Research Question

T r

W a

b

a

A R R A

K D S C S F F H S

1

t s r e l s a s B R e d c s i

v

o T

0 d

Journal of Anxiety Disorders 26 (2012) 215– 224

Contents lists available at SciVerse ScienceDirect

Journal of Anxiety Disorders

he hierarchical model of social interaction anxiety and depression: The critical oles of fears of evaluation

ei-Ting Wanga, Wen-Yau Hsua,b,∗, Yu-Chen Chiua, Chi-Wen Lianga

Department of Psychology, National Chengchi University, 64, Sec. 2, Zhinan Road, Wenshan District, Taipei City 11605, Taiwan Research Center for Mind, Brain, and Learning, National Chengchi University, Taiwan

r t i c l e i n f o

rticle history: eceived 16 April 2011 eceived in revised form 31 October 2011 ccepted 7 November 2011

eywords: epression ocial phobia

a b s t r a c t

In this paper, we articulate a hierarchical model of social interaction anxiety (SIA) and depression to account for their comorbidity and the uniqueness of SIA. First, negative affect (NA) and positive affect (PA) are conceptualized as general factors shared by SIA and depression; the fear of negative evaluation (FNE) is operationalized as the specific factor, which accounts for more of the variance in SIA than in depression, and the fear of positive evaluation (FPE) is operationalized as the factor unique to SIA. FPE is the key feature that differentiates SIA from depression. Second, the proposed hierarchical model describes structural relationships among these factors, in which the higher-level factors (i.e., high NA and low PA)

omorbidity ocial interaction anxiety ear of negative evaluation ear of positive evaluation ierarchical model tructural model of affect

represent the vulnerability markers of both SIA and depression and the lower-level factors (i.e., FNE and FPE) are the dimensions of specific cognitive features. In addition, an alternative model, in which all of the relationships are the same, except that FPE is operationalized as a specific factor, is proposed to clarify the role of FPE. The results from the hierarchical regression and the structural equation modeling support the hypothesized hierarchical model. Further theoretical and practical implications for FPE and the multilevel model are discussed.

. Introduction

Among anxiety disorders, social anxiety disorder is second only o general anxiety disorder in terms of comorbidity with depres- ion (Kessler, Chiu, Demler, Merikangas, & Walters, 2005). The emarkable rate of comorbidity that exists between social anxi- ty and depression has also been documented in a large body of iterature (e.g., Belzer & Schneier, 2004; Brunello et al., 2000). In pite of extensive empirical covariation between social anxiety nd depression, they are viewed as theoretically distinct con- tructs (Gibb, Coles, & Heimberg, 2005; Krueger, 1999; Sellbom, en-Porath, & Bagby, 2008; Watson, 2005; Watson & Clark, 2006). ecent psychopathological studies have shown a growing inter- st in discriminating phenotypic features between highly comorbid iagnostic classes, and explicating the underlying mechanisms of omorbidity and heterogeneity between social anxiety and depres- ion could advance our knowledge of the taxonomic and etiological

ssues.

Structural models of affect have been developed to pro- ide a theoretical framework for explaining the complexity of

∗ Corresponding author at: National Chengchi University, Department of Psychol- gy, 64, Sec. 2, Zhinan Rd., Wenshan District, Taipei City 11605, Taiwan. el.: +886 2 29387379; fax: +886 2 29366725.

E-mail addresses: [email protected], [email protected] (W.-Y. Hsu).

887-6185/$ – see front matter © 2011 Elsevier Ltd. All rights reserved. oi:10.1016/j.janxdis.2011.11.004

© 2011 Elsevier Ltd. All rights reserved.

comorbidity and, most importantly, the distinguishability between depression and anxiety disorders. The tripartite model of anxiety and depression (Clark & Watson, 1991), the hierarchical model of anxiety disorders (Zinbarg & Barlow, 1996), the integrative hierar- chical model of anxiety and depression (Mineka, Watson, & Clark, 1998), and the multilevel trait predictor model of anxiety dis- orders (Kotov, Watson, Robles, & Schmidt, 2007) share the core assumptions that mood and anxiety disorders both subsume a common component, which contributes to the comorbidity, and a unique component, which makes them distinguishable. Never- theless, structural models of depression and anxiety have reached a bottleneck in differentiation in that two well-established charac- teristics, positive affectivity (PA) and physiological hyperarousal, which differentiate anxiety disorders from major depression, have failed to explain the difference between social anxiety and depres- sion (Brown, Chorpita, & Barlow, 1998; Hughes et al., 2006). Following the rationale of the structural models of affect, the cur- rent study aims to add new components to the structural model of depression and social anxiety to clarify relationships of their correlates and address the issue of discriminability.

1.1. Structural models of affect

According to the structural models of affect, three distinguish- ing components can be used to explain the relationships between

2 nxiety

s w v ( f e u t T c

t r v a t t a e f p t H s t e v p t

1

e u s r n N W a t 2 ( 2 i f d m s M

e t s t i t r L 2 s ( i C

16 W.-T. Wang et al. / Journal of A

ocial interaction anxiety (SIA) and depression: (1) general factors, hich are higher-order trait variables responsible for the common

ariance shared by all disorders, such as negative affectivity (NA) Brown et al., 1998); (2) specific factors, which account for dif- erent attributable variances among distinct syndromes (Mineka t al., 1998); and (3) a unique factor, which is the unique contrib- tor to one disorder and the only component for distinguishing he disorder among the observed comorbidity (Kotov et al., 2007). he specific and unique factors might be lower-order psychological orrelates.

The structural models of affect, as mentioned above, implicate wo sets of hypotheses—generality and causality. First, generality efers to the extent to which a component accounts for the shared ariance among the observed syndromes. That is, general, specific, nd unique factors reflect the degree of the generality-specificity of heir contributions to the observed emotional disorders. An emo- ional disorder is viewed as a combination of different variances ttributable to the general, specific, and unique factors (Mineka t al., 1998). Moreover, the so-called general, specific, and unique actors are relative terms, defined in terms of specificity in com- arison to other given disorders. Second, relationships among hese factors can be hypothetically presented in a structural order. igher-order factors are reflective of dispositional characteristics,

uch as the traits of NA and PA, which reflect heritable and rela- ively stable trait dimensions. At the second hierarchical level are motional disorders that are, to a certain degree, influenced by the ulnerable dimensions. At the lowest level, certain symptoms or sychological features, such as physiological hyperarousal, charac- erize emotional disorders (Brown et al., 1998).

.2. The general factors

Based on extensive evidence from the structural model lit- rature, negative affectivity, a basic personality dimension, is ndoubtedly a common factor between social anxiety and depres- ion. In addition, as same as what has found in depression, recent esearch has found that social anxiety is characterized by anhedo- ia (Brown et al., 1998; Hughes et al., 2006; Kashdan, 2002, 2004; aragon-Gainey, Watson, & Markon, 2009; Sellbom et al., 2008; atson, Gamez, & Simms, 2005). A meta-analysis revealed that

reliable negative association between social anxiety and posi- ive emotions even holds after controlling for depression (Kashdan, 007). According to the self-regulatory model proposed by Kashdan 2007) and colleagues (Kashdan & Breen, 2008; Kashdan & Steger, 006), excessive concerns about potential social threats exhaust

ndividuals’ finite self-regulatory resource, eroding opportunities or pursuing reward-seeking activities, and resulting in hedonic eficiency. In addition, social activities and interpersonal engage- ents, which individuals with social anxiety avoid, have been

hown to be vital sources of positive emotion (Clark, Watson, & ineka, 1994; Watson, Clark, & Tellegen, 1988). In short, although these shared characteristic patterns might

xplain the high comorbidity between the two syndromes, rela- ively sparse evidence exists regarding the discrimination between ocial anxiety and depression. Past research has indicated two dis- inguishable types of social anxiety: performance anxiety and social nteraction anxiety. Compared with performance anxiety, the fea- ures of SIA overlap more extensively with depression due to its elative unique relation to anhedonia (Alden, Taylor, Mellings, & aposa, 2008; Brown et al., 1997; Hughes et al., 2006; Kashdan, 002, 2007). This makes sense because SIA interferes with a perva-

ive range of social encounters, which serve as sources of pleasure Kashdan, 2002, 2007). Moreover, compared to performance anx- ety, SIA exhibits higher comorbidity with depression (Stein & havira, 1998) and more severely impairs daily life, chronicity, and

Disorders 26 (2012) 215– 224

the recovery process. For these reasons, we restricted our focus in this study to SIA rather than the broad category of social anxiety.

