Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Dec 1.
Published in final edited form as: Transcult Psychiatry. 2022 Jul 25;59(6):878–888. doi: 10.1177/13634615221111629

The Relationship between Interdependent and Independent Self-Construals and Social Anxiety Symptom Severity in a Clinical Sample of Treatment-Seeking Patients

Antonia N Kaczkurkin a, Savannah Simon b, Lily Brown b, Anu Asnaani c
PMCID: PMC10540138  NIHMSID: NIHMS1929226  PMID: 35876512

Abstract

Differences in cultural orientations, such as interdependent and independent self-construals, may influence social anxiety disorder (SAD) symptom presentations. However, prior research on the association between interdependent/independent self-construals and SAD was limited to non-clinical samples. Using a treatment-seeking population with clinical levels of anxiety, the current study extends prior research by examining whether the relationship between interdependent/independent self-construals and SAD is specific to SAD or indicative of a broader relationship with anxiety or depression more generally. We also expand upon prior work by examining the effect of self-construals on treatment outcomes and whether self-construals change over time. The results showed that endorsing a less independent self-construal was associated with greater SAD symptoms specifically, and was not associated with other anxiety or depression symptom measures. Additionally, while interdependent and independent self-construals did not moderate SAD treatment outcomes, there was a decrease in interdependent self-construal and increase in independent self-construal over a course of cognitive behavioral therapy. Notably, this change over time was tied to specific items that correlated strongly with SAD symptoms. Together, these results increase our understanding of the relationship between interdependent/independent self-construals and SAD symptoms in treatment-seeking anxiety patients.

Keywords: cultural orientation, interdependence, independence, social anxiety disorder, anxiety, depression

Social anxiety disorder (SAD) is characterized by a persistent fear of negative evaluation, social scrutiny, or rejection (American Psychiatric Association, 2013) and is associated with significant distress and impairment. SAD has a lifetime prevalence rate of 12.1% in the United States (Kessler et al., 2005) and is associated with moderate to severe impairment in social, family, and/or work life (McKnight et al., 2016; Patel et al., 2002). SAD is also highly comorbid with other psychiatric symptoms including other anxiety disorders, depressive disorders, substance use, and suicidal ideation (Fehm et al., 2008; Sareen et al., 2005). Thus, SAD is associated with a significant burden on individuals and society more broadly.

Culture may play an important role in influencing the development of SAD symptoms, since fear of negative evaluation by others is related to culturally-defined social standards regarding role expectations. A commonly reported cultural orientation framework in the literature is that of collectivism versus individualism. Collectivism refers to cultural norms that promote relationships and the collective well-being, while individualism stresses individual autonomy and independence of the self (Vandello & Cohen, 1999). Another theoretical framework for cultural orientation is based on an individual’s cultural self-construal, that is, the extent to which one defines oneself in terms of independence from others or interdependence with others (Markus & Kitayama, 1991; Singelis, 1994). Using this framework, interdependent self-construal is characterized by an image of the self that emphasizes connectedness, social context, and relationships, while independent self-construal focuses on the separateness, internal attributes, and uniqueness of individuals (Singelis, 1994). The interdependent and independent self-construals are not mutually exclusive constructs (Dinnel et al., 2002; Markus & Kitayama, 1991). Additionally, these concepts are dynamic and active, with the ability to change over time as a result of adjustments in self-worth, continuity, belonging, distinctiveness, and coherence (Pilarska, 2014). Notably, interdependent and independent self-construals are developed in the context of one’s cultural surroundings, with various cultures having divergent views of the self (Kashima et al., 2011; Kitayama & Park, 2007).

The cultural context should be considered when examining the relationship between cultural orientation and SAD symptoms. For example, the U.S.-American culture values individualism, where greater emphasis is placed on self-reliance and self-determination (Vandello & Cohen, 1999). Shaping one’s own destiny through effort, choice, and ability is also of high importance. This cultural context is closely associated with an independent self-construal. Research demonstrates that American students who endorse an interdependent self-construal report significantly higher SAD symptom severity than those who endorse an independent self-construal (Okazaki, 1997). However, this previous work was limited to a non-clinical sample of college students. The current study builds upon prior work by examining the relationship between interdependent/independent self-construals and SAD symptoms in a sample with clinical levels of anxiety symptoms.

The majority of studies on interdependent and independent self-construals focus on SAD symptoms primarily (Krieg & Xu, 2015; Norasakkunkit & Kalick, 2009; Okazaki, 1997). It is unclear whether the relationship with cultural orientation is specific to SAD or whether it is indicative of a broader relationship with other anxiety or depressive symptoms, which are highly comorbid with SAD (Fehm et al., 2008). For example, collectivism has been shown to be positively correlated with depression, social anxiety, and obsessive-compulsive disorder symptoms in students residing in a highly individualistic society (Caldwell-Harris & Ayçiçegi, 2006). Previous work has relied on correlational or univariate regression analyses, where each class of symptoms is tested separately or where symptoms like depression were controlled for rather than examined as an effect of interest. Performing separate analyses for each symptom class does not account for the high correlations between disorders. One statistical method that considers the associations between highly correlated symptom measures is canonical correlation analysis (CCA). This approach allows us to better understand the relationships between interdependent/independent self-construals and symptom measures, while accounting for the high degree of overlap between diagnoses. Thus, the current study will expand prior work by using CCA to examine whether the relationship between self-construals and symptoms is specific to SAD or apparent for anxiety and depression more generally.

Additionally, there are currently no studies examining SAD treatment outcomes in those with different self-construals. A small number of studies have examined race/ethnicity as predictors of treatment outcomes in patients with SAD. A review of these studies concluded that race/ethnicity does not predict response to psychological or pharmacological treatments (Hofmann et al., 2010). However, race/ethnicity is a limited proxy for cultural orientation, and does not necessarily capture individual differences in cultural values (Asnaani et al., 2010; Heinrichs et al., 2006). Thus, the current study will examine whether interdependent and independent self-construals moderate SAD treatment outcomes. This would allow us to test whether levels of self-construal are associated with divergent trajectories in symptom reduction, which is important for determining who will benefit from treatment. Finally, it is unclear whether interdependent or independent self-construals change over time. There has been considerable disagreement regarding whether self-construals are relatively stable or malleable, as well as how this concept should be operationalized and measured (Gudykunst & Lee, 2003; Levine et al., 2003). Therefore, the current study will contribute to the literature by examining change in self-construals over the course of treatment.

