-
7967(94)00075-U
Lovibond, S.H., & Lovibond, P. F. (1995). Manual for the Depression Anxiety and Stress scales
(2nd ed.). Sydney, AU: Psychological Foundation.
Lutwak, N., Razzino, B.E., & Ferrari, J.R. (1998). Self-perceptions and moral affect: An
exploratory analysis of subcultural diversity in guilt and shame emotions. Journal of
Social Behavior & Personality, 13, 333-348. Retrieved from
http://search.proquest.com.proxy.lib.iastate.edu/docview/619354139?accountid=10906
MacKinnon, D.P., Lockwood, C.M., & Williams, J. (2004). Confidence limits for the indirect
effect: Distribution of the product and resampling methods. Multivariate Behavioral
Research, 39, 99-128. doi:10.1207/s15327906mbr3901_4
Mallinckrodt, B., Abraham, T.W., Wei, M., & Russell, D. W. (2006). Advances in testing
statistical significance of mediation effects. Journal of Counseling Psychology, 53, 372-
378. doi:10.1037/0022-0167.53.3.372
Markus, H.R., & Kitayama, S. (1991). Culture and the self: Implications for cognition, emotion,
and motivation. Psychological Review, 98, 224-253. doi: 10.1037/0033-295X.98.2.224
Maslow, A.H. (1954). Motivation and personality. Oxford, England: Harpers.
National Alliance on Mental Illness (2003). Asian American and Pacific Islander Community
and Mental Health [Fact sheet]. Retrieved from
http://www.nami.org/Template.cfm?Section=Depression&Template=/ContentManageme
nt/ContentDisplay.cfm&ContentID=89034
96
Neely, M.E., Schallert, D.L., Mohammed, S.S., Roberts, R.M., & Chen, Y-J. (2009). Selfkindness when facing stress: The role of self-compassion, goal regulation, and support in
college students’ well-being. Motivation and Emotion, 33, 88-97. doi: 10.1007/s11031-
008-9119-8
Neff, K.D. (2003). The development and validation of a scale to measure self-compassion. Self
and Identity, 2, 223-250. doi: 10.1080/15298860309027
Neff, K.D. (2004). Self-compassion and psychological well-being. Constructivism in the Human
Sciences, 9, 27-37. Retrieved from
http://search.proquest.com.proxy.lib.iastate.edu/docview/620731491?accountid=10906
Neff, K.D. (2008). Self-compassion: Moving beyond the pitfalls of a separate self-concept. In
H.A. Wayment & J. Bauer (Eds.), Transcending self-interest: Psychological explorations
of the quiet ego (pp. 95-105). Washington, DC: Help write my thesis – APA Books.
Neff, K.D., Kirkpatrick, K.L., & Rude, S.S. (2007). Self-compassion and adaptive psychological
functioning. Journal of Research in Personality, 41, 139–154. doi:
10.1016/j.jrp.2006.03.004
Neff, K.D., Pisitsungkagarn, K., & Hsieh, Y.P. (2008). Self-compassion and self-construal in the
United States, Thailand, and Taiwan. Journal of Cross-Cultural Psychology, 39, 267–
285. doi: 10.1177/0022022108314544
Neff, K.D., Rude, S.S., & Kirkpatrick, K.L. (2007). An examination of self-compassion in
relation to positive psychological functioning and personality traits. Journal of Research
in Personality, 41, 908-916. doi: 10.1016/j.jrp.2006.08.002
Nolen-Hoeksema, S. (1991). Responses to depression and their effects on the duration of
depressive episodes. Journal of Abnormal Psychology, 100, 569-582. doi: 10.1037/0021-
843X.100.4.569
Norton, P.J. (2007). Depression Anxiety and Stress Scales (DASS-21): Psychometric analysis
across four racial groups. Anxiety, Stress, & Coping, 20, 253-265. doi:
10.1080/10615800701309279
97
Ortega, N.E., Wang, K.T., Slaney, R.B., Hayes, J.A., & Morales, A. (2014). Personal and
familial aspects of perfectionism in Latino/a students. The Counseling Psychologist, 42,
406-427. doi: 10.1177/0011000012473166
Perera, M.J., & Chang, E.C. (2015). Ethnic variations between Asian and European Americans
in interpersonal sources of socially prescribed perfectionism: It’s not just about parents!
