University of Southern California Dissertations and Theses

The impact of child maltreatment on the mental health of adolescents: a longitudinal study of social anxiety and self-perception

(USC Thesis Other)

The impact of child maltreatment on the mental health of adolescents: a longitudinal study of social anxiety and self-perception

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content THE IMPACT OF CHILD MALTREATMENT
ON THE MENTAL HEALTH OF ADODLESCENTS:
A LONGITUDINAL STUDY OF SOCIAL ANXIETY AND SELF-PERCEPTION

by
Ju Ye Ji

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

August 2009

Copyright 2009 Ju Ye Ji
ii

DEDICATION

This dissertation is dedicated to my parents, who have given me unconditional love and
personal sacrifices; and Dr. John L. Horn, who instilled in me the inspiration to set high
academic goals and the confidence to achieve them.

iii

ACKNOWLEDGEMENTS

I would like to gratefully and sincerely thank Dr. Penelope Trickett for her
guidance, understanding, and patience during my doctoral studies at the University of
Southern California. Her mentorship was paramount in providing a well rounded
experience consistent my long-term career goals.
This is a great opportunity to express my gratitude and respect to my dissertation
committee. I would like to thank Drs. Ferol Mennen, Devon Brooks and John J. McArdle
for their sincere caring, hand-on supervision and guidance, accessibility, academic
challenges, and valuable discussions throughout my doctoral studies at USC.
Most of all, I would like to give special thanks to Dr. John L. Horn for his
guidance and lessons in getting my academic career started on the right foot. He was a
true model of academic scholar and a mentor in my life. He inspired me to have
passionate academic motivation and confidence to pursue academic career.
I am pleased to thank USC school of social work for its unlimited supports and
academic opportunities. In particular, I would like to thank Dr. Bruce Jansson, a chair of
doctoral program and Ms. Malinda Sampson, a program manager for their sincere
consideration on doctoral students and efforts to create the best academic environment.
In addition, I would like to thank my friends and colleagues for their great support,
encouragement, particularly some much needed humor and entertainment in what could
have otherwise been a somewhat stressful academic environment. I would like to thank
Dick Smoak, Marissa Hansen, Hansung Kim, Dennis Kao, Gretchen Heideman,
Kyunghyea Lee, Jaehee Yi, , Minkyung Rhee, Jina Sang, Sangmi Cho, Kris Stevenson,
iv

Sonya Negriff, Matt Brensilver, May Guo, Lu Zhou, Inhee Choi, Seiyoung Hwang, and
Youngjoo Han.
Finally, and most importantly, I would like to thank my parents, brother and sister.
Their faith, encouragement, quiet patience and unwavering love are undeniably the
bedrock upon which my life has been built.
v

TABLE OF CONTENTS

DEDICATION---------------------------------------------------------------------------------
ACKNOWLEDGEMENTS----------------------------------------------------------------
LIST OF TABLES
LIST OF FIGURES
ABSTRACT------------------------------------------------------------------------------------
CHAPTER 1: RESEARCH PARADIGM-------------------------------------------------
1.1 Resilience in Maltreated Children----------------------------------------------------
1.2 Resilience as a Process: Importance of Longitudinal Research------------------
CHAPTER 2: LITERATURE REVIEW--------------------------------------------------
2.1 Social Anxiety --------------------------------------------------------------------------
2.2 Self-Perception as a Key Resilience Factor-----------------------------------------
2.3 Conceptual Framework: The Impact of Child Maltreatment on Social
Anxiety and Self-Perception in Adolescence---------------------------------------

CHAPTER 3: RESEARCH QUESTIONS -----------------------------------------------
CHAPTER 4: METHOD --------------------------------------------------------------------
4.1 Sample------------------------------------------------------------------------------------
4.2 Procedure---------------------------------------------------------------------------------
4.3 Measures---------------------------------------------------------------------------------
4.4 Analysis ----------------------------------------------------------------------------------

CHAPTER 5: RESULTS--------------------------------------------------------------------
5.1 Preliminary Analysis-------------------------------------------------------------------
5.2 Substantive Structural Analysis-------------------------------------------------------
5.3 Summary of Findings-------------------------------------------------------------------

CHAPTER 6: DISCUSSION----------------------------------------------------------------
REFERENCES---------------------------------------------------------------------------------
APPENDIX A----------------------------------------------------------------------------------
APPENDIX B----------------------------------------------------------------------------------
ii

iii

vi

vii

ix

1
1
4

7
7
13

19

25

26
26
27
29
31

41
41
48
71

74

90

105

106

vi

LIST OF TABLES
Table 1 Sample Characteristics-------------------------------------------------------------
Table 2 Confirmatory Factor Analyses of Humiliation Anxiety and Performance
Anxiety-------------------------------------------------------------------------------

Table 3 Standardized Factor Loadings and Factor Correlation of Two Factor
Model of Social Anxiety----------------------------------------------------------

Table 4 Standardized Factor Loadings and Factor Correlations of Social Self-
Perception Factors------------------------------------------------------------------

Table 5 Mean and Standard Deviation of Study Variables across Age---------------
Table 6 Mean and Standard Deviation of Study Variables by the Maltreatment
and the Comparison Groups------------------------------------------------------

Table 7 Results of the Latent Growth Curve Analyses of Humiliation Anxiety-----
Table 8 Results of the Latent Growth Curve Analyses of Performance Anxiety----
Table 9 Results of the Latent Growth Curve Analyses of Social Self-Perception---
Table 10 Results of the Univariate Latent Difference Score Models of Humiliation
Anxiety, Performance Anxiety and Social Self-Perception------------------

Table 11 Model fit Indices of the Bivariate Latent Difference Score Models --------
Table 12 Results of the Bivariate Latent Difference Score Model of Humiliation
Anxiety and Social Self-Perception----------------------------------------------

Table 13 Results of the Bivariate Latent Difference Score Model of Performance
Anxiety and Social Self-Perception ---------------------------------------------

27
42
43
44
45
46
60
61
62
64
66
67
70

vii

LIST OF FIGURES
Figure 1 Mediation of Self-Perception in the Relationship between Child
Maltreatment and Social Anxiety----------------------------------------------

Figure 2 Longitudinal Dynamic Changes between Social Anxiety and
Self-Perception-------------------------------------------------------------------

Figure 3 Test of Discretion of the Study Constructs-----------------------------------
Figure 4 Analytical Model: Mediation of Social Self-Perception in the
Relationship between Child Maltreatment and Social Anxiety------------

Figure 5 Analytical Model: Latent Growth Curve Model----------------------------
Figure 6 Analytical Model: Bivariate Latent Difference Score Model ------------
Figure 7 Mean Scores of Humiliation Anxiety, Performance Anxiety and Social
Self-Perception Across Age----------------------------------------------------

Figure 8 Full Mediation of Social-Self-Perception between Child
Maltreatment and Humiliation Anxiety--------------------------------------

Figure 9 Full Mediation of Social-Self-Perception between Child
Maltreatment and Performance Anxiety-------------------------------------

Figure 10 Intra-individual Trajectories of Humiliation Anxiety with Age---------
Figure 11 The Estimated Growth Trajectory of Humiliation Anxiety by the
Maltreatment and the Comparison Groups ---------------------------------

Figure 12 Intra-individual Trajectories of Performance Anxiety---------------------
Figure 13 The Estimated Growth Trajectory of Performance Anxiety by the
Maltreatment and the Comparison Groups ---------------------------------

Figure 14 Intra-individual Trajectories of Social Self-Perception--------------------
Figure 15 The Estimated Growth Trajectory of Social Self-Perception by the
Maltreatment and the Comparison Groups----------------------------------

Figure 16 Results of the Bivariate Latent Difference Score Model of Humiliation
Anxiety and Social Self-Perception--------------------------------------------

23
24
33
35
36
40
47
50
52
54
55
57
57
59
59
68
viii

Figure 17 Results of the Bivariate Latent Difference Score Model of
Performance Anxiety and Social Self-Perception--------------------------- 71
ix

ABSTRACT
The purpose of the current study is to advance our understanding of the
development of social anxiety—with particular attention to humiliation anxiety and
performance anxiety—and social self-perception in maltreated and nonmaltreated
comparison adolescents, and to provide prospective empirical evidence supporting a
cognitive model of social anxiety. Specifically, the study aimed 1) to test mediation of
social self-perception in the relationship between child maltreatment experience and
social anxiety over time; 2) to understand the developmental trends of social anxiety and
self-perception across age; 3) to investigate the impact of child maltreatment on the
developmental trajectories of social anxiety and self-perception; 4) and to examine the
longitudinal dynamic mechanisms between social anxiety and self-perception across age.
This study was part of a larger longitudinal prospective study of the impact of
maltreatment on the development of adolescents. The ethnically diverse sample consisted
of 303 maltreated and 151 comparison children. Maltreated children were recruited from
active cases in the Los Angeles County Department of Children and Family Services and
the demographically similar comparison sample was recruited from the same zip-codes as
the maltreated sample. The subjects were evaluated three times with 1-year interval. A
series of longitudinal analyses techniques were used within a framework of Structural
Equation Modeling including: a cross-lagged model, a latent growth curve model, and a
univariate and a bivariate latent difference score model.
The results supported full mediation of social self-perception between child
maltreatment experience and social anxiety in adolescence. Child maltreatment
x

experience predicated lower social self-perception which in turn predicted higher
humiliation anxiety and performance anxiety at later time. Social anxiety tended to
decrease and social self-perception tended to increase with age, but there was no
significant difference in developmental trend of social anxiety between the maltreated
and the comparison adolescents. The results from the bivariate latent difference score
models reveal that social self-perception is a leading factor influencing change in
performance anxiety over time.
The findings lend strong empirical support for a cognitive model of social anxiety
in relations to self-perception. This study provides important empirical data that can be
used to guide the development and provision of effective intervention for social anxiety
and social self-perception in maltreated adolescents.

1

CHAPTER 1: RESEARCH PARADIGM
1.1 Resilience in Maltreated Children
Child maltreatment is a critical social problem that negatively influences the
psychosocial functioning of adolescent victims. In 2005, more than 3.5 million children
in the United States were investigated for alleged maltreatment, a rate of 48.3 children
per thousand. From the investigations, an estimated 899,000 children were determined to
be victims of abuse or neglect (US Department of Health and Human Services, 2007).
Although most frequently the victims of neglect, some of these children also suffered
physical, sexual, and/or emotional abuse, which is not surprising since most maltreated
children experience more than one type of maltreatment (Clausen & Crittenden, 1991;
McGee, Wolfe, Yuen, Wilson, & Carnochan, 1995).
Child development normally follows a predictable and organized pathway.
However, in an abnormal environment such as that involving child abuse and neglect, the
predictability and organization of child development are disrupted and deviate from the
normal developmental pathway. During last four decades, a good deal of research has
revealed the negative impacts of maltreatment on child development (Trickett &
McBride-Chang, 1995; Zielinski & Bradshaw, 2006), in particular, mental health
problems exhibited by the maltreated victims have been well recognized in child
maltreatment and mental health literature (Cohen, Brown, & Smailes, 2001; Kaplow &
Widom, 2007; MacMillan et al., 2001).
Contemporary understanding of the relationship between maltreatment of children
and psychopathology has become increasingly complex because a substantial number of

2

victims appear to be relatively asymptomatic or to exhibit healthy functioning (Collishaw,
Pickles, Messer, Rutter, Shearer, & Maughan, 2007; Lynsky & Fergusson, 1997;
McGloin & Widom, 2001). These individuals are termed “resilient” because of their
ability to overcome and withstand the adverse impact of childhood trauma (Garmezy,
1993). Resilience has been defined as a “successful adaption or competent functioning”
despite adversity or traumatic stresses (Cicchetti & Garmezy, 1993; Masten, 2001;
Masten, Hubbard, Gest, Telegen, Garmezy, & Ramirez, 1999). A review of the available
research on resilience among individuals who have experienced traumatic or life-
threatening events suggests that resilience phenomena are relatively common (Bonnano,
2004).
Given the great variation in the developmental outcomes of maltreated children,
victims who are resilient to the effects of child abuse raise important research questions:
what are the etiological pathways from child maltreatment to psychopathology, and why
are all children who experienced maltreatment not similarly affected? The investigation
of factors that bridge (mediate) or attenuate (moderate) the relationship between
maltreatment and the developmental outcomes of maltreated children is critical to
informing theories of etiology in psychopathology and to guiding clinical intervention
and preventive efforts for maltreated children.
Developmental psychopathology is a theoretical framework with which to explain
variability in the developmental outcomes of children, the processes that account for that
variation and how positive adaptations are made in the presence of adversity (Luthar,
Cicchetti, & Becker, 2000; Masten, 2004). The concept of "multifinality" (Cicchetti &

3

Rogosch, 1996) specifically describes a phenomenon in which similar adversity leads to
different outcomes among individuals. This theoretical perspective emphasizes the
differences in children's vulnerability and resiliency and their interactions with the social
environment, which interactions may be either risky or protective. Among the theories
based on developmental psychopathology, the ecological-transactional model of
maltreatment of children (Cicchetti & Lynch, 1998) posits that risk and protective factors
in all levels of the ecological environment contribute to the great variability in the
developmental outcomes of maltreated children.
A resilience factor can be any positive attribute or strength that leads to relatively
better adaptation and functioning over time (Leon, Ragsdale, Miller, & Spaccarelli, 2008).
Review of the existing literature reveals three sets of resilience factors: 1) individual
attributes; 2) family characteristics, and 3) a wider social environment or a relationship
with the external environment (Hasket, Nears, Ward, & McPhersin, 2006; Heller, Larrieu,
D’Impero, & Borris, 1999; Iwaniec, Larkin, & Higgins, 2006; Luthar & Zigler, 1991;
Masten & Garmezy, 1985; Perkins & Jones, 2004; Polk, 1997). It is possible that,
because no single factor may predict multiple aspects of resilience outcomes, the study of
resilience should move beyond focusing on a single resilience factor to consider dynamic
processes that promote adaptive functioning. However, exploration of multiple protective
or risk factors together may result in little specificity in predicting resilience outcomes. In
this respect, resilience research should accept the notion that there are multiple pathways
to developmental outcomes of maltreated victims (Fischer, Ayoub, Singh, Noam,
Maraganore, & Raya, 1997), as people with different psychopathologies all develop

4

along unique pathways to psychopathology. Critical factors and mechanisms under the
resilience phenomenon vary depending on the population (e.g., maltreated children),
developmental domains (e.g., social), developmental stages (e.g., adolescence), and types
of resilience factor (i.e., individual, familial or environmental). Shonk and Cicchetti
(2001) found unique resilience mechanisms in different developmental domains and
outcomes; in particular, they found that the effects of maltreatment on academic
adjustment were partially mediated by academic engagement while behavioral
adjustment was fully mediated by social competency and ego-resiliency. Therefore,
research on resilience should explore specificity and more detailed variations in resilience
patterns and mechanisms to identify unique developmental pathways to psychopathology
(Luthar et al., 2000).

1.2 Resilience as a Process: Importance of Longitudinal Research
Moving beyond identification of resilience factors, research has endeavored to
reveal the underlying mechanisms and processes of resilience: how such factors promote
positive outcomes. Resilience is adaptation through many complex dynamic interactions
between individual and environment, as well as intra-individual interactions (Cicchetti &
Rogosch, 1997; Egeland, Carlson, & Sroufe, 1993; Masten et al., 1999). Rutter (1987)
also defined resilience as “a process incorporating protective factors.” Thus, resilience
can be understood as a transactional process which negotiates the impact of adverse
situations, products of transactions within and between multiple systemic levels, and
changes in outcomes over time.

5

Taken as whole, the dynamic and transactional nature of resilience process and
phenomena require longitudinal investigation of children who experience traumatic
adversity such as child maltreatment (Cicchetti & Rogosch, 1997; Luthar et al., 2000;
Masten et al., 1999). Determining when to measure resilience could highly influence the
result of a research study; the status of resilience may change over time, so a single
measure of resilience at one point in time may be misleading. Children may indicate
positive adaptation at one point in time, but their level of adaptation may deteriorate
subsequently. Therefore, changes in and the stability of resilience over time are critical to
understanding of the long-term effect of maltreatment on the development of children.
However, many existing research studies of resilience in maltreated children have
employed a cross-sectional design (Collinshaw et al., 2007; Cohen, Adler, Kaplan,
Pelcovitz, & Mandel, 2002; Flores, Cicchetti & Rogosch, 2005; Perkins & Jones, 2004;
Rajendran & Videka, 2006; Spaccarelli, & Kim, 1995), while few have used a
longitudinal design. The major limitation of cross-sectional design is that the study could
be confounded by the age or developmental stages of abuse victims (Trickett, Kurtz, &
Pizzigati, 1998). The retrospective approach has also been used frequently in the existing
literature of child maltreatment (Himelein, & McElrath, 1996; Muller & Lemieux, 2000).
Although some research studies have employed a longitudinal research design
(DuMont, Widom, Czaja, 2007; Herrenkohl & Herrenkohl, 1994; Jaffee, Caspi, Moffitt,
Polo-Tomas, Taylor, 2007; Leon, Ragdsale, Miller, & Spaccarelli, 2008; Lynskey &
Fergusson, 1997; McGloin & Widom, 2001; Moran & Eckenrode, 1992), few studies
have used a longitudinal data analysis strategy or investigated longitudinal change or

6

developmental trajectories of maltreated children (Lansford, Malone, Stevens, Dodge,
Bates & Petti, 2006). It is important to use rigorous analytic techniques to examine the
resilience process and mechanisms related to the development of maltreated children.
With the recent development of more user-friendly statistical software, sophisticated
longitudinal data analyses techniques (e.g., latent growth curve modeling and hierarchical
linear modeling) have become more accessible. These statistical analyses help
researchers understand more comprehensively the long-term developmental trajectories
of maltreated children.

7

CHAPTER 2: LITERATURE REIVEW
2.1 Social Anxiety
2.1.1 Social Anxiety in Adolescence
Social anxiety is a severe fear of social interactions and/or situations that involve
social evaluation by others and/or performance before others, with irrational expectation
of psychological harm and desire to escape or avoid these situations (American
Psychiatric Association, 2000; La Greca, & Stone, 1993; Schlenker & Leary, 1982).
Social anxiety is a normal human reaction in many situations; however, some children
and adolescents experience extreme forms of social anxiety that is highly distressing, and
this extreme anxiety may interfere with their ability to maintain regular social
relationships and cope with social situations. In particular, adolescence is a period during
which social interaction and participation in social activities are more frequently expected
and encouraged; therefore, social anxiety can be a critical mental health issue influencing
many areas of adolescents’ lives. Over the past several decades, a substantial amount of
research has investigated social anxiety disorder in adulthood (Beidel, Turner, Dancu,
1985), but there has been relatively little exploration of social anxiety in adolescence
(Beidel, Turner, Morris, 1999; Kashdan & Herbert, 2001; Rao, Beidel, Turner,
Ammerman, Crosby, & Sallee, 2007). Social anxiety in adolescents should be explored
and understood better for several reasons. First, the mean age for a diagnosis of social
anxiety is middle adolescence, from 11 to 15 years old (Last, Perrin, Hersen, & Kazdin,
1992; Schneier, Johnson, Hornig, Liebowitz, & Weissman, 1992), and documented
prevalence estimates of social anxiety in adolescence ranges between 5 and 15% of the

8

population, placing it among the most common psychiatric disorders (Essau, Conradt, &
Peterman, 1999; Heimberg, Stein, Hiripi, & Kessler, 2000; Wittchen, Stein, & Kessler,
1999). Second, recent literature on social anxiety addresses a wide range of functional
impairments caused by social anxiety, including low school performance and educational
underachievement (Woodward & Fergusson, 2001), fear of joining social and athletic
activities (Caspi, Elder, & Bem, 1988), physiological symptoms (Storch, Brassard,
Masia-Warner, 2003), difficulty in social/peer relationships (Ginsburg, La Greca,
Silverman, 1998; Inderbitzen, Walters, & Bukowski, 1997; La Greca & Lopez, 1998;
Storch et al., 2003), low quality of life (Safren, Heimberg, Brown, & Holle. 1996), and
alcohol or drug dependence (DeWit, MacDonald, & Offord, 1999). Third, social anxiety
is comparatively stable and has chronic impacts across the sufferer’s lifetime. Adults with
social anxiety often report that their symptoms began in childhood or adolescence and
persisted into adulthood (Davidson, Hughes, George, & Blazer, 1993). Long-term
outcomes of social anxiety in adulthood include low levels of work productivity (van
Velzen & Emmelkamp, Scholing, 2000), impaired functioning in romantic and marital
relationships (Forthofer, Kessler, Story, & Gotlib, 1996), low subjective well being
(Mendlowicz & Stein, 2000; Stein & Kean, 2000), and financial dependency (Kessler,
2003; Lipsitz, & Schneier, 2000). Fourth, there is some evidence that social anxiety in
adolescence is predictive of secondary mental disorders in adulthood, particularly
depression and other types of anxiety disorder (Kessler, 2003; Stein, Fuetsch, Muller,
Höfler, Lieb, & Wittchen 2001; Woodward & Fergusson, 2001). A longitudinal
investigation of the causal relationship between anxiety and depression (Cole, Peeke,

9

Martin, Truglio, & Seroczynski, 1998) found that anxiety leads to the development of
depression and that this pattern was consistent across time. Finally, individuals with
social anxiety are less likely to recognize themselves as having a mental disorder and
much less likely to seek treatment than people with other disorders of comparable
severity. (Kessler, 2003). The median time between the age at which social anxiety
begins and the age at which the sufferer first seeks professional help is more than 15
years. In general, adolescents do voluntarily seek professional help without parents’
initiation and it makes even more difficult for them to be treated.
Thus, given the importance of social anxiety in adolescence, more scientific evidence
would be useful in better understanding and treating social anxiety in adolescence.
Ultimately, the investigation of social anxiety in adolescents will contribute to their well
being and will help prevent further mental health problems in adulthood.