The purpose of the current study was to test whether interdependent and independent self-construals are associated with SAD symptoms in a U.S.-American context. Specifically, we aimed to examine: 1) whether greater interdependent self-construal and lower independent self-construal would be associated with a greater severity of SAD symptoms, given the overarching independent cultural orientation in the U.S., and 2) whether the relationship between interdependent/independent self-construals and symptoms is specific to SAD or also apparent in other anxiety and depressive measures. To fill the gap in research on cultural self-construals and treatment outcomes, we had two additional exploratory aims: 3) we examined whether interdependent and independent self-construals moderate the reduction in symptom severity during SAD treatment, and 4) we examined whether interdependent and independent self-construals change over the course of treatment.

Methods

Participants

Participants were 321 adults (age: M = 31.23 years, SD = 11.83 years, range: 18–70) seeking treatment at an outpatient specialty anxiety clinic in Philadelphia, PA. All participants received a primary diagnosis of a Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) (American Psychiatric Association, 2013) anxiety or anxiety-related disorder. Frequently diagnosed primary disorders included obsessive-compulsive disorder (OCD), generalized anxiety disorder (GAD), social anxiety disorder (SAD), posttraumatic stress disorder (PTSD), panic disorder (PD), specific phobia, and other/unspecified anxiety disorder. Patients also frequently had a secondary diagnosis of major depressive disorder. Exclusion criteria are covered in detail elsewhere (Asnaani, Benhamou, Kaczkurkin, Turk-Karan, & Foa, 2019). Briefly, exclusion criteria included a primary diagnosis other than an anxiety-related disorder, imminent suicide risk, active or uncontrolled psychosis, evidence of intellectual disability or severe autism, and evidence of a primary substance or alcohol disorder.

Demographics of the sample are shown in Table 1. Participants primarily identified as White with the next most common racial/ethnic categories being Asian, multi-racial, Hispanic, and African American. Given the clinic’s setting near downtown Philadelphia on the University of Pennsylvania campus, participants included primarily a mixture of local residents and students from campus. The city of Philadelphia is a large, multicultural city whose population includes a diverse combination of residents including many immigrants, racial/ethnic minorities, and students from abroad. The sample included slightly more females than males, with the percentages of highest level of education similar to the general U.S. population. Bachelor’s degree was the most common level of education in the current sample. The most commonly endorsed relationship status was single followed by married. In terms of employment, most participants indicated either full-time employment or student status. The percentages of primary, secondary, and tertiary diagnoses are presented in Table 2. Of the total sample size, 53.9% reported clinical levels of SAD symptoms, as measured by scores greater than or equal to 19 on the Social Phobia Inventory (Connor et al., 2002).

Table 1.

Sample Demographics.

N %
Gender
 Female 172 53.6%
 Male 145 45.2%
 Transgender 1 .3%
 Other 3 .9%
Race
 White 259 80.7%
 Asian 26 8.1%
 Multiracial 15 4.7%
 Hispanic 9 2.8%
 African American 8 2.5%
 Other 4 1.2%
Level of Education
 No degree 9 2.8%
 High school diploma/GED 102 31.8%
 Associates degree 11 3.4%
 Bachelor’s degree 121 37.7%
 Master’s degree 51 15.9%
 Doctoral degree 27 8.4%
Relationship status
 Single 147 45.8%
 In a relationship 53 16.5%
 Living with partner 34 10.6%
 Married 77 24.0%
 Separated 4 1.2%
 Divorced 6 1.9%
Employment status
 Full-time 127 39.5%
 Part-time 41 12.8%
 Retired 10 3.1%
 Student 92 28.7%
 Disabled 7 2.2%
 Unemployed, looking for work 16 5.0%
 Unemployed, not looking for work 20 6.2%
 Other 8 2.5%
Open in a new tab

Table 2.

Primary, Secondary, and Tertiary Diagnoses

Primary Diagnosis
Secondary Diagnosis
Tertiary Diagnosis
N % N % N %
Obsessive-Compulsive Disorder 112 38.1% 6 3.5% 5 6.0%
Generalized Anxiety Disorder 42 14.3% 22 12.7% 8 9.6%
Social Anxiety Disorder 38 12.9% 19 11.0% 7 8.4%
Posttraumatic Stress Disorder 34 11.6% 8 4.6% 3 3.6%
Panic Disorder 25 8.5% 10 5.8% 3 3.6%
Specific Phobia 11 3.7% 5 2.9% 1 1.2%
Depressive Disorder 5 1.7% 53 30.6% 21 25.3%
Other 27 9.2% 50 28.9% 35 42.2%
Open in a new tab

Procedures

All study procedures were approved by the Institutional Review Board at the University of Pennsylvania and all participants provided informed consent. Participants were screened through an initial 20 to 30-minute phone screen by a trained research assistant who assessed for primary anxiety symptoms and exclusion criteria (see above). For additional details on recruitment and screening, see the Supplement. Eligible patients participated in a 2-hour intake assessment with a trained clinician and completed a pre-treatment self-report questionnaire packet using a secure online database: REDCap (Harris et al., 2009) at baseline, mid-treatment, and post-treatment (see Treatments for details). Study procedures, including clinician experience and designation of primary, secondary and tertiary diagnoses are described in more detail elsewhere (Asnaani et al., 2019).

Treatments

Treatment procedures are described in detail elsewhere (Asnaani et al., 2019). Briefly, patients were treated for their primary diagnosis using evidence-based cognitive behavioral therapy (CBT) and related treatment protocols such as exposure and response prevention (Ex/RP) for OCD (Foa et al., 2012), prolonged exposure (PE) for PTSD (Foa et al., 2007), CBT for panic (Craske et al., 2009) and SAD (Hofmann & Otto, 2008), and mindfulness for GAD (Roemer & Orsillo, 2012). Patients typically completed 10–20 sessions of CBT therapy, with data collection timepoints chosen to correspond to baseline (week 0), mid-treatment (week 7), and post-treatment (week 19). In terms of dropouts, 26.9% (n = 79) of those who initiated therapy did not complete treatment and were defined as non-completers, while 73.1% (n = 215) were completers (for details on the determination of dropout, see the Supplement). An additional 27 individuals completed the intake but did not initiate therapy; therefore, they are included only in baseline analyses, but not in treatment outcome analyses. Completers and non-completers did not significantly differ on age, interdependent/independent self-construals, or any of the symptom measures (see Supplemental Table 1).