Asian American Journal of Psychology, 6, 31-37. doi: 10.1037/a0036175
Pisitsungkagarn, K., Taephant, N., & Attasaranya, P. (2014). Body image satisfaction and selfesteem in Thai female adolescents: The moderating role of selfcompassion. International Journal of Adolescent Medicine and Health, 26, 333-338. doi:
10.1515/ijamh-2013-0307
Raes, F. (2011). The effect of self-compassion on the development of depression symptoms in a
non-clinical sample. Mindfulness, 2, 33-36. doi: 10.1007/s12671-011-0040-y
Ribeiro, J.D., & Joiner, T.E. (2009). The interpersonal-psychological theory of suicidal behavior:
Current status and future directions. Journal of Clinical Psychology, 65, 1291-1299. doi:
10.1002/jclp.20621
Rice, K.G., Ashby, J.S., & Preusser, K.J. (1996). Perfectionism, relationships with parents, and
self-esteem. Individual Psychology, 52, 246-260. Retrieved from
http://search.proquest.com.proxy.lib.iastate.edu/docview/618887490?accountid=10906
Rosenberg, M., & McCullough, B.C. (1981). Mattering: Inferred significance and mental health
among adolescents. Research in Community & Mental Health, 2, 163-182. Retrieved
from
http://search.proquest.com.proxy.lib.iastate.edu/docview/616708816?accountid=10906
Sangalang, C.C., & Gee, G.C. (2012). Depression and anxiety among Asian Americans: The
effects of social support and strain. Social Work, 57, 49-60. doi: 10.1093/sw/swr005
Schafer, J.L., & Graham, J.W. (2002). Missing data: Our view of the state of the art.
Psychological Methods, 7, 147–177. doi: 10.1037/1082-989X.7.2.147
98
Schlomer, G.L., Bauman, S., & Card, N.A. (2010). Best practices for missing data management
in counseling psychology. Journal of Counseling Psychology, 57, 1-10. doi:
10.1037/a0018082
Shafran, R., & Mansell, W. (2001). Perfectionism and psychopathology: A review of research
and treatment. Clinical Psychology Review, 21, 876-906. doi: 10.1016/S0272-
7358(00)00072-6
Shrout, P.E., & Bolger, N. (2002). Mediation in experimental and nonexperimental studies: New
procedures and recommendations. Psychological Methods, 7, 422-445. doi:
10.1037/1082-989X.7.4.422
Slaney, R.B., Rice, K.G., Mobley, M., Trippi, J., & Ashby, J.S. (2001). The revised Almost
Perfect Scale. Measurement and Evaluation in Counseling and Development, 34, 130-
145. Retrieved from
http://search.proquest.com.proxy.lib.iastate.edu/docview/619599462?accountid=10906
Tangney, J.P. (1996). Conceptual and methodological issues in the assessment of shame and
guilt. Behaviour Research and Therapy, 34, 741-754. doi: 10.1016/0005-7967(96)00034-
4
Tangney, J.P., & Dearing, R.L. (2002). Shame and guilt. New York, NY: Guildford Press.
Taylor, J., & Turner, R.J. (2001). A longitudinal study of the role and significance of mattering
to others for depressive symptoms. Journal of Health and Social Behavior, 42, 310-325.