2.1.2 Definition of Social Anxiety
Several different psychological terms have been used to describe fear or
avoidance of social interactions and situations. The most frequently mentioned terms
demonstrating fear of social interaction and situations are social anxiety, social phobia,
shyness, and social withdrawal (Kearney, 2005). Among those, social phobia and social
anxiety are the terms most widely used in the literature. In general, “social phobia” refers
to a clinical diagnostic category, and “social anxiety” refers to levels along a normally
distributed continuum (Rapee & Spence, 2004). Fear and avoidance of social situations
occurs broadly and normally across population, but a diagnosis of social phobia will be

10

associated with higher levels of social anxiety on this continuum (Rapee, 1995). The
current investigation uses the term “social anxiety,” and the focus of the study will be to
understand the mechanisms that contribute to individuals’ location or movement along
this distributional construct. Two subclusters of social anxiety may exist: humiliation
anxiety and performance anxiety. Humiliation anxiety is a sever fear of evaluation,
rejection or humiliation by others, particularly peers in adolescence. Performance anxiety
is persistent anxiety or fears experienced when an individual is required to perform in
public.

2.1.3 Child Maltreatment and the Development of Social Anxiety
Anxiety is a commonly reported mental health outcome in the study of maltreated
victims (Cohen et al., 2001; Kaplow & Widom, 2007; MacMillan et al., 2001; Stein et al.,
1996). Although anxiety disorder is a broad-spectrum disorder ranging from general
anxiety to post-traumatic stress disorder, the previous research on child maltreatment has
seldom articulated the relationship between maltreatment in childhood and specific types
of anxiety; instead, it has applied the general term of “anxiety.” Consequently, there is
little empirical evidence about the relationship between child maltreatment and social
anxiety.
A couple of studies have reported a significant relationship between child
maltreatment and the risk of social anxiety. First, Chartier, Walker and Stein’s (2001)
retrospective study of social phobia and childhood risk factors reported a significant
relationship between childhood physical and sexual abuse and social phobia. Similarly,

11

DeWit et al. (2005) studied the effects of family adversity on the onset of social phobia
and found that physical abuse in childhood by a father figure increased the risk of social
phobia in the female group. However, both studies had significant methodological
limitations, including cross-sectional designs and a wide age range in their samples of
adolescents and adults. Magee (1999) also investigated the effects of negative life
experiences on phobia onset and found that violence in childhood at the hands of adults
had a significant relationship with social phobia.
Although there is little direct evidence on the effect of child maltreatment on the
development of social anxiety, research has demonstrated a substantial contribution of
familial and socio-environmental risk factors to the development of social anxiety.
Theoretically, the etiology of social anxiety may begin with the disapproval of significant
others (Rapee & Spence, 2004). Since the most salient feature of social anxiety is a fear
of unsatisfactory evaluations from others in social situations, rejection or disproval by
caregivers and family members in early life plays an important role in shaping social
anxiety. Experiences and interactions with the environment and others are stored in a
memory system and this internalized information is used to construct an image of one’s
observable self. The internalized image of self determines an individual’s expectations,
beliefs, feelings, and rules of behavior.
Empirically, parenting style is one of the most commonly studied familial risk
factors for social anxiety (Bögels, van Osten, Muris, & Smulders, 2001; Lieb, Wittchen,
Höfler, Fuetsch, Stein & Merikangas, 2000; Masia & Morris, 1998). Parental rejection,
lack of emotional warmth and parental control are known to be associated with

12

development of social anxiety. Evidence from observational studies of parent-child
interactions have supported the idea that parents of socially anxious children tend to exert
greater parental control and less emotional warmth (Dadds, Barrett, Rapee, & Ryan,
1996). Research on withdrawn preschool children revealed that their parents tend to
believe more strongly that children should be raised using higher parental power and
coercion (Rubin & Mills, 1990). Parents of socially anxious children tend to be less
sociable and to isolate themselves from social situations (Bruch & Heimberg, 1994).
Therefore, their children are less likely to have social interactions and modeling
regarding adaptive social and cognitive skills, which, in turn, contribute to the
development of social anxiety.
Then, lack of a close relationship with a caregiver or an adult is also known as a
risk factor for social anxiety (Chartier et al., 2001). The interpersonal model of
psychopathology identifies social anxiety as a dysfunctional interpersonal pattern, the
“result of an ongoing interaction between the individual and social environment” (Alden
& Taylor, 2004). The central feature of the interpersonal model of social anxiety is a
“self-perpetuating cycle,” that is, the memory of past social experiences may be laid
down in the form of an image of the social self, and this is reactivated later in life when a
person faces social interactions and situations.
Negative parenting and lack of a close relationship with a caregiver are major
characteristics of childhood experiences of maltreatment; thus, these findings indirectly
explain the impact of child maltreatment on the development of social anxiety. However,
most have been cross-sectional (Chartier et al., 2001; Lieb et al., 2000; Masia & Morris,

13

1998) or based on adults’ retrospective reports on childhood experiences (Bögels et al.,
2001; Magee, 1999).
Summarizing the direct and the indirect research evidence together, the existing
literature provides some support for the probable relationship between child maltreatment
and social anxiety. However, more prospective evidence is needed; in particular, an
investigation of the long-term effects of child maltreatment on social anxiety
development is required for greater understanding of the etiology of social anxiety and to
development of more effective preventive and intervention strategies.

2.2 Self-Perception as a Key Resilience Factor
2.2.1 Definition of Self-Perception
Previous research has demonstrated that the self-system is a key factor in
promoting resilience in maltreated children and adolescents (Cicchetti & Rogosch, 1997;
Toth & Cicchetti, 1996; Romans, Martin, Anderson, Shea, & Mullen, 1995; Shonk &
Cicchetti, 2001). The psychological functioning of the self-system, particularly
individuals’ cognitive perception or subjective evaluation of self, has been a major
interest of social science research for decades. Many psychological terms have been used
to describe this cognitive construct, including self-concept (Shavelson, 1976; Marsh &
Ayotte, 2003), perceived self-efficacy (Bandura, 1993), self-image (Petersen, 1984), self-
criticism (Blatt, D'Afflitti, & Quinlan, 1976), perceived competence (Harter, 1982), and
self-perception (Harter, 1988, 1999). Self-efficacy is individuals’ “sense of agency” or
the belief about one’s ability to perform certain actions, and the cognitive representation

14

that those actions can be performed successfully. Self-criticism is a construct related to
“feelings of inferiority and worthlessness by a sense that one has failed to live up to
expectations and standards” (Blatt et al., 1976, p.383). Self-perception is an evaluative
component of self and the individual’s belief or cognitive refection of self-competence
and ability in producing desired behaviors and outcomes. Harter also contended that self-
competence constitutes the fundamental basis of a child’s self-perception and plays an
important role in the self-system. While there are some subtle differences between the
concepts and theories, these constructs are similar in that they all describe the extent to
which an individual perceives or evaluates himself or herself and his or her ability as
competent. The current study uses the term “self-perception” to indicate individuals’
cognitive reflection of self and the ability to competently perform desired behaviors and
reach desired outcomes.

2.2.2 Role of Self-perception in the Development of Psychopathology
Previous research has emphasized the critical role of self-perception in the
development of psychopathology (Cole, 1990; Cole, Peeke, Dolezal, Murray, &
Canzoniero, 1999; DuBois, Felner, Brand, & George, 1999; Marsh, Parada, Yeung, &
Healey, 2001). A compelling body of theory and empirical research has proposed that a
child’s distorted self-perception is an important pathway between negative experience in
childhood and the development of maladaptive psychosocial functioning (Bailey &
McCloskey, 2005; Bruce, Cole, Dallaire, Jacquez, & Pineda, 2006; Gibb et al., 2001;
Grills & Ollendick,2002; Jacquez, Cole, & Serale, 2004; Cole et al., 1996, 1998, 1999;

15

Kim & Cicchetti, 2004, 2006; Koss & Figueredo, 2004; Raboteg-Saric, Rijavec, &
Brajsa-Zganec, 2001; Romans, et al., 1995; Sachs-Ericsson et al., 2006; Tram & Cole,
2000; Wyatt & Newcomb, 1990). That is, adversity experienced in childhood diminishes
the positive sense of self which, in turn, is a risk or protective factor for the development
of psychopathology.
Although a number of research studies have paid attention to the relationship
between self-perception and psychopathology, only a few studies have investigated the
causal directionality between self-perception and psychopathology using rigorous
analytic techniques (Cole & Tram, 2000; Cole et al., 1996, 1999). Moreover, the extant
literature provides inconsistent findings on the directionality between the two constructs.
There has been disagreement in the literature as to whether negative self-perception leads
to psychopathology (Cole et al., 1996, 1999; Sachs-Ericsson et al., 2006; Tram & Cole,
2000) or psychopathological symptoms cause a negative view of the self (McGrath &
Reptti, 2002). There is even a study suggesting a reciprocal relationship between self-
perception and psychopathology (Nelson, Rubin & Fox, 2005). Most psychopathology
studies have tested whether a negative view of self causes psychopathology, but few have
tested the reverse causal relationship to determine whether psychopathology influences
negative self-perception. Also, because the investigation of the causal mechanism
between self-perception and psychopathology has mainly focused on depression (Cole,
1990; Cole et al., 1996, 1999; Tram & Cole, 2000), evidence elucidating the directional
relationship between self-perception and other types of psychopathology is lacking.

16

2.2.3 Multidimensional (domain-specific) Model of Self-Perception
Self-perception research has debated two theoretical approaches to self-
perception: situation-specific self-image (Marsh et al., 2003) and global beliefs about self
(Rosenberg, 1979). Recent theoretical conceptualizations of self-perception have paid
great attention to a multidimensional (domain-specific) conceptualization of self-
perception, although the uni-dimensional approach is still dominant in mental health and
child abuse literature. The conventional model of development has been criticized for an
oversimplification of and a homogeneous approach to developmental processes and
pathways. Complexity and specificity in pathological development of children can be
better understood when alternative pathways and alternative factors comprising the
pathways are considered (Fischer et al., 1997). The multifaceted approach to self-
perception proposes an idea that self-perception in different developmental domains may
play a unique role in the development of psychopathology and lead to distinctive
developmental outcomes (DuBois et al., 1999; Marsh, Parada, & Ayotte, 2004).
Prospective investigation of the relationship between negative affect and academic and
social competence (Cole et al., 1999) revealed interesting causal relationships between
and among these constructs. Social competence predicted a negative affect, while a
negative affect predicted academic competence. This finding supports the idea that self-
perception in different domains may have a unique relationship with psychopathology.
According to the multidimensional perspective on self-perception, the focus of the
current investigation is particularly on the self-perception in social domain, which are

17

one of the most salient developmental features in adolescence and closely related to
social anxiety.

2.2.4 Child Maltreatment and the Development of Self
Development of a positive sense of self emerges from the context of social
relationships as children learn to regard themselves in ways that reflect how others regard
them (Harter, 1999). If children receive negative feedback from multiple sources, they
become cognitively cornered into adopting relatively global, negative views of
themselves. In addition, the sense of self develops based on a sense of trust and security
in social relationships (Cicchetti, 1991). Similarly, the competency-based model of
psychopathology explains the influence of caregivers and family on self-competence
(Cole, Martin, & Powers, 1997). Children and adolescents use information from the
environment, particularly feedback from significant others, as they construct beliefs about
their competencies. Feedback that conveys criticism of or dissatisfaction with their
character or behavior negatively affects adolescents’ construction of self-concept.
Children and adolescents internalize this competency-relevant feedback to conceptions of
themselves as competent or incompetent in various domains.
Child maltreatment is often extreme form of negative care-giving which increases
the risk for the formation of negative self-perception (Cole & Putnam, 1992; Mannarino,
Cohen, & Gregor, 1989; Tong, Oates & McDowell, 1987; Toth & Cicchetti, 1996; Toth,
Cicchetti, Mcfie & Emde, 1997). The developmental psychopathology model (Beeghly &
Cicchetti, 1994; Cole & Putnam, 1992; Toth, Cicchetti, Mcfie & Emde, 1997) suggests

18

that the traumatic nature of child maltreatment can interfere with the ongoing
development of self and can cause deviations in defining and integrating aspects of self.
In a similar vein, drawing from attachment theory, abusive and neglectful care-giving
often leads to development of negative representations of others, consequently
contributing to a negative internal scheme of self (Toth, Cicchetti, Macfie, Maughan, &
Vameenen, 2000). In addition, a developmental framework for conceptualizing the
effects of sexual abuse also demonstrates that sexual abuse trauma diminishes the
victim’s positive sense of self (Cole & Putnam, 1992; Diehl & Prout, 2002; Spaccarelli,
1994).
An impaired self-system caused by child maltreatment is often described as “self-
disorder” or “self-pathology” (Toth & Cicchetti, 1996). Maltreated children tend to have
more negative view of self than do non-maltreated children. These negative views often
include low self-esteem (Bolger, Patterson, & Kupersmidt, 1998; Kaufman & Cicchetti,
1989) less self-competence (Cicchetti, Beeghly, Carlson, & Toth, 1990; Toth & Cicchetti,
1996; Vondra, Barnett, Cicchetti, 1989), a negative representation of self (Fischer et al,
1997; Toth et al, 1997; Toth et al., 2000), low self-perception (Barnett, Vondra, & Shonk,
1996), higher self-criticism (Sachs-Ericsson et al., , 2006), self-blame (Wyatt &
Newcomb, 1990) and negative self-statement (Calverley, Fischer, & Ayoub, 1994).
The theory of resilience in maltreated children highlights the important role
played by self-perception in psychosocial functioning of maltreated children (Masten et
al., 1999; Romans et al., 1995; Sachs-Ericsson et al., 2006; Shonk & Cicchetti, 2001;
Toth & Cicchetti, 1996). Although the transactional nature of resilience requires a

19

longitudinal investigation of self-perception (Cicchetti & Rogosch, 1997), only a handful
of studies have taken a longitudinal view of the relationship between child maltreatment
and self-perception (Cicchetti & Rogosch, 1997; Tram & Cole, 2000; Kim & Cicchetti,
2004, 2006; Koss & Figueredo, 2004). In particular, it is not well understood whether
child maltreatment influences long-term changes in or trajectory of self-perception. In
addition to the primary negative effects of child maltreatment on self-perception,
previous research has suggested that the resilience mechanism in relation to self-
perception may differ between maltreated and non-maltreated children. Cicchetti and
Rogosch (1997) reported that interpersonal relationships act as a salient resilience factor
for non-maltreated children, while self-reliance and self-confidence were more important
resilience factor for maltreated children. This finding suggests an idea that child
maltreatment may moderates the relationship between self-perception and later
psychopathology. However, whether the resilience mechanism in terms of self-perception
and its relationship to psychopathology differs between maltreated and non-maltreated
children has not yet been investigated.

2.3. Conceptual Framework: The Impact of Child Maltreatment on Social Anxiety
and Self-Perception in Adolescence
2.3.1 Cognitive Model of Social Anxiety: Social Anxiety and Self-Perception
A cognitive model of social anxiety has been the most widely used model for
treatment and research (Clark & Wells, 1995; Gould, Buckminster, Pollack, Otto, &
Massachusetts, 1997; Rapee & Heimberg, 1997). The model emphasizes the centrality of

20

the mediating role played by negative cognition (Alfano, Beidel, & Turner, 2006; Bogels
& Zitgerman, 2000; Clark & Wells, 1995; Hofmann, 2000; Hofmann, Moscovitch, Kim
& Taylor, 2004). Specifically, the cognitive mechanism in social anxiety is explained by
the underestimation of one’s abilities and the overestimation of danger, threat, and fear.
People with social anxiety tend to have low expectations regarding their ability to
produce preferred impression-relevant reactions and to cope in the face of obstacles. Also,
socially anxious individuals anticipate negative consequences of social events; typically,
prior to a social event, they become anxious and tend to be dominated by recollections of
past failure, by negative self-images and by predictions of poor performance and
rejection by others.
The cognitive model of social anxiety emphasizes the individual’s negative self-
perception in the etiological and intervening mechanism of the disorder (Alfano et al.,
2006; Clark & Wells, 1995; Ginsburg, La Greca, & Silverman, 1998; Hofmann, 2004;
Hofmann et al., 2004; Rapee, & Heimberg, 1997). Self-perception is an important
psychological vulnerability leading to the development and maintenance of social
phobia—perhaps even for the likelihood of development future comorbidity, including
major depression (Cole, Peeke, Martin, Truglio, & Seroczynski, 1998). Negative self-
image—the perceived inability to predict, control and obtain desired outcomes—has been
viewed as a major antecedent of anxiety (Schlenker & Leary, 1982). It is often expressed
as negative cognition, irrational beliefs and low expectations of performance (Alfano et
al., 2006). Moreover, socially anxious individuals tend to be excessively self-focused, so
their attention is shifted away from external social cues (Clark, 2001), impairing

21

engagement and performance in social situations (Hope & Heimberg, 1988) and
contributing to further development of a negative sense of self (Clark & Wells, 1995;
Alfano et al., 2006).
Terms used to describe negative self-perception in relation to social anxiety
include self-criticism (Cox et al., 2004), negative self-schema (Epkins, 1996), self-
perception (Ginsburg, La Greca, & Silverman, 1998; Hofmann et al., 2004; Mallet,
Rodriguez-Tome, 1999), self-imagery (Hirsch, Meynen, & Clark, 2004), self-competence
(Chansky & Kendall, 1997; Smari, Petursdottir, & Porsteinsdottir, 2001), self-
representation (Schlenker & Learly, 1982), negative self-talk (Spence, Donovan, &
Brechmann-Toussaint, 1999), and self-impression (Harvey, Clark, Ehlers, Rapee, 2000).
Substantial empirical evidence has supported the relationship between self-
perception and anxiety in childhood (Kendall & Treadwell, 2007; Scheier, Carver, &
Bridge, 1994; Schiefelben, Susman, & Dorn, 2006; Chansky & Kendall, 1997; Weems,
Berman, Silverman, & Saavedra, 2001), and research on social anxiety in adulthood has
also provided empirical evidence on the relationship between cognitive self-view and
social anxiety (Cox et al., 2004; Harvey, Clark, Ehlers, Rapee, 2000; Hirsch et al., 2004;
Hofmaan et al., 2004; Schlenker & Learly, 1982). However, the investigation of
childhood social anxiety has begun only recently, so only a handful of studies have
examined self-perception in relation to social anxiety in children and adolescents (Alfano
et al., 2006; Epkins, 1996; Ginsburg, La Greca, & Silverman, 1998; Smari et al., 2001).
For example, Alfano et al. (2006) investigated the cognitive correlates of social phobia
among children and adolescents and found that children and adolescents with social

22

phobia tend to expect to perform poorly and to evaluate their performance as inferior to
those of their peers who have no psychiatric disorder. Similarly, Epkins (1996) revealed
higher levels of negative self-schemas in socially anxious children than in a control group,
and Ginsburg et al. (1998) reported that socially anxious children perceived their social
acceptance and general self-worth to be low.
Although cognitive-behavioral therapy has been identified as most likely to be
efficacious (Gould et al., 1997; Hofmann, 2004), the cognitive mechanisms under social
anxiety in childhood and adolescence are not well understood. More importantly, most
studies have employed a cross-sectional design, so longitudinal research on the role of
self-perception in the development of social anxiety is lacking. Additional exploration of
the cognitive mechanism in childhood social anxiety is required to better understand the
etiology of social anxiety and to provide evidence-based treatment for children and
adolescents with social anxiety.
2.3.2 Child Maltreatment, Social Anxiety and Self-Perception
A substantial body of empirical research has shown that maltreated adolescents
are at risk for a wide range of psychopathological outcomes, but the literature has also
revealed resilience in maltreated adolescents. Unfortunately, little is known about the
developmental course of psychopathology and the resilience mechanism in maltreated
children and adolescents. Investigation of factors contributing to variability in mental
health outcomes and resilience in maltreated victims elucidates the etiological mechanism
of psychopathology and eventually help to develop better preventive and intervening
strategies.