To resolve the predicament of the discriminability among social anxiety and depression in structural models, Watson and Clark (1997) have noted that, as NA and PA are both higher-order per- sonality traits that do not serve for discrimination, studies should investigate the lower-order traits influenced by these syndromes. Naragon-Gainey et al. (2009) investigated the relationships among lower facets of positive emotionality with respect to social anxiety and depression. They found that depression is related to positive emotionality; social anxiety, however, is related to all four facets of PA (positive emotionality, sociability, ascendance, and fun-seeking) and, in particular, is strongly associated with sociability. Never- theless, this warrants further investigation because the concepts between positive emotionality and depression and between socia- bility and social anxiety overlap. In light of these propositions, lower-order correlates subsumed into a structural model could be considered as an improvement over the existing structural models. Based on structural models of affect, we attempted to synthesize other crucial lower-level factors to shed light on the specific and unique factors between SIA and depression.

1.3. Factors specific and unique to SIA

Cognitive-behavioral models have been one of the major the- oretical conceptualizations for social anxiety. In these models, social evaluative fears have been theorized to be the central cogni- tive features of socially anxious individuals (Clark & Wells, 1995; Rapee & Heimberg, 1997). They perceive themselves as not meet- ing the expectations of others and believe that others will evaluate them negatively, which leads to undesirable social consequences. Similarly, from the interpersonal point of view, social anxiety is conceptualized as a self-protective interpersonal pattern in which socially anxious individuals possess an overriding fear of disgrace and strongly attempt to avoid evaluation from others (Stravynski, 2007, p. 9). Moreover, according to the DSM-IV-TR (American Psychiatric Association, 2000), marked and persistent fear of evalu- ation and rejection are characterized as the cognitive symptoms of social anxiety. In the current study, we attempt to propose a hier- archical model of SIA and depression with fears of evaluation as lower-order specific or unique factors.

1.3.1. Fear of negative evaluation The fear of negative evaluation (FNE) is among the best estab-

lished cognitive-behavioral features of social anxiety (Clark & Wells, 1995; Collins, Westra, Dozois, & Stewart, 2005; Kocovski & Endler, 2000; Rapee & Heimberg, 1997). Studies indicate that indi- viduals with social anxiety show a more extreme level of FNE than normal controls (Weeks et al., 2005). Additionally, it is noteworthy that FNE shows considerable specificity to social anxiety compared to other anxiety disorders, such as panic disorder and obsessive- compulsive disorder (Ball, Otto, Pollack, Vccello, & Rosenbaum, 1995; Kotov et al., 2007; Saboonchi, Lundh, & Ost, 1999).

With respect to depression, although there is no direct theo- retical insight into the relationship between FNE and depression, empirical data have revealed moderate correlations between FNE and depression (Collins et al., 2005; Duke, Krishnan, Faith, & Storch, 2006; Kocovski & Endler, 2000; O’Connor, Berry, Weiss, & Gilbert, 2002; Sato, McCann, & Ferguson-Isaac, 2004). Some indirect evi- dence has suggested that symptoms of depression are related to the fear of rejection, disapproval, and criticism (Clark, Steer, Haslam, Beck, & Brown, 1997; Sato, 2003). According to the prominent cog-

nitive theory of depression by Beck (1976), negative thought is central to depression. It makes sense that FNE is related to depres- sion because negative evaluation from others verifies the biased schema of personal failure and deteriorates their depressive mood.

nxiety Disorders 26 (2012) 215– 224 217

F e t v

1

2 t p u o s a t s a a F f f i

o S s s c s 2 R u S c a t r W s b a e i &

s a t J c s 1 d W a u d r

r i u o t fi h

The alternative model The hypothesized hierarchical model

PA NA

SIA Dep.

FNE FPE

PA NA

SIA Dep.

FNE FPE

W.-T. Wang et al. / Journal of A

or these reasons, we hypothesized that FNE, which is to some xtent related to depression, can be formulated as a specific fac- or that is shared by SIA and depression but accounts for larger ariances in SIA than in depression.

.3.2. Fear of positive evaluation As supported by the psycho-evolutionary point of view (Gilbert,

001), Weeks, Heimberg, and Rodebaugh (2008) have advocated hat socially anxious individuals have a tendency to disqualify their ositive experiences and have found that the fear of positive eval- ation (FPE) is a distinguishable and important cognitive feature f social anxiety. The intent of social anxiety is to avoid unneces- ary conflicts and maintain safe interactions. Individuals with social nxiety regard themselves as incapable or inferior, and they believe hat favorable appraisal and receiving acclaim elevates their social tatus. This elevation is believed to induce clash and competition, s well as higher expectations, which they are unable to reach (see lso Wallace & Alden, 1995, 1997). Weeks et al. demonstrated that NE and FPE are independent dimensions and that FPE accounted or unique variances in the prediction of SIA beyond that accounted or by FNE, implying that FPE is a critical feature of SIA that is ndependent from FNE.

Do depressed individuals also show characteristics of FPE? To ur knowledge, no studies have directly addressed this issue. ome relevant evidence has shown that FPE is significantly more trongly related to SIA than to depression; one study, however, has hown that there is no significant difference between these two orrelations and that the relationship between FPE and depres- ion remained significant upon controlling for SIA (Fergus et al., 009; Weeks, Heimberg, & Rodebaugh, 2008; Weeks, Heimberg, odebaugh, & Norton, 2008). As a result of these data, it remains nclear whether FPE is related to depression or is a unique quality of IA because these findings were derived from correlation (or partial orrelation) analyses, which were confounded by other shared vari- nces. Thus, although correlation evidence has indicated a possible rend suggesting that FPE is specific to SIA, multivariate analyses are equired to depict the nature of its uniqueness (Naragon-Gainey &

atson, 2010). In the current study, we adopt hierarchical regres- ions and structural equation modeling to answer this question ecause the multilevel structural approach enables us to model nd control for the correlations among shared constructs, thereby nabling us to identify specific and unique constructs that show ncremental associations with the target disorder (Naragon-Gainey

Watson, 2010). Theoretically, on the basis of the interpersonal theory of depres-

ion, depressed individuals are characterized by an overwhelming nd constant demand for reassurance of their self-worth and attrac- ion, even when the guarantee has been offered (Coyne, 1976; oiner, Alfano, & Metalsky, 1992). This quality contradicts the con- ept of FPE. Secondly, according to behavioral approach/inhibition ystems (BAS/BIS; Carver & Scheier, 1998; Carver & White, 994), anxiety-related emotions are linked to avoidance, whereas epression-related emotions are linked to approach behaviors. hile the underlying drive for SIA is fear and attempts to avoid

ny kind of perceived social threats, including FNE and FPE, the nderlying tendency for depression is the inability to move toward esired goals. Therefore, positive evaluation, a type of favorable eward, might not be a threat to depressed individuals.

To summarize, according to empirical data and relevant theo- etical speculation, the current study aims to clarify whether FPE s a factor unique to SIA. Another possibility is that FPE is not a nique factor, but a factor specific to SIA, accounting for more

f the variance in SIA than in depression. This possibility echoes he assertion of Mineka and colleagues (1998), “Symptom speci- city must be viewed in relative rather than absolute terms.” It is ighly unlikely to find a unique factor that correlate only to a single

Fig. 1. The hypothesized hierarchical model of social interaction anxiety and depression and the alternative model. NA: negative affect; PA: positive affect; SIA: social interaction anxiety; Dep: depression; FPE: fear of positive evaluation; FNE: fear of negative evaluation.

disorder. That is to say, we could only find specific factors that contribute more variance to one disorder than to another in the hierarchical model.