doi: 10.2307/3090217
Taylor, S.E., Sherman, D.K., Kim, H.S., Jarcho, J., Takagi, K., & Dunagan, M.S. (2004). Culture
and social support: Who seeks it and why? Journal of Personality and Social
Psychology, 87, 354-362. doi: 10.1037/0022-3514.87.3.354
Terry, M.L., Leary, M.R., & Mehta, S. (2012). Self-compassion as a buffer against
homesickness, depression, and dissatisfaction in the transition to college. Self and
Identity, 1, 1-13. doi: 10.1080/15298868.2012.667913
99
Triandis, H.C., Bontempo, R., & Villareal, M.J. (1988). Individualism and collectivism: Crosscultural perspectives on self-ingroup relationships. Journal of Personality and Social
Psychology, 54, 323-338. doi: 10.1037/0022-3514.54.2.323
Van Dam, N.T., Sheppard, S.C., Forsyth, J.P., & Earleywine, M. (2010). Self-compassion is a
better predictor than mindfulness of symptom severity and quality of life in mixed
anxiety and depression. Journal of Anxiety Disorders, 25, 123-130. doi:
10.1016/j.janxdis.2010.08.011
Van Orden, K.A., Cukrowicz, K.C., Witte, T.K., & Joiner, T.E. (2012). Thwarted belongingness
and perceived burdensomeness: Construct validity and psychometric properties of the
Interpersonal Needs Questionnaire. Psychological Assessment, 24, 197-215. doi:
10.1037/a0025358
Van Orden, K.A., Lynam, M.E., Hollar, D., & Joiner, T.E. (2006). Perceived burdensomeness as
an indicator of suicidal symptoms. Cognitive Therapy and Research, 30, 457-467. doi:
10.1007/s10608-006-9057-2
Van Orden, K.A., Witte, T.K., Cukrowicz, K.C., Braithwaite, S., Selby, E.A., & Joiner, T.E.
(2010). The interpersonal theory of suicide. Psychological Review, 117, 575-600.
doi:10.1037/a0018697
Van Orden, K.A., Witte, T.K., Gordon, K.H., Bender, T.W., & Joiner, T.E. (2008). Suicidal
desire and the capability for suicide: Test of the interpersonal-psychological theory of
suicidal behavior among adults. Journal of Consulting and Clinical Psychology, 76, 72-
83. doi: 10.1037/0022-006X.76.1.72
Wang, K.T. (2010). The Family Almost Perfect Scale: Development, psychometric properties,
and comparing Asian and European Americans. Asian American Journal of Psychology,
1, 186-199. doi: 10.1037/a0020732
Wang, K.T. (2012). Personal and family perfectionism of Taiwanese college students:
Relationships with depression, self-esteem, achievement motivation, and academic
grades. International Journal of Psychology, 47, 305-314. doi:
10.1080/00207594.2011.626050
100
Wang, K.T., Wong, Y.J., Fu, C-C. (2013). Moderation effects of perfectionism and
discrimination on interpersonal factors and suicidal ideation. Journal of Counseling
Psychology, 60, 367-378. doi: 10.1037/a0032551
Wei, M., Mallinckrodt, B., Russell, D.W., & Abraham, W.T. (2004). Maladaptive perfectionism
as a mediator and moderator between attachment and depressive mood. Journal of
Counseling Psychology, 51, 201-212. doi: 10.1037/0022-0167.51.2.201
Wong, C.C.Y., & Mak, W.W.S. (2012). Differentiating the role of three self-compassion
components in buffering cognitive-personality vulnerability to depression among
Chinese in Hong Kong. Journal of Counseling Psychology, 1, 162-169. doi:
10.1037/a0030451
Wong, Y.J., Brownson, C., & Schwing, A. E. (2011a). Risk and protective factors associated
with Asian American students’ suicidal ideation: A multicampus, national study. Journal
of College Student Development, 52, 396-408. doi: 10.1353/csd.2011.0057
Wong, Y.J., Kim, B.S.K., Nguyen, C.P., Cheng, J.K.Y., & Saw, A. (2014). The Interpersonal
Shame Inventory for Asian Americans: Scale development and psychometric properties.
Journal of Counseling Psychology, 61, 119-132. doi: 10.1037/a0034681
Wong, Y.J., Koo, K., Tran, K.K., Chiu, Y-C., & Mok, Y. (2011b). Asian American college
students’ suicide ideation: A mixed methods study. Journal of Counseling Psychology,
58, 197-209. doi: 10.1037/a0023040
Wong, Y.J., & Poon, M-L. (2010). Counseling Asian American men who demonstrate suicidal
behavior. In W.M. Liu, D. Iwamoto, & M. Chae (Eds.), Culturally responsive counseling
with Asian American men (pp. 279–298). New York, NY: Routledge.