Measures

Interdependent and independent self-construals were measured with the Self-Construal Scale (Singelis, 1994). Symptom measures included the Social Phobia Inventory (SPIN), the Generalized Anxiety Disorder, 7-item scale (GAD-7), the Panic Disorder Severity Scale (PDSS), the Posttraumatic Diagnostic Scale for DSM-5 (PDS-5), the Obsessive-Compulsive Inventory (OCI-R), and the Beck Depression Inventory II (BDI-II). For details on these measures see the Supplement.

Data Analysis

All analyses were conducted using IBM SPSS Statistics, version 26. We first report the correlations between interdependent and independent self-construals and the symptom measures. We then wanted to examine the specificity of the relationship between interdependent and independent self-construals and SAD symptoms. In particular, we were interested in whether the relationship between interdependent/independent self-construals and SAD symptom severity was specific to SAD or apparent for other anxiety or depressive symptom measures as well. To do this, we used a canonical correlation analysis (CCA) with interdependent and independent self-construals as predictors of the symptom measures (SPIN, GAD-7, PDSS, PDS, OCI-R, and BDI-II).

CCA investigates commonalities between two sets of intercorrelated variables while controlling for multiple comparisons. CCA is especially useful when dealing with multiple independent variables or predictors (such as interdependent and independent self-construals) and multiple dependent variables or criterion variables (such as SPIN, GAD-7, PDSS, PDS, OCI-R, and BDI-II). To evaluate the relationship between multiple predictor and criterion variables, CCA creates a single synthetic (or latent) variable from the predictor variables and a single synthetic variable from the criterion variables using linear equations (Sherry & Henson, 2005). CCA then performs a bivariate correlation on these two synthetic variables (see Figure 1), which attempts to maximize the correlation between the two synthetic variables. In other words, this approach identifies unique relationships between the predictor and criterion variables while controlling for all variables in the model. A CCA approach is appropriate when the goal is to explain the relationship between two sets of correlated variables while limiting the probability of committing Type I errors (Sherry & Henson, 2005). CCA is also advantageous because it takes into account complex relationships between variables better than performing separate analyses for each disorder. In CCA, the overall model significance is reported using Wilks’ lambda. Based on the recommendation of Sherry and Henson (2005), we interpreted canonical functions that could explain at least 10.0% of the variance in their respective function. For all predictor and criterion variables, standardized canonical function coefficients are reported. We also report the structure coefficient (rs), which is the bivariate correlation between an observed variable and a synthetic variable. Squared canonical structure coefficients (rs2) represent the proportion of the variance shared between an observed variable and the synthetic variable. The squared canonical correlation (Rc2) is the proportion of variance shared by two synthetic variables and represents a measure of effect size (Vacha-Haase & Thompson, 2004).

Figure 1. Canonical Correlation Diagram between Interdependent/Independent Self-Construals and the Symptom Measures.

Figure 1

Open in a new tab

Note. Diagram illustrating the first function in a canonical correlation analysis with two predictors (interdependent self-construal and independent self-construal) and six symptom measure criterion variables (SPIN, GAD-7, PDSS, PDS-5, OCI-R, and BDI-II). Pearson’s Rc2 represents the correlation between the synthetic predictor and criterion variables that result from a linear combination of the observed variables. SPIN = Social Phobia Inventory; GAD-7 = Generalized Anxiety Disorder 7-item scale; PDSS = Panic Disorder Severity Scale; PDS-5 = Posttraumatic Diagnostic Scale for the DSM-5; OCI-R = Obsessive-Compulsive Inventory-Revised; BDI-II = Beck Depression Inventory, Second Edition.

Finally, mixed-effects models were used to examine the effects of interdependent and independent self-construals on treatment outcomes. Specifically, we examined whether interdependent or independent self-construals moderated the reduction in SPIN scores over the course of treatment. We also examined whether interdependent or independent self-construals changed over the three time points. Analyses were intent-to-treat and included both completers and non-completers. All models were tested with and without gender and race/ethnicity added as covariates. Little’s MCAR test demonstrated that non-completer data were missing at random (interdependent self-construal: χ2 = 3.35, p = .646; independent self-construal: χ2 = 6.79, p = .236; SPIN: χ2 = 3.91, p = .563). For additional details on the mixed effects modeling, see the Supplement.

Results

Correlations with measures of interest

First, we examined the relationship between interdependent and independent self-construals and the symptom measures. As predicted, interdependent self-construal was positively associated with SAD symptoms while independent self-construal was negative associated with SAD symptoms (see Table 3 for correlations, means, and standard deviations at baseline; Supplemental Tables 5 and 6 for these same measures at mid-treatment and post-treatment). Independent self-construal was also negatively associated with depressive symptoms at baseline. There were no other significant correlations between interdependent or independent self-construals and the other symptom measures at baseline. However, individual correlations are limited because they do not take the interrelationships among variables into account.

Table 3.

Means, Standard Deviations, and Intercorrelations between Variables at Baseline

Variable 1 2 3 4 5 6 7 8
1. Interdependent self-construal ―
2. Independent self-construal .02 ―
3. SPIN .18** −.40*** ―
4. GAD-7 .04 −.06 .37*** ―
5. PDSS −.04 .07 .28*** .51*** ―
6. PDS-5 .03 −.001 .30*** .34*** .40*** ―
7. OCI-R .11 −.03 .28*** .40*** .18** .21*** ―
8. BDI-II .05 −.18** .48*** .65*** .47*** .47*** .43*** ―
M 4.64 4.53 23.11 12.53 7.85 12.84 16.41 21.33
SD 0.68 0.80 16.88 5.86 7.05 20.20 12.73 12.62
α .71 .79 .95 .89 .94 .96 .89 .93
Open in a new tab

Note.