1.4. The current study and hypotheses

FPE and FNE represent two distinct constructs, and this dis- tinction is the prerequisite for testing the hypothesized model. Although Weeks, Heimberg, and Rodebaugh (2008) and Fergus et al. (2009) provided evidence for the distinctness of FNE and FPE, these results have not been replicated in a Taiwanese sample. Therefore, before testing the main hypotheses in the current study, we con- duct a confirmatory factor analysis to examine the validity of the two-factor structure of FNE and FPE in our sample.

Building on and expanding the structural models of Mineka et al. (1998) and Brown et al. (1998), we propose a hierarchical model of SIA and depression. We assume that NA and PA are the higher-order general factors, FNE is the lower-order specific factor that accounts for more variance in SIA than in depression, and FPE is the lower- order unique factor, which is the key feature that differentiates SIA from depression. With all of the relationships among these factors being equal, except for FPE being a specific factor rather than a unique factor, an alternative model was also tested (see Fig. 1 for the hypothesized hierarchical model and the alternative model).

To examine these two models, we first test the incremental vari- ance contributed by FPE, upon the variance explained by NA, PA, and FNE. We expected that the incremental variance in the SIA could be accounted for by FPE, whereas the incremental variance in depression could not be accounted for by FPE. If this were true, FPE would represent a unique contributor to SIA rather than a specific factor.

Secondly, we organized these factors by their structural rela- tionships and tested the hierarchical model in which NA and PA are considered to be higher-order factors, and FNE and FPE are considered to be lower-order factors. In the proposed hierarchical model, we assumed significant paths from NA and PA to both SIA and depression, significant paths from SIA and depression to FNE, and a significant path from SIA to FPE. However, what if FPE is not unique to SIA but specific to SIA? To test this, we evaluate the alter- native model with all the above relationships being equal, but we add a path from depression to FPE. We predicted that the proposed hierarchical model would fit the data better than the alternative model.

2. Methods

2.1. Participants

The participants were 612 undergraduates who were enrolled in introductory psychology classes at two universities in Taiwan;

2 nxiety

5 ( t r

2

t m S n ( o c b M t d c i o t w p p O

2 e

a o f o 5 a c

2 S e

r A 5 C u l e v u n

2 H N

t r t c a w

18 W.-T. Wang et al. / Journal of A

66 of them provided a complete set of answers, and of these, 344 60.8%) were female, 220 (38.9%) were male, and 2 failed to report heir gender. The average age was 20.14 years (S.D. = 1.34) with a ange of 18–31 years.

.2. Measures and procedures

With the teachers’ approval, the participants completed a bat- ery of seven measures during class. We translated all of the

easures into Chinese except for the Beck Depression Inventory, econd Edition (BDI-II); for this measure, we used a published Chi- ese version (see Chang, 2005). Following published guidelines Brislin, 1986), these measures were translated into Chinese by ne of the authors of this manuscript with a Master’s degree in linical psychology. The Chinese items were then back-translated y a bilingual professional and an experienced translator with a aster’s degree in psychology. Discrepancies between the back-

ranslated English version and the original English version were iscussed carefully by the two translators and the second author, a linical psychology professor. At this stage, the items that might be nappropriate at a conceptual level and the cognitive equivalence f the translations were discussed. Each item was discussed until he three translators agreed that the original meaning of the item as appropriately retained in the Chinese item. At the final stage, a retest was conducted on undergraduate students to ensure com- rehensibility and to refine the statements to be reader friendly. nly minor changes were made according to students’ feedback.

.2.1. Positive and Negative Affect Schedule (PANAS; Watson t al., 1988)

The PANAS consists of two 10-item scales that measure positive ffect (PANAS-P) and negative affect (PANAS-N). The trait version f the PANAS, in which responses are made according to the general eelings, was adopted in the current study. The ratings were made n a 5-point scale ranging from 1 = “very slightly or not at all” to

= “extremely.” In the current study, both the PANAS-P ( ̨ = .83) nd the PANAS-N ( ̨ = .87) subscales exhibited very good internal onsistency.

.2.2. The Brief Fear of Negative Evaluation Scale traight-forward score (BFNE-S; Rodebaugh et al., 2004; Weeks t al., 2005)

The BFNE (Leary, 1983) consists of 12 items describing wor- ying thoughts about receiving negative evaluations from others. ll of the items are rated on a 5-point scale (1 = not at all to

= extremely). Rodebaugh et al. (2004), Weeks et al. (2005), and arleton, Collimore, McCabe, and Antony (2011) have suggested tilizing the eight straightforward BFNE items (BFNE-S) to calcu-

ate the total score for parsimony because BFNE-S demonstrated xcellent reliability (all ˛s > .92), factorial validity, and construct alidity comparable to the BFNE. Consequently, the BFNE-S were tilized in the present analyses and demonstrated excellent inter- al consistency ( ̨ = .88).

.2.3. The Fear of Positive Evaluation Scale (FPES; Weeks, eimberg, & Rodebaugh, 2008; Weeks, Heimberg, Rodebaugh, & orton, 2008)

The FPES is a 10-item scale that ranges from 0 (not at all true) o 9 (very true). Items 5 and 10 are designed to avoid acquiescence esponse styles and are not utilized in the calculation of the FPES

otal score. The FPES has been shown to have acceptable internal onsistency ( ̨ = .80) and 5-week test-retest reliability (r = .70), and dequate convergent validity. Cronbach’s alpha for the FPES items as very good in the current sample ( ̨ = .82).

Disorders 26 (2012) 215– 224

2.2.4. The Social Interaction Anxiety Scale (SIAS; Mattick & Clarke, 1998)

The SIAS is a 20-item measure of social interaction anxiety. The respondents were asked to indicate the degree to which the description of the items on the SIAS is true. The items are rated using a 0 (not at all characteristic or true of me) to 4 (extremely character- istic or true of me) scale. Mattick and Clarke (1998) have provided evidence demonstrating that the SIAS exhibits high test-retest reli- ability (r = .92), excellent internal consistency ( ̨ = .94), and strong convergent and discriminant validity. Cronbach’s alpha for the SIAS was very good in the current sample ( ̨ = .88). In the current study, the SIAS was used as one of two indicators of social interaction anxiety.

2.2.5. The Liebowitz Social Anxiety Scale-Self Report Format-Social Interaction Subscale (LSAS-SR-SI; Safren et al., 1999)

The LSAS-SR-SI was used as the second indicator of social inter- action anxiety in the current study. The original Liebowitz Social Anxiety Scale (LSAS; Liebowitz, 1987) is a clinician-administered measure designed to assess social anxiety. Currently, the LSAS has been made into a self report version (LSAS-SR) for convenience in research settings, and each subscale of the LSAS-SR has been shown to possess good internal consistency (Cox, Ross, Swinson, & Direnfeld, 1998). The 12-item LSAS Fear of Social Interaction Sub- scale (LSAS-SR-SI), which assesses fear and avoidance in 12 social interaction situations, was used in the current study (Safren et al., 1999). Internal consistency for the LSAS-SR-SI was very good in the current sample ( ̨ = .88).

2.2.6. The BDI-II (Beck, Steer, & Brown, 1996) The BDI-II is a widely used 21-item measure that assesses

the severity of depressive symptoms. High internal consistency ( ̨ = .91–.93) and robust construct validity have been observed (Beck et al., 1996; Buckley, Parker, & Heggie, 2001). The Chinese version of the BDI-II was translated from the English BDI-II by the Chinese Behavioral Science Corporation in Taiwan. The Chinese ver- sion of the BDI-II has been shown to possess adequate internal consistency in a college student sample ( ̨ = .86) and good construct validity (Chang, 2005). Data from our sample indicated that the Chinese BDI-II possesses very good internal consistency ( ̨ = .89).

2.2.7. The Short-Form Version of the Depression Anxiety Stress Scales (DASS-21; Lovibond & Lovibond, 1995)

The DASS-21 is a 21-item, 4-point instrument measuring the severity of the core symptoms of depression, anxiety, and stress. The depression scale of DASS-21 (DASS-21-Dep) was used in the current study as an indicator for depression. The DASS-21-Dep consists of seven items emphasizing dysphoria, hopelessness, self- deprecation, lack of interest and involvement, and so forth. The responses range from 0 (did not apply to me at all) to 3 (applied to me very much or most of the time). Strong psychometric prop- erties have been reported for the DASS-21-Dep ( ̨ = 83–.88; Henry & Crawford, 2005; Norton, 2007). The internal consistency for the DASS-21-Dep was very good in the current sample ( ̨ = .82).