Woods, H., & Proeve, M. (2014). Relationships of mindfulness, self-compassion, and meditation
experience with shame-proneness. Journal of Cognitive Psychotherapy: An International
Quarterly, 28, 20-33. doi: 10.1891/0889-8391.28.1.20
Wyatt, L.C., Ung, T., Park, R., Kwon, S.C., & Trinh-Shevrin, C. (2015). Risk factors of suicide
and depression among Asian American, Native Hawaiian, and Pacific Islander Youth: A
systematic literature review. Journal of Health Care for the Poor and Underserved, 26,
191-237. doi: 10.1353/hpu.2015.0059
101
Yoon, J., & Lau, A.S. (2008). Maladaptive perfectionism and depressive symptoms among Asian
American college students: Contributions of interdependence and parental relations.
Cultural Diversity and Ethnic Minority Psychology, 14, 92-101. doi: 10.1037/1099-
9809.14.2.92
Zhang, J., Lester, D., Zhao, S., & Zhou, C. (2013). Suicidal ideation and its correlates: Testing
the interpersonal theory of suicide in Chinese students. Archives of Suicide Research, 17,
236-241. doi: 10.1080/13811118.2013.805643
102
APPENDIX A. IRB APPROVAL
103
APPENDIX B. INVITATION EMAIL
Dear Student,
My name is Stephanie Carrera, and I am a doctoral student at Iowa State University. I am
conducting my dissertation study related to interpersonal connections and well-being. I am
emailing you now because I need your help to complete this important task. The survey will take
between 20-30 minutes to complete.
You must be at least 18 years old, self-identify as Asian American, and attend [name of
university] to participate in this study.
I will contact you again at two other points (i.e., at the end of the spring semester and middle of
the fall semester) to complete a shorter version of this survey. Each survey at each time point
will take between 20-30 minutes or less to complete. For your time and effort, each time you
participate, you will have a chance to enter a random drawing for a $25 VISA gift card.
If you are interested in this research, please click the link below.
https://iastatepsychhci.az1.qualtrics.com/SE/?SID=SV_80KqEJ2KoTueZUN
I truly appreciate your help with my dissertation study. Thank you.
Best,
Stephanie Carrera
104
APPENDIX C. UNIVERSITY RECRUITMENT EMAILS
To whom it may concern,
My name is Stephanie Carrera, and I am currently a researcher at Iowa State University. I hope
this email finds you well.
This semester, I will be conducting my dissertation study related to interpersonal connections
and well-being among currently enrolled undergraduate students who identify as Asian
American. The purpose of this study is to better understand how relations with others, family,
and personal strengths affect well-being in Asian American college students. We would like to
collect data from universities in the Midwest, including the [name of university].
We have already received IRB approval from Iowa State University (attached) to conduct this
study. Per the advice of the [name of university] IRB office, I have emailed you to ask if it is
possible to gain access to a list of first names and email addresses for currently enrolled
undergraduate students who identify as Asian American. We hope to contact these students via
email invitation about this opportunity to participate.
I appreciate your help. If should you need more information about this study, I would be happy
to give it to you. I look forward to hearing from you.
Cordially,
Stephanie Carrera
Graduate Assistant
scarrera@iastate.edu
105
Dear XX,
My name is Stephanie Carrera, and I am a doctoral student at Iowa State University. I hope this
email finds you well.
I am contacting you for help with recruitment of Asian American students (undergraduate
students) for my dissertation study. I am especially interested in better understanding how
relations with others, family, and personal strengths affect well-being among Asian American
students studying at Midwestern universities like [name of university]. This is an online study
that would take approximately 20-30 minutes to complete.
Would it be possible for you to forward an email invitation for this study to your members in
[name of organization]? I have included it below for your convenience.
I truly appreciate your help with this! Please let me know if you have any questions.