*

p < .05;

**

p ≤ .01;

***

p ≤ .001; SPIN = Social Phobia Inventory; GAD-7 = Generalized Anxiety Disorder 7-item scale; PDSS = Panic Disorder Severity Scale; PDS-5 = Posttraumatic Diagnostic Scale for the DSM-5; OCI-R = Obsessive-Compulsive Inventory-Revised; BDI-II = Beck Depression Inventory, Second Edition; M = mean; SD = standard deviation; α = Cronbach’s alpha.

Canonical correlation analysis between interdependent and independent self-construals and symptom measures

To account for the interrelationships among the measures, we used canonical correlation analysis (CCA). The CCA results using the interdependent and independent self-construal scales as predictors of the symptom measures (SPIN, GAD-7, PDSS, PDS, OCI-R, and BDI-II) yielded a significant overall model (Wilks’s λ = .74, p < .001) and explained 26.2% of the variance shared between the variable sets. The hierarchical testing of the functions revealed that only one function was statistically significant (F(12, 626) = 8.54, p < .001). Examination of the squared canonical correlation or Rc2 effects revealed that this function explained 25.4% of the variance in the variable sets. The standardized canonical function coefficients, structure coefficients (rs), and squared structure coefficients (rs2) are presented in Table 4.

Table 4.

Canonical Solution for Interdependent and Independent Self-Construals as Predictors of the Symptom Measures

Function 1
Coef rs rs2 (%)
Predictor Variables
 Interdependent Self-Construal .383 .369 13.6
 Independent Self-Construal −.929 −.924 85.4
Criterion Variables
 SPIN .975 .868 75.3
 GAD-7 −.102 .119 1.4
 PDSS −.436 −.181 3.3
 PDS-5 −.194 .003 <0.1
 OCI-R −.096 .139 1.9
 BDI-II .285 .355 12.6
Rc2 -- -- 25.4
Open in a new tab

Note. Coef = standardized canonical function coefficient; rs = structure coefficient; rs2 = squared structure coefficient; Rc2 = squared canonical correlation; SPIN = Social Phobia Inventory; GAD-7 = Generalized Anxiety Disorder 7-item scale; PDSS = Panic Disorder Severity Scale; PDS-5 = Posttraumatic Diagnostic Scale for the DSM-5; OCI-R = Obsessive-Compulsive Inventory-Revised; BDI-II = Beck Depression Inventory, Second Edition.

Examination of the structure coefficients (rs) in Table 4 shows that the primary predictor variable was independent self-construal with interdependent self-construal making less of a contribution to the synthetic criterion variable. The relative contribution of each self-construal variable can also be seen in the squared structure coefficients (rs2), which indicate the proportion of the variance shared between an observed variable and the synthetic variable. The criterion variables are shown in the lower half of Table 4. As can be seen by rs and rs2, the primary criterion variable was SPIN, with BDI-II possibly making secondary contributions to the synthetic criterion variable. Using the convention of interpreting structure coefficients (rs) greater than |.45| (Sherry & Henson, 2005), the results of this CCA analysis show that the primary relationship to emerge from these variables was between independent self-construal and SPIN scores, with less independent self-construal being associated with greater SAD symptoms. Using a less stringent interpretative cutoff, a less robust relationship may also be present for BDI-II scores, with less independent self-construal and more interdependent self-construal being associated with both greater SAD symptoms and greater depressive symptoms. Of note, these results remain similar regardless of whether gender or race/ethnicity were included as covariates in the model.

Interdependent and independent self-construals as moderators of reduction in SAD symptom severity

Given the association between interdependent and independent self-construals and SAD symptoms, we next examined whether either of these measures moderated the reduction in SPIN scores over the course of treatment in a subset of patients with a primary, secondary, or tertiary diagnosis of SAD with complete data on the variables of interest (n = 64). SPIN scores significantly decreased over the course of treatment in patients diagnosed with SAD (p < .001). Supplemental Table 7 shows the multilevel model estimates, standard errors, t values, p values, and model fit indices. Neither interdependent self-construal nor independent self-construal significantly moderated the reduction in SAD symptoms (p-values ≥ .490). The lack of moderation was apparent whether or not gender or race/ethnicity were included as covariates. A significant main effect of independent self-construal was found, suggesting that higher independent self-construal was associated with fewer social anxiety symptoms throughout treatment.

Change over time in interdependent and independent self-construals

Finally, we examined whether interdependent or independent self-construal scores change over time in the same patients with a primary, secondary, or tertiary diagnosis of SAD. Interdependent self-construal significantly decreased over time, while independent self-construal increased. See supplemental Table 8 for estimates, standard errors, t values, p values, and model fit indices. The results did not substantially differ whether or not the model controlled for gender or race/ethnicity. Of note, changes in self-construals were tied to specific items that correlated strongly with SAD symptoms. Inspection of the individual items of the independent self-construal scale revealed high negative correlations between SPIN scores and items such as, “Speaking up during a class (or a meeting) is not a problem for me” (r = −.56, p < .001) and “I can talk openly with a person who I meet for the first time, even when this person is much older than I am” (r = −.50, p < .001). Increased endorsement of these items would be expected after SAD treatment and likely drives the observed increase in independent self-construal over time. Likewise, some interdependent self-construal items showed high positive correlations with SPIN scores such as, “I usually go along with what others want to do, even when I would rather do something different” (r = .35, p < .001) and “Even when I strongly disagree with group members, I avoid an argument” (r = .31, p < .001). Endorsement of such items might be expected to decrease with SAD treatment. Thus, the changes over time in interdependent and independent self-construals are possibly attributable to the high overlap between particular self-construal items and SPIN scores.