2.3. Analysis plans

To address the first research question, hierarchical regression analyses were used to assess the incremental validity of specific fac- tors over more general traits. Because these factors are invariably

correlated, Kotov et al. (2007) suggested that their independent contributions can only be evaluated appropriately by partialling out the variances of correlated factors in the order from general to unique.

W.-T. Wang et al. / Journal of Anxiety Disorders 26 (2012) 215– 224 219

NAo

PAo

NA

PA FPESIA

FNEDep.

FPEo

FNEo

SIAS

.17

.21

.91

.89

.33

.49

.44

-.45

-.37

.15

.47

NAe

PAe

.85

.78

LSAS-SR SI

.39

.28

.80

BDI-II

.29

.84

DASS-21 - Dep.

.87

.25

.35

.32

.36

.97

.82

.05FNEe

.83 .32 .65

.82 .80

FPEe

Fig. 2. The completely standardized solution of the hypothesized structural model. All of the regression coefficients of the paths in the model reached significance at the ˛ = .05 level. NA: negative affect; PA: positive affect; SIA: social interaction anxiety; Dep: depression; FPE: fear of positive evaluation; FNE: fear of negative evaluation. The s ven it

t L F h e v d l w

3

3

a f r e o ( o H N f r o o H a 2 c

t s

ubscripts “o” and “e” indicate that the observed variables are the sum of odd and e

Secondly, to evaluate the proposed hierarchical structural rela- ionships, we adopted a structural equation model (SEM) using isrel 8.7 (Jöreskog & Sörbom, 2004). The models are illustrated in ig. 1. The model testing also provides further confirmation of the ypotheses of the general, specific, and unique factors. As noted arlier, the SIAS and the LSAS-SR-SI are indicators of the latent ariable SIA, and the BDI-II and the DASS-21-Dep are markers of epression. Because there should be at least two indicators for each

atent variable, the PANAS-P, the PANAS-N, the BFNE-S and the FPES ere even-odd halved (see Fig. 2).

. Results

.1. Descriptive statistics and correlation analyses

Table 1 presents descriptive statistics and correlations among ll of the study measures. The means and the standard deviations or the FPES, the LSAS-SR-SI, and the BDI-II were similar to those eported in other undergraduate samples in the literature. Of inter- st, participants had elevated score on the SIAS compared to the ther undergraduate samples reported by Gilbert (2001), Kashdan 2002), and Schofield, Coles, and Gibb (2009) and an elevated score n the BFNE-S compared to those reported by Weeks et al. (Weeks, eimberg, & Rodebaugh, 2008; Weeks Heimberg, Rodebaugh, & orton, 2008; Weeks et al., 2009). Moreover, the average score

or the SIAS was more than one standard deviation above those eported by Mattick and Clarke (1998). It is worth noting that 45.4% f the participants had a SIAS score above the recommended cut- ff value of 34 or higher (Brown et al., 1997; Rodebaugh, Woods, eimberg, Liebowitz, & Schneier, 2006), 48.1% had a BFNE-S score bove the cut-off value of 25 or higher (Carleton et al., 2011), and 8.1% of the participants had a BDI-II above the mild depression

ut-off value of 14.

All of the correlation coefficients were significant, except for he correlation between PANAS-N and PANAS-P (r = −.08). This uggests that NA and PA are two independent dimensions. As

ems, respectively.

expected, PANAS-P was negatively correlated with the FPES, the BFNE-S, the SIAS, the LSAS-SR-SI, the BDI-II, and the DASS-21-Dep, whereas PANAS-N was positively correlated with these measures. Most notably, the correlations of the FPES and BFNE-S scores with makers of SIA were larger than those with markers of depression. Furthermore, although indicators of SIA significantly correlated with indicators of depression, these coefficients did not exceed .30.

3.2. Confirmatory factor analysis (CFA)

We conducted CFA in LISREL to ensure the construct of FPE was separable from the construct of FNE. That is, we examine whether the two-factor model (the BFNE-S and FPES items loaded onto sep- arate latent factors) fit the data better than the unitary factor model (all of the BFNE-S and FPES items load onto a single latent factor). According to Hu and Bentler (1999), several indicators of model fit were used to assess the congruence between the data and the hypothesized model. These include the Bentler–Bonett Normed Fit Index (NFI), the Non-Norm Fit Index (NNFI), the Comparative Fit Index (CFI), the Standardized Root-Mean-Square Residual (SRMR), and the Root-Mean-Square Error of Approximation (RMSEA).

According to Hu and Bentler’s (1999), criteria, the results showed that the hypothesized two-factor model fit the data well (�2 (103) = 460.76, �2/df = 4.47; CFI = .95; NFI = .94; NNFI = .94; RMSEA = .078; SRMR = .063). Relative to the two-factor model, the unitary factor model showed a significantly worse fit (�2 (104) = 2622.30, �2/df = 25.21; CFI = .78; NFI = .77; NNFI = .75; RMSEA = .21; SRMR = .16). These results support the discriminant validity of FNE and FPE in the Taiwanese sample. The factor load- ings for the BFNE-S items ranged from .52 to .83, and those for the FPES items ranged from .51 to .81, except for item 7, which had weak

but significant (.21, p < .001) factor loading on the FPES. Moreover, the two fear of evaluation factors only showed low positive corre- lation (r = .24, p < .001), indicating that they are quite independent from each other.

220 W.-T. Wang et al. / Journal of Anxiety Disorders 26 (2012) 215– 224

Table 1 Descriptive statistics and correlations among study measures.

M (S.D.) Skewness Kurtosis 1 2 3 4 5 6 7

1. PANAS-N 21.84 (6.61) .34 −.29 2. PANAS-P 30.77 (5.69) −.26 .66 −.08 3. BFNE-S 24.19 (5.95) −.01 −.54 .34*** −.10*

4. FPES 26.61 (12.63) .08 −.57 .25*** −.23*** .25***

5. SIAS 32.54 (9.27) .12 −.32 .42*** −.30*** .45*** .48***

6. LSAS-SR-SI 13.19 (5.70) .29 .36 .31*** −.35*** .37*** .48*** .69***

7. DASS-21-Dep 3.54 (3.55) 1.25 1.37 .41*** −.38*** .30*** .25*** .23*** .27***

8. BDI-II 9.73 (7.84) 1.39 3.44 .43*** −.35*** .27*** .25*** .29*** .29*** .73***

Note: N = 566. PANAS-N: Positive and Negative Affect Schedule, negative affect subscale; PANAS-P: Positive and Negative Affect Schedule, positive affect subscale; BFNE-S: The Brief Fear of Negative Evaluation Scale; FPES: The Fear of Positive Evaluation Scale; SIAS: The Social Interaction Anxiety Scale; LSAS-SR-SI: The Liebowitz Social Anxiety Scale, self report format, social interaction subscale; DASS-21-Dep: The short-form version of the Depression Anxiety Stress Scales, depression subscale; BDI-II: The Beck D

3

e o a F

p o e t ( r i

a t e

3

w m h

T T

N F S T

epression Inventory, Second Edition. * p < .05.

*** p < .001

.3. Hierarchical regression analyses

The SIAS, the LSAS-SR-SI, the DASS-21-Dep, and the BDI-II were xamined individually as dependant variables. The entering steps f the independent variables were as follows: NA and PA were ssessed first, the contributions of FNE were estimated next, and PE was examined last. The results are presented in Table 2.

As can be seen in Table 2, PANAS-N and PANAS-P significantly redicted all of the dependent variables, accounting for 20.3–29.8% f the variance. Step 2 revealed that the BFNE-S had significant ffects associated with all of the symptom indicators. Specifically, he amount of variance explained by the BFNE-S was larger for SIA 10.7% and 6.8%) than for depression (1.8% and 1.1%). The final step evealed that the FPES made an independent contribution to both ndicators of SIA but not to those of depression.

Consistent with our expectation, the results revealed that NA nd PA are general factors. FNE was found to be a specific fac- or because it had a larger effect on SIA than on depression. FPE merged as a unique factor linked only to SIA.