Stephanie Carrera
106
APPENDIX D. INFORMED CONSENT FORMS
Informed Consent Form (ISU Students)
Title of Study: Interpersonal Connections and Well-being
Investigators: Stephanie Carrera (Principal Investigator) and Meifen Wei (Faculty Supervisor)
This is a research study. Please take your time in deciding if you would like to participate. You
must be at least 18 years old and self-identify as Asian American to participate in this study.
INTRODUCTION
The purpose of this study is to better understand how relations with others, family, and personal
strengths affect well-being in Asian American college students. You are invited to participate in
this study because you attend Iowa State University and self-identify as Asian American.
DESCRIPTION OF PROCEDURES
Participation in this research is completely voluntary. If you agree to participate in this study,
your participation will take approximately 20-30 minutes for each session (i.e., three sessions
total). During the study you may expect to complete three surveys related to interpersonal
connections and well-being. The first survey will be administered to you at the beginning of the
spring semester (Year 1), the end of the spring semester (Year 1), and the third survey in the
middle of the fall semester (Year 1).
RISKS
While participating in this study, you may experience very mild personal discomfort when you
respond to questions related to interpersonal connections and well-being. If you experience
personal discomfort and do not wish to continue completing the survey, you may exit the survey
at any time. Should you wish to talk about your discomfort, we will provide you with contact
information for Student Counseling Services at the end of our survey.
BENEFITS
If you decide to participate in this study, there will be no direct benefit to you. It is hoped that the
results of this study will help us to understand how relations with others, family, and personal
strengths affect well-being in Asian American college students.
COSTS AND COMPENSATION
You will not have any costs from participating in this study. For your time and effort, we will
offer you a chance at each time point to enter a random drawing to win a $25 VISA gift card (the
odds of winning are 1 in 25). The drawing will be held after all the data has been collected. The
winner will be notified via email and will need to fill out the research participation receipt.
Information regarding documentation required for participant compensation may be obtained
from the Controller’s Department at (515) 294-2555 or www.controller.iastate.edu. You will not
be monetarily compensated for participating in this study.
107
PARTICIPANT RIGHTS
Your participation in this study is completely voluntary and you may refuse to participate or
leave the study at any time. If you decide not to participate in the study or leave the study early,
it will not result in any penalty or loss of benefits to which you are otherwise entitled. You may
skip any question you do not wish to answer or that makes you feel uncomfortable without
receiving any penalty. For the information to be useful to us, we encourage you to complete all
the items.
CONFIDENTIALITY
Records identifying participants will be kept confidential to the extent permitted by applicable
laws and regulations and will not be made publicly available. However, federal government
regulatory agencies and the Institutional Review Board (a committee that reviews and approves
human subject research studies) may inspect and/or copy your records for quality assurance and
data analysis. These records may contain private information.
To ensure confidentiality to the extent permitted by law, the following measures will be taken:
No personal identifying information (e.g., ID number) will be asked on the questionnaire
responses. Your questionnaire responses will be assigned an arbitrary number (i.e., 001, 002,
003, etc.) for the purpose of creating a data file. In addition, because this is a 3-time point study,
the principal investigator (PI) and supervisor will use the last four digits of your phone numbers,
the first three letters of your street address (i.e., permanent address), and the day of your birth
(e.g., 01, 17) that you directly enter into the survey to pair your responses at Times 1-3. While
your name and email address will be accessed after Time 1 of data collection to contact you
again at Time 2 and Time 3, your data will remain confidential. Only the PI and her supervisor
will have access to the data. The data will be encrypted and stored on the PI’s computer with a
specific password in order to access the computer. Once the data from Times 1-3 have been
merged, all names and email addresses will be removed. If the results are published, your
identity will remain confidential.
QUESTIONS OR PROBLEMS
You are encouraged to ask questions at any time during this study. For further information about
the study contact Stephanie Carrera at (630) 536-9604 and scarrera@iastate.edu, or Dr. Meifen
Wei at (515) 294-7534 and wei@iastate.edu. If you have any questions about the rights of
research subjects or research-related injury, please contact the IRB Administrator at (515) 294-
4566, irb@iastate.edu, or the Director, at (515) 294-3115, Office for Responsible Research, 1138
Pearson Hall, Ames, IA 50011.