Discussion

The first purpose of the current study was to examine whether the relationship between interdependent/independent self-construals and symptoms is specific to SAD or indicative of a broader relationship with anxiety or depression more generally. Using a canonical correlation analysis (CCA) to control for the overlapping variance between measures, the results showed that those endorsing less independent self-construals reported greater SAD symptoms but not other anxiety symptoms. Furthermore, a less robust relationship was also present for depressive symptoms, with those endorsing less independent self-construals and more interdependent self-construals showing both greater SAD symptoms and greater depressive symptoms. The second aim of the current study was to examine interdependent and independent self-construals as moderators of the change in SAD symptom severity over the course of treatment and to examine whether interdependent or independent self-construals changed over time. While interdependent and independent self-construals did not moderate SAD treatment outcomes, there was a decrease in interdependent self-construal and an increase in independent self-construal over a course of cognitive behavioral therapy. Finally, we show that the change over time in interdependent and independent self-construals was tied to specific items that correlated strongly with SAD symptoms.

The finding that less independent self-construal was associated with greater SAD symptoms is consistent with prior research in non-clinical samples (Krieg & Xu, 2015; Norasakkunkit & Kalick, 2009; Okazaki, 1997). Specifically, Okazaki (1997) showed that American students who endorse less independent self-construals report significantly greater social avoidance and distress. Additionally, meta-analytic findings show that the higher SAD symptom severity found in those with Asian heritage residing in Western countries was dependent, in part, on lower independent self-construal in particular (Krieg & Xu, 2015). The current study builds on this prior work by showing a similar relationship between independent self-construal and SAD symptoms in a treatment-seeking clinical sample. We also extend prior work by demonstrating that the relationship between independent self-construal and SAD symptoms is not found for other anxiety symptoms by using CCA to control for the variance shared across measures. The CCA results revealed a specific relationship between independent self-construal and SAD symptoms that was not found for OCD, generalized anxiety, panic, or PTSD and only weakly for depression. This suggests that lower endorsement of independent self-construal is related to greater SAD symptoms in particular, rather than anxiety or depression more generally, in this U.S.-American context.

These findings must be interpreted against the backdrop of the larger cultural context of where this clinic was located: the U.S., a cultural context which generally favors independent self-construals. In a place where individualism is highly valued, it would be expected that one would be more anxious about interacting with others if he or she adheres less to independent cultural values (Hofmann et al., 2010; Markus & Kitayama, 1991). In contrast, interdependent self-construal showed a less robust relationship with SAD symptoms in the current study. The weaker relationship between interdependent self-construal and SAD symptoms is consistent with prior studies (Krieg & Xu, 2015; Norasakkunkit & Kalick, 2009; Okazaki, 1997) and may be due to the type of social phobia measured in the current study, which was self-focused (fear of embarrassing oneself in front of others) rather than other-focused (fear of offending others). Studies have shown a stronger relationship between interdependent self-construal and other-focused social anxiety symptoms (Dinnel et al., 2002; Norasakkunkit et al., 2012). Given these different manifestations of SAD symptoms, social anxiety measures commonly used in Western countries may not capture culturally variant symptoms of SAD and may conceivably pathologize interdependent self-construals (Hofmann et al., 2010; Norasakkunkit & Kalick, 2002). For instance, a higher degree of fear of negative evaluation (a core symptom of SAD in Western conceptualizations of the disorder) may be culturally appropriate in an interdependent, closely connected social network, and not at all indicative of a disorder in non-Western countries. Indeed, the prevalence rates of SAD are actually lower in many non-Western countries.

The research on the impact of cultural orientation on SAD symptoms reveals an initially paradoxical relationship with those from collectivist countries endorsing greater SAD symptoms than those from individualist countries (Heinrichs et al., 2006; Hong & Woody, 2007; Wong & Moulds, 2014), while at the same time, collectivist countries show a lower prevalence of SAD diagnoses than individualist countries (Hofmann et al., 2010). There are multiple possible explanations for this paradox. One possibility is that this apparent discrepancy may be explained by the degree of cultural acceptance of SAD symptoms. Having SAD symptoms (e.g., being easily embarrassed, having significant concerns about being evaluated negatively) may be more acceptable in a collectivist culture, and thus would not necessarily lead to a diagnosis of SAD, keeping the prevalence low. For instance, shy and inhibited children in collectivist countries are more accepted by peers, are more likely to be considered for awards and leadership positions, and are regarded as more competent by their teachers (Chen et al., 1995). Conversely, showing these same tendencies in an individualist culture may lead to rejection, causing distress/impairment and increasing the likelihood that these traits will be labeled as pathological, thereby increasing the prevalence of SAD in Western countries. Thus, the lower prevalence of SAD in collectivist cultures may be due to greater cultural acceptance of these symptoms.

Another possible explanation for this paradox is the “reference-group effect” proposed by Heine and colleagues that posits that individuals make social comparisons relative to similar others and that the reference group for this comparison is key to enhancing or minimizing perceived differences (Heine et al., 2002). The authors suggest that the reference-group effect is especially prevalent when using subjective Likert scales for cross-cultural comparisons. Thus, it is possible that the paradoxical relationship where collectivistic cultural contexts tend to show lower SAD prevalence rates than individualistic cultural contexts may be due to making comparisons relative to similar others (i.e., “my fear of embarrassment is about the same as my neighbor”) rather than dissimilar others.

Finally, a third possibility comes from Chiao and Blizinsky’s culture-gene coevolutionary theory, which posits that cultural values operate under the same principles of evolution and adaptation as genetic selection (Chiao & Blizinsky, 2010). They showed that collectivistic cultures were significantly more likely to carry the short (S) allele of the serotonin transporter functional polymorphism (5-HTTLPR); however, despite this genetic vulnerability, the prevalence rates of anxiety and mood diagnoses remains lower in collectivist countries. The authors used a mediation analysis to demonstrate that the relationship between higher genetic vulnerability and lower prevalence rates may be due to increased collectivist cultural values (Chiao & Blizinsky, 2010). Thus, the culture-gene coevolutionary theory suggests that cultural values that promote social support may mitigate this genetic risk, thereby leading to lower prevalence rates of SAD.

Additionally, given the lack of research on interdependent/independent self-construals and treatment outcomes, we examined two additional exploratory aims. First, we analyzed the impact of interdependent and independent self-construals on the change in SAD symptom severity over the course of cognitive behavioral treatment. And second, we examined whether interdependent or independent self-construals changed over time. Using longitudinal analyses, we found that neither interdependent nor independent self-construals moderated SAD symptom reduction during the course of CBT therapy; however, those who endorsed a greater level of independent self-construal showed lower social anxiety symptoms throughout treatment.

