Shorter communication
Threat-related automatic associations in socially anxious adolescents

https://doi.org/10.1016/j.brat.2011.05.008Get rights and content

Abstract

Threat-related automatic associations are assumed to play an important role in the development and maintenance of social anxiety. We tested whether threat-related automatic associations are already evident in high socially anxious adolescents, by comparing a group of adolescents (age 12–15) with subclinical levels of social anxiety (n = 170) to a group of low socially anxious adolescents (n = 193). We used a single-target implicit association test to measure threat-related automatic associations to social cues. Results showed that indeed in high socially anxious adolescents social cues automatically elicited relatively strong threat-related associations. Supporting the relevance of differentiating between automatic and more explicit measures, both automatic and explicit associations were independently associated with adolescents’ level of self-reported social anxiety. The present pattern of findings is not only consistent with the view that automatic and more deliberate threat-related associations are both involved in the etiology of social-anxiety symptoms, but also suggest that both types of associations are proper targets for early intervention programs.

Highlights

► We examined automatic associations in high vs. low socially anxious adolescents. ► In socially anxious adolescents, social cues elicit strong threat-related associations. ► Automatic and explicit associations are both independently related to self-reported social anxiety.

Introduction

Social anxiety typically develops during adolescence and elevated levels of social anxiety may have a large impact on current and future functioning (Wittchen, Nelson, & Lachner, 1998). For example, it may prevent adolescents from developing their social skills, hinder school performance (Wittchen, Stein, & Kessler, 1999), and may give rise to substance abuse, depression, social isolation, and poor functioning in work or studies (Essau, Conradt, & Petermann, 1999).

Current cognitive models of anxiety in children (Daleiden and Vasey, 1997, Muris and Field, 2008) stress the importance of threat-confirming information processing biases in the development and maintenance of childhood anxiety. It has been proposed that this biased information processing elicits feelings of fear and anxiety, which in turn enhances the occurrence of cognitive biases and further strengthens the associations between the threatening situation and anxious feelings in memory (Huijding, Wiers, & Field, 2010). Recent dual process models (e.g., Ouimet, Gawronski, & Dozois, 2009), emphasize the importance to differentiate between initial automatic associations elicited by a threatening stimulus and explicit associations, formed afterwards in the deliberate, rule-based memory system. Automatic associations and explicit associations may diverge, when the initial automatic associations are being discarded as irrelevant or untrue. The more deliberate, reflective explicit associations are assumed to guide the more controllable behaviors, whereas the initial, reflexive automatic associations are more important for driving spontaneous feelings and behaviors (Back, Schmukle, & Egloff, 2009). In support of this, it has been shown that indeed automatic associations displayed differential predictive validity for relatively automatic fear behaviors (e.g., Egloff and Schmukle, 2002, Huijding and de Jong, 2006). Attesting to the relevance of automatic associations in the context of fear and anxiety, previous research showed that parent-reported anxiety symptoms of clinically anxious children were best predicted by the children’s automatic affective associations elicited by generally threatening pictures (Vervoort et al., 2010).

Supporting the view that automatic associations are also involved in social anxiety, socially anxious adults showed stronger threat-related associations towards social stimuli than low-anxious participants (de Jong, Pasman, Kindt, & van den Hout, 2001). Since social anxiety typically starts in early adolescence (Rapee & Spence, 2004) it is important to examine whether early adolescents who show symptoms of social anxiety are similarly characterized by fear-confirming automatic associations. If automatic associations are critically involved in the origin of social anxiety, one would expect that threat-related automatic associations are already evident in adolescents with an elevated level of social anxiety. That is, in those adolescents who are shy and show symptoms of social anxiety, but do not (yet) suffer from a clinical disorder (e.g., Neal & Edelmann, 2003). Therefore, the present study tested the strength of automatic associations between social cue words (e.g., conversation, exam) and socially threatening negative outcome words (e.g., stupid, failure) in adolescents with subclinical social anxiety compared to low socially anxious adolescents. We also assessed more deliberate (explicit) associations to examine the predictive validity of automatic associations over and above explicitly assessed associations with respect to adolescents’ current level of social anxiety.

Section snippets

Participants

Participants in this study were selected from two different groups. The high socially anxious participants (n = 179, 12–15 years) were selected on the basis of a screening in a large research project on the prevention of social and test anxiety in adolescents (Project Pasta; www.projectpasta.nl). After approval for this study by the Medical Ethical Committee of the University Medical Center, a screening took place in the first two grades of 25 high schools in the Northern part of the Netherlands.

Social anxiety

The Revised Child Anxiety and Depression Scale (RCADS; Chorpita et al., 2000) is a revised version of the Spence Children’s Anxiety Scale (SCAS; Spence, 1998). It is a 47-item self-report instrument, with items rated on a 4-point scale ranging from 0 to 3. In psychometric research the internal consistency of the scale and subscales were found to be good (Chorpita et al., 2000). In the current study only the social phobia subscale (9 items; α = .88) will be reported.