If you would like to retain a copy of the consent form for your records, please print this page.
By clicking the “>>” button below, you indicate that you have read the informed consent form
and agree to participate in this study.
108
Informed Consent Form (All other students)
Title of Study: Interpersonal Connections and Well-being
Investigators: Stephanie Carrera (Principal Investigator) and Meifen Wei (Faculty Supervisor)
This is a research study. Please take your time in deciding if you would like to participate. You
must be at least 18 years old and self-identify as Asian American to participate in this study.
INTRODUCTION
The purpose of this study is to better understand how relations with others, family, and personal
strengths affect well-being in Asian American college students. You are invited to participate in
this study because you are an undergraduate college student who self-identifies as Asian
American.
DESCRIPTION OF PROCEDURES
Participation in this research is completely voluntary. If you agree to participate in this study,
your participation will take approximately 20-30 minutes for each session (i.e., three sessions
total). During the study you may expect to complete three surveys related to interpersonal
connections and well-being. The first survey will be administered to you at the beginning of the
spring semester (Year 1), the second survey at the end of the spring semester (Year 1), and the
third survey in the middle of the fall semester (Year 1).
RISKS
While participating in this study, you may experience mild personal discomfort when you
respond to questions related to interpersonal connections and well-being. If you experience
personal discomfort and do not wish to continue completing the survey, you may exit the survey
at any time. Should you wish to talk about or consult with someone about your discomfort, we
will provide you with contact information for counseling services at the end of our survey.
BENEFITS
If you decide to participate in this study, there will be no direct benefit to you. It is hoped that the
results of this study will help us to understand how relations with others, family, and personal
strengths affect well-being in Asian American college students.
COSTS AND COMPENSATION
You will not have any costs from participating in this study. For your time and effort, we will
offer you a chance at each time point to enter a random drawing to win a $25 VISA gift card (the
odds of winning are 1 in 25). The drawing will be held after all the data has been collected. The
winner will be notified via email and will need to fill out the research participation receipt.
Information regarding documentation required for participant compensation may be obtained
from the Controller’s Department at (515) 294-2555 or www.controller.iastate.edu. You will not
be monetarily compensated for participating in this study.
109
PARTICIPANT RIGHTS
Your participation in this study is completely voluntary and you may refuse to participate or
leave the study at any time. If you decide not to participate in the study or leave the study early,
it will not result in any penalty or loss of benefits to which you are otherwise entitled. You may
skip any question you do not wish to answer or that makes you feel uncomfortable without
receiving any penalty. For the information to be useful to us, we encourage you to complete all
the items.
CONFIDENTIALITY
Records identifying participants will be kept confidential to the extent permitted by applicable
laws and regulations and will not be made publicly available. However, federal government
regulatory agencies and the Institutional Review Board (a committee that reviews and approves
human subject research studies) may inspect and/or copy your records for quality assurance and
data analysis. These records may contain private information.
To ensure confidentiality to the extent permitted by law, the following measures will be taken:
No personal identifying information (e.g., ID number) will be asked on the questionnaire
responses. Your questionnaire responses will be assigned an arbitrary number (i.e., 001, 002,
003, etc.) for the purpose of creating a data file. In addition, because this is a 3-time point study,
the principal investigator (PI) and supervisor will use the last four digits of your phone numbers,
the first three letters of your street address (i.e., permanent address), and the day of your birth
(e.g., 01, 17) that you directly enter into the survey to pair your responses at Times 1-3. While
your name and email address will be accessed after Time 1 of data collection to contact you
again at Time 2 and Time 3, your data will remain confidential. Only the PI and her supervisor
will have access to the data. The data will be encrypted and stored on the PI’s computer with a
specific password in order to access the computer. Once the data from Times 1-3 have been
merged, all names and email addresses will be removed. If the results are published, your
identity will remain confidential.