23

Self-perception is a key resilience factor which plays a critical role in the
development of psychopathology. In specific, the literature specifies that self-perception
mediates the relationship between childhood adversity and later mental health outcomes
(see Figure 1). The experience of child maltreatment contributes to the development of
negative self-perception, which, in turn, leads to the development of psychopathology;
the cognitive model of social anxiety emphasizes the mediating role of self-perception in
the etiological and the intervening mechanisms of social anxiety. Social anxiety is a
critical mental health problem which can significantly impair adolescents’ psychosocial
functioning and cognitive behavioral therapy as the most popular treatment modality for
social anxiety disorder. However, an extensive review of the literature revealed that
prospective evidence on the relationships between and among child maltreatment, social
anxiety and self-perception is sparse.
Figure 1 Mediation of Self-Perception in the Relationship between Child Maltreatment
and Social Anxiety

Furthermore, research on a cognitive model of childhood social anxiety only
began recently, and there is very little prospective evidence explaining the developmental
course of social anxiety (Hale III, Raaijmakers, Muris, van Hoof, & Meeus, 2008) and
self-perception (Cole et al., 2001) and elucidating the causal mechanism between anxiety

24

and self-perception in childhood and adolescence (Bosque & Egeland, 2006; Tram &
Cole, 2000). Particularly, only a handful of studies have employed longitudinal data
analysis techniques to investigate relationships between and among child maltreatment,
self-perception and psychosocial outcomes (Kim & Cicchetti, 2004; 2006; Koss &
Figueredo, 2004). Figure 2 depicts a mode of longitudinal dynamic changes between
social anxiety and self-perception. Particularly, the primary focus of the current study is
to understand whether self-perception is a leading factor influencing change in social
anxiety. In other words, it is tested whether change in social anxiety is predicted by
previous state of self-perception over time (the blue one headed arrow).

Figure 2 Longitudinal Dynamic Changes between Social Anxiety and Self-Perception

25

CHAPTER 3: RESEARCH QUESTIONS

This is a study to examine the long-term impact of child maltreatment on the
development of social anxiety—with particular attention to two specific types of social
anxiety, humiliation anxiety and performance anxiety— and self-perception in social
domain using a series of longitudinal data analysis techniques. Key research questions for
the study are as follows:
1. Does self-perception mediate relationship between child maltreatment
experience and social anxiety?
2. What are developmental trends found in social anxiety and self-perception over
age?
3. Does child maltreatment experience have significant impact on the
developmental trajectories of social anxiety and self-perception?
4. What are longitudinal dynamic mechanisms between social anxiety and self-
perception?

26

CHAPTER 4: METHOD

4.1 Sample
Recruitment. Participants are adolescents enrolled in the Young Adolescent
Project (YAP), an ongoing longitudinal study of the impact of child maltreatment on the
development of young adolescents. The initial sample consists of 454 ethnically diverse,
9- to 12-year-old boys and girls. The maltreated sample (N=303) was recruited from
active substantiated cases in the Los Angeles County Department of Children and Family
Services (LADCFS). To control for the confounding effects of diverse conditions
inherent in the large metropolitan area served by LADCFS, participants were selected
from only ten zip codes in Los Angeles County. Inclusion criteria for the maltreated
group at initial assessment were 1) a new substantiated report of any type of maltreatment
to the LADCFS in the preceding month; 2) a child aged 9 to 12 years; 3) a child
identified as African-American, Hispanic (non-white), or Caucasian; and 4) a child
residing in one of ten zip-codes in Los Angeles County at the time of referral to LADCFS.
According to information abstracted from case records, most children in the
maltreated group experienced multiple forms of maltreatment (66%) and had multiple
referrals as well. The majority of the maltreatment sample experienced neglect in some
form (78%), about half of the sample experienced physical abuse (48%) and/or emotional
abuse (48%), and approximately one fifth experienced sexual abuse (21%). On average,
the participants had experienced two types of maltreatment and four referrals to CFS. A
demographically similar comparison sample of nonmaltreated children (N=151) was

27

recruited from the same zip codes via school lists of families with 9- to 12-year-old
children.
Upon enrollment in the study both the maltreatment and comparison groups were
compared on a number of demographic variables. The two groups were similar on age,
gender, race/ethnicity, and neighborhood characteristics (based on Census block
information). However they were different in terms of living arrangements. In the
comparison group 93% lived with a biological parent, whereas this was the case for only
52% of the maltreated group. The remainders were living in foster care, which is not
unusual for those adolescents involved with social services.
Table 1 Sample Characteristics
Maltreatment group Comparison group
n % n %
Gender
Male
Female

152
151

50
50

90
61

60
40
Ethnicity
African American
Latino
White
Mixed/Biracial

123
106
35
39

40
35
12
13

48
71
15
17

32
47
10
11
Living Arrangements
With Biological Parents
Foster care/Kin

155
148

52
48

118
33

93
7
n M(SD) n M(SD)
Age at Time 1 303 10.84 (1.15) 151 11.11 (1.15)
Age at Time 2 249 12.00 (1.17) 142 12.28 (1.26)
Age at Time 3 192 13.56 (1.38) 128 13.83 (1.45)

4.2. Procedure
Recruitment proceeded with approvals from LADCFS, the courts and the

28

University Institutional Review Boards. Caregivers of families were contacted via mail
and asked to return a postcard indicating their willingness to participate. The recruitment
of the maltreated sample began in spring 2002 and ended in fall 2004, and recruitment of
the comparison sample began in 2002 and ended in fall 2005. To date, the subjects have
been evaluated twice, and the attrition rate between Wave 1 and Wave 2 is 13.4% (n=61)
and between Time 1 and Time 3 in this study is 29.5% (n=134). Two separate binary
logistic regression analyses were performed to test whether attrition at Time 2 and Time 3
was random. The dependent variable for the attrition analysis was a dichotomous variable
(yes/no) indicating attrition at Time 2 and Time 3. Time 1 humiliation anxiety,
performance anxiety, social self-perception and demographic variables were entered in
order to predict the dropout of participants during the longitudinal assessment. The
results of attrition analyses indicated that the participants who were not seen at Time 2 or
at Time 3 were more likely to be in the maltreatment group (odd ratio for Time 2 =2.11,
odd ratio for Time 3= 2.09 )
A trained research assistant interviewed the caretakers and the adolescents
separately using a broad range of measurements: biological, observational and self-
reported. Caretakers complete standardized measurements, assessing various attitude and
behaviors about themselves, their children and their families. Adolescent participants
respond to a testing battery composed of 1) social-cognitive tasks and saliva collection
designed to measure stress-reactivity; 2) a health assessment, including physical and
pubertal development; and 3) standardized measures of cognitive ability, behavioral and
emotional problems, family functioning and trauma history. All modules are completed

29

in a one-day session which lasts about 4 hours. After the interview, a debriefing session is
followed by the disbursement of payments to family based on rates specified by the
NIMH Normal Volunteer Program.
4.3 Measures
Child Maltreatment. Although the YAP study was designed specifically to
examine the effects of neglect on development, participants who had experienced all
types of maltreatment, so long as they were reported to DCFS and the report was
substantiated, were recruited. This is in part because it would have been difficult to amass
a large sample of children who had experienced only neglect, especially given evidence
which indicates that most maltreated children are likely to experience more than one type
of maltreatment (Clausen & Crittenden, 1991; McGee et al., 1994). In fact, 66% of the
maltreatment sample participating in this study had experienced multiple forms of
maltreatment; thus, the maltreatment variable is positive if a child experienced any type
of maltreatment (0=comparison group, 1=maltreatment group).
Social Anxiety. The Multidimensional Anxiety Scale for Children (MASC; March,
Parker, Sullivan, Stallings, & Conners, 1997) is a self-report questionnaire developed to
measure different types of anxiety in children and adolescents. The instrument consists of
39 items in 4 subscales: physical symptoms (12 items), harm avoidance (9 items), social
anxiety (9 items) and separation-panic scale (9 items). The present study utilizes the
social anxiety subscale, which has 4 items for performance anxiety (e.g., “I get nervous if
I have to perform in public”) and 5 items for humiliation anxiety (e.g., “I worry about
what other people think of me”). Each statement is rated on 4-point continuum from “0”

30

(never true about me) to “3” (often true about me). Documented reliability for the social
anxiety scale was 0.82. In the present study, Crobach’s alpha coefficients at Time 1, Time
2 and Time 3 ranged from .75 to .80 for humiliation anxiety subscale and ranged from .62
to .71 for the performance anxiety subscale.
Self-Perception. The Self-Perception Profile for Adolescents (SPPA; Harter,
1988) is one of the most widely used self-reported measures of self-perception. The
SPPA is a 45-item questionnaire containing one general self-worth subscale and nine
domain-specific subscales (scholastic, social, friendship, romantic appearance, job
competence, athletic, physical appearance and behavioral conduct, general self-worth).
Each subscale is composed of five items, with three worded in the positive direction and
two in the negative direction. The items are structured in a 4-point alternative format in
order to offset the tendency for adolescents to respond in a socially desirable manner. In
particular, the proposed study utilized social acceptance which assesses adolescents’
perception regarding peer acceptance and likability. Adolescents respond to each item by
deciding which of two statements (e.g., “Some people feel that they are socially
accepted” or “Other people wish that more people their age accepted them.”) best
describes them; then they decide whether the statement they have chosen is very true for
them or just sort of true for them. The documented reliability coefficients for the SPPA
subscales ranged from 0.80 to 0.85 (Harter, 1988) and Cronbach’s alpha coefficient in
this study at Time 1, Time 2 and Time 3 ranged from .60 to .68. The validity research
reported in the SPPA manual (Harter, 1988) suggest a discrete factor structure of SPPA.

31

4.4 Analysis
4.4.1 Preliminary Analyses
Incomplete Data. As with most prospective studies, data for some participants are
not available in all three waves of assessment. Analyzing only those cases with complete
data has the potential to produce biased results; therefore, the study will analyze the total
sample to reduce potential bias in its findings. First, in order to obtain the maximum
number of cases, item-level missing data were imputed with a multiple imputation
method using the NORM software program (Schafer, 1997). Multiple imputation replaces
missing data with an estimated value “representing a distribution of possibilities” through
iteration processes (Rubin, 1987). In order to address issues regarding missing data, the
full-information maximum likelihood estimation (FIML; Arbuckle, 1996) was used to
incorporate the incompleteness of data in estimating parameters. FIML does not impute
data but breaks the likelihood function down into components based on patterns of
missing data, allowing estimation to proceed using all available data. To implement
FIML, intercepts of observed variable and means of latent factors are estimated in the
model. FIML estimation of SEM parameters requires data with normal distribution.
Factor Analysis of Study Construct. Factor structure of the study constructs was
tested to support the reliability of measurement. In order to accomplish this, the factor
structure of the two social anxiety outcomes (i.e. humiliation anxiety, performance
anxiety) and social self-perception was examined using confirmatory factor analysis
(restricted structural factor analysis; McArdle, 1996).

32

4.4.2. Substantive Structural Analysis
Goodness-of-fit criteria. To compare the models, the relative fit of the nested
competing models was examined with chi-square difference tests and goodness-of-fit
indices, including the root mean square error of approximation (RMSEA; Steiger, 1990)
and the comparative fit index (CFI; Bentler, 1990). RMSEA refers to the lack of fit of the
model to the population covariance matrix. An RMSEA of .05 or less indicates a close fit
between a model and data (Browne & Cudeck, 1993). The 90% confidence interval and
probability that the RMSEA is less than .05 (PCLOSE) provide additional information on
how closely a model fits the data. The 90% confidence interval around the RMSEA
should contain .05 and PCLOSE should be nonsignificant at .05 level to indicate a close
fit to the data. The CFI ranges from 0 to 1, where 0 represents

the goodness of fit
associated with the null model and 1 represents

the goodness of fit associated with a
saturated model. A value above .95 suggests an acceptable fit of the data to the model.

Research Question 1. Mediation of self-perception in the relationship between
child maltreatment and social anxiety
James, Mulaik, and Brett (2006) articulated the testing of mediation in a
framework of SEM where X is a predictor, M is a mediator, and Y is an outcome. There
are two endogeneous variables, M and Y and two structural equations explaining the
relationship between X and M and between M and Y. In the case of full mediation, the
relationship between X and Y is set to 0. The partial mediation model differs from the
full mediation model by the addition of a direct path from X to Y. Although the

33

traditional method for testing mediation (Baron & Kenny, 1986; Holmbeck, 1997)
requires a significant relationship between predictor X and outcome Y as a necessary step
for establishing mediation, recently a number of studies have demonstrated that this
relationship is not necessary (James et al., 2006; MacKinnon, Lockwood, Hoffman, West,
& Sheets, 2002; Shrout & Bolger, 2002). In this study, X is the history of child
maltreatment, M is social self-perception and Y is social anxiety (either humiliation
anxiety or performance anxiety).
Figure 3 Test of Discretion of the Study Constructs

Note. HUM=humiliation anxiety; PERF=performance anxiety; SSP=social self-
perception

Longitudinal meditation of social self-perception in between child maltreatment
and social anxiety was tested with data collected at Time 1, Time 2 and Time 3. First of
all, it was examined whether three study constructs, humiliation anxiety, performance
anxiety and social self-perception are discrete from each other before testing a
meditational model (See, Figure 4). Then, the meditational models of social self-
perception between child maltreatment and social anxiety outcomes were tested. A cross-
lagged modeling strategy in the SEM framework fits all reasonable alternative models
and examines and compares goodness of fit. Therefore, the study tested two nested
models of mediation for each humiliation anxiety and performance anxiety (see Figure 4).

34

Model 1 tested the full meditation of the relationship between child maltreatment
experience, and the anxiety outcomes by social self-perception. Model 2 tested the partial
meditation of the relationship between child maltreatment experience, and the anxiety
outcomes by social self-perception. The direct paths from child maltreatment experience
to the anxiety outcomes and social self-perception at Time 2 and Time 3 were set to be
zero in the full meditation models and were freely estimated in the partial mediation
models. The model fit between the Model 1 and the Model 2 were examined to determine
the best-fitting model using the 1-degree-of-freedom nested chi-square difference test.
Although the main interest of the study was to test mediation of social self-
perception between child maltreatment and the social anxiety outcomes, it was expected
that there would be a reciprocal effect between social self-perception and social anxiety
over time. Therefore, cross-lagged effects between social self-perception and social
anxiety outcome across Time 1, 2 and 3 were also included in the models. The basic idea
of a cross-lagged modeling is to identify the causal priority
1
of the variables by
comparing the crossed and lagged paths between constructs. The crossed path coefficient
represents the longitudinal prediction of one construct by the other, above and beyond the
autoregressive prediction of that construct from the earlier measure of itself (Curran &
Bollen, 2001). In addition, both models included gender, interaction term between
maltreatment experience and gender, minority status and household income (as a proxy
for socioeconomic status) as covariates.

1
If X is said to be the effect of Y, establishment of causality requires; 1) X and Y must have temporal and
spatial continuity; 2) X must precede Y; and 3) There must, if X then Y for all occasion.

35

Figure 4 Analytical Model: Mediation of Social Self-Perception in the Relationship
between Child Maltreatment and Social Anxiety

Note. HUM=humiliation anxiety; PERF=performance anxiety; SSP=social self-
perception; T1= Time 1; T2=Time 2; T3=Time 3.

Research Question 2 & 3. Longitudinal Trajectories of Social Anxiety and Self-
Perception
This study utilized latent growth curve (LGC) analysis, applying structural
equation modeling techniques (SEM). As shown in our structural model (Figure 5), this
approach assumes that the repeated measures of a given construct are explained by two
common latent factors—the intercept and slope—which together represent the underlying
growth trajectories for each person across time (McArdle & Epstein 1987). Based on this
model, we can estimate the mean intercept ( μ
0
), or the average initial score at baseline,
and mean slope ( μ
s
), or the average rate of change in the study constructs over time or
age, as well as their respective variances ( σ
0
and σ
s
), which reflects the variability around
each value. The factor loadings reflect the relationship between the two latent factors and
the manifested scores of the study constructs for each time point. The loadings for the
intercept factor are always fixed at 1, while the loadings for the slope factor ( λ) can be

36

either fixed or estimated based on the data. The linear model fixes λ
9-16
to be 0 through 7.
More technical details of LGC modeling are available elsewhere, e.g. in Bollen and
Curran (2006), Duncan et al. (2006), Meredith and Tisak (1990), and McArdle (2004).
Social anxiety and social self-perception are expected to change systematically over age
during adolescence, and thus chronological ages of adolescents were used as the axis of
change instead of times at assessment (Time 1, 2, 3). Time 1, 2 and 3 data were re-sorted
by the age of the adolescents and employment of age as the axis of change allowed us to
explore the developmental trajectories without confounding effects of age. When the x
variable is age and the y variable is social anxiety, a change in social anxiety is explained
by the function of age, and age is considered to be the axis of change.

Figure 5 Analytical Model: Latent Growth Curve Model

Note. μ
0
=Mean of intercept (initial score at age 9); μ
s
= mean of slope (average rate of
change); λ[t] = fixed to be 0 through 8; σ
e
= variance of residual.

37

Second, a series of multiple-group analysis were conducted to compare the
baseline and rate of change in humiliation anxiety, performance anxiety and social self-
perception across the groups: the maltreated adolescents vs. the comparison adolescents.
To determine their significance, model parameters were restricted to be equal across the
two groups in the following successive order: 1) intercept; and 2) slope. Likelihood ratio
tests were used to assess whether the restricting of specific parameters as being equal
resulted in a significant worsening in the model goodness-of-fit. A statistically-significant
χ
2
difference would indicate that the equality constraint on a specific parameter was not
consistent with the data, suggesting that the parameter was significantly different across
the groups. The analysis was conducted using Mplus 4.0 (Muthén & Muthén, 2006).

Research Question 4. Longitudinal dynamic changes of social self-perception and
social anxiety
The third aim of the current study was achieved by applying latent difference
score modeling for the dynamics between social anxiety and social self-perception over
time (LDS; Ferrer & McArdle, 2004; McArdle, 2001; McArdle, Hamagami, Meredith,
Bradway, 2000). LDS modeling not only examines how development progresses in a
time-ordered manner but also investigates how two related psychological constructs are
intertwined over time (Ferrer & McArdle, 2004). The LDS modeling provide information
on which factor is a leading factor of the other. The Figure 5 is a visual description of
bivariate LDS modeling. According to the conceptualization by McArdle (2001), first,
LDS models define true scores by separating them from measurement errors based on a

38

classical true score model. The observed scores of two variables Y and X for person i at
time t are explained by combination of true score y and x and error e
y
and e
x
as
Y
it
= y
it
+ e
yit
and X
it
= x
it
+ e
xit.
And the true scores can now be used to express the current score of each variable as a
function of its previous score plus change as
Y
it
= y
it-1
+∆ y
it
and
X
it
= x
it-1
+∆ x
it.
From this basic equation of two measurement occasions, the trajectory equation of each
variable can be written as a linear accumulation of latent changes as
t
Y
it
= y
i0
+ ( ∑ ∆
yki
) + e
yit
and
k-1
t
X
it
= x
i0
+ ( ∑ ∆
xki
) + ex
xit,
k-1
and the model for the latent change scores can then be written as a function of several
components as
∆y
it
= α
y
· y
is
+ β
y
· y
it-1
+ γ
y
· x
it-1
+ ζ
∆yit
and
∆x
it
= α
x
· x
is
+ β
x
· x
it-1
+ γ
x
· y
it-1
+ ζ
∆xit.
α is the coefficient associated with the slopes y
is
and x
is
; β is a self-feedback
coefficient representing the effect of the same variable at the previous time on the change
of the variable; γ is the coupling coefficients, representing the effect of the other variable
measured at previous time on the change of the variable; ζ
∆
represents the unexplained
variance in the changes. Thus, according to this model, change in the true scores for each
variable ∆y
t
and ∆x
t,
is a function of three components: a constant slope α, the scores on

39

the same variable at the previous occasion β and the scores on the other variable at
previous occasion γ.
Before testing bivariate LDS models, univariate LDS model was estimated for
each study construct (i.e. humiliation anxiety, performance anxiety and social self-
perception) and then bivariate LDS dynamic models were applied for each pair of study
constructs: humiliation anxiety and social self-perception; performance anxiety and social
self-perception; and humiliation anxiety and performance anxiety. A series of nested
bivariate LDS models was tested in order to determine the importance of the coupling
parameters—whether dynamic influences between the two constructs have meaningful
contribution to the model and the inclusion of the coupling parameters significantly
improve the model fit. For each analysis, a bivariate LDS model started estimating a
model with no coupling parameter between two constructs, added either a coupling
parameter from x to y (y regressed on x) or from y to x (x regressed on y), and then
finally included both coupling parameters from x to y and from y to x.