.4. SEM analyses

To address the hierarchical relationships and the role of FPE, e tested two SEM models utilizing the maximum likelihood esti- ation method. Fig. 2 presents the standardized solutions of the

ypothesized model. The hypothetical model provided an excellent

able 2 he results of the hierarchical regression analyses.

Dependant variable SIAS

Step 1 ˇ

Step 2 ˇ

Step 3 ˇ

NA .36*** .24*** .18***

PA −.25*** −.23*** −.17***

FNE .35*** .30***

FPE .30***

Dependant variable BDI-II

Step 1 ˇ

Step 2 ˇ

Step 3 ˇ

NA .41*** .36*** .35***

PA −.31*** −.30*** −.28***

FNE .15*** .14***

FPE .06

ote: PANAS-N: Positive and Negative Affect Schedule, negative affect subscale; PANAS-P ear of Negative Evaluation Scale Straight-forward score; FPES: The Fear of Positive Evalu ocial Anxiety Scale, self report format, social interaction subscale; DASS-21-Dep: The sho he Beck Depression Inventory, Second Edition. ** p < .01.

*** p < .001.

fit to the data [�2 (47) = 99.67, �2/df = 2.12, less than 3, indicat- ing very good model fit (Carmines & McLver, 1981)]. The NFI (.98), the NNFI (.98), the Goodness-Of-Fit Index (GFI = .97), the Adjusted Goodness-Of-Fit Index (AGFI = .95), and the CFI (.99) were all above .90, whereas the SRMR (.043) and the RMSEA (.045) were both far below .05 (Hu & Bentler, 1999; McDonald & Ho, 2002). The alterna- tive model resulted in a very close fit [�2 (46) = 97.33, �2/df = 2.12, NFI (.98), NNFI (.98), GFI (.97), AGFI (.95), CFI (.99), SRMR (.041), and RMSEA (.044)]. The difference between these two models [��2

(1) = 2.34, p > .05] was not significant. For the principle of parsi- mony, we retained the hypothesized model. Moreover, the sole additional path in the alternative model – depression to the FPES – was not significant (standardized path coefficient = .08, p > .05). Therefore, we concluded that the hypothesized model is more acceptable than the alternative model.

In addition, it is worth noting that, again, SIA showed a stronger association to FNE than depression (standardized path coeffi- cient = .47 and .16, respectively).

4. Discussion

The hierarchical model proposed in this study was confirmed

with hierarchical regression and SEM analyses. First, in agreement with the findings of Brown et al. (1998), Hughes et al. (2006), and Watson et al. (2005), our results revealed that NA and PA are the higher-order general factors (i.e., shared features) of SIA and

LSAS-SR-SI

�R2 Step 1 ˇ

Step 2 ˇ

Step 3 ˇ

�R2

20.3 .29*** .20*** .13*** 20.7 −.33*** −.31*** −.24***

10.7 .28*** .22*** 6.8 8.0 .36*** 11.0

DASS-21-Dep

�R2 Step 1 ˇ

Step 2 ˇ

Step 3 ˇ

�R2

27.8 .38*** .34*** .33*** 29.8 −.36*** −.36*** −.35***

1.8 .11** .10** 1.1 .3 .05 .3

: Positive and Negative Affect Schedule, positive affect subscale; BFNE-S: The Brief ation Scale; SIAS: The Social Interaction Anxiety Scale; LSAS-SR-SI: The Liebowitz

rt-form version of the Depression Anxiety Stress Scales, depression subscale; BDI-II:

nxiety

d ( s i a s a h T t R i & u r t o C a

e t h u 2 R v S u a c f w i

r a F t m m t m s F a . c n e a o e

a c i a s 1 i d t i s a

W.-T. Wang et al. / Journal of A

epression. Second, consistent with the findings by Kotov et al. 2007), FNE accounted for a fair amount of variance in SIA mea- ures above that accounted for by NA and PA, indicating that FNE s a significant characteristic of SIA. Moreover, the fact that FNE ccounted for a larger amount of variance in SIA than in depres- ion (in both hierarchical regressions and SEM path coefficients),

pattern also demonstrated by Weeks et al. (2005), validated the ypothesis that FNE is the specific factor in the hierarchical model. he results of the present study are in accordance with the notion hat FNE is the crucial feature of social anxiety (Clark & Wells, 1995; apee & Heimberg, 1997), and FNE is more important in SIA than

n depression. As suggested by cognitive-behavioral models (Clark Wells, 1995; Rapee & Heimberg, 1997), socially anxious individ- als have a negative self-image and mainly allocate their attention esources to external threats that result from their biased men- al representation. Therefore, they view others critically and focus n any indicator of negative evaluation in a social environment. ollectively, the comorbidity between SIA and depression may be ccounted for by NA and PA and partly by FNE.

Finally, the major contribution of the current study is the discov- ry that FPE is significantly associated with SIA, but not depression; his verifies its role as a unique predictor of SIA. Past research as tested the relationship between SIA or depression and FPE sing correlation (or partial correlation) analyses (Fergus et al., 009; Weeks, Heimberg, & Rodebaugh, 2008; Weeks Heimberg, odebaugh, & Norton, 2008), and the findings merely and conser- atively concluded that FPE was less related to depression than to IA (Weeks et al., 2010). This raised the question of whether FPE is nique to SIA. The multivariate analyses (hierarchical regressions nd SEM) adopted in the current study are improvements over orrelation analysis with respect to controlling for more general actors in the model, specifying the contribution of FPE. Therefore, e are more confidently able to argue that FPE plays a unique role

n SIA. With respect to the discriminant validity of FNE and FPE, the

esults of the CFA in the present study corroborate the argument nd the findings of Weeks, Heimberg, and Rodebaugh (2008) and ergus et al. (2009) suggesting that FPE is a characteristic dis- inct from FNE. This distinction is also supported by the underlying

echanisms argued in the literature. While FNE is related to biased ental representation and attentional allocation, FPE underscores

heir discomfort in response to positive social feedback, leading to inimization of positive social experience (Weeks et al., 2010). It

hould be noted that the results of the CFA showed that FNE and PE are only mildly correlated (r = .24), whereas the results from the forementioned studies showed moderate correlations (r = .59 and

74, respectively). This implies that the constructs of FNE and FPE ould be more independent among the Taiwanese. Because the Chi- ese culture emphasizes humility, this value might permeate the nvironment of social interaction. Therefore, FPE not only served s part of the fears of evaluation, but also represented a concept f manner, making it more separable from FNE. Nonetheless, this xplanation needs more empirical support from future studies.

The present findings especially highlight the importance of FPE s a unique factor in SIA, which has implications for the theoretical onceptualization of SIA. This argument is endorsed by the find- ngs of Weeks (2010) demonstrating that positive minimizing is

core feature of social anxiety. Several lines of theoretical per- pectives echo this quality in SIA. Disqualifying the positive (Beck, 976) is a cognitive bias that emphasizes that socially anxious

ndividuals tend to attribute their success to external factors and isqualify their own self-worth. Alden et al. (2008) have claimed

hat socially anxious individuals are prone to negatively interpret- ng positive social events as threatening. In accordance with these tudies, findings by Kashdan (2007) also underscore the diminished nd impaired positive psychological experiences in SIA.

Disorders 26 (2012) 215– 224 221

There are two theoretical perspectives that describe the under- lying mechanisms for FPE, which have been raised by Gilbert (2001) and Wallace and Alden (1995, 1997). Gilbert’s psycho-evolutionary model highlights the cognitive bias of “fear of doing well”. Accord- ing to Gilbert, individuals with social anxiety consider themselves disadvantaged, but believe that improving their social status will incite competition and conflicts with more dominant individu- als, which will, in turn, lead to the loss of the benefits gained by being ordinary. Socially anxious individuals would rather main- tain nonviolent social interactions and keep a low profile. Indeed, Weeks, Heimberg, Rodebaugh, and Norton (2008) found a mod- erate correlation between FPE and the scores on the Submissive Behavior Scales, which are an indication of keeping oneself in com- pliance with group rules and avoiding conflicts with people high on the social hierarchy. Similarly, Naragon-Gainey et al. (2009) have presented evidence that social anxiety is significantly neg- atively related to ascendance, an indicator of exhibitionism and dominance. Furthermore, while social anxiety had a remarkable association with ascendance, depression did not. These findings echo our assertion that FPE is unique to SIA. Collectively, find- ings from Weeks and Naragon-Gainey and their colleagues have paralleled the view of Gilbert that individuals with social anxi- ety perceive that positive evaluation would lead to threatening outcome of conflicts, which would afflict them by losing benefits, which may lead to social exclusion. Secondly, Wallace and Alden have proposed an explanation similar to the above-mentioned etiology, albeit somewhat different: socially anxious individuals regard the price of success as being scrutinized more critically and others having higher expectations of them. While they consider their success as contingent and their typical performance of not being praiseworthy, they dread higher expectations and failure.