Finally, there was a decrease in interdependent self-construal and an increase in independent self-construal over time. These changes were tied to specific items that correlated strongly with SAD symptoms. Thus, it is possible that these observed changes over time are attributable to overlapping items on the self-construal and SAD symptom measures, which has been highlighted as a possible explanation in other studies examining self-construals in SAD (Norasakkunkit & Kalick, 2002). Alternatively, if interdependent and independent self-construals are malleable constructs, the current results may suggest that cognitive behavioral therapy not only impacts SAD symptom reduction, but may also influence cultural interpretations and perceptions. This is consistent with the process of socialization in therapy, where the therapist and patient continually negotiate a shared understanding of the therapeutic course and goals within the context of cultural norms (Daniels & Wearden, 2011; Fong et al., 2016). Such micro-cultures created by the therapeutic setting may provide safe spaces for the exploration of cultural contingencies, potentially leading to shifts in perceptions. Likewise, it is possible that CBT increases independent self-construals and decreases interdependent self-construals. CBT aims to “correct” negative biases that are contrary to self-enhancement or positive self-regard. Others have noted that the emphasis on positive self-regard is not universal; rather, it is more common in cultures that favor an independent self-construal (Heine et al., 1999). Thus, it may not be surprising that CBT decreased interdependent self-construals and increased independent self-construals in the current study. Notably, there may be an important distinction between an individual’s pre-existing level of self-construal and those activated during the therapeutic process, as prior work suggests that priming for independent self-construal reduces SAD symptoms in both the U.S. and Japan (Norasakkunkit & Kalick, 2009). However, it remains unclear whether interdependent and independent self-construals would have changed in the absence of therapy. To disentangle the effects of SAD treatment and spontaneous change over time in interdependent and independent self-construals, a longitudinal study that measures self-construals repeatedly outside the context of treatment is needed.

A limitation of the current study was the presence of race/ethnicity as the only proxy for cultural background due to a lack of data on participants’ countries of origin or cultural community affiliation, both of which would have allowed us to test whether those endorsing greater interdependent self-construals were from more collectivistic cultural contexts. Another limitation of the current study includes the reliance on self-report measures for the variables of interest (interdependent/independent self-construals, SAD, other anxiety symptoms, and depression). As noted previously, there is currently no consensus on how interdependent and independent self-construals should be operationalized and measured (Gudykunst & Lee, 2003; Levine et al., 2003). The two-dimensional interdependent/independent self-construal measure used in the current study treats independence and interdependence as separate and unitary dimensions, which has been brought into question (Dinnel et al., 2002; Vignoles et al., 2016). While our measure of interdependent/independent self-construals has considerable psychometric support for its use, adequate operationalizing and testing of this construct remains an important consideration when interpreting results regarding interdependent/independent self-construals and mental health/treatment outcomes. Additionally, SAD measures may be biased towards pathologizing interdependent self-construals based on a Western conceptualization of SAD (Norasakkunkit & Kalick, 2002). Future studies should consider use of a battery of cultural measures to test consistency and deviations in observations in order to develop more comprehensive measures of interdependent and independent self-construals (Asnaani et al., 2010). The field would also likely benefit more broadly from an increase in the routine incorporation of measures of cultural orientation across cross-sectional and longitudinal clinical studies both in naturalistic samples and randomized controlled trials (Hofmann et al., 2010).

Together, these results increase our understanding of the relationship between interdependent/independent self-construals and SAD symptoms in treatment-seeking anxiety patients in a U.S.-American context. The current study found considerable variability in interdependent and independent self-construals even within a single cultural context. This study suggests that different types of self-construal are important cultural constructs to consider in mental health research.

Supplementary Material

Supplement

Acknowledgements

The authors would like to first express their sincerest appreciation to Dr. Edna B. Foa, who is the Director and Founder of the Center for the Treatment and Study of Anxiety (CTSA) at the University of Pennsylvania, who has generously supported the integration of research into the CTSA treatment clinic and has ensured that the entire clinical team at CTSA continues to provide evidence-based treatments to patients seeking treatment at our clinic. The authors would also like to acknowledge Jody Zhong, the research assistant who assisted heavily on the original creation of the REDCap database and data infrastructure, coordinated data collection and patient reminders, and assisted with creation of the database for analysis, and Kathy Benhamou, the subsequent research assistant who enhanced this infrastructure and greatly streamlined our processes to make the data collection and database creation what it is today. We would also like to thank the current research assistant on the project, Jesse McCann, who assisted in the cleaning and set-up of the full data examined in the current study and Jeremy Tyler, who oversees the management of the study. The authors would like to acknowledge the creators of REDCap (Harris et al., 2009), the data acquisition program used to collect the data analyzed in the present study. This software reduced burden on patients by allowing them to provide responses at home, and facilitated easier and more accurate access of treatment outcome data compared to traditional paper and pencil formats. Finally, we would also like to deeply thank all the patients seeking treatment at our Center who were willing to allow us to analyze their deidentified data in order to better understand the efficacy of our treatments on symptom reduction and other constructs of interest throughout their treatment at our facility.

Dr. Kaczkurkin’s contribution is supported in part by the National Institute of Mental Health (grant number: R00MH117274), a NARSAD Young Investigator Award from the Brain & Behavior Foundation, and a Sloan Research Fellowship.

Biographies

Dr. Antonia Kaczkurkin is an Assistant Professor in the Department of Psychology at Vanderbilt University. Her research focuses on identifying neurobiological markers of internalizing disorders to develop a comprehensive understanding of the basic mechanisms underlying anxiety and depressive disorders.

Dr. Anu Asnaani is an Assistant Professor in the Department of Psychology at the University of Utah, where she directs the Treatment Mechanisms, Community Empowerment, and Technology Innovations (TCT) Lab. Her research examines the mechanisms underlying effective treatments for fear-based disorders, and ways to optimize such treatments for diverse local and global communities.

Dr. Lily Brown is the Director of the Center for the Treatment and Study of Anxiety and an Assistant Professor in the Department of Psychiatry at the University of Pennsylvania. Her research focuses on anxiety-related disorders and suicide risk.