40

Figure 6 Analytical Model: Bivariate Latent Difference Score Model

Note. This graph was the revised reproduction of the graph presented in Ferrer &
McArdle (2004). A path diagram of a bivariate latent difference score model. Y[t] and
X[t]=observed scores at age t; y[t] and x[t]=latent true scores at age t ; ∆y[t] and
∆x[t]=latent chances at age t; y
0
and x
0
= slopes; y
0
* and x
0
*=standardized slope scores;
triangle=constant (=1); α=slope parameter(=10); β=self-feedback parameter; γ=coupling
parameter; μ
y0
and μ
x0
=mean of initial level scores at age 9 (Age was centered at 9); μ
ys
and μ
xs
=mean of slope score; σ
y0
and σ
x0
=deviation of level scores; σ
ys
and σ
xs
=deviation
of slope scores; ρ
x0xs
and

ρ
y0ys
=correlation between level and slope; ρ
ysxs
=correlation
between slopes; ρ
xsy0
and

ρ
ysx0
=correlation between level and slope; σ
ey
and σ
ex
=variance
of residual scores; σ
eyex
=covariance between residual scores. All paths without a specified
parameter are fixed to 1.

41

CHAPTER 5: RESULTS
5.1 Preliminary Analysis
5.1.1 Confirmatory factor Analysis
Confirmatory factor analyses (CFAs) were performed to examine the factorial
validity of one and two factor model of social anxiety and a one factor model of social
self-perception at Time 1, Time 2, and Time 3. Separate CFAs were conducted to
validate for each social anxiety and social self-perception. Because the responses on the
MASC and the SPPC scales were categorical in nature measured on a 4-point Likert scale,
the weighted least squares mean and variance adjusted (WLSMV) estimator in Mplus 4.0
software (Muthén & Muthén, 2006) was used. The WLSMV is robust against the
violation of normality and can be used with ordered categorical data and smaller samples
sizes (Brown, 2006; Muthén & Muthén, 2006). Table 2 summarizes the model fit indices
of the one factor model and the two factor model of social anxiety. Both models fitted the
data very well but the two factor model provided slightly better fit over the one factor
model with lower chi-square and RMSEA. Table 3 shows the standardized factor
loadings for humiliation anxiety and performance anxiety factors at Time 1, Time 2, and
Time 3, and correlations between the factors. The items loaded on humiliation anxiety
factor with loadings ranging from 0.45 to 0.90 and on performance anxiety factor with
loadings ranging from 0.53 to 0.74 across different time points. The correlations
humiliation anxiety and performance anxiety factors at each time point ranged from 0.79
to 0. The correlations of humiliation anxiety factors between different time points and of
performance anxiety factors between different time points ranged from 0.39 to 0.51 and

42

from 0.56 to 0.61. Table 4 summarizes results for the confirmatory factor analysis of
social self-perception factors at Time 1, Time 2 and Time 3. The model fit the data very
well. Factor loadings across different time points ranged from 0.37 to 0.77 and the
correlation of the factor between time points ranged from 0.48-0.68.

Table 2 Confirmatory Factor Analyses of Humiliation Anxiety and Performance Anxiety

χ² Df
a
RMSEA TLI

One Factor Model of
Social Anxiety
234 102 0.05 0.97

Two factor Model of
Social Anxiety
(Humiliation and Performance Anxiety)
178 102 0.04 0.98
Note. Weighted Least Square Mean and Variance Adjusted (WLSMV) estimator was
utilized and the chi-square values cannot be used for the chi-square difference test

43

Table 3 Standardized Factor Loadings and Factor Correlation of Two Factor Model of
Social Anxiety (Humiliation and Performance Anxiety)
Items of
Social Anxiety
T1 HUM T2 HUM T3 HUM T1PERF T2 PERF T3 PERF
T1 HUM 1
0.62
T1 HUM 2 0.81
T1 HUM 3 0.77
T1 HUM 4 0.75
T1 HUM 5 0.48
T2 HUM 1 0.78
T2 HUM 2 0.83
T2 HUM 3 0.79
T2 HUM 4 0.77
T2 HUM 5 0.47
T3 HUM 1 0.79
T3 HUM 2 0.90
T3 HUM 3 0.80
T3 HUM 4 0.83
T3 HUM 5 0.41
T1 PERF 1 0.59
T1 PERF 2 0.61
T1 PERF 3 0.55
T1 PERF 4 0.67
T2 PERF 1 0.61
T2 PERF 2 0.53
T2 PERF 3 0.66
T2 PERF 4 0.73
T3 PERF 1 0.67
T3 PERF 2 0.69
T3 PERF 3 0.74
T3 PERF 4 0.72
Factor Correlations

T1HUM T2HUM T3HUM T1 PERF T2 PERF T3 PERF
T1 HUM 1
T2 HUM 0.51 1
T3 HUM 0.39 0.51 1
T1 PERF 0.87 0.51 0.52 1
T2 PERF 0.45 0.86 0.61 0.56 1
T3 PERF 0.33 0.38 0.79 0.56 0.61 1
Note. All factor loadings and correlations were significant at 0.01 level;
HUM=humiliation anxiety; PERF=performance anxiety

44

Table 4 Standardized Factor Loadings and Factor Correlations of Social Self-Perception
Factors
Items of Social Self-Perception
T1 SSP
Factor
T2 SSP
Factor
T3 SSP
Factor
Time 1 SSP 1 0.66
Time 1 SSP 2 0.70
Time 1 SSP 3 0.43
Time 1 SSP 4 0.63
Time 1 SSP 5 0.51
Time 2 SSP 1 0.66
Time 2 SSP 2 0.68
Time 2 SSP 3 0.37
Time 2 SSP 4 0.73
Time 2 SSP 5 0.59
Time 3 SSP 1 0.66
Time 3 SSP 2 0.75
Time 3 SSP 3 0.39
Time 3 SSP 4 0.77
Time 3 SSP 5 0.72
Factor Correlation

T1 SSP
Factor
T2 SSP
Factor
T3 SP Factor
Time 1 SSP Factor 1
Time 2 SSP Factor 0.64 1
Time 3 SSP Factor 0.48 0.68 1
Note. SSP= Social Self-Perception; χ²/df = 91/46 ; RMSEA=.05; TLI=.972; SSP=Social
Self-Perception; All factor loadings and correlations were significant at 0.01 level.

5.1.2. Descriptive
The meditational model was tested across assessment time points. However, the
LGC and LDS models were estimated based on adolescents’ age. Therefore, the data at
Time 1, Time 2, and Time 3 were re-sorted by adolescents’ age for the LGC and the LDS
models. Table 5 shows sample size, means, and standard deviations of all variables by
age. Table 6 and Figure 7 present the descriptive statistics by both the maltreatment and

45

the comparison groups. The mean scores on humiliation anxiety tended to decrease with
age in both the maltreatment and the comparison groups. Different pattern of change was
observed in performance anxiety. From age 9 through 13, performance anxiety decreased
in both groups. However, during the period between the ages 14 and 16, performance
anxiety increased for the comparison group, but decreased for the maltreatment group.
The mean scores on social self-perception tended to increase with age. Although a high
correlation between humiliation anxiety and performance anxiety was observed in the
confirmatory factor analysis, each construct was treated as a separate social anxiety
outcome because the examination of descriptive statistics suggested different
developmental pattern for each construct.

Table 5 Mean and Standard Deviation of Study Variables

Age 9
n=53
Age 10
n=166
Age 11
n=227
Age 12
n=266
Age 13
n=224
Age 14
n=131
Age 15
n=56
Age 16
n=17
Age 17
n=13
Humiliation
Anxiety
Mean (SD)
6.28
(3.87)
5.89
(4.08)
5.45
(4.02)
5.19
(4.00)
4.76
(3.83)
4.98
(3.86)
5.20
(3.76)
5.06
(3.65)
3.54
(3.21)
Performance
Anxiety
Mean (SD)
6.30
(2.98)
5.55
(3.25)
5.20
(3.30)
5.06
(3.19)
4.73
(3.13)
4.92
(3.10)
5.13
(3.22)
5.12
(3.55)
5.69
(3.68)
Social
Self-Perception
Mean (SD)
13.16
(3.15)
13.39
(3.62)
14.36
(3.56)
14.54
(3.79)
15.43
(3.28)
15.89
(3.11)
16.18
(3.26)
15.71
(2.99)
14.46
(3.33)

46
Table 6 Mean and Standard Deviation of Study Variables by the Maltreatment and the Comparison Groups

Note. HUM=humiliation anxiety; PERF=performance anxiety; SSP=social self-perception. 2 individuals at age 18 in the
maltreatment group were excluded since no 18 year-old was in the comparison group.

Study Variables Across Age
HUM 9 HUM 10 HUM 11 HUM 12 HUM 13 HUM 14 HUM 15 HUM 16 HUM 17
Comparison n 17 47 82 98 75 47 20 8 7
Mean 5.94 6.04 5.49 4.09 4.13 5.02 4.45 5.75 4.57
SD 4.18 3.86 3.68 3.83 3.34 4.15 2.91 3.77 3.41
Maltreated n 36 119 145 168 149 84 36 9 6
Mean 6.44 5.83 5.43 5.83 5.07 4.95 5.61 4.44 2.33
SD 3.76 4.18 4.21 3.98 4.03 3.72 4.14 3.64 2.73
PERF 9 PERF 10 PERF 11 PERF12 PERF 13 PERF 14 PERF 15 PERF 16 PERF 17
Comparison n 17 47 82 98 75 47 20 8 7
Mean 6.18 5.62 4.76 4.69 4.47 5.45 5.95 6.75 6.86
SD 2.24 2.82 3.11 3.24 3.05 3.22 2.46 3.01 2.04
Maltreated n 33 119 145 168 149 84 36 9 6
Mean 6.36 5.53 5.46 5.27 4.86 4.63 4.67 3.67 4.33
SD 3.33 3.42 3.39 3.16 3.18 3.01 3.52 3.50 4.84
SSP 9 SSP 10 SSP 11 SSP 12 SSP 13 SSP 14 SSP 15 SSP 16 SSP 17
Comparison n 16 47 82 98 75 47 20 8 7
Mean 12.56 13.77 15.07 15.12 15.59 15.66 16.50 15.13 13.57
SD 3.03 3.75 3.37 3.53 3.28 3.40 3.98 3.44 4.20
Maltreated n 33 118 144 167 149 84 35 9 6
Mean 13.45 13.23 13.95 14.20 15.35 16.02 16.00 16.22 15.50
SD 3.60 3.58 3.61 3.91 3.29 2.95 2.82 2.64 1.76

47

Figure 7 Mean Scores of Humiliation Anxiety, Performance Anxiety and Social Self-
Perception Across Age

48

5.2 Substantive Structural Analysis
5.2.1 Research Question 1. Social Self-Perception as a Mediator between Child
Maltreatment and Social Anxiety
The model designed to test whether three study constructs, humiliation anxiety,
performance anxiety and social self-perception are discrete from each other very poorly
fit the data ( χ²/df=188/25, CFI=.85 TLI=.78, RMSEA=.12 PCLOSE=0.00). This showed
that humiliation anxiety, performance anxiety and social self-perception are discrete
construct from each other and the study proceeded to test a mediation model of social
self-perception.
First, social self-perception was assessed as a mediator between child
maltreatment experience and humiliation anxiety. The partial mediational model fit the
data reasonably well ( χ²/df=13/4, TLI=0.77, CFI=0.98, RMSEA=0.07 PCLOSE=0.20, C.I.
of RMSEA=0.03 -0.11), but there was no significant direct effect of child maltreatment
experience on humiliation anxiety at Time 2 and Time 3. The full meditational model
provided slightly better model fit than the partial meditational model with more degree of
freedom ( χ²/df=13/6, TLI=0.87, CFI=0.99, RMSEA=0.05 PCLOSE=0.44 C.I. of
RMSEA=0.01-0.08). The results from the analyses of both models supported that social
self-perception mediated the relationship between child maltreatment experience and
humiliation anxiety and provided the same results regarding the relationships among the
study variables. Considering parsimoniousness of the model and degrees of freedom, the
full meditational model was selected for the final model. Figure 8 presents the results
from the full meditational model. The maltreated adolescents had a higher humiliation

49

anxiety and a lower social self-perception at Time 1 compared to the comparison group.
Significant effects of gender were found for humiliation anxiety at Time 1 and Time 3
and social self-perception at Time 1. Specifically, girls exhibited higher humiliation
anxiety and lower social self-perception. All autoregressive coefficients and correlations
between the residuals estimated in the model were statistically significant.
Child maltreatment experiences significantly and negatively predicted social self-
perception at Time 1 ( (b) = -0.19 (-1.5), p<0.01). Reciprocal dynamic relationships
existed between social self-perception and humiliation anxiety at Time 1 and at Time 2.
Social self-perception at Time 1 predicted humiliation anxiety at Time 2 ( (b) = -0.18 (-
0.19), p<0.01). Humiliation anxiety at Time 1 predicted social self-perception at Time 2
(( (b) = -0.1 (-0.19), p<0.01. Further, social self-perception at Time 2 had significant
effect on humiliation anxiety at Time 3 ( (b) = -0.18 (-0.22), p<0.01), but humiliation
anxiety at Time 2 did not affect social self-perception at Time 3.

50

Figure 8 Full Mediation of Social-Self-Perception between Child Maltreatment and
Humiliation Anxiety

Note. Model showing the results of the full mediation analysis for the total sample, only
significant standardized path coefficients were shown for simplicity. No significant effect
of CM*gender was found and this variable and the control variables are not shown for
simplicity; CM=child maltreatment experience; HUM=humiliation anxiety; SSP= social
self-perception. * p<.05, **p<.01

Second, the study tested social self-perception as a mediator between child
maltreatment experience and performance anxiety (see Figure 9). The partial mediation
model fit the data reasonably well ( χ²/df=15/4, TLI=0.71, CFI=0.98, RMSEA=0.08
PCLOSE=0.11 C.I. of RMSEA=0.04 -0.12) but no significant direct effect of child
maltreatment on performance anxiety at Time 2 and Time 3 was found. The full
meditational model also fit the data reasonably well, but with more degrees of freedom
than the partial meditational model ( χ²/df=17/6, TLI=0.80, CFI=0.98, RMSEA=0.07
PCLOSE=0.20 C.I. of RMSEA=0.03 - 0.10). Child maltreatment predicted performance
anxiety and social self-perception at Time 1. There was significant gender difference on
performance anxiety at Time 1 and Time 3 and social self-perception at Time 1. Girls
reported higher performance anxiety and lower social self-perception. Further, there was
a significant interaction effect between child maltreatment experience and gender in

51

predicting performance anxiety at Time 3. When post-hoc comparison between
the maltreated and the comparison boys and girls was conducted, mean score of
performance anxiety was higher for the comparison girls compared with the comparison
boys and the maltreated girls or boys. In addition, the adolescents from lower income
families exhibited lower social self-perception. All autoregressive path coefficients and
correlations of residuals were significant.
Regarding the dynamics between performance anxiety and social self-perception,
reciprocal influences were found between performance anxiety and social self-perception
at Time 1 and Time 2. Performance anxiety at Time 1 predicted social self-perception at
Time 2 and social self-perception at Time 1 predicted performance anxiety at time 2.
However, while performance anxiety at Time 2 had significant effect on social self-
perception at Time 3, social self-perception at Time 3 did not affect performance anxiety
at Time 2. The findings suggest that the mediation of social self-perception between child
maltreatment and performance anxiety was only supported in early adolescence.

52

Figure 9 Full Mediation of Social-Self-Perception between Child Maltreatment and
Performance Anxiety

Note. Model showing the results of the full mediation analysis excluding direct paths
from CM to T2PERF and T3PERF; Only significant standardized path coefficients were
shown for simplicity. Control variables are not shown for simplicity; CM=child
maltreatment experience; CM*Gender= interaction term between child maltreatment and
gender; PERF=performance anxiety; SSP= social self-perception. * p<.05, **p<.01

5.2.2 Research Question 2 & 3. Developmental Trajectories of Social Anxiety And Social
Self-Perception
First, the study estimated three separate growth trajectory for each humiliation
anxiety, performance anxiety, and social self-perception data on the total sample in order
to examine the developmental trend of social anxiety and social self-perception. Then, it
was tested whether child maltreatment experience and gender had a significant impact on
the intercept ( μ
0,
average baseline score at age 9) and the slope ( μ
s,
average rate of
change) of these three constructs by regressing the intercept and the slope on child
maltreatment experience and gender variables. As the last step of LGC, a series of
multiple-group LGC analyses was performed with each study construct in order to
explore whether the developmental trends differ between the maltreatment and the

53

comparison groups. All multi-group analyses started with the freely estimated means of
the intercept and the slope. In order to test whether the average level of humiliation
anxiety at age 9 was different between the maltreatment and the comparison groups, the
means of intercepts were additionally constrained to be equal across the groups. The
study also tested whether average rate of change was different between the groups by
additionally constraining the mean of slopes to be equal across the groups.
Latent Growth Curve Analysis of Humiliation anxiety (see Table 7). The
estimated linear growth trajectory model of humiliation anxiety fit data very well
( χ²/df=55/34, CFI=0.98, RMSEA=0.04). The estimated initial level of humiliation anxiety
at age 9 was 6.18 with significant individual variation in the initial score ( μ
0
=6.18,
p<0.01 σ
0
=7.00, p<0.01). There was a significant change in humiliation anxiety across
age in that humiliation anxiety tended to decrease with age ( μ
s
= -0.30, p<0.01), but the
variance of the slope was not significant ( σ
0
=0.16, p>0.05). It is suspected that
insignificance of the slope variance may be due to smaller number of cases in the side-
tails of the humiliation anxiety distribution (i.e. at age 9, 15, and 16). However, it could
not be ruled out that variation might exist in the central part of the distribution. Therefore,
despite of the insignificant slope variance, both the intercept and the slope of humiliation
anxiety were regressed on the child maltreatment experience, gender, and interaction
between child maltreatment experience and gender variables. Results of the analysis
showed significant gender effect on the intercept with girls experiencing higher
humiliation anxiety at age 9 than boys (b=0.23, p<0.05) but no significant effect of
gender on the slope were found. Neither child maltreatment experience nor the

54

interaction term between child maltreatment experience and gender predicted the
intercept and the slope of humiliation anxiety.
Multi-group analysis of LGC was conducted in the next step. Constraining either
the mean of the intercept factor or the slope factor separately did not significantly reduce
the model fit, indicating that the developmental trends in humiliation anxiety are similar
between the maltreatment and the comparison groups. The declining tendency in
humiliation anxiety with age was observed in both groups (see, Table 7). The mean of
initial score at age 9 were similar between the groups ( μ
0
= 6.10 for the comparison group,
μ
0
= 6.19 for the maltreatment group). The average decline in humiliation anxiety for the
maltreatment group ( μ
s
=-0.24, p<0.01) was less negative than the rate of change for the
comparison group ( μ
s
= -0.40, p<0.01), suggesting that level of humiliation anxiety
experienced by the maltreated adolescents over time may be higher compared to the
comparison adolescents. Figure 10 and Figure 11 depicts the developmental trajectories
of humiliation anxiety by the maltreatment group and the comparison group.

Figure 10 Intra-individual Trajectories of Humiliation Anxiety

55

Figure 11 The Estimated Growth Trajectory of Humiliation Anxiety by the Maltreatment
and the Comparison Groups
0
1
2
3
4
5
6
7
9 10 1112 131415 16
Age
Humiliation Anxiety
Maltreated Comparison

Latent Growth Curve Analysis of Performance Anxiety (see Table 8). The linear
LGC model of performance anxiety provided excellent fit to the data ( χ²/df=36/34,
CFI=0.99, RMSEA=0.01). The mean of the intercept factor ( μ
0
) was 5.78 (p<0.01) with
significant individual variations in the intercept score ( σ
0
=4.76, p<0.01). The trajectory of
performance anxiety estimated by LGC demonstrated that performance anxiety
significantly decreased with age ( μ
s
= -0.22, p<0.01). As observed in the analysis of
humiliation anxiety, variance in the intercept was significant but variance in the slope
was not significant. When the initial score and the rate of change were regressed on the
child maltreatment, gender, and the interaction between child maltreatment and gender
variables, gender had a significant effect on the intercept but insignificant effect on the
slope. Specifically, girls reported higher performance anxiety at age 9 (b=1.05, p<0.01).
Child maltreatment experience predicted neither the intercept nor the slope of
performance anxiety.

56

The initially estimated multi-group LGC of performance anxiety across the
maltreatment and the comparison groups included data from age 9 through age 16, but
this model did not converge due to lack of sufficient information to estimate missingness
in the data. The second model with the data at age 9 through age 15 (excluding data at
age 16) was successfully estimated but indicated non-positive definite in the covariance
matrix involving variance of the slope factor. For the model to convergence, insignificant
correlation between the slope and the intercept factors of performance anxiety was fixed
to be zero. When testing the difference in the intercept and the slope between the two
groups, constraining either the mean of the intercept or the slope did not significantly
reduce the model fit, indicating that the developmental trends in performance anxiety are
not significantly different between the maltreatment and the comparison groups.
However, further examination of the mean estimates suggested that developmental
pattern of performance anxiety may differ between the maltreatment group and the
comparison group (See Figure 12, Figure 13). The average initial score of performance
anxiety at age 9 was higher for the maltreated group ( μ
0
=6.04, p<0.01) compared with
the comparison group ( μ
0
=5.34, p<0.01).The mean of the slope was not significant for
the comparison group ( μ
s
= -0.11 p >0.05) but it was significant for the maltreatment
group ( μ
s
= -0.29, p<0.01). This indicated that performance anxiety significantly
decreased with age in the maltreatment group and the meaningful pattern of change in
performance anxiety was established only for the maltreatment group.