With respect to depressed people, a nonsignificant relation- ship between FPE and depression has suggested that depressed individuals do not show FPE. Consistent with the interpersonal point of view, their constant demands for reassurance of their self- worth might not lead to FPE. Moreover, this difference is in line with the perspective that social anxiety and depression are dis- tinct emotional distress that results from BIS and BAS, respectively. The inability of depressed individuals to pursue desired rewards should not result in the refusal of positive feedback. Also in line with Beck’s cognitive-content specificity model, anxiety and depression can be discriminated in terms of unique cognitive content (Beck, 1976; Beck & Emery, 1985); therefore, we can speculate that dis- qualifying the positive is specific to social anxiety. While anxious individuals allocate extensive attention to future threats, depressed people are more focused on self-depreciation. It therefore makes sense that the attentional bias for threatening information directly and strongly links FNE and FPE to social anxiety. On the contrary, depressed people are characterized by biases for self-referential negative information (Mineka et al., 1998). Negative evaluation by others might verify their biased schema of personal failure, thereby, to a lesser degree, linking depression to FNE. However, FPE is not related to depressive self-depreciation. Collectively, findings on the specific role of FNE and the unique role of FPE echo the conceptu- alization of the psychopathology of social anxiety and depression in Beck’s theory, the interpersonal model, and the BIS/BAS model.

In contrast, an alternative perspective involving the goals of the defense system has been proposed by Trower and Gilbert (1989). They have suggested that social anxiety and depression are states on a continuous dimension, in which depression is an end state of despair at failure of avoidance (anxious state). Based on this per- spective, socially anxious individuals adopt submissive strategies

to avoid harm or rejection. If unsuccessful, they will escape from or avoid perceived threats. When all of these strategies are unsuc- cessful, depression is experienced by socially anxious individuals (Weeks et al., 2009). According to the current findings, FPE can

2 nxiety

d m r t i d

r b h e e t s f a u p w b c ( a e d B J f t t s s i e c

A h i a r t t p s c s M i a a e

r t d t p s b a s h r e v

22 W.-T. Wang et al. / Journal of A

ifferentiate SIA from depression. However, Trower and Gilbert’s odel implies that SIA utilizes strategies that involve FNE- and FPE-

elated cognition in the beginning stage, and if these strategies fail, hey experience depressive mood in a later stage. Future research s required to clarify whether social anxiety and depression are two istinct categorizations or different stages on the same continuum.

Because FPE is the unique contributor to SIA, this implies a ole for FPE in psychological interventions. Two major cognitive- ehavioral models (Clark & Wells, 1995; Rapee & Heimberg, 1997) ave underscored the importance of excessive worry over negative valuation and its detrimental consequences. Our findings, how- ver, indicated that the significance of FPE might be greater than hat of FNE, as the results of the SEM analysis revealed that the tandardized path coefficient from SIA to FPE was .65, while that rom SIA to FNE was only .47. This implies that, for assessment nd treatment, focusing on the negative impact of FNE on individ- als with social anxiety would be unsatisfactory. Merely providing ositive social feedback or social experiences (e.g., interactions ith friendly individuals) does not diminish their social anxiety

ecause they typically feel uncomfortable toward compliments and onsider them inaccurate, which foreshadow their future failure Weeks, Heimberg, & Rodebaugh, 2008). Taken together, socially nxious individuals do not absorb positive information into their xisting schema, and even worse, this feedback causes remarkable istress. This has been an obstacle in psychotherapy (Heimberg & ecker, 2002). According to the self-verification theory (Giesler,

osephs, & Swann, 1996), people desire and look for evaluative eedback that they believe to be accurate because this enhances heir senses of predictability and controllability. It is possible hat socially anxious individuals are not confident in their social kills, and as a result, positive feedback is inconsistent with their elf-concept and produces discomfort. To reduce this discomfort, ndividuals doubt of the accuracy of the information because they xpected negative feedback. Future work will hopefully clarify this ognitive-interpersonal mechanism.

Gilboa-Schechtman, Franklin, and Foa (2000), Wallace and lden (1997), and Weeks, Heimberg, and Rodebaugh (2008) ave suggested efficiently incorporating FPE-related elements

n the contemporary cognitive-behavioral therapy of social nxiety. Examples include cognitive reconstructions of FPE- elated automatic thinking and exposure therapy directed at he agony associated with positive social feedback. According to he present findings, cognitive behavioral intervention involving sycho-education of FPE-related thoughts and challenging their elf-schema by addressing positive social experiences and the dis- repancy between their self-evaluation and their actual ability can ystematically reduce the barrier in cognitive-behavior therapy. oreover, introducing FPE-related cognitive change would inval-

date their tendency to disqualify positive feedback from others nd their exaggeration of others’ expectations of them, and in turn, meliorating the negative feelings toward positive evaluations may ventually lessen social interaction anxiety.

The current study is not without limitations. First, although the esults in the current study consistently indicate the unique con- ribution of FPE to SIA, the current study adopted a correlational esign. To obtain more reliable data, further research is needed o replicate our findings and improve our methodology, for exam- le, by conducting experiments. Second, we used an undergraduate ample rather than a clinical sample to address the differences etween SIA and depression. However, many researchers have dvocated conceptualizing psychopathology in a dimensional con- truct (e.g., Kotov et al., 2007; Sellbom et al., 2008), and they

ave argued that the results obtained from clinical and nonclinical espondents are highly consistent (O’Connor et al., 2002; Sellbom t al., 2008; Watson et al., 2005). Moreover, our sample had ele- ated scores on the SIAS and the BFNE-S; approximately half of

Disorders 26 (2012) 215– 224

the participants reached the cut-off scores. Therefore, our sample might share similar characteristics with respect to the SIAS and the BFNE-S. Although our results provide strong support for all of these hypotheses, replicating the hierarchical model of SIA and depression in clinical samples is required. Third, the proposed hier- archical model primarily focuses on the roles of FNE and FPE but does not attempt to elaborate all of the related correlates concern- ing SIA and depression. Thus, the current model may have ignored several related correlates, such as blushing propensity, rejection sensitivity, and rewards of social attention. We speculate that these constructs may be related to the concept of fears of evaluation, and their relationships warrant further studies. In addition, the rela- tion between PA and SIA in the current model might be simplistic. Research has shown that this association is not straightforward and that affect regulatory strategies serve as moderators (Kashdan & Breen, 2008; Kashdan & Steger, 2006). Again, future studies that aim to clarify the structural model of SIA and depression should take this delicate association into consideration. Fourth, the present data were obtained exclusively from self-report measures, which may be suspect due to shared method effects. It is noteworthy that the findings have been paralleled by results obtained from interviews for DSM diagnoses (e.g., Brown et al., 1998). However, examina- tion of the generalizability using multiple forms and modalities of measurements is required. Last, because the current study mainly focused on the unique feature of SIA, the unique factor for depres- sion remains an issue for future exploration.

5. Conclusions

Despite these limitations, our findings are encouraging in that they are among the first to address the unique role of FPE in SIA relative to depression and to examine their hierarchical relation- ships. We developed a hierarchical model of SIA and depression based not only on the specificity/generality of the correlated fea- tures but also on their hierarchical relationships. The results from the hierarchical regression and SEM analyses support our hypoth- esized hierarchical model. These structural relationships provide further evidence with respect to the underlying etiology of these disorders and should facilitate the development of interventions for social anxiety.