Savannah Simon contributed to this project through the Penn Undergraduate Research Mentorship (PURM) 10-week summer program funded by the Center for Undergraduate Research and Fellowships (CURF) at the University of Pennsylvania.

References

  1. American Psychiatric Association. (2013). Diagnostic and Statistical Manual of Mental Health Disorders. In American Psychiatric Association (5th ed.). American Psychiatric Association. [Google Scholar]
  2. Asnaani A, Benhamou K, Kaczkurkin AN, Turk-Karan E, & Foa EB (2020). Beyond the constraints of an RCT: Naturalistic treatment outcomes for anxiety-related disorders. Behavior Therapy, 51, 434–446. 10.1016/j.beth.2019.07.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Asnaani A, Richey JA, Dimaite R, Hinton DE, & Hofmann SG (2010). A cross-ethnic comparison of lifetime prevalence rates of anxiety disorders. Journal of Nervous and Mental Disease, 198(8), 551–555. 10.1097/NMD.0b013e3181ea169f [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Caldwell-Harris CL, & Ayçiçegi A (2006). When Personality and Culture Clash: ThE Psychological Distress of Allocentrics in an Individualist Culture and Idiocentrics in a Collectivist Culture. Transcultural Psychiatry, 43(3), 331–361. 10.1177/1363461506066982 [DOI] [PubMed] [Google Scholar]
  5. Chen X, Rubin KH, & Li B (1995). Social and school adjustment of shy and aggressive children in China. Development and Psychopathology, 7(2), 337–349. 10.1017/S0954579400006544 [DOI] [Google Scholar]
  6. Chiao JY, & Blizinsky KD (2010). Culture-gene coevolution of individualismcollectivism and the serotonin transporter gene. Proceedings of the Royal Society B: Biological Sciences, 277(1681), 529–537. 10.1098/rspb.2009.1650 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Connor KM, Davidson JRT, Churchill LE, Sherwood A, Weisler RH, & Foa E (2002). Psychometric properties of the Social Phobia Inventory (SPIN). British Journal of Psychiatry, 176(4), 379–386. 10.1192/bjp.176.4.379 [DOI] [PubMed] [Google Scholar]
  8. Craske MG, Roy-Byrne PP, Stein MB, Sullivan G, Sherbourne C, & Bystritsky A (2009). Treatment for anxiety disorders: Efficacy to effectiveness to implementation. Behaviour Research and Therapy, 47(11), 931–937. 10.1016/j.brat.2009.07.012 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Daniels J, & Wearden AJ (2011). Socialization to the model: The active component in the therapeutic Alliance? A preliminary study. Behavioural and Cognitive Psychotherapy, 39(2), 221–227. 10.1017/S1352465810000792 [DOI] [PubMed] [Google Scholar]
  10. Dinnel DL, Kleinknecht RA, & Tanaka-Matsumi J (2002). A cross-cultural comparison of social phobia symptoms. Journal of Psychopathology and Behavioral Assessment, 24(2), 75–84. 10.1023/A:1015316223631 [DOI] [Google Scholar]
  11. Fehm L, Beesdo K, Jacobi F, & Fiedler A (2008). Social anxiety disorder above and below the diagnostic threshold: Prevalence, comorbidity and impairment in the general population. Social Psychiatry and Psychiatric Epidemiology, 43, 257–265. 10.1007/s00127-007-0299-4 [DOI] [PubMed] [Google Scholar]
  12. Foa EB, Hembree EA, & Rothbaum B (2007). Prolonged exposure therapy for PTSD. Oxford University Press. [Google Scholar]
  13. Foa EB, Yadin E, & Lichner TK (2012). Exposure and response (ritual) prevention for obsessive-compulsive disorder: Therapist guide (2nd ed.). Oxford University Press. [Google Scholar]
  14. Fong EH, Catagnus RM, Brodhead MT, Quigley S, & Field S (2016). Developing the Cultural Awareness Skills of Behavior Analysts. Behavior Analysis in Practice, 9(1), 84–94. 10.1007/s40617-016-0111-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Gudykunst WB, & Lee CM (2003). Assessing the Validity of Self Construal Scales. Human Communication Research, 29(2), 253–274. 10.1111/j.1468-2958.2003.tb00838.x [DOI] [Google Scholar]
  16. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, & Conde JG (2009). Research electronic data capture (REDCap)-A metadata-driven methodology and workflow process for providing translational research informatics support. Journal of Biomedical Informatics, 42(2), 377–381. 10.1016/j.jbi.2008.08.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Heine SJ, Lehman DR, Peng K, & Greenholtz J (2002). What’s wrong with cross-cultural comparisons of subjective Likert scales?: The reference-group effect. Journal of Personality and Social Psychology, 82(6), 903–918. 10.1037/0022-3514.82.6.903 [DOI] [PubMed] [Google Scholar]
  18. Heine SJ, Markus HR, Lehman DR, & Kitayana S (1999). Is there a universal need for positive self-regard? Psychological Review, 106(4), 766–794. 10.1037/0033-295X.106.4.766 [DOI] [PubMed] [Google Scholar]
  19. Heinrichs N, Rapee RM, Alden LA, Bögels S, Hofmann SG, Ja Oh K, & Sakano Y (2006). Cultural differences in perceived social norms and social anxiety. Behaviour Research and Therapy, 44, 1187–1197. 10.1016/j.brat.2005.09.006 [DOI] [PubMed] [Google Scholar]
  20. Hofmann SG, Asnaani A, & Hinton DE (2010). Cultural aspects in social anxiety and social anxiety disorder. Depression and Anxiety, 27(12), 1117–1127. 10.1002/da.20759 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Hofmann SG, & Otto MW (2008). Cognitive behavioral therapy for social anxiety disorder: Evidence-based and disorder-specific treatment techniques. Routledge. 10.4324/9780203927526 [DOI] [Google Scholar]