An important aspect of social

Anxiety measures

Table 1 shows means and standard deviations for all anxiety measures. Independent sample t tests indicated that the scores on social phobia (M difference = 9.21, t[361] = 28.53, p < .001), test anxiety (M difference = 15.44, t[361] = 13.81, p < .001), and fear of negative evaluation (M difference = 16.86, t[361] = 18.25, p < .001), were significantly higher for the high socially anxious group than the low socially anxious group.

Data inspection

The average overall percentage of errors made on the stIAT was 11.71% (SD = 5.07). This

Discussion

This study was designed to test whether high socially anxious adolescents are characterized by threat-related automatic associations to social stimuli. The major results can be summarized as follows: (i) high socially anxious adolescents indeed showed relatively strong automatic associations between social cue words and negative outcome words and (ii) automatic and explicit associations with social cues were independently associated with adolescents’ current self-reported level of social

Acknowledgements

This research was supported by grants from ZonMw, nr. 62200027, the Netherlands.

References (41)

  • S.H. Spence

    A measure of anxiety symptoms among children

    Behaviour Research and Therapy

    (1998)
  • C. Thush et al.

    Explicit and implicit alcohol-related cognitions and the prediction of future drinking in adolescents

    Addictive Behaviors

    (2007)
  • M.J. Voncken et al.

    Not self-focused attention but negative beliefs affect poor social performance in social anxiety: an investigation of pathways in the social anxiety–social rejection relationship

    Behaviour Research and Therapy

    (2010)
  • L.S. Aiken et al.

    Multiple regression – Testing and interpreting interactions

    (1991)
  • A.M. Albano et al.

    Social anxiety disorder

  • J.B. Asendorpf

    Implicit representations and personality

    International Journal of Psychology

    (2007)
  • M.D. Back et al.

    Predicting actual behavior from the explicit and implicit self-concept of personality

    Journal of Personality and Social Psychology

    (2009)
  • M. Bluemke et al.

    Reliability and validity of the single-target IAT (ST-IAT): assessing automatic affect towards multiple attitude objects

    European Journal of Social Psychology

    (2008)
  • R.N. Carleton et al.

    Brief fear of negative evaluation scale—revised

    Depression & Anxiety

    (2006)
  • J. De Houwer et al.

    What are implicit measures and why are we using them

    The handbook of implicit cognition and addiction

    (2006)
  • Cited by (20)

    • Do implicitly measured math–anxiety associations play a role in math behavior?

      2019, Journal of Experimental Child Psychology
      Citation Excerpt :

      The overall error percentages were 9.3%, 9.8%, and 9.8% for math, English, and trait anxiety ST-IATs, respectively, which are relatively low in comparison with other studies using the ST-IAT in adolescents (11.7%: de Hullu, de Jong, Sportel, & Nauta, 2011; 17.7% Thush & Wiers, 2007). Reliabilities, as indexed by the average Spearman–Brown corrected intercorrelations between split-half D scores for 10 random sets of trials (de Hullu et al., 2011; Karpinski & Steinman, 2006), indicated acceptable internal consistency (α = .70, α = .73, and α = .72 for math, English, and trait anxiety ST-IATs, respectively), in line with previous studies (α = .76: de Hullu et al., 2011; αs = .43–.52: Thush & Wiers, 2007). Randomization checks revealed no significant effects of key response assignment on ST-IAT scores (all ps ≥ .503).

    • Cognitive bias modification and CBT as early interventions for adolescent social and test anxiety: Two-year follow-up of a randomized controlled trial

      2017, Journal of Behavior Therapy and Experimental Psychiatry
      Citation Excerpt :

      We, therefore, expected that CBT in particular would have a beneficial effect on prosocial behaviors. Third, we included an implicit performance measure to examine whether the current interventions were effective in modifying participants’ automatic associations between social situations and indicators of a negative outcome (de Hullu, de Jong, Sportel, & Nauta, 2011). Attesting to the importance of automatic threat associations in the persistence of fear, it has been shown that residual automatic associations predict return of fear after exposure therapy for speech anxiety (Vasey, Harbaugh, Buffington, Jones, & Fazio, 2012).

    • “What I believe is true”: Belief-confirming reasoning bias in social anxiety disorder

      2016, Journal of Behavior Therapy and Experimental Psychiatry
    • Social anxiety disorder and stuttering: Current status and future directions

      2014, Journal of Fluency Disorders
      Citation Excerpt :

      In particular, evidence suggests that social anxiety and fear of negative evaluation for some individuals who stutter may be higher than community controls but less severe than clinically anxious patients (Iverach, O’Brian, et al., 2009; Kraaimaat et al., 1991; Mahr & Torosian, 1999). Given the potential for subclinical fears and anxious associations to contribute to the development of anxiety disorders (Balazs et al., 2013; de Hullu, de Jong, Sportel, & Nauta, 2011; Glashouwer, de Jong, & Penninx, 2011), it would appear relevant to consider early psychological intervention in such cases. Addressing subclinical fears before they become clinically elevated may also play a role in curtailing some of the negative consequences associated with stuttering.

    View all citing articles on Scopus
    View full text