References

Alden, L. E., Taylor, C. T., Mellings, T. M. B., & Laposa, J. M. (2008). Social anxiety and the interpretation of positive social events. Journal of Anxiety Disorders, 22, 577–590.

American Psychiatric Association. (2000). Diagnostic and Statistical Manual of Mental Disorders (4th ed.). Washington, DC: Am. Psychiatr. Assoc. Text Version.

Ball, S. G., Otto, M. W., Pollack, M. H., Uccello, R., & Rosenbaum, J. F. (1995). Differen- tiating social phobia and panic disorder: a test of core beliefs. Cognitive Therapy and Research, 19, 473–781.

Beck, A. T. (1976). Cognitive therapy and the emotional disorders. New York: Interna- tional Universities Press.

Beck, A. T., & Emery, G. (1985). Anxiety disorders and phobias: a cognitive perspective. New York: Basic Books.

Beck, A. T., Steer, R. A., & Brown, G. K. (1996). Manual for the Beck Depression Inventory- II. San Antonio, TX: Psychological Corporation.

Belzer, K., & Schneier, F. R. (2004). Comorbidity of anxiety and depressive disorders: issues in conceptualization, assessment, and treatment. Journal of Psychiatric Practice, 10, 296–306.

Brislin, R. W. (1986). The wording and translation of research instruments. In: W. L. Lonner, & J. W. Berry (Eds.), Field methods in cross-cultural research (pp. 137–164). Newbury Park: Sage Publications.

Brown, E. J., Turovsky, J., Heimberg, R. G., Juster, H. R., Brown, T. A., & Barlow, D. H. (1997). Validation of the Social Interaction Anxiety Scale and the Social Phobia Scale across the anxiety disorders. Psychological Assessment, 9, 21–27.

Brown, T. A., Chorpita, B. F., & Barlow, D. H. (1998). Structural relationships among dimensions of the DSM-IV anxiety and mood disorders and dimensions of

negative affect, positive effect, and autonomic arousal. Journal of Abnormal Psy- chology, 107, 179–192.

Brunello, N., den Boer, J. A., Judd, L. L., Kasper, S., Kelsey, J. E., Lader, Y., et al. (2000). Social phobia: diagnosis and epidemiology, neurobiology and pharmacology, comorbidity and treatment. Journal of Affective Disorders, 60, 61–74.

nxiety

B

C

C

C

C

C

C

C

C

C

C

C

C

D

F

G

G

G

G

H

H

H

H

J

J

K

K

K

K

K

K

W.-T. Wang et al. / Journal of A

uckley, T. C., Parker, J. D., & Heggie, J. (2001). A psychometric evaluation of the BDI- II in treatment-seeking substance abusers. Journal of Substance Abuse Treatment, 20, 197–204.

arleton, R. N., Collimore, K. C., McCabe, R. E., & Antony, M. M. (2011). Addressing revisions to the Brief Fear of Negative Evaluation Scale: measuring fear of nega- tive evaluation across anxiety and mood disorders. Journal of Anxiety Disorders, 25, 822–828.

armines, E. G., & McIver, J. P. (1981). Analyzing models with unobserved variables: analysis of covariance structures. In: G. W. Bhrnstedt, & E. F. Borgatta (Eds.), Social measurement: current issues (pp. 65–115). Beverly Hills, CA: Sage Publications.

arver, C. S., & Scheier, M. F. (1998). On the self-regulation of behavior. New York: Cambridge University Press.

arver, C. S., & White, T. L. (1994). Behavioral inhibition, behavioral activation, and affective responses to impending reward and punishment: the BIS/BAS scales. Journal of Personality and Social Psychology, 67, 319–333.

hang, H. W. (2005). Dimensions of the Chinese beck depression inventory-II in a university sample. Individual Differences Research, 3, 193–199.

lark, D. A., Steer, R. A., Haslam, N., Beck, A. T., & Brown, G. K. (1997). Personal- ity vulnerability, psychiatric diagnoses, and symptoms: cluster analyses of the Sociotropy-Autonomy Subscales. Cognitive Therapy and Research, 21, 267–283.

lark, L. A., & Watson, D. (1991). Tripartite model of anxiety and depression: psy- chometric evidence and taxonomic implications. Journal of Abnormal Psychology, 100, 316–336.

lark, D. M., & Wells, A. (1995). A cognitive model of social anxiety. In: R. G. Heim- berg, M. R. Liebowitz, D. A. Hope, & F. R. Schneier (Eds.), Social phobia: diagnosis, assessment, and treatment (pp. 69–93). New York: The Guilford Press.

lark, L. A., Watson, D., & Mineka, S. (1994). Temperament, personality, and the mood and anxiety disorders. Journal of Abnormal Psychology, 103, 103–116.

ollins, H. A., Westra, H. A., Dozois, D. J. A., & Stewart, S. H. (2005). The validity of the brief version of the fear of negative evaluation scale. Anxiety Disorders, 19, 345–359.

ox, B. J., Ross, L., Swinson, R. P., & Direnfeld, D. M. (1998). A comparison of social phobia outcome measures in cognitive-behavioral group therapy. Behavior Mod- ification, 22, 285–297.

oyne, J. C. (1976). Toward an interactional description of depression. Psychiatry, 39, 28–40.

uke, D., Krishnan, M., Faith, M., & Storch, E. A. (2006). The psychometric properties of the Brief Fear of Negative Evaluation Scale. Anxiety Disorders, 20, 807–817.

ergus, T. A., Valentiner, D. P., McGrath, P. B., Stephenson, K. L., Gier, S. L., & Jencius, S. (2009). The Fear of Positive Evaluation Scale: psychometric properties in a clinical sample. Journal of Anxiety Disorders, 23, 1177–1183.

ibb, B. E., Coles, M. E., & Heimberg, R. G. (2005). Differentiating symptoms of social anxiety and depression in adults with social anxiety disorder. Journal of Behavior Therapy and Experimental Psychiatry, 36, 99–109.

iesler, R. B., Josephs, R. A., & Swann, W. B., Jr. (1996). Self-verification in clinical depression: the desire for negative evaluation. Journal of Abnormal Psychology, 105, 358–368.

ilbert, P. (2001). Evolution and social anxiety: the role of attraction, social compe- tition, and social hierarchies. Psychiatric Clinics of North America, 24, 723–751.

ilboa-Schechtman, E., Franklin, M. E., & Foa, E. B. (2000). Anticipated reactions to social events: differences among individuals with generalized social phobia, obsessive compulsive disorder, and nonanxious controls. Cognitive Therapy and Research, 24, 731–746.

eimberg, R. G., & Becker, R. E. (2002). Cognitive-behavioral group therapy for Social Phobia: Basic mechanisms and clinical strategies. New York: Guilford Press.

enry, J. D., & Crawford, J. R. (2005). The 21-item version of the Depression Anxiety Stress Scales (DASS-21): normative data and psychometric evaluation in a large non-clinical sample. British Journal of Clinical Psychology, 44, 227–239.

u, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance struc- tural analysis: conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1–55.

ughes, A. A., Heimberg, R. G., Coles, M. E., Gibb, B. E., Liebowitz, M. R., & Schneier, F. R. (2006). Relations of factors of the tripartite model of anxiety and depression to types of social anxiety. Behaviour Research and Therapy, 44, 1629–1641.

oiner, T. E., Jr., Alfano, M. S., & Metalsky, G. I. (1992). When depression breeds con- tempt: reassurance seekings, self-esteem, and rejection of depressed college students by their roommates. Journal of Abnormal Psychology, 101, 165–173.

öreskog, K. G., & Sörbom, D. (2004). LISREL 8: user’s reference guide. Lincolnwood, IL: Scientific Software International, Inc.

ashdan, T. B. (2002). Social anxiety dimensions, neuroticism, and the contours of positive psychological functioning. Cognitive Therapy and Research, 26, 789–810.

ashdan, T. B. (2004). The neglected relationship between social interaction anx- iety and hedonic deficits: differentiation from depressive symptoms. Anxiety disorders, 18, 719–730.

ashdan, T. B. (2007). Social anxiety spectrum and diminished positive experiences: theoretical synthesis and meta-analysis. Clinical Psychology Review, 27, 348–365.

ashdan, T. B., & Breen, W. E. (2008). Social anxiety and positive emotions: a prospec- tive examination of a self-regulatory model with tendencies to suppress or express emotions as a moderating variable. Behavior Therapy, 39, 1–12.

ashdan, T. B., & Steger, M. (2006). Expanding the topography of social anxiety: an experience sampling assessment of positive emotions and events, and emotion

suppression. Psychological Science, 17, 120–128.

essler, R. C., Chiu, W. T., Demler, O., Merikangas, K. R., & Walters, E. E. (2005). Prevalence, severity, and comorbidity of 12-month DSM-IV disorders in the national comorbidity survey replication. Archives of General Psychiatry, 62, 617–627.