  22. Hong JJ, & Woody SR (2007). Cultural mediators of self-reported social anxiety. Behaviour Research and Therapy, 45(8), 1779–1789. 10.1016/j.brat.2007.01.011 [DOI] [PubMed] [Google Scholar]
  23. Kashima Y, Koval P, & Kashima ES (2011). Reconsidering Culture and Self. Psychological Studies, 56(1), 12–22. 10.1007/s12646-011-0071-4 [DOI] [Google Scholar]
  24. Kessler RC, Berglund P, Demler O, Jin R, Merikangas KR, & Walters EE (2005). Lifetime Prevalence and Age-of-Onset Distributions of DSM-IV Disorders in the National Comorbidity Survey Replication. Arch Gen Psychiatry, 62, 593–602. 10.1001/archpsyc.62.6.593 [DOI] [PubMed] [Google Scholar]
  25. Kitayama S, & Park H (2007). Cultural Shaping of Self, Emotion, and Well-Being: How Does It Work? Social and Personality Psychology Compass, 1(1), 202–222. 10.1111/j.1751-9004.2007.00016.x [DOI] [Google Scholar]
  26. Krieg A, & Xu Y (2015). Ethnic differences in social anxiety between individuals of Asian Heritage and European Heritage: A meta-analytic review. Asian American Journal of Psychology, 6(1), 66–80. 10.1037/a0036993 [DOI] [Google Scholar]
  27. Levine TR, Bresnahan MJ, Park HS, Lapinski MK, Wittenbaum GM, Shearman SM, Lee SY, Chung D, & Ohashi R (2003). Self-construal scales lack validity. Human Communication Research, 29(2), 210–252. 10.1093/hcr/29.2.210 [DOI] [Google Scholar]
  28. Markus HR, & Kitayama S (1991). Culture and the self: Implications for cognition, emotion, and motivation. Psychological Review, 98(2), 224–253. 10.1037/0033-295X.98.2.224 [DOI] [Google Scholar]
  29. McKnight PE, Monfort SS, Kashdan TB, Blalock DV, & Calton JM (2016). Anxiety symptoms and functional impairment: A systematic review of the correlation between the two measures. Clinical Psychology Review, 45, 115–130. 10.1016/j.cpr.2015.10.005 [DOI] [PubMed] [Google Scholar]
  30. Norasakkunkit V, & Kalick SM (2002). Culture, ethnicity, and emotional distress measures: The role of self-construal and self-enhancement. Journal of Cross-Cultural Psychology, 33(1), 56–70. 10.1177/0022022102033001004 [DOI] [Google Scholar]
  31. Norasakkunkit V, & Kalick SM (2009). Experimentally detecting how cultural differences on social anxiety measures misrepresent cultural differences in emotional well-being. Journal of Happiness Studies, 10(3), 313–327. 10.1007/s10902-007-9082-1 [DOI] [Google Scholar]
  32. Norasakkunkit V, Kitayama S, & Uchida Y (2012). Social Anxiety and Holistic Cognition: Self-Focused Social Anxiety in the United States and Other-Focused Social Anxiety in Japan. Journal of Cross-Cultural Psychology, 43(5), 742–757. 10.1177/0022022111405658 [DOI] [Google Scholar]
  33. Okazaki S (1997). Sources of ethnic differences between Asian American and White American college students on measures of depression and social anxiety. Journal of Abnormal Psychology, 106(1), 52–60. 10.1037/0021-843X.106.1.52 [DOI] [PubMed] [Google Scholar]
  34. Patel A, Knapp M, Henderson J, & Baldwin D (2002). The economic consequences of social phobia. Journal of Affective Disorders, 68(2–3), 221–233. 10.1016/S0165-0327(00)00323-2 [DOI] [PubMed] [Google Scholar]
  35. Pilarska A (2014). Self-Construal as a Mediator Between Identity Structure and Subjective Well-Being. Current Psychology, 33(2), 130–154. 10.1007/s12144-013-9202-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Roemer L, & Orsillo SM (2012). Anxiety disorders: Acceptance, compassion and wisdom. In Germer C & Siegel R (Eds.), Wisdom and Compassion in Psychotherapy: Deepening Mindfulness in Clinical Practice (pp. 234–248). The Guilford Press. [Google Scholar]
  37. Sareen J, Cox BJ, Afifi TO, de Graaf R, Asmundson GJG, ten Have M, & Stein MB (2005). Anxiety Disorders and Risk for Suicidal Ideation and Suicide Attempts. Archives of General Psychiatry, 62(11), 1249. 10.1001/archpsyc.62.11.1249 [DOI] [PubMed] [Google Scholar]
  38. Sherry A, & Henson RK (2005). Conducting and interpreting canonical correlation analysis in personality research: A user-friendly primer. Journal of Personality Assessment, 84(1), 37–48. 10.1207/s15327752jpa8401_09 [DOI] [PubMed] [Google Scholar]
  39. Singelis T (1994). The measurement of independent and interdependent self-construals. Personality and Social Psychology Bulletin, 20(5), 580–591. [Google Scholar]
  40. Vacha-Haase T, & Thompson B (2004). How to estimate and interpret various effect sizes. Journal of Counseling Psychology, 51(4), 473–481. 10.1037/0022-0167.51.4.473 [DOI] [Google Scholar]
  41. Vandello JA, & Cohen D (1999). Patterns of Individualism and Collectivism Across the United States. Journal of Personality and Social Psychology, 77(2), 279–292. 10.1037/0022-3514.77.2.279 [DOI] [Google Scholar]
  42. Vignoles VL, Owe E, Becker M, Smith PB, Easterbrook MJ, Brown R, González R, Didier N, Carrasco D, Cadena MP, Lay S, Schwartz SJ, Des Rosiers SE, Villamar JA, Gavreliuc A, Zinkeng M, Kreuzbauer R, Baguma P, Martin M, … Bond MH (2016). Beyond the “East-West” dichotomy: Global variation in cultural models of selfhood. Journal of Experimental Psychology: General, 145(8), 966–1000. 10.1037/xge0000175 [DOI] [PubMed] [Google Scholar]
  43. Wong QJJ, & Moulds ML (2014). An Examination of the Measurement Equivalence of the Brief Fear of Negative Evaluation Scale Across Individuals Who Identify With an Asian Ethnicity and Individuals Who Identify With a European Ethnicity. Assessment, 21(6), 713–722. 10.1177/1073191114528570 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplement
NIHMS1929226-supplement-Supplement.docx (65.5KB, docx)

RESOURCES