57

Figure 12 Intra-individual Trajectories of Performance Anxiety

Figure 13 The Estimated Growth Trajectory of Performance Anxiety by the Maltreatment
and the Comparison Groups
0
1
2
3
4
5
6
7
9 101112 131415 16 Age
Performance Anxiety
Maltreated Comparison

Latent Growth Curve Analysis of Social self-perception (see Table 9). The first
estimated LGC from age 9 through age 16 did not converge due to lack of sufficient
information to estimate incompleteness in the data. The LGC model with social self-
perception at age 9 through 15 successfully converged and provide very good mode fit ( χ²
(27) =46, CFI=0.99, RMSEA=0.04). The average score of social self-perception at age 9
was 12.73 (p<0.01) and social self-perception significantly increased with age ( μ
s
=0.65,

58

p<0.01). There is statistically significant individual variation in the initial score ( σ
0
=9.34,
p<0.01) but not in the rate of change. Results from the regression of the intercept and the
slope factors on the child maltreatment, gender and the interaction term of child
maltreatment and gender variables revealed the significant effects of child maltreatment
experience on the initial score of social self-perception at age 9 with the maltreated
adolescents reporting lower level of social self-perception than the comparison
adolescents (b= -1.37 p<0.01). No significant gender difference was found in predicting
the intercept or the slope of social self-perception.
Then, the study conducted a series of multi-group analyses of the growth
trajectory of social self-perception across the maltreatment group and the comparison
group. No statistically significant difference in the model fit was yielded when the equal
constraint was imposed on either the mean of the intercept or the slope. Despite the
insignificance difference, the comparison group reported more positive social self-
perception score at age 9 ( μ
0
=13.15, p<0.01) than the maltreatment group ( μ
0
=12.53,
p<0.01). Conversely, the maltreatment group showed more rapid increase in social self-
perception ( μ
s
=0.67, p<0.01) than the comparison group ( μ
s
=0.61, p<0.01) (See Figure
14 and Figure 15). However, when fixing the mean of intercept to be equal across the two
groups—setting the initial score at age 9 to be same across the groups ( μ
0
=12.74, p<0.01),
increasing rate of change in the social self-perception was higher for the comparison
group ( μ
s
=0.71, p<0.01) than the maltreatment group ( μ
s
=0.62, p<0.01).

59

Figure 14 Intra-individual Trajectories of Social Self-Perception

Figure 15 The Estimated Growth Trajectory of Social Self-Perception by the
Maltreatment and the Comparison Groups
0
2
4
6
8
10
12
14
16
18
20
9 10 1112 1314 1516
Age
Social Self‐Perception
Maltreated Comparison

60
Table 7 Results of the Latent Growth Curve Analyses of Humiliation Anxiety
χ² df ∆χ²/ ∆df RMSEA CI RMSEA μ
0
μ
s
σ
0
σ
s
ρ
0,S

Single group LGC 55 34 0.04 0.02-0.05
6.18**
(2.33)
-0.30**
(-0.76)
7.00*
(1.00)
0.16
(1.00)
-0.44
(-.42)
Multi-group LGC
Freed μ
0
and μ
s
95 68 0.04 0.02-0.06 - - - - -
Maltreated - - - - -
6.19**
(2.31)
-0.24**
(-0.55)
7.2**
(1.00)
0.19
(1.00)
-0.49
(-0.41)
Comparison - - - - -
6.10**
(2.31)
-0.40**
(-1.00)
6.96**
(1.00)
0.15
(1.00)
-0.51
(-0.50)
Fixed μ
0
to be equal 95 69 0/+1 0.04 0.02-0.06 - - - - -
Maltreated - - - - -
6.16**
(2.29)
-0.24**
(-0.54)
7.21**
(1.00)
1.90
(1.00)
-0.49
(-0.54)
Comparison - - - - -
6.16**
(2.34)
-0.41**
(-1.06)
6.95*
(1.00)
0.15
(1.00)
-0.51
(-0.50)
Fixed μ
s
to be equal 96 69 +1/+1 0.04 0.02-0.06 - - - - -
Maltreated - - - - -
6.34**
(2.37)
-0.30**
(-0.69)
7.16**
(1.00)
0.19
(1.00)
-0.47
(-0.41)
Comparison - - - - -
5.79**
(2.18)
-0.30**
(-0.80)
7.03**
(1.00)
0.14
(1.00)
-0.52
(-0.52)
Effects of Child Maltreatment and Gender on Intercept Mean and Slope Mean
Intercept Mean ( μ
0
) Slope Mean ( μ
s
)
Estimate Est/S.E. Estimate Est/S.E.
Child Maltreatment 0.48 (0.09) 0.87 -0.05 (-0.06) -0.33
Gender 1.19 (0.23) 2.30* -.0.04 (-0.05) -0.28
CM*Gender 0.90 (0.08) 0.82 -0.08 (-0.05) -0.26
Note. The single group model and the multi-group models included the data at age 9 through 15; RMSEA=Root Mean
Square Error of Approximation; C.I. RMSEA= 90% confidence interval of RMSEA; * p<0.05 ** p<0.01

61
Table 8 Results of the Latent Growth Curve Analyses of Performance Anxiety
χ² df ∆χ²/ ∆df RMSEA CI RMSEA μ
0
μ
s
σ
0
σ
s
ρ
0,S

Single group LGC
a
36 34 0.01 0.00-0.04
5.78**
(2.65)
-0.22**
(-0.60)
4.76**
(1.00)
0.13
(1.00)
-0.30**
(-0.30)
Multi-group LGC
b

Freed μ
0
and μ
s
93 56 0.05 0.03-0.07
Maltreated - - - - -
6.04**
(1.13)
-0.29**
(-1.29)
3.72**
(1.00)
0.05
(1.00)
0.00
c

Comparison - - - - -
5.34**
(2.65)
-0.11
(0.83)
4.05**
(1.00)
0.20
(1.00)
0.00
c

Fixed μ
0
to be equal 95 57 +2/+1 0.06 0.03-0.07
Maltreated
5.81**
(3.02)
-0.23**
(-1.05)
3.78**
(1.00)
0.05
(1.00)
0.00
c

Comparison
5.81**
(2.87)
-0.22**
(-1.47)
4.09**
(1.00)
0.02
(1.00)
0.00
c

Fixed μ
s
to be equal 95 57 +2/+1 0.05 0.03-0.07 - - - - -
Maltreated
5.87**
(3.03)
-0.23**
(-1.21)
3.40**
(1.00)
0.05
(1.00)
0.00
c

Comparison
5.67**
(2.86)
-0.23**
(-1.21)
3.93**
(1.00)
0.04
(1.00)
0.00
c

Effects of Child Maltreatment and Gender
Intercept Mean ( μ
0
) Slope Mean ( μ
s
)
Estimate Est/S.E. Estimate Est/S.E.
Child Maltreatment 0.77 (0.16) 1.76† -0.14 (-0.17) -1.10
Gender 1.05 (0.24) 2.51** -0.02 (-0.19) -0.19
CM*Gender -0.40 (-0.04) -0.45 0.28 (0.17) 0.28
Note. a= The model included data at age 9 through 16; b = The model included only data at age 9 through 15 for
convergence; c= ρ
0,S
was not significant and fixed to be zero for convergence; RMSEA=Root Mean Square Error of
Approximation; C.I. RMSEA= 90% confidence interval of RMSEA; †= p<0.1, **p<0.01.

62
Table 9 Results of the Latent Growth Curve Analyses of Social Self-Perception
χ² df ∆χ²/ ∆df RMSEA C.I. RMSEA μ
0
μ
s
σ
0
σ
s
ρ
0,S

Single group LGC
a
46 27 0.04 0.02-0.06
12.73**
(4.16)
0.65**
(1.31)
9.34**
(1.00)
0.25
(1.00)
-1.06
(-0.70)
Multi-group LGC
a

Freed μ
0
and μ
s
73 54 0.04 0.01-0.06
Maltreated
12.53**
(4.03)
0.67**
(1.48)
9.68**
(1.00)
0.20
(1.00)
-1.06
(-0.76)
Comparison
13.15**
(4.52)
0.61**
(1.08)
8.45**
(1.00)
0.32
(1.00)
-1.02
(-0.62)
Fixed μ
0
to be equal 75 55 +1/+1 0.01-0.06
Maltreated
12.74**
(4.12)
0.62**
(1.40)
9.55**
(1.00)
0.19
(1.00)
-1.03
(-0.76)
Comparison
12.74**
(4.33)
0.71**
(1.26)
8.56**
(1.00)
0.32
(1.00)
-1.03
(-0.62)
Fixed μ
s
to be equal 73 55 +1/+1 0.04 0.01-0.06
Maltreated
12.59**
(4.06)
0.65
(1.45)
9.62
(1.00)
0.20
(2.00)
-1.05
(-0.76)
Comparison
13.04**
(4.49)
0.65
(1.15)
8.46**
(1.00)
0.32
(1.00)
-1.02
(-0.62)
Effects of Child Maltreatment and Gender
Intercept Mean ( μ
0
) Slope Mean ( μ
s
)
Estimate Est/S.E. Estimate Est/S.E.
Child Maltreatment -1.37 (-0.21) -2.57** 0.17 (0.15) 1.18
Gender -0.98 (-1.56) 1.93† 0.26 (0.24) 1.86†
CM*Gender -.0.58 (-0.04) -0.54 0.12 (0.05) 0.43
Note. a= The models included only the data at age 9 through 15 for convergence; RMSEA=Root Mean Square Error of
Approximation; C.I. RMSEA= 90% confidence interval of RMSEA; †= p< 0.10, ** p<0.01
63

5.2.3. Research Question 4: Longitudinal Dynamic Changes between Social Anxiety
(Humiliation Anxiety and Performance Anxiety) and Social Self-Perception

Univariate Latent Difference Score Model. In univariate LDS model, latent
change is explained by a function of a constant trend ( α, a mean of slope μ
s
) and the
effect of previous state ( β, proportional change) as
Δy
[t]
= α
y
+ β
y
y
[t-1].
The univariate LDS models of humiliation anxiety, performance anxiety, and social self-
perception provided excellent fit to data. Table 10 summarizes the parameter estimates
and fit of the models. Results of univariate LDS model of humiliation anxiety showed
significant change over time with a significant constant slope ( μ
s
=1.88, p<0.01) and a
significant effect of humiliation anxiety score at a previous time ( β= -0.41, p<0.01). This
shows significant decline in humiliation anxiety score over age. Therefore, change in
humiliation anxiety is defined as
ΔHA
[t]
=1.88 ± [1.31] -0.41 · HA
[t-1].
Application of univariate LDS model to performance anxiety also indicated its significant
change over time with a significant constant slope ( μ
s
=2.10, p<0.01) and significant
effect of humiliation anxiety score at a previous time ( β= -0.45, p<0.01). As seen in the
analysis of humiliation anxiety, this shows significant decline in performance anxiety
score over age. Therefore, change in performance anxiety is expressed as
ΔPA
[t]
=2.10 ± [1.03] - 0.45 · PA
[t-1].

64

However, the results from the univariate model of social self-perception showed no
significant constant change ( μ
s
=1.40, p>0.05).The effect of previous social self-
perception state on its change was also not significant ( β= -0.05, p>0.05). In other words,
a score at previous time did not systematically influence change in social self-perception.
Although there was no significant change in social self-perception, the study
proceeded to test bivariate LDS models in order to examine whether social self-
perception predicted change in humiliation anxiety or performance anxiety and whether
either of the social anxiety outcomes predicted change in social self-perception.

Table 10 Results of the Univariate Latent Difference Score Models of Humiliation
Anxiety, Performance Anxiety and Social Self-Perception
Parameters &
Fit Indices
Humiliation
Anxiety
Performance
Anxiety
Social
Self-Perception
Proportional β -0.41** -0.45** -0.05
Initial Mean μ
0
6.88** 6.45** 12.60**
Slope Mean μ
s
( α)

1.88** 2.10** 1.40
Initial Deviation σ
0
8.06** 4.58* 9.74**
Slope Deviation σ
s
1.31† 1.03† 0.23†
Correlation ρ
0,S
1.13 1.26† -0.77
Error Deviation σ
e
9.14** 5.73** 6.55**
χ²/df 48/33 28/33 46/26
RMSEA ε
a
0.03 0.00 0.04
Note. †= p<0.1, *p<0.05, **p<0.01.

65

Bivariate Latent Difference Score Model. A series of nested bivariate LDS models
were tested for each pair of social anxiety and social self-perception dynamic:
humiliation anxiety and social self-perception; performance anxiety and social self-
perception; and humiliation anxiety and performance anxiety. For each analysis, a
bivariate LDS model started by estimating a model with no coupling parameter between
two constructs. Either a coupling parameter from x to y (y regressed on x) or from y to x
(x regressed on y) was then added. Finally, both coupling parameters from x to y and
from y to x were included.
Humiliation anxiety (y)and social self-perception (x). The bivariate LDS model
with no coupling between humiliation anxiety and social self-perception fit the data very
well ( χ²/df=121/92. RMSEA=0.03, PCLOSE=1.00). Then, the LDS model added a
coupling effect of social self-perception in predicting change in humiliation anxiety. This
model also showed an excellent fit to the data ( χ²/df=120/91. RMSEA=0.03,
PCLOSE=1.00) but there was no significant improvement of model fit ( Δχ²/ Δdf = -1/-1).
However, neither the LDS model with a coupling parameter predicting change in social
self-perception by humiliation anxiety at a previous time nor the model with both
coupling parameters properly converged. This non-convergence was interpreted as
improper model specification regarding a coupling effect of humiliation anxiety on
change in social self-perception. Also, it is very important to note that univariate latent
difference score model of social self-perception did not show any significant pattern of
change, therefore it may not be appropriate to test coupling effect of social anxiety
outcomes to predict change in social self-perception. Therefore, no coupling effect of

66

humiliation anxiety on change in social self-perception was assumed in the study. Table
12 summarizes the parameter estimates and model fit indices from the bivariate LDS
model with a coupling effect of social self-perception on change in humiliation anxiety (a
coupling effect of humiliation anxiety on social self-perception set to be zero). An initial
mean for humiliation anxiety was 6.72 (p<0.01) and for social self-perception was 12.54
(p<0.01), which represent the average starting points at age 9 with substantial individual
variation around both means ( σ
y0
= 7.30, σ
x0
=9.99). The growth estimates indicate that
change in humiliation anxiety was significantly influenced by negative auto-proportion
(self-feedback from its previous state; β= -0.58). The coupling parameter predicting
change in humiliation anxiety was negative, but not statistically significant ( γ = -0.06, p
>0.05) and suggested that social self-perception does not have a significant influence in
the development of humiliation anxiety (See Figure 16).

Table 11 Model Fit Indices of the Bivariate Latent Difference Score Models
χ²/df ∆χ²/ ∆df RMSEA PCLOSE
Dynamics between Humiliation Anxiety (y) and Social Self-Perception (x)
No coupling ( γ) 121/92 reference 0.03 1.00
γ (x →y) only 120/91 -1/-1 0.03 1.00
γ (y →x) only No convergence - - -
Full Model No convergence - - -
Dynamics between Performance Anxiety (z) and Social Self-Perception (x)
No coupling ( γ) 122/92 reference 0.02 1.00
γ (x →z) only 108/91 -14/-1** 0.02 1.00
γ (z →x) only No convergence - - -
Full Model No convergence - - -
Note. **p<0.01.

67

Table 12 Results of the Bivariate Latent Difference Score Model of Humiliation Anxiety
and Social Self-Perception (with x →y coupling model)
Parameter & Fit Indices
Social Self-Perception
(x)

Humiliation Anxiety
(y)
Proportional β -0.06

-0.58**
Coupling γ -0.21
(x→y)

0.00
a

(y→x)
Initial Mean μ
0
a
12.54**

6.72**
Slope Mean μ
s
1.60

5.74
Initial Deviation σ
0
9.99**

7.30*
Slope Deviation σ
s
0.26

1.76
Error Deviation σ
e
6.51**

8.89**
Correlation:
Intercept & Slope ( ρ
x0,xS,
ρ
y0,yS
)
-0.77

1.36
Correlations between x and y
ρ
x0,y0
-2.18
ρ
xs,ys
-0.13
ρ
xs,y0
0.67
ρ
ys,x0
-0.95
ρ
ex
,
ey
-0.99**
Note.All parameters are raw maximum likelihood estimates fitted using Mplus4.0; Initial
Mean indicates score at Age 9; a= fixed to be 0;*p<0.05 **p<0.01.

68

Figure 16 Results of the Bivariate Latent Difference Score Model of Humiliation Anxiety
and Social Self-Perception

69

Performance anxiety (z) and social self-perception (x). A series of nested
bivariate LDS models was estimated with performance anxiety and social self-perception
data (see Table 11). A model with no coupling had excellent fit to the data ( χ²/df=122/92.
RMSEA=0.02, PCLOSE=1.00). When the LDS model added a coupling effect of social
self-perception on change in performance anxiety, the model provided superior fit over
the model with no coupling ( χ²/df=108/92. RMSEA=0.02, PCLOSE=1.00) and
substantial improvement in model fit was observed ( Δχ²/ Δdf = -14/-1). This improvement
clearly suggested significant influence of social self-perception on change in performance
anxiety over age. As observed in the analyses of humiliation anxiety, neither a model
with a coupling parameter predicting change in social self-perception nor a model with
both coupling parameters properly converged. This non-convergence was also interpreted
as improper model specification and no coupling effect of humiliation anxiety on social
self-perception was assumed.
Table 13 and Figure 17 summarize the parameter estimates and model fit indices
of a bivariate model with a coupling effect of social self-perception on change in
performance anxiety. An initial mean for performance anxiety and social self-perception
were 6.10 (p<0.01) and 12.40 (p<0.01) respectively, with substantial individual variation
around the means ( σ
z0
= 6.37, σ
x0
=10.32). The growth estimate indicated that change in
performance anxiety was influenced by a positive constant slope ( μ
zs
= 2.44, p<0.01), a
negative auto-proportion (self-feedback from its previous state; β
z
= -1.75, p<0.05), and a
negative coupling parameter ( γ = -0.68, p<0.01).
ΔPA
[t]
=2.44 ± [10.32] -1.75 · PA
[t-1].
-0.68 · SSP
[t-1].

70

Positive social self-perception tends to decrease the level of performance anxiety
over time and acts as a protective factor. This finding suggested that social self-
perception may be a leading factor influencing the development of performance anxiety
and supported a cognitive model of social anxiety.

Table 13 Results of the Bivariate Latent Difference Score Model of Performance Anxiety
and Social Self-Perception (with x →z coupling model)
Parameter & Fit Indices
Social Self-Perception
(x)

Performance Anxiety
(z)
Proportional β -0.12 -1.75*
Coupling γ -0.68**
(x →z)
0.00
a

(z →x)
Initial Mean μ
0
a
12.40** 6.10**
Slope Mean μ
s
18.73** 2.44**
Initial Deviation σ
0
10.32** 6.37**
Slope Deviation σ
s
0.29* 10.32
Error Deviation σ
e
6.50** 5.5**
Correlation:
Intercept & Slope ( ρ
x0,xS,
ρ
z0,zS
)
1.73 15.04*
Correlations: x and z
ρ
x0,z0
-5.99**
ρ
xs,zs
-40.10
ρ
xs,z0
-13.30
ρ
zs,x0
-2.31
ρ
ex
,
ez
-0.67**
Note. All parameters are raw maximum likelihood estimates fitted using Mplus4.0; Initial
Mean indicates score at Age 9; a= fixed to be 0; *p<0.05 **p<0.01.