Disorders 26 (2012) 215– 224 223

Kocovski, N., & Endler, N. S. (2000). Social anxiety, self-regulation, and fear of nega- tive evaluation. European Journal of Personality, 14, 347–358.

Kotov, R., Watson, D., Robles, J. P., & Schmid, N. B. (2007). Personality traits and anxiety symptoms: the multilevel trait predictor model. Behaviour Research and Therapy, 45, 1485–1503.

Krueger, R. F. (1999). The structure of common mental disorders. Archives of General Psychiatry, 56, 921–926.

Leary, M. R. (1983). A brief version of the Fear of Negative Evaluation Scale. Person- ality and Social Psychology Bulletin, 9, 371–375.

Liebowitz, M. R. (1987). Social phobia. Modern Problems of Pharmacopsychiatry, 22, 141–173.

Lovibond, S. H., & Lovibond, P. F. (1995). Manual for the Depression Anxiety Stress Scales (2nd ed.). Sydney, Australia: Psychological Foundation of Australia.

Mattick, R. P., & Clarke, J. C. (1998). Development and validation of measures of social phobia scrutiny fear and social interaction anxiety. Behaviour Research and Therapy, 36, 455–470.

McDonald, R. P., & Ho, M. R. (2002). Principles and practice in reporting structural equation analysis. Psychometrika, 58, 431–443.

Mineka, S., Watson, D., & Clark, L. A. (1998). Comorbidity of anxiety and unipolar mood disorders. Annual Review Psychology, 49, 377–412.

Naragon-Gainey, K., & Watson, D. (2010). Clarifying the dispositional basis of social anxiety: a hierarchical perspective. Personality and Individual Differences, 49, 926–934.

Naragon-Gainey, K., Watson, D., & Markon, K. E. (2009). Differential relations of depression and social anxiety symptoms to the facets of extraversion/positive emotionality. Journal of Abnormal Psychology, 118, 299–310.

Norton, P. J. (2007). Depression Anxiety and Stress Scales (DASS): psychometric anal- ysis across four racial groups. Anxiety, Stress, and Coping: An International Journal, 20, 253–265.

O’Connor, L. E., Berry, J. W., Weiss, J., & Gilbert, P. (2002). Guilt, fear, submission, and empathy in depression. Journal of Affective Disorders, 71, 19–27.

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

Rodebaugh, T. L., Woods, C. M., Heimberg, R. G., Liebowitz, M. R., & Schneier, F. R. (2006). The factor structure and screening utility of the Social Interaction Anxiety Scale. Psychological Assessment, 18, 231–237.

Rodebaugh, T. L., Woods, C. M., Thissen, D. M., Heimberg, R. G., Chambless, D. L., & Rapee, R. M. (2004). More information from fewer questions: the factor structure and item properties of the original and Brief Fear of Negative Evaluation Scale. Psychological Assessment, 16, 169–181.

Saboonchi, F., Lundh, L., & Ost, L. (1999). Perfectionism and self-consciousness in social phobia and panic disorder with agoraphobia. Behaviour Research and Ther- apy, 37, 799–808.

Safren, S. A., Heimberg, R. G., Horner, K. J., Juster, H. R., Schneier, F. R., & Liebowitz, M. R. (1999). Factor structure of social fears: The Liebowitz Social Anxiety Scale. Journal of Anxiety Disorders, 13, 253–270.

Sato, T. (2003). Sociotropy and autonomy: the nature of vulnerability. The Journal of Psychology, 137, 447–466.

Sato, T., McCann, D., & Ferguson-Isaac, C. (2004). Sociotropy-autonomy and situation-specific anxiety. Psychological Reports, 94, 67–76.

Schofield, C. A., Coles, M. E., & Gibb, B. E. (2009). The Fear of Positive Evaluation Scale: psychometric properties in a clinical sample. Journal of Anxiety Disorders, 23, 884–890.

Sellbom, M., Ben-Porath, Y. S., & Bagby, R. M. (2008). On the hierarchical structure of mood and anxiety disorders: confirmatory evidence and an elab- orated model of temperament markers. Journal of Abnormal Psychology, 117, 576–590.

Stein, M. B., & Chavira, D. A. (1998). Subtypes of social phobia and comorbidity with depression and other anxiety disorders. Journal of Affective Disorders, 50, S11–S16.

Stravynski, A. (2007). Fearing others: the nature and treatment of social phobia (chap. 1). Cambridge, UK, Cambridge University Press.

Trower, P., & Gilbert, P. (1989). New theoretical conceptions of social anxiety and social phobia. Clinical Psychology Review, 9, 19–35.

Wallace, S. T., & Alden, L. E. (1995). Social anxiety and standard setting following social success or failure. Cognitive Therapy and Research, 19, 613–631.

Wallace, S. T., & Alden, L. E. (1997). Social phobia and positive social events: the price of success. Journal of Abnormal Psychology, 106, 416–424.

Watson, D. (2005). Rethinking the mood and anxiety disorders: a quantita- tive hierarchical model for DSM–V. Journal of Abnormal Psychology, 114, 522–536.

Watson, D., & Clark, L. A. (1997). Extraversion and its positive emotional core. In: R. Hogan, J. Johnson, & S. Briggs (Eds.), Handbook of personality psychology (pp. 767–793). San Diego, CA: Academic Press.

Watson, D., & Clark, L. A. (2006). Clinical diagnosis at the crossroads. Clinical Psychol- ogy: Science and Practice, 13, 210–215.

Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect: the PANAS scales. Journal of Personality and Social psychology, 54, 1063–1070.

Watson, D., Gamez, D., & Simms, W. (2005). Basic dimensions of temperament and

their relation to anxiety and depression: a symptom-based perspective. Journal of Research in Personality, 39, 46–66.

Weeks, J. W. (2010). The Disqualification of Positive Social Outcomes Scale: a novel assessment of a long-recognized cognitive tendency in social anxiety disorder. Journal of Anxiety disorders, 24, 856–865.

2 nxiety

W

W

W

avoid evaluation, withdraw: Fears of evaluation and depressive cognitions lead

24 W.-T. Wang et al. / Journal of A

eeks, J. W., Heimberg, R. G., Fresco, D. M., Hart, T. A., Turk, C. L., Schneier, F. R., et al. (2005). Empirical validation and psychometric evaluation of the Brief Fear of Negative Evaluation Scale in patients with social anxiety disorder. Psychological Assessment, 17, 179–190.

eeks, J. W., Heimberg, R. G., & Rodebaugh, T. L. (2008). The Fear of Positive Evalua-

tion Scale: assessing a proposed cognitive component of social anxiety disorder. Journal of Anxiety Disorders, 22, 44–55.

eeks, J. W., Heimberg, R. G., Rodebaugh, T. L., & Norton, P. J. (2008). Exploring the relationship between fear of positive evaluation and social anxiety. Journal of Anxiety disorders, 22, 386–400.

Disorders 26 (2012) 215– 224

Weeks, J. W., Jakatdar, T. A., & Heimberg, R. G. (2010). Comparing and contrasting fears of positive and negative evaluation as facets of social anxiety. Journal of Social and Clinical Psychology, 29, 68–94.

Weeks, J. W., Rodebaugh, T. L., Heimberg, R. J., Norton, P. J., & Jakatdar, T. A. (2009). To

to social anxiety and submissive withdrawal. Cognitive Therapy and Research, 33, 375–389.

Zinbarg, R. E., & Barlow, D. H. (1996). Structure of anxiety and the anxiety disorders: a hierarchical model. Journal of Abnormal Psychology, 105, 181–193.