QUESTIONS OR PROBLEMS
You are encouraged to ask questions at any time during this study. For further information about
the study contact Stephanie Carrera at (630) 536-9604 and scarrera@iastate.edu, or Dr. Meifen
Wei at (515) 294-7534 and wei@iastate.edu. If you have any questions about the rights of
research subjects or research-related injury, please contact the IRB Administrator at (515) 294-
4566, irb@iastate.edu, or the Director at (515) 294-3115, Office for Responsible Research, 1138
Pearson Hall, Ames, IA 50011.
If you would like to retain a copy of the consent form for your records, please print this page.

By clicking the “>>” button below, you indicate that you have read the informed consent form
and agree to participate in this study.
110
APPENDIX E. DEBRIEFING FORMS
Debriefing Form (All ISU students)
Thank you very much for participating in this study. This project seeks to understand the
associations among interpersonal relations, family influences, personal strengths, and well-being
in Asian American college students. The information gained from this study will provide with
information regarding these processes.
If you have any questions or would like to learn about the study, please contact Stephanie
Carrera, W183 Lagomarcino Hall (office), via phone at (630) 536-9604, or via email at
scarrera@iastate.edu or Dr. Meifen Wei, W214 Lagomarcino Hall (office), via phone at (515)
294-7534, or via email at wei@iastate.edu. Furthermore, if you experienced any discomfort
while completing this survey and would like to talk about your reactions, you may contact
Mental Health America (800-969-6642), the National Alliance on Mental Illness (800-950-6264)
for support, information, or referrals to a local mental health agency, or Student Counseling
Services, Student Services Building, 3rd floor, at (515) 294-5056. Free counseling is available to
ISU students. If you have any questions about the rights of research subjects or research related
injury, please contact the IRB Administrator at (515) 294-4566 and irb@iastate.edu, or the
Director at (515) 294-3115, Office for Responsible Research, 1138 Pearson Hall, Ames, IA,
50011.
Please click on “>>” below to finish the survey and enter a drawing for a $25 VISA gift card.
Debriefing Form (All other students)
Thank you very much for participating in this study. This project seeks to understand the
associations among interpersonal relations, family influences, personal strengths, and well-being
in Asian American college students. The information gained from this study will provide with
information regarding these processes.
If you have any questions or would like to learn about the study, please contact Stephanie
Carrera, W183 Lagomarcino Hall (office), via phone at (630) 536-9604, or via email at
scarrera@iastate.edu or Dr. Meifen Wei, W214 Lagomarcino Hall (office), via phone at (515)
294-7534, or via email at wei@iastate.edu. Furthermore, if you experienced any discomfort
while completing this survey and would like to talk about your reactions, you may contact
Mental Health America (800-969-6642), the National Alliance on Mental Illness (800-950-6264)
for support, information, or referrals to a local mental health agency, or counseling services on
your campus. If you have any questions about the rights of research subjects or research related
injury, please contact the IRB Administrator at (515) 294-4566 and irb@iastate.edu, or the
Director at (515) 294-3115, Office for Responsible Research, 1138 Pearson Hall, Ames, IA,
50011.
Please click on “>>” below to finish the survey and enter a drawing for a $25 VISA gift card.
111
ACKNOWLEDGMENTS
I share my deepest gratitude and appreciation with the following individuals who have
supported me throughout this project. First, I would like to thank my mother, father, and partner
for believing in my potential and encouraging me to persevere throughout graduate school. Their
love and support carried me during the challenging years that I had to be away from home.
Second, I would like to thank Dr. William (“Todd”) Abraham for his expertise and consultation
time for the statistical analyses of this project. Third, I would like to express my gratitude to Drs.
David Vogel, Frederick Lorenz, Karen Scheel, and Norman Scott for their support and guidance
as members of my dissertation committee. Lastly, I would like to thank my advisor, Dr. Meifen
Wei, for taking me on as an advisee all those years ago and guiding me throughout my journey.
Dr. Wei, you have always believed in me and challenged me to grow as a scientist, counselor,
and Ecuadorian-American woman. As I begin to write the final chapter of my graduate training,
my memory of your optimism, faith in me, and presence will always remain close to my heart.
Thank you.

Published by

Essays

View all posts