71

Figure 17 Results of the Bivariate Latent Difference Score Model of Performance
Anxiety and Social Self-Perception

5.3 Summary of Findings
Research Question 1. Does self-perception mediate relationship between child
maltreatment experience and social anxiety?
The investigation of the partial and full meditational models based on a
longitudinal cross-lagged design supported full meditational model over the partial
mediation model. The results supported the full mediation of social self-perception for
both humiliation anxiety and performance anxiety outcomes between Time 1 and Time 2.
The child maltreatment experience was associated with a lower social self-perception at
Time 1, which in turn predicted higher humiliation anxiety and performance anxiety at
Time 2. Further, both humiliation anxiety and performance anxiety at Time 1 had
significant effect on social self-perception at Time 2, showing reciprocal influences

72

between the social anxiety and self-perception. However, the mediation of self-perception
was established only for humiliation anxiety between Time 2 and Time 3 but not for
performance anxiety.
Research Question 2. What are developmental trends found in humiliation anxiety,
performance anxiety and social self-perception over age?
The result of LGC analyses indicated that significant patterns of change in
humiliation anxiety, performance anxiety and social self-perception over age.
Humiliation anxiety and performance anxiety increased with age and social self-
perception decreased with age.
Research Question 3. Does child maltreatment experience have significant impact
on the developmental trajectories of social anxiety and self-perception?
No statistically significant difference was observed in the baseline score of and
rate of change in humiliation anxiety and performance anxiety between the maltreatment
and the comparison groups. However, substantial difference between the maltreated and
the nonmaltreated comparison adolescents existed in the baseline score of social self-
perception at age nine but the two groups did not differ regarding its rate of change. The
maltreatment group reported significantly lower social self-perception at age nine than
did the comparison adolescents. Nonetheless, results from the multi-group LGC analysis
of performance anxiety revealed that significant pattern of change is observed only in the
maltreatment group and no significant pattern of change existed in the comparison group.
Research Question 4. What are longitudinal dynamic mechanisms between social
anxiety and self-perception?

73

There was no coupling effect between humiliation anxiety and social self-
perception and social self-perception did not predict change in humiliation anxiety over
time. The significant coupling effects of social self-perception on change in performance
anxiety showed that social self-perception significantly influence the change in
performance anxiety over time. The results suggest that social self-perception is a leading
factor influencing change in performance anxiety.

74

CHAPTER 6: DISCUSSION
There are several issues to be considered when interpreting the results presented
here. One limitation is the possible confounding effect of age or assessment time on the
results of the study. The mediational analyses were modeled with data across three
assessment points (i.e. Time 1, Time 2 and Time 3). Although age was included as a
covariate to control its effect on social anxiety and social self-perception, confounding
effect of age may still be present in the relationships between the social anxiety outcomes
and social self-perception. Different from the meditational analyses, the latent growth
curve, the univariate, and the multivariate latent difference score models were estimated
with age as the axis of change, which may be the most appropriate approach to
understand the development of social anxiety and self-perceptions of adolescents over
time. However, adolescents might be more anxious when they visited the laboratory for
the first time, and their anxiety levels decreased over time with multiple visits to the
laboratory. Similarly, testing or practice effect may have partially contributed to declines
in the anxiety scores. Therefore, the results of this study should be cautiously interpreted.
The second limitation of the study was a smaller number of cases in two-side tails
of the data distribution (i.e. age 9, 15 and 16), and smaller sample size in the comparison
group. This limited sample size in the tails may have contributed to insignificant
individual variation in the slope (a rate of change) estimated by latent growth curve
analyses. Smaller sample size in the comparison group may have limited the power to
detect differences between the maltreated adolescents and comparison adolescents.
Moreover, in the growth curve analyses, the number of variables predicting the intercept

75

and the slope was restricted due to relatively small sample size to the number of
parameters to be estimated (when a number of predictor increased, the models did not
properly converged).
Third, there was also substantial attrition at Time 2 and 3 of the study. The
majority of those who did not return for Time 2 and Time 3 assessments were in the
maltreated group. Because this group may be at an increased risk for mental health
problems, we may be losing some variance at the higher end of the social anxiety or
lower end of social self-perception outcomes. However, due to the family disruption and
poor family functioning of maltreated children, it is extremely difficult to keep track of
their whereabouts. Since many of the maltreated adolescents did not return for Time 3,
there may be an underestimation of social anxiety and an overestimation of social self-
perception in this sample either at later time points or at older age.
Fourth, the present study focused on broader child maltreatment experience
including neglect, physical abuse, sexual abuse, and emotional abuse together. Therefore,
the results should not be generalized to a specific type of child maltreatment. Future
research should further explore unique effects of different types of child maltreatment
practices on the development of social anxiety and self-perception as well as the
dynamics between these constructs. In addition, the study only used the Multidimesional
Anxiety Scale for Children (March et al., 1997) to measure social anxiety and the Self-
Perception Profile for Adolescents (Harter, 1985) to assess the self-perception construct.
Therefore, the interpretation of the results may be restricted to these specific measures.

76

Despite these limitations, our study yields important findings. The purpose of the
present study was to advance our understanding of the development of social anxiety and
social self-perception in maltreated and nonmaltreated comparison adolescents, and to
provide prospective empirical evidence supporting a cognitive model of social anxiety in
relations to social self-perception.
Research Question 1. The first aim was to examine whether social self-perception
mediates the relationship between child maltreatment experience and social anxiety. The
results supported the mediation of social self-perception for both humiliation anxiety and
performance anxiety outcomes between Time 1 and Time 2. However, the mediation of
self-perception was established only for humiliation anxiety between Time 2 and Time 3
but not for performance anxiety. Bosquet and Egeland (2006) designed a longitudinal
study that followed participants from infancy through adolescence in order to examine
the relationship between social competence and anxiety symptoms in a high-risk
community sample. Social incompetence in childhood predicted anxiety symptoms in
preadolescence, conversely, anxiety symptoms in preadolescence predicted social
incompetence in adolescence. Considering the results of both the present study and the
Bosquet and Egeland’s study, research on the relationship between social anxiety and
self-perception suggests that developmental mechanism between social anxiety and social
self-perception may differ across developmental stages. In other words, as discussed by
Bosquet and Egeland, developmental stage may moderate the relationships between
social competence and psychopathological symptoms. However, there is not sufficient
empirical evidence that would determine how the meditational mechanism differs across

77

developmental stages. Therefore, a methodologically sound longitudinal investigation of
social anxiety and self-perception mechanisms across developmental span is warranted.
Moreover, it should be further explored how differently social self-perception play a role
between humiliation anxiety development and performance anxiety development and
whether developmental stage explain the difference.
Many of the previous studies on the anxiety in maltreated children and
adolescents have used broad definitions of anxiety and have not differentiated specific
types of anxiety such as social anxiety. This study is one of the few studies that
investigated the impact of child maltreatment on social anxiety longitudinally including
both humiliation anxiety and performance anxiety. Results of the meditational analyses
revealed a significant main effect of child maltreatment on performance anxiety and
social self-perception at Time 1 and indirect effect of child maltreatment on the
humiliation anxiety through social self-perception at Time 1. This finding demonstrates
the critical influence of child maltreatment experience on the maltreated adolescents in
early to middle adolescence. In general, social interaction and participation in social
activities become more frequently expected and encouraged in early to middle
adolescence. Particularly, adolescents often need to take an initiative to build their own
independent peer relationships without parents’ involvement. In this sense, social anxiety
in early adolescence may have a profound influence on adolescents’ lives.
Substantial associations between social anxiety scores at different times (auto-
regressive paths) were also observed, and the significant associations suggest the stability
of social anxiety symptoms over time. Stability of social anxiety symptoms has been

78

documented in the literature (Beidel, Fink, & Turner, 1996). The present study also
supported stability of social anxiety symptoms over time. However, the support from the
present study may be limited to because it only shows moderate associations between
social anxiety scores between different times (correlations ranging from 0.35 to 0.51).
Another important finding from the meditational analyses was significant gender
difference in humiliation anxiety at time 1 and performance anxiety at Time 3.
Specifically, girls exhibited higher humiliation anxiety and performance anxiety than did
boys. The previous literature has pointed to gender difference in the prevalence of anxiety
and suggested that girls are at a greater risk for anxiety disorder (Gater, Tansella, Korten,
Tienmens, Mavreas, & Olatawura, 1998). Therefore, it is important to consider gender
differences when studying development of anxiety disorder in adolescence. Ohannessian,
Lerner, Lerner, and von Eye (1999) examined whether self-competence predicts the
emergence of gender differences in depression and anxiety in a sample of six- and seven-
graders. The study reported that boys had significantly higher levels of self-competence
than did girls. In addition, boys were significantly less anxious compared to girls in 7
th
-
grade, but not in 6
th
-grade. Finally, the results from their regression analyses support the
hypothesis that self-competence is partially responsible for the emergence of gender
differences in depression and anxiety during early adolescence. Further research is
needed to understand whether there is gender difference in developmental mechanism of
social anxiety in adolescence.
Furthermore, the results of the meditational analyses indicated significant
interaction effects between child maltreatment and gender on performance anxiety at

79

Time 3. Given the significant effects of gender on the performance anxiety at Time 3, the
significant interaction effect indicates that child maltreatment experience moderated the
relationship between gender and performance anxiety, with the comparison girls
reporting higher performance anxiety than the maltreated girls. The possible explanation
for the reason why the comparison girls reported higher performance anxiety is
demonstrated within the section discussing research question 3.
Research Question 2. The purpose of the study was to explore the developmental
trajectories of humiliation anxiety, performance anxiety, and social self-perception using
latent growth curve modeling. This is one of the few studies providing empirical evidence
on the growth pattern of humiliation anxiety and performance anxiety across
developmental stages. The anxiety in childhood and adolescence has been studied in the
past, but with the recent advancement of statistical software to estimate the
developmental trajectory, research has just begun to address developmental trends in
anxiety across different developmental stages. To our best knowledge, there are only few
studies that examined adolescent developmental trajectories of anxiety symptoms in a
longitudinal design (Cole, Martin, Peeke, Seroczynski, & Fier, 1999; Hale III et al., 2008).
Hale and colleagues studied the developmental trajectories of various anxiety disorders
(i.e. generalized anxiety disorder, social anxiety disorder, separation anxiety disorder,
panic disorder, school anxiety) in a large community sample of adolescents over the
course of 5 years. They reported that social anxiety disorder symptoms decreased over
time, and this decrease was significantly stronger for the early adolescent cohorts.
Consistent with the findings in the study by Hale et al. (2008), the current study also

80

found that social anxiety symptoms significantly decreased. In addition, a positive
increase in social self-perception across age is reported in the current study in agreement
with findings from a previous study that has examined the growth trajectories of self-
perception. Cole et al. (2001) investigated the developmental trajectories of self-
perceptions in different developmental domains measured by the Self-Perception Profile
for Adolescents (Harter, 1988) in a sample of adolescents aged 8 to 16. Overall, the study
reported significant increase in social self-perception from middle to late childhood
through early to middle adolescence. However, no significant pattern of change was
found during the middle to late adolescence period, and this research finding suggests
possible moderating effects of developmental stage in the development of social self-
perception. Despite the evidence that demonstrates a decrease in social anxiety and an
increase in social self-perception across age, there is still not enough empirical evidence
to document the increase and decrease of social anxiety or social self-perception in
adolescence. Therefore, future research should further explore the developmental course
of social anxiety and social self-perception, and replicate the findings from this study.
Research Question 3. There are several plausible explanations for non-significant
effect of child maltreatment experience on the baseline score or the rate of change in
humiliation anxiety and performance anxiety. First, variability in child maltreatment
experiences (e.g. severity, duration) may have contributed to the finding. As discussed
previously, the majority of the maltreated adolescents in the current sample experienced
multiple forms of child maltreatment that varied in the nature, onset, chronicity of child
maltreatment, all of which may have differential effects on the development of social

81

anxiety and social self-perception. Further, it cannot be ruled out that there may be
subgroups whose social anxiety or social self-perception get worsen as they grow older.
Future research should continue to explore variability and heterogeneity in the social
anxiety and self-perception outcomes of the maltreated adolescents. Application of
advanced statistical methods such as growth mixture modeling would contribute to
identify unique subgroups of the maltreated adolescents with regard to development of
social anxiety and self-perception. Most importantly, it should be noted that smaller
sample size in the tails of the data distribution may have resulted in non-significant
individual variations in the slope parameters and relatively smaller sample size in the
comparison group may have limited our power to detect the group differences in the
intercept of social anxiety outcomes and the slopes of the social anxiety outcomes and
social self-perception. Despite the insignificant differences between the groups, average
rate of change estimated by multi-group analyses of latent growth curve models depicted
that declining rate of humiliation anxiety in the maltreatment group is smaller than the
comparison group. Based on this simple comparison of mean estimates, there is a
possibility that the level of humiliation anxiety in the maltreatment group may stay higher
than in the comparison group, and in turn, the gaps between the maltreatment and the
comparison group may become larger across age groups. Future research should continue
to explore a developmental course of humiliation anxiety with a larger sample and more
power in order to determine this difference between the maltreated and nonmaltreated
adolescents.

82

In the multi-group analyses of the LGC of performance anxiety, an interesting
difference was observed with regard to change in performance anxiety. Decline in the
level of performance anxiety over age was only found in the maltreatment group and no
significant pattern of change in performance anxiety existed in the comparison group.
This insignificant pattern of change may be associated with the interaction effect between
child maltreatment and gender on the performance anxiety at Time 3 observed in the
meditational analyses. As noted above, girls in the comparison group tended to report
higher performance anxiety at Time 3 and this may create heterogeneity in the
developmental trend in performance anxiety with insignificant mean of slope. Cole et al.
(1999) investigated a longitudinal developmental trajectory of anxiety in a sample of
normal boys and girls in a school setting. This study reported that the level of anxiety
tends to decrease for boys but to increase for girls with age. According to a traditional
inverted U-curve relationship between anxiety and performance (Yerkes & Dodson,
1908), some anxiety may be beneficial to performance only to a certain point after which
anxiety begins to undermine performance, in particular, this may be more true for girls. It
would be important to explore how the elevated level of performance anxiety in the
comparison girls is associated with their academic performance to determine whether this
indicates positive or negative implication.
Research question 4. The last goal of the study was to understand developmental
processes and dynamics between social anxiety outcomes and cognitive social self-
perception by applying a bivariate latent difference score modeling. First, there was no
significant dynamic prediction between humiliation anxiety and social self-perception.

83

The direction of a coupling parameter predicting humiliation anxiety by social self-
perception was negative, which was expected, but the parameter was not statistically
significant. Unfortunately, this finding is not in line with our previous analyses, which
examined cross-lagged effects between humiliation anxiety and social self-perception
over three assessment times. As described above, the results of mediational analyses
showed that social self-perception at Time 1 and Time 2 predicted humiliation anxiety at
Time 2 and Time 3. One possible explanation for this discrepancy is that the bivariate
latent difference score analysis estimates only one equal coupling parameter representing
all dynamics at different time points, thus, if there is heterogeneity in the dynamic
relationships across time, the coupling parameter can be insignificant. As noted above,
dynamic relationship between social anxiety and social self-perception may vary at
different developmental stages and it is important to investigate whether dynamics
between humiliation anxiety and self-perception are moderated by developmental stages.
However, given the methodological limitation, the present study is limited in
comprehensibly capturing heterogeneity in dynamic relationships between humiliation
anxiety and social perception at different developmental stages.
Perhaps the most important of our findings is significant effect of social self-
perception on the change in performance anxiety, demonstrating positive (higher) social
self-perception significantly reduces the level of performance anxiety over time.
Conversely, previous state of performance anxiety did not affect change in social self-
perception over time. This finding shows that the dynamics between performance anxiety
and social self-perception are not symbiotic and highlight a critical role played by social

84

self-perception in the development of performance anxiety. This finding is consistent
with hypothesized relationship based on a cognitive model of social anxiety and
demonstrates self-perception as a resilience or protective factor that decreases the level of
performance anxiety.

Research Implications
Theoretical consideration. The current study framed self-perception as a mediator
between child maltreatment experience and social anxiety. Although the findings of the
current study supported a cognitive model of social anxiety, there has been an important
theoretical debate regarding the role of self-perception regarding role played by self-
perception; the debate is on a question whether it acts as a mediator, a moderator, or both.
Self-perception has often been identified as a mediator in the literature and there are
many empirical and theoretical research studies supporting it. However, meditational and
moderational processes are not necessarily mutually exclusive, hence some studies tested
both mediation and moderation together (Tram & Cole, 2000), while other studies tested
self-perception as either mediator (Hirsch et al., 2003) or moderator (Cole e al. 2008).
Overall, the findings from the present study support a meditational role of self-perception
in the development of social anxiety. However, discrepancy between findings from the
meditational analyses and the latent difference score analyses raises an issue whether
self-perception acts as a moderator under certain conditions or at certain developmental
stage. This discrepancy between findings may be better explained when a meditational
causal role and a moderational buffering role are further explored in future research. This

85

investigation would advance our understanding of etiological mechanisms underlying
social anxiety in adolescence.
Second, in order to understand better the etiology of social anxiety, we must
examine diverse aspects of self-perception (i.e. general self-worth, academic self-
perception, physical appearance self-perception). It would be useful to determine if there
were particular types of self-perception linked to social anxiety that were more persistent
or led to secondary mental disorder such as depression. In addition, critical contextual
experiences can alter the mechanisms that underlie the association between social anxiety
and self-perception. Thus, it is necessary to examine the environment and other
individual experiences when examining the relationship between child maltreatment, self-
perception, and social anxiety in order to understand fully the complexity of these
relationships and to identify factors that may contribute to maintaining the stability of
social anxiety symptoms.
Methodological Implications. The most important methodological implication of
the current study is that it offers developmental perspective within social work research.
Developmental sequences and dynamics in human development have been of major
interests within social science research. Despite of the strong interest, previous research
has failed to explore the dynamic mechanisms in child development with
methodologically sound approach (Ferrer et al., 2007). Latent difference score dynamic
modeling is one of few methods that can longitudinally examine the change in a
psychological construct, and causal dynamics between two constructs simultaneously.
Application of this new approach would provide a better way to depict dynamic and

86

interactive nature of child development and significantly contribute to further
advancement of theories of child development and etiology of psychopathology.

Social work practice implication
Social anxiety in the maltreated children has not been understood well, although,
anxiety may be even more common than depression during adolescence. More clinical
attention should be paid to anxiety disorder symptoms in the maltreated children and
adolescents. Although a cognitive model of social anxiety has been a dominant
theoretical framework for the treatment of anxiety, there has been surprisingly little
evidence on social anxiety in childhood and adolescence. Results of our study indicate a
critical role played by individual’s self-perception in the development of social anxiety
and provide prospective empirical evidences supporting a cognitive model of social
anxiety. Hence, this study makes substantial contribution to advancement of etiological
theory of social anxiety in adolescents and eventually guides the development and
provision of effective interventions.
Of particular interest is the finding that social self-perception plays a key role in
protecting maltreated adolescents from social anxiety and promoting resilience among
them. Specifically, given the important role of self-perception in the development of
psychopathology, it is critical to develop an assessment strategy to better understand
profiles of self-perception in the maltreated children and adolescents. All too often,
clinical assessment for a maltreated victim focuses on psychopathological symptoms such
as depressive symptoms, but a critical role of self-perception is equally important in

87

assessing a victim’s perception of self. Our study lends strong empirical support for the
development and routine provision of mental health services and interventions designed
to strengthen positive view of self, particularly of those who have diminished social self-
perception. Interventions for maltreated children and adolescents should be designed to
improve self-perception as well as psychopathological symptoms simultaneously in order
to produce positive changes in both constructs. An intervention model targeting early
adolescence would be particularly important given the critical impact of child
maltreatment on social self-perception in early to middle adolescence. A recent review of
intervention designed to enhance self-perception revealed striking findings (O’Mara,
Marsh, Craven, & Debus, 2006). O’Mara and colleagues examined impact of self-
perception intervention on children in 145 primary studies and the study supported
substantial effectiveness of the interventions promoting positive self-perception were
significantly effective. However, approximately 80 percent of the intervention studies
were conducted in school settings (2% early childhood school, 52% primary school, and
22% secondary school) and only 6% of the intervention studies were conducted in mental
health settings. These findings clearly show that there is a lack of research on self-
perception intervention within a mental health setting, and also it may indicate self-
perception intervention may not be popular in a mental health setting. More importantly,
the results of the analyses showed that the lowest mean effect size was found in studies
with a mental health professional administering the intervention (i.e. social worker,
psychiatrist, and psychologist) and the highest mean effects size was found in studies
with teachers administering the intervention followed by school counselors. This finding

88

shows the least effectiveness of self-perception intervention provided by a mental health
professional including social work. Mental health professionals tend to be more
pathology-oriented and intervention aimed at increasing positive self-perception may be
less accepted by social worker and mental health professionals. However, an importance
of self-perception in mental health of adolescents and intervention model to enhance
positive self-perception should be acknowledged in social work and mental health field.
Lastly, although there have been tremendous efforts to restore impaired
contextual environments of maltreated adolescents (e.g. abuse family environment, lack
of social resources), all too often the environmental characteristics are not easily
transformed. It is a sad reality, that many maltreated children have to live in foster homes
or other institutions and are less likely to get enough support. Definitely, we should not
stop social efforts to restore family, community, and social environment to provide better
nurturing environment to maltreated children and adolescents. In addition, it is very
important to help maltreated children and adolescents maintain positive sense of self and
perceive themselves as capable and competent in order to increase their resiliency in an
adverse environment. Hence, intervention strategy to promote positive sense of self
should be emphasized more in social work practice.
In conclusion, the current study provides empirical support for the cognitive
model of social anxiety and indicates significant effects of child maltreatment on
humiliation anxiety, performance anxiety, and self-perception. More clinical attention
should be given to social anxiety symptoms and social self-perception in maltreated
children and adolescents, in particular in early to middle adolescence. In order to enhance

89

understanding of mental health of maltreated children and adolescents, research studies
should apply developmental perspective and should employ rigorous research
methodology which appropriately captures developmental dynamics in childhood and
adolescence. Further, future research should explore how different types of self-
perception mediate the relationship between child maltreatment and different types of
psychopathology, or conversely, how psychopathology influences domain-specific self-
perception. The implications of these findings are far-reaching and may inform future
research on etiology of social anxiety, development of maltreated children and adolescent,
and self-perception as a resilience or protective factor in intervention. These findings also
contribute to our understanding of the adaptive functioning of maltreated adolescents, and
guide intervention efforts aimed at improving mental health of maltreated adolescents.

90

REFERENCES

Alden, L. E., & Taylor, C. T. (2004). Interpersonal processes in social phobia. Clinical
Psychology Review, 24, 857-882.

Alfano, C. A., Beidel, D. C., & Turner
,
S. M. (2006). Cognitive correlates of social
phobia among children and adolescents. Journal of Abnormal Child Psychology, 2,
189-201.

American Psychiatric Association (2000). Diagnostic and Statistical Manual of Mental
Disorders, Fourth Edition, Text Revision (DSM-IV-TR). Washington, DC:
American Psychiatric Association.

Bailey, J. A., & McCloskey, L. A. (2005). Pathways to adolescent substance use among
sexually abused girls. Journal of Abnormal Child Psychology, 33, 39-53.

Bandura, A. (1993). Perceived self-efficacy in cognitive development and functioning.
Educational Psychologist, 28, 117 – 148.

Banjeree, R., & Henderson, L. (2001). Social-cognitive factors in childhood social
anxiety: A preliminary investigation. Social Development, 10, 558–572.

Barnett, D., Vondra, J. I., & Shonk, S. M. (1996). Self-perceptions, motivation, and
school functioning of low-income maltreated and comparison children. Child
Abuse & Neglect, 20, 397-410.

Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in
social psychological research: Conceptual, strategic, and statistical considerations.
Journal of Personality and Social Psychology, 51(6), 1173-1182.

Beeghly, M., & Cicchetti, D. (1994). Child maltreatment, attachment and the self system:
Emergence of an internal state lexicon in toddlers at high social risk. Development
and Psychopathology, 6, 5–30.

Beidel, D. C., Fink, C. M., Turner, S. M. (1996) Stability of Anxious Symptomatology in
Children, Journal of Abnormal Psychology, 24, 257-269.

Beidel, D. C., Turner, S. M., & Dancu, C. V. (1985). Physiological, cognitive and
behavioral aspects of social anxiety. Behavior Research and Therapy, 23, 109-17.

Beidel, D. C., Turner, S. M., & Morris, T. L. (1999). Psychopathology of childhood
social phobia. Journal of the American Academy of Child & Adolescent
Psychiatry, 38, 643-650.

91

Blatt, S. J., D'Afflitti, J. P., & Quinlan, D. M. (1976). Experiences of depression in
normal young adults. Journal of Abnormal Psychology, 85, 383-389.

Bögels, S. M.,van Oosten, M., Muris, P., & Smulders, D. (2001). Familial correlates of
social anxiety in children and adolescents. Behaviour Research and Therapy, 39,
273-287.

Bögels, M., & Zitgerman, D. (2000). Dysfunctional cognitions in children with social
phobia, separation anxiety disorder, and generalized anxiety disorder. Journal of
Abnormal Child Psychology, 28, 205–211.

Bolger, K. E., Patterson, C. J., & Kupersmidt, J. B. (1998). Peer relationships and self-
esteem among children who have been maltreated. Child Development, 69, 1171-
1197.

Bollen, K. A., & Curran, P. J. (2006). Latent Curve Models: A Structural Equation
Approach. New York: Wiley-Interscience.
Bonanno, G. (2004). Loss, trauma, and human resilience: have we underestimated the
human capacity to thrive after extremely aversive events. American Psychologist,
59, 20-28.

Bosquet, M., & Egeland, B. (2006). The development and maintenance of anxiety
symptoms from infancy through adolescence in a longitudinal sample.
Development and Psychopathology, 18, 517-550.

Bruce, A. E., Cole, D. A., Dallaire, D. H., Jacquez, F. M., Pineda, A. Q., & LaGange, B.
(2006). Relations of parenting and negative life events to cognitive diatheses for
depression in children. Journal of Abnormal Child Psychology, 34, 321-333.

Bruch, M. A., & Heimberg, R. G. (1994). Differences in perceptions of parental and
personal characteristics between generalized and nongeneralized social phobics.
Journal of Anxiety Disorders, 8, 155-168.

Calverley, R., Fischer, K. W., & Ayoub, C. (1994). Complex splitting of self-
representations in sexually abused adolescent girls. Development and
Psychopathology, 6, 195–213.

Caspi, A., Bem, D. J., Elder, G. H. (1989). Continuities and consequences of interactional
styles across the life course. Journal of Personality 57, 375–406.

Cicchetti, D. (1991). Fractures in the crystal: Developmental psychopathology and the
emergence of self. Developmental Review, 11, 271–287.

92

Cicchetti, D., Beeghly, M., Carlson, V., & Toth, S. (1990). The emergence of the self in
atypical populations. In D. Cicchetti & M. Beeghly (Eds.), The self in transition:
Infancy to childhood (pp. 309–344). Chicago: University of Chicago Press.

Cicchetti, D., & Garmezy, N. (1993). Prospects and promises in the study of resilience.
Development and Psychopathology, 5, 497-502.
Cicchetti, D., & Rogosch, F. A. (1996). Equifinality and multifinality in developmental
psychopathology. Development and Psychopathology, 8, 597-600.

Chartier, M. J., Walker, J. R., & Stein, M. B. (2001). Social phobia and potential
childhood risk factors in a community sample. Psychological Medicine, 31, 307-
315.

Chou, C. P., Bentler, P. M., & Satorra A. (1991). Scaled test statistics and robust standard
errors for non-normal data in covariance structure analysis: a Monte Carlo study.
The British Journal of Mathematical and Statistical Psychology, 44, 347-357.

Cicchetti, D., & Rogosch, F. A. (1997). The role of self-organization in the promotion of
resilience in maltreated children. Development and Psychopathology, 9, 799–817.

Clark, D. M. (2001). A cognitive perspective on social phobia. In W. R. Crozier, & L. E.
Alden (Eds.), International handbook of social anxiety: concepts, research and
interventions relating to the self and shyness. Chichester, UK: Wiley.

Clark, D. M., & Wells, A. (1995). A Cognitive model of social phobia. In R. Heimberg,
M. Liebowitz, D. A. Hope, & F. R. Schneier (Eds.). Social Phobia: Diagnosis,
Assessment and Treatment. New York: Guilford Press.

Clausen, A. H., & Crittenden, P.M. (1991). Physical and psychological maltreatment:
Relations among types of maltreatment, Child Abuse and Neglect, 15, 5–18.

Cohen, A. J., Adler, N., Kaplan, S. J., Pelcovitz, D., & Mandel. F. S. (2002). Interactional
effects of marital status and physical abuse on adolescent psychopathology. Child
Abuse and Neglect, 26, 277-288

Cohen, P., Brown, J., & Smailes, E. (2001). Child abuse and neglect and the development
of mental disorders in the general population. Development and Psychopathology,
13, 981–999.

Cole, D. A. (1990). Relation of social and academic competence to depressive symptoms
in childhood. Journal of Abnormal Psychology, 99, 422-429.

93

Cole, D. A., Martin, J. A., Peeke, L.A., Seroczynski, A. D., & Fier, J. (1999). Children’s
over- and underestimation of academic competence: A longitudinal study of
gender differences, depression and anxiety. Child Development, 70, 459-473.

Cole, D. A., Martin, J. M., & Powers, B. (1997). A competency-based model of child
depression: A longitudinal study of peer, parent, teacher, and self-evaluations. The
Journal of Child Psychology, Psychiatry and Allied Disciplines, 38, 505-514.

Cole, D. A., Martin, J. M, Power, B., & Truglio, R. (1996). Modeling causal relations
between academic and social competence and depression: a multitrait-
multimethod longitudinal study of children. Journal of Abnormal Psychology, 105,
258-270.

Cole, D. A., Maxwell, S. E., Martin, J. A., Peeke, L. G., Seroczynski, A. D., Tram, J. M.,
Hoffman, K. B., Ruiz, M. D., Jacquez, F., & Maschman, T. (2001). The
development of multiple domains of child and adolescent self-concept: A cohort
sequential longitudinal design. Child Development, 72, 1723-1746.

Cole, D. A., Peeke, L., Dolezal, S., Murray, N., & Canzoniero, A. (1999). A longitudinal
study of negative affect and self-perceived competence in young adolescents.
Journal of Personality and Social Psychology, 77, 851-861.
Cole, D. A., Peeke, L. G., Martin. J. M., Truglio, R., & Seroczynski, A. D. (1998). A
longitudinal look at the relation between depression and anxiety in children and
adolescents. Journal of Consulting and Clinical Psychology, 66, 451-160.
Cole, P. M., & Putnam, F. W. (1992). Effect of incest on self and social functioning: A
developmental psychopathology perspective. Journal of Consulting and Clinical
Psychology, 60, 174-184.
Collishaw, S., Pickles, A., Messer, J., Rutter, M., Shearer, C., & Maughan, B. (2007).
Resilience to adult psychopathology following childhood maltreatment: Evidence
from a community sample. Child Abuse and Neglect, 31, 211-229.
Cox, B. J., Fleet, C., & Stein, M. B. (2004). Self-criticism and social phobia in the US
national comorbidity study. Journal of Affective Disorders, 82, 227-234.
Dadds, M. R., Barrett, P. M., Rapee, R. M., & Ryan, S. (1996). Family process and child
anxiety and aggression: An observational analysis. Journal of Abnormal Child
Psychology, 24, 715-734.
Davidson, J. R., Hughes, D. L., George, L. K., & Blazer, D. G. (1993). The epidemiology
of social phobia: findings from the Duke epidemiological catchment
area study. Psychological Medicine, 23, 709-718.
DeWit, D. J., Ogborne, A. Offord, D. R., & MacDonald, K.(1999).Antecedents of the
risk of recovery from DSM-III-R social phobia. Psychological Medicine, 29, 569-
582.
Diehl, A. S., & Prout, M. F. (2002). Effects of posttraumatic stress disorder and child
sexual abuse on self-efficacy development. American Journal of Orthopsychiatry,
72, 262-265.

94

DuBois, D. L., Felner, R. D., Brand, S., & George, G. R. (1999). Profiles of self-esteem
in early adolescence: Identification and investigation of adaptive correlates.
American Journal of Community Psychology, 27, 899-931.
DuMont, K. A., Widom, C. S., & Czaja, S. J. (2007). Predictors of resilience in abused
and neglected children grown-up: The role of individual and neighborhood
characteristics. Child Abuse and Neglect, 31, 255-274.
Duncan, T. E., Duncan, S. C. , Strycker, L. A., Li, F., & Alpert. A. (2006). An
Introduction to Latent Variable Growth Curve Modeling: Concepts, Issues, and
Applications. Mahwah: Lawrence Erlbaum Associates.
Egeland, B., Carlson, E., & Sroufe, L. A. (1993). Resilience as process. Development and
Psychopathology, 5, 517–528.
Epkins, C. C. (1996). Cognitive specificity and affective confounding in social anxiety
and dysphoria in children. Journal of Psychopathology and Behavioral
Assessment, 18, 83-101.

Essau, C., Conradt, J., & Petermann, F. (1999). Frequency and comorbidity of social
phobia and social fears in adolescents. Behaviour Research and Therapy, 37,
831–843.

Ferrer, E., & McArdle, J. J. (2004). A experimental analysis of dynamics hypotheses
about cognitive abilities and achievement from childhood to early adulthood.
Developmental Psychology, 40, 935-952.

Ferrer, E., McArdle, J. J, Shaywitz, B. A., Holahan, J. M., Marchione, K., & Shaywitz, S.
E. (2007). Longitudinal models of developmental dynamics between reading and
cognition form childhood to adolescence. Developmental Psychology, 43, 1460-
1473.

Fischer, K. W., Ayoub, C., Singh, I., Noam, G., Maraganore, A., & Raya, P. (1997).
Psychopathology as adaptive development along distinctive pathways.
Development and Psychopathology, 9, 749-779.

Flores, E., Cicchetti, D., & Rogosch, F. A. (2005). Predictors of Resilience in maltreated
and nonmaltreated Latino children. Developmental Psychology, 41, 238-351.

Forthofer, M. S., Kessler, R. C., Story, A. L., & Gotlib, I. H. (1996). The effects of
psychiatric disorders on the probability and timing of first marriage. Journal of
Health and Social Behavior, 37, 121-132.

Gater, R., Tansella, M., Korten, A., Tiemens, B. G., Mavreas, V. G., & Olatawura, M. O.
(1998). Sex differences in the prevalence and detection of depressive and anxiety
disorders in general health care settings, report from the World Health
Organization Collaborative Study on psychological problems in general health
care. Archives of General Psychiatry, 55, 405-413.

95

Garmezy, N. (1993). Vulnerability and resilience. In d. C. Funder. R. D. Parke, C.
Tomlinso-Keesey, & K. Widaman (Eds.), Studying lives through time”
Approaches to personality and development (pp. 377-398). Washington, DC:
American Psychological Association.

Ginsburg, G. S., La Greca, A. M., & Silverman, W. K. (1998). Social anxiety in children
with anxiety disorders: Relation with social and emotional Functioning. Journal
of Abnormal Child Psychology, 26, 175-185.

Gould, R. A., Buckminster, S., Pollack, M. H., Otto, M. W., & Massachusetts, L. Y.
(1997). Cognitive-behavioral and pharmacological treatment for social phobia: A
meta-analysis. Clinical Psychology: Science and Practice, 4, 291–306.
Grills, A. E., & Ollendick, T. H. (2002). Peer victimization, global self-worth, and
anxiety in middle school children. Journal of Clinical Child & Adolescent
Psychology, 31, 59 – 68.

Hale III, W. W., Raaijmakers, Q., Muris, P., Hoof, A., & Meeus, W. (2008).
Developmental trajectories of adolescent anxiety disorder symptoms: A 5 year
prospective community study. Journal of American Academy of Child and
Adolescent Psychiatry, 47, 556-564.

Harter, S. (1982). The perceived competence scale for children. Child Development, 53,
87-97.

Harter, S. (1988). Manual for the Self-Perception profile for adolescents. Denver, CO:
University of Denver.

Harter, S. (1999). The construction of the self: A developmental perspective. New York,
NY : Guilford Press.

Harvey, A. G., Clark, D. M., Ehlers, A., & Rapee, R. M. (2000). Social anxiety and self-
impression: cognitive preparation enhances the beneficial effects of video
feedback following a stressful social task. Behavior Research and Therapy, 38,
1183-1192.

Hasket, M. E., Nears, K., Ward, C. S., & McPhersin, A. V. (2006). Diversity in
adjustment of maltreated children: Factors associated with resilient functioning.
Clinical Psychology Review, 26, 796-812.

Heimberg, R. G., Stein, M. B., Hiripi, E., & Kessler, R. C. (2000). Trends in the
prevalence of social phobia in the United States: a synthetic cohort analysis of
changes over four decades. European Psychiatry, 15, 29-37.

96

Heller, S. S., Larrieu, J. A., D’Imperio, R., & Boris, N. W. (1999). Research on resilience
to child maltreatment: empirical comparisons. Child Abuse and Neglect, 23, 213-
338.

Himelein, M. J. & McElrath, J. V. (1996). Resilient child sexual abuse survivors:
Cognitive coping and illusion. Child Abuse and Neglect, 20, 747-458.

Hirsch, C. R., Meynen, T., & Clark, D. M. (2004). Negative self-imagery in social
anxiety contaminates social interactions. Memory, 12, 496-506.

Hofmann. S. G. (2000). Treatment of social phobia: Potential mediators and moderators.
Clinical Psychology: Science and Practice 7, 3–16.

Hofmann, S. G. (2004). Cognitive mediation of treatment change in social phobia.
Journal of Consulting and Clinical Psychology, 72, 393-399.
Hofmann, S. G., Moscovitch, D. A., Kim, H. J., & Taylor, A. N. (2004). Changes in self-
perception during treatment of social phobia. Journal of Consulting and Clinical
Psychology, 72, 588-596.

Holmbeck, G. N. (1997). Toward terminological, conceptual, and statistical clarity in the
study of mediators and moderators: Examples from the child-clinical and pediatric
psychology literatures. Journal of Consulting and Clinical Psychology, 65(4),
599-610.

Hope, D. A., & Heimberg, R.G. (1988). Public and private self-consciousness and social
phobia. Journal of Personality Assessment, 52, 626–639.

Inderbitzen, H. M., Walters, K. S., & Bukowski, A. L. (1997). The role of social anxiety
in adolescent peer relations: Differences among sociometric status groups and
rejected subgroups. Journal of Clinical Child and Adolescent Psychology, 26,
338-348.

Iwaniec, D., Larkin, E., & Higgins, S., (2006). Research review: Risk and resilience in
case of emotional abuse. Child and Family Social Work, 11, 73-82.

Jaffee, S. R., Caspi, A., Moffitt, T. E., Polo-Tomas, M., & Taylor
,
A. (2007). Individual,
family, and neighborhood factors distinguish resilient from non-resilient
maltreated children: A cumulative stressors model. Child Abuse and Neglect, 31,
231-253.

Jacquez, F., Cole, D. A., & Serale, B. (2004). Self-perceived competence as a mediator
between maternal feedback and depressive symptoms in adolescents. Journal of
Abnormal Child Psychology, 32, 355-367.

97

James, L. R., Mulaik, S. A., & Brett, J. M. (2006). A Tale of Two Methods.
Organizational Research Methods, 9(2), 233-244.

Kaplow, J. B., & Widom, C. S. (2007). Age of onset of child maltreatment predicts long-
term mental health outcomes. Journal of Abnormal Psychology, 116,176-187.

Kashdan, T. B., & Herbert, J. D. (2001). Social anxiety disorder in childhood and
adolescence: Current status and future directions. Clinical Child and Family
Psychology, 4, 37-61.

Kaufman, J., & Cicchetti, D. (1989). Effects of maltreatment on school-age children's
socioemotional development: Assessments in a day-camp setting. Developmental
Psychology, 25, 516-524.

Kendall, P. C., & Treadwell, K. R. H. (2007). The Role of Self-Statements as a Mediator
in Treatment for Youth with Anxiety Disorders. Journal of Consulting and
Clinical Psychology, 75, 380-389.
Kessler, R. C. (2003). The impairments caused by social phobia in the general
population: implications for intervention. Acta Psychiatrica Scandinavia, 108
(s417), 19-27.

Kearney, C. A. (2005). Social Anxiety and Social Phobia in Youth: Characteristics,
Assessment, and Psychological Treatment. New York: Springer.

Kim, J., & Cicchetti, D. (2004). A longitudinal study of child maltreatment, mother–child
Relationship quality and maladjustment: The role of self-esteem and social
competence. Journal of Abnormal Child Psychology, 32, 341-354.

Kim, J., & Cicchetti, D. (2006). Longitudinal trajectories of self-system processes and
depressive symptoms among maltreated and nonmaltreated children. Child
Development, 77, 624–639.

Koss, M. P., & Figueredo, A. J. (2004). Change in cognitive mediators of rape's impact
on psychosocial health across 2 Years of recovery. Journal of Consulting and
Clinical Psychology, 72, 1063-1072.

La Greca, A. M., & Lopez, N. (1998). Social anxiety among adolescents: Linkages with
peer relations and friendships. Journal of Abnormal Child Psychology, 26, 83-94.

La Greca, A. M., & Stone, W. L. (1993). Social anxiety scale for children-revised: factor
structure and concurrent validity. Journal of Clinical Child Psychology, 22, 17-27.

Lansford, J. Malone, P, Stevens, K. I., Dodge, K. A., & Bates, J. E., & Pettit, G. S. (2006).
Developmental trajectories of externalizing and internalizing behaviors: Factors

98

underlying resilience in physically abused children. Development and
Psychopathology, 18, 35-55.

Last, C. G., Perrin, S., Hersen, M., & Kazdin, A. E. (1992). DSM-III-R anxiety disorders
in children: Sociodemographic and clinical characteristics. Journal of the
American Academy of Child & Adolescent Psychiatry, 31, 1070-1076.

Leon, S. C., Ragsdale, B., Miller, S. A., & Spacarelli, S.(2008). Trauma resilience among
youth in substitute care demonstrating sexual behavior problems. Child Abuse and
Neglect, 32, 67-81.

Lieb, R., Wittchen, H., Höfler, M., Fuetsch, M., Stein, M., & Merikangas, K. R. (2000).
Parental Psychopathology, Parenting Styles, and the Risk of Social Phobia in
Offspring : A Prospective-Longitudinal Community Study. Archives of General
Psychiatry. 57, 859-866.

Lipsitz, J. D., & Schneier, F. R. Social phobia: Epidemiology and cost of illness.
Pharmacoeconomics, 18, 23-32.
Luthar, S. S., & Zigler, E. (1991). Vulnerability and competence: A review of research on
resilience in childhood. American Journal of Orthopsychiatry, 61, 6-22.

Luthar, S. S., Cicchetti, D., & Becker, B. (2000). The Construct of resilience: A critical
evaluation and guidelines for future work. Child Development, 71, 543–562.

Lynch, M., & Cicchetti, D. (1998). An ecological-transactional analysis of children and
contexts: The longitudinal interplay among child maltreatment, community
violence, and children's symptomatology. Development and Psychopathology, 10,
235-257.

Lynskey, M. T. & Fergusson, D. M. (1997). Factors protecting against the development
of adjustment difficulties in young adults exposed to childhood sexual abuse.
Child Abuse and Neglect, 2,, 1177-1190.

MacMillan, H. L., Fleming, J. E., Streiner, D. L., Lin, E., Boyle, M. H., Jamieson, E., M.
Duku, E. K., Walsh, C. A., Wong, M. Y., & Beardslee, W. R. (2001). Childhood
abuse and lifetime psychopathology in a community Sample. American Journal of
Psychiatry, 158, 1878-1883.

MacKinnon, D. P., Lockwood, C. M., Hoffman, J. M., West, S. G., & Sheets, V. (2002).
A comparison of methods to test mediation and other intervening variable effects.
Psychological Methods, 7(1), 83-104.

Magee, W. J. (1999). Effects of negative life experiences on phobia onset. Social
Psychiatry and Psychiatric Epidemiology, 34, 343-351.

99

Mallet, P., & Rodriguez-Tome, G. (1999). Social anxiety with peers in 9- to 14-year-olds.
developmental process and relations with self-consciousness and perceived peer
acceptance. European Journal of Psychology of Education, 14, 387-402.

Mannarino, A. P., Cohen, J. A., & Gregor, M. (1989). Emotional and behavioral
difficulties in sexually abused girls. Journal of Interpersonal Violence, 4, 437-
451.

March, J. S., Parker, J. D. A., Sullivan, K., Stallings, P., & Conners, K. (1997). The
multidimensional anxiety scale for children (masc): Factor structure, reliability,
and validity. Journal of the American Academy of Child & Adolescent Psychiatry,
36, 554-565.

Marsh, H. W., & Ayotte, V. (2003). Do multiple dimensions of self-concept become
more differentiated with age? the differential distinctiveness hypothesis. Journal
of Educational Psychology, 95(4), 687-706.

Marsh, H. W., Parada, R. H., & Ayotte, V. (2004). A multidimensional perspective of
relations between self-concept (self description questionnaire II) and adolescent
mental health (Youth Self-Report). Psychological Assessment, 16, 27-41.

Marsh, H. W., Parada, R. H., Yeung, A. S., & Healey, J. (2001). Aggressive school
troublemakers and victims: A longitudinal model examining the pivotal role of
self-concept. Journal of Educational Psychology, 93, 411-419.

Masten, A. S. (2001). Ordinary Magic: Resilience Process in Development. American
Psychologist, 56(3), 227-238.

Masten, A. S. (2004) Regulatory processes, risk, and resilience in adolescent
development. Annals of the New York Academy of Sciences, 1021, 310–319.

Masten, A. S., & Garmezy, N. (1985). Risk, vulnerability, and protective factors in
developmental psychopathology. Advances in Clinical Child Psychology, 8, 1-52.

Masten, A. S., Hubbard, J. J., Gest, S. D., Tellegen, A., Garmezy, N., & Ramirez, M. L.
(1999). Adaptation in the context of adversity: Pathways to resilience from
childhood through adolescence. Development and Psychopathology, 11, 143–169.

McArdle, J. J. (1996). Current directions in structural factor analysis. Current Directions
in Psychological Science, 5, 10–17.

McArdle, J. J. (2001). A latent difference score approach to longitudinal dynamic
structural analyses. In R. Cudeck, S. du Toit & D. Sorbom (Eds.). Structural

100

Equation Modeling: Present and Future. A Festschrift in honor of Karl Jöreskog
(pp. 341-380). Lincolnwood, IL: SSI.

McArdle, J. J. (2004). Latent Growth Curve Analysis Using Structural Equation
Modeling Techniques. In D. M. Teti & M. A. Malden (Eds.), Handbook of
Research Methods in Developmental Science. Blackwell Publishing.
McArdle, J. J. & Epstein D. (1987). Latent Growth Curves within Developmental
Structural Equation Models. Child Development, 58, 110-33.
McArdle, J. J. Hamagami, F., Meredith, W., & Bradway, K. P. (2000). Modeling the
dynamic hypotheses of Gf-Gc theory using longitudinal life-span data. Learning
and Individual Differences, 12, 53-79.

McGee, R.A., Wolfe, D. A., Yuen, S. A., Wilson, S. K., & Carnochan, K. (1995). The
measurement of maltreatment: A comparison of approaches. Child Abuse and
Neglect, 19, 233-249.

McGloin, J. M., & Widom, C. S. (2001). Resilience among abused and neglected
children grown up. Development and Psychopathology, 13, 1021-1038.
McGrath, E. P., & Repetti, R. L.(2002). A longitudinal study of children’s depressive
symptoms, self-perceptions and cognitive distortions about the self. Journal of
Abnormal Psychology, 111, 77-87.

Mendlowicz, M. V., & Stein, M. B. (2000). Quality of life in individuals with anxiety
disorders. American Journal of Psychiatry, 157, 669-682.

Moran, P., & Eckenrode, J. (1992). Protective personality characteristics among
adolescent victims of maltreatment. Child Abuse and Neglect, 16, 743-754.

Meredith, W., & Tisak, J. (1990). Latent Curve Analysis. Psychometrika 55,107-22.
Nelson, L. J., Rubin, K. H., & Fox. N. A. (2005). Social withdrawal, observed peer
acceptance, and the development of self-perceptions in children ages 4 to 7 years.
Early Childhood Research Quarterly, 20, 185-200.

Ohannessian, C., Lerner, R. M., Lerner, J. V., & von Eye, A. (1999). Does self-
competence predict gender differences in adolescent depression and anxiety?
Journal of Adolescence, 22, 397-411.

O’Mara, A. J., Marsh, H. W., Craven, R. G., & Debus, R. (2006). Do self-concept
treatments make a difference? A synergistic blend of construct validation and
meta analysis. Educational Psychologist, 41, 181-206.

101

Petersen, A. C., Schulenberg, J., Abramowitz, R. H., Offer, D., & Jarcho, H. D. (1984). A
self-image questionnaire for young adolescents (siqya): Reliability and validity
studies. Journal of Youth and Adolescents, 13, 93-111.

Perkins, D. F., & Jones, K. R. (2004). Risk behaviors and resiliency within physically
abused adolescents. Child Abuse and Neglect, 28, 547-563.

Polk, L. V. (1997). Toward a middle range theory of resilience. Advanced Nursing
Science, 19(3), 1-13.

Raboteg-Saric, Z., Rijavec, M., & Brajsa-Zganec, A. (2001). The relation of parental
practices and selfconceptions to young adolescent problem behaviors and
substance use. Nordic Journal of Psychiatry, 55, 203–209.

Rajendran, K., & Videka, L. (2006). Relational and academic components of resilience in
maltreated adolescents. Annals of the New York Academy of Sciences, 1094, 345–
349.

Rao, P. A., Beidel, D. C., Turner, S. M., Ammerman, R. T., Crosby, L. E., & Sallee, F. R.
(2007). Social anxiety disorder in childhood and adolescence: Descriptive
psychopathology. Behavior Research and Therapy, 45, 1181-1191.
Rapee, R. M. (1995). Descriptive psychopathology of social phobia. In R. G. Heimberg,
M. R. Liebowitz, D. A. Hope, & F. R. Schneier (Eds.), Social phobia: Diagnosis,
assessment and treatment (pp. 41-66). New York: The Guilford Press.

Rapee, R. M., & Heimberg, R. G. (1997). A cognitive-behavioral model of anxiety in
social phobia. Behavior Research and Therapy, 35, 741-756.

Rapee, R. M., & Spence, S. H. (2004). The etiology of social phobia: empirical evidence
and an initial model. Clinical Psychology Review, 24, 737-767.

Romans, S. E., Martin, J. L., Anderson, J. C., O'Shea, M. L., & Mullen, P. E. (1995).
Factors that mediate between child sexual abuse and adult psychological outcome.
Psychological Medicine, 25, 127-142.

Rosenberg, M. (1979). Conceiving the Self. New York: Basic Books.
Rubin, D. B. (1987). Multiple imputation for Nonresponse in Surveys. New York: J
Wiley & Sons.

Rubin, K. H., & Mills, R. S. (1990). Maternal beliefs about adaptive and maladaptive
social behaviors in normal, aggressive, and withdrawn preschoolers. Journal of
Abnormal Child Psychology, 18, 419-435.

102

Rutter, M. (1987). Psychosocial resilience and protective mechanisms. American Journal
of Orthopsychiatry, 57, 316-331.

Safren, S. A., Heimberg, R. G., Brown, E. J., Holle, C. (1996). Quality of life in social
phobia. Depression and Anxiety, 4, 126-133.

Schafer, J. L. (1997). Analysis of Incomplete Multivariate Data, London: Chapman &
Hall.

Scheier, M. F., Carver, C. S., & Bridge, M. W. (1994). Distinguishing optimism from
neuroticism (and trait anxiety, self-mastery and self-esteem): A Reevaluation of
the life orientation test. Journal of Personality and Social Psychology, 67, 1063-
1078.

Schiefelbein, V. L., Susman, E. J., Dorn, L. D. (2005). Self-competence mediates earlier
and later anxiety in adolescent mothers: a 3-year longitudinal perspective. Journal
of Research on Adolescence, 15, 625-655.

Schlenker, B. R., & Leary, M. R. (1982). Social anxiety and self-presentation: a
conceptualization and model. Psychological Bulletin, 92, 641-669.

Schneier, F. R., Johnson, J., Hornig, C. D., Liebowitz, M. R, & Weissman, M. (1992).
Social phobia. Comorbidity and morbidity in an epidemiologic sample. Archives
of General Psychiatry, 49, 282-288.
Sachs-Ericsson , N., Verona , E., Joiner, T., & Preacher, K., (2006). Parental verbal
abuse and the mediating role of self-criticism in adult internalizing
disorders. Journal of Affective Disorders, 93, 71 – 78.

Shavelson, R. J., Hubner, J. J., & Stanton, G. C. (1976). Self-concept : Validation of
construct interpretation. Review of Educational Research, 46, 407-441.

Shonk, S. M., & Cicchetti, D. (2001). Maltreatment, competency deficits, and risk for
academic and behavioral maladjustment. Developmental Psychology, 37, 3–14.

hrout, P. E., & Bolger, N. (2002). Mediation in experimental and nonexperimental
studies: New procedures and recommendations. Psychological Methods, 7, 422-
445.

Smari, J., Petursdottir, G., & Porsteinsdottir, V. ( 2001). Social anxiety and depression in
adolescents in relation to perceived competence and situational appraisal. Journal
of Adolescence, 24, 199– 207.

Spaccarelli, S. (1994). Stress, appraisal, and coping in child sexual abuse: A theoretical
and empirical review. Psychological Bulletin, 116(2), 340-362.

103

Spaccarelli, S., & Kim, S. (1995). Resilience criteria and factors associated with
resilience in sexually abused girls. Child Abuse and Neglect, 19, 1171-1182.

Spauwen, J., Krabbendam, L., Lieb, R., Wittchen,H., & Van Os, J. (2006). Impact of
psychological trauma on the development of psychotic symptoms: relationship
with psychosis proneness. The British Journal of Psychiatry, 188, 527-533.

Spence, S. H., Donovan, C., & Brechman-Toussaint, M. (2000). The treatment of
childhood social phobia: The effectiveness of a social skills training-based,
cognitive-behavioral intervention, with and without parental involvement. The
Journal of Child Psychology and Psychiatry and Allied Disciplines, 41, 713-726.

Stein, M. B., Fuetsch, M., Müller, N., Höfler, M., Lieb, R., & Wittchen, H. (2001). Social
anxiety disorder and the risk of depression: A prospective community study of
adolescents and young adults. Arch Gen Psychiatry, 58, 251-256.

Storch, E. A., Brassard, M. R., & Masia-Warner, C. L. (2003). The relationship of peer
victimization to social anxiety and loneliness in adolescence. Child Study Journal,
33, 1-18.

Tram, J. M., & Cole, D. A. (2000). Self-perceived competence and the relation between
life events and depressive symptoms in adolescence. Journal of Abnormal
Psychology, 109, 753-760.

Trickett, P. K., Kurtz, D. A., & Noll, J. G. (2005).The Consequences of child sexual
abuse for female development. In D. J. Bell, S. L. Foster, & E. J. Mash (Eds.),
Handbook of Behavioral and Emotional Problems in Girls. New York: Kluwer
Academic / Plenum Publishers.

Trickett, P. K., & McBride-Chang, C. (1995). The developmental impact of different
forms of child abuse and neglect. Developmental Review, 15, 311-337.

Trickett, P. K., McBride-Chang, C., & Putnam, F. W. (1994). The classroom
performance and behavior of sexually abused girls. Development and
Psychopathology, 6, 183-194.

Tong, L., Oates, K., & McDowell, M. (1987). Personality development following sexual
abuse. Child Abuse & Neglect Special Issue; Child Abuse and Neglect, 11(3),
371-383.

Toth, S. L., & Cicchetti, D. (1996). Patterns of relatedness, Depressive symptomatology,
and perceived competence in maltreated children. Development and
Psychopathology, 64, 32-41.

104

Toth, S. L., Cicchetti, D., Macfie, J., & Emde, R. M. (1997). Representations of self and
other in the narratives of neglected, physically abused, and sexually abused
preschoolers. Development and Psychopathology, 9, 781-796.

Toth, S. L., Cicchetti, D., MacFie, J., Maughan, A., & Vanmeenen, K. (2000). Narrative
representations of caregivers and self in maltreated pre-schoolers. Attachment &
Human Development, 2, 271 – 305.

U.S. Department of Health and Human Services, Children’s Bureau. (2007). Child
maltreatment 2005. Washington, DC: U.S. Government Printing Office.
Retrieved February 2008 from
http://www.acf.hhs.gov/programs/cb/pubs/cm05/index.htm.

Van Velzen, C., Emmelkamp, C., & Scholing, A. (2000). Generalized social phobia
versus avoidant personality disorder differences in psychopathology, personality
traits, and social and occupational functioning. Journal of Anxiety Disorders, 14,
395-411.

Vondra J., Barnett, D., & Cicchetti, D. (1989). Self-concept, motivation, and competence
among preschoolers from maltreating and comparison families, Child Abuse and
Neglect, 14, 525-540.

Weems, C. F., Berman, S. L., Silverman W. K, & Saavedra, L. M. (2001). Cognitive
errors in youth with anxiety disorders: The linkage between negative cognitive
errors and anxious symptoms. Cognitive Therapy and Research, 25, 559-575.
Wittchen, H. U., Stein, M. B., & Kessler, R. C. (1999). Social fears and social phobia in a
community sample of adolescents and young adults: prevalence, risk factors and
co-morbidity. Psychological Medicine, 29, 309-323.

Woodward, L. J., & Fergusson, D. M. (2001). Life course outcomes of young people
with anxiety disorders in adolescence. Journal of the American Academy of Child
& Adolescent Psychiatry, 40, 1086-1093.

Wyatt, G. E., & Newcomb, M. (1990). Internal and External Mediators of Women's
Sexual Abuse in Childhood. Journal of Consulting and Clinical Psychology, 58,
758-767.

Yerkes, R. M., & Dodsonm J. D. (1908). The relation of strength of stimulus to rapidity
of habit formation. Journal of Comparative Neurology and Psychology, 18, 458-
482.

Zielinski, D. S., & Bradshaw, C. P. (2006). Ecological Influences on the Sequelae of
Child Maltreatment: A Review of the Literature, Child Maltreatment, 11, 49-62.

105
APPENDIX A

Social Anxiety Scale of Multidimensional Anxiety Scale for Children (March et al., 1997)

Never
True About
Me (0)
Rarely
True About
Me (1)
Sometimes
True About
Me (2)
Often
True About
Me (3)
Humiliation Anxiety
3. I worry about other people laughing at me.
10. I’m afraid that other kids will make fun of me.
16. I’m afraid other people will think I’m stupid
22. I worry about what other people think of me
29. I worry about doing everything exactly right
Performance Anxiety
14. I worry about getting called on in class
33. I get nervous if I have to perform in public
37. I have trouble asking other kids to play with me
39. I feel shy

106
APPENDIX B

Social Self-Perception (Self Perception Profile For Adolescents, Peer Acceptance Subscale; Harter, 1988)
Really
True for
Me
Sort of
True for
Me

Sort of
True for
Me
Really
True
for Me

1
 
Some people find it hard to
make friends
BUT
For other people it’s pretty easy.
 
2
 
Some people have a lot of
friends
BUT
Other people don’t have very
many friends.
 
3
 
Some people are very hard
to like
BUT
Other people are really easy to
like.
 
4
 
Some people are popular
with others their age
BUT
Other people are not very
popular.
 
5
 
Some people feel that they
are socially accepted
BUT
Other people wish that more
people their age accepted them.
 

Abstract (if available)

Abstract The purpose of the current study is to advance our understanding of the development of social anxiety -- with particular attention to humiliation anxiety and performance anxiety -- and social self-perception in maltreated and nonmaltreated comparison adolescents, and to provide prospective empirical evidence supporting a cognitive model of social anxiety. Specifically, the study aimed 1) to test mediation of social self-perception in the relationship between child maltreatment experience and social anxiety over time

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Untangling the developmental relations between depression and externalizing behavior among maltreated adolescents

PDF

Examining the longitudinal relationships between community violence exposure and aggressive behavior among a sample of maltreated and non-maltreated adolescents

PDF

Trajectories of internalizing problems among maltreated girls and boys: differences by maltreatment type and developmental timing

PDF

Maltreated adolescents and their families: a longitudinal examination of family functioning, parenting attitudes, & youth mental health

PDF

Attachment, maltreatment and autonomic nervous system responsivity as predictors of adolescent anxiety and depression

PDF

The relationship between pubertal timing and delinquent behavior in maltreated male and female adolescents

PDF

Parenting practices among non-offending mothers of sexually abused girls and its impact on the abused girls' behavioral adjustment: perspectives from a multigenerational, longitudinal study

PDF

Observed and self-reported childrearing in mothers with a history of sexual abuse

PDF

The role of depression symptoms on social information processing and tobacco use among adolescents

PDF

A longitudinal study of job resources, burnout, and turnover among social workers: implications for evidence-based organizational practices

PDF

The effects of math anxiety and low self-efficacy on students’ attitudes and interest in STEM

PDF

Neighborhood context and adolescent mental health: development and mechanisms

PDF

Evaluating social-cognitive measures of motivation in a longitudinal study of people completing New Year's resolutions to exercise

PDF

The effects of time and space on health status of older adults in China

PDF

Examining the longitudinal influence of the physical and social environments on social isolation and cognitive health: contextualizing the role of technology

PDF

“You move, therefore I am”: the combinatorial impact of kinesthetic motion cues on social perception

PDF

The formation and influence of online health social networks on social support, self-tracking behavior and weight loss outcomes

PDF

The impact of the mindful method Youth Empowerment Seminar (YES!) on students' self-efficacy, self-regulation, and academic performance for becoming college- and career-ready

PDF

Self-perceptions of Aging in the Context of Neighborhood and Their Interplay in Late-life Cognitive Health

PDF

Mechanisms of the effect of involuntary retirement on older adults' health and mental health

Asset Metadata

Creator Ji, Ju Ye (author)

Core Title The impact of child maltreatment on the mental health of adolescents: a longitudinal study of social anxiety and self-perception

School School of Social Work

Degree Doctor of Philosophy

Degree Program Social Work

Publication Date 08/06/2009

Defense Date 05/07/2009

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

Tag bivariate latent difference score model,Child abuse,latent growth curve model,longitudinal study,OAI-PMH Harvest,self-perception,social anxiety

Place Name Los Angeles (city or populated place)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Trickett, Penelope K. (committee chair), Brooks, Devon (committee member), McArdle, John J. (committee member), Mennen, Ferol E. (committee member)

Creator Email jji@usc.edu,juye.ji@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m2508

Unique identifier UC1212119

Identifier etd-Ji-3053 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-180028 (legacy record id),usctheses-m2508 (legacy record id)

Legacy Identifier etd-Ji-3053.pdf

Dmrecord 180028

Document Type Dissertation

Rights Ji, Ju Ye

Type texts

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

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu