SPINning factors: Factor analytic evaluation of the Social Phobia Inventory in clinical and nonclinical undergraduate samples

https://doi.org/10.1016/j.janxdis.2009.09.003Get rights and content

Abstract

The Social Phobia Inventory (SPIN) was designed to assess three dimensions of social anxiety (i.e., fear, avoidance, and physiological arousal) as posited by the scale authors (Connor et al., 2000). Despite expectations of a 3-factor solution, analyses of the SPIN to date have provided support for 3- and 5-factor solutions (Radomsky et al., 2006). Moreover, a 3-item version, the Mini-SPIN (Connor et al., 2001), has good sensitivity and specificity for generalized social anxiety disorder (SAD), implying some item redundancy. Another recent psychometric analysis of the SPIN was performed in a diagnostically diverse clinical sample (Antony et al., 2006); however, the study did not include a comprehensive evaluation of the factor structure. The current study was designed to comprehensively assess the SPIN factor structure using exploratory (EFA) and confirmatory (CFA) factor analyses in undergraduate (N = 227) and clinical samples (N = 355) using current recommendations for factor analyses (Osborne et al., 2008). Results suggest a 10-item 3-factor solution may be an ideal fit for clinical samples; however, using the undergraduate sample, the same solution was significantly better than precedent solutions but nonetheless not ideal. Implications and recommendations for future research are discussed.

Introduction

Social anxiety (SA) refers to anxiety or apprehension experienced in social or performance situations. People with high levels of SA fear being negatively evaluated by others, making a bad impression, or acting in a way that might be embarrassing. Social anxiety disorder (SAD; also called social phobia) refers to disabling levels of SA (American Psychiatric Association, 2000) and is the fourth most common psychiatric disorder in American samples (Kessler, Chiu, Demler, Merikangas, & Walters, 2005). Lifetime prevalence rates of SAD range from 7 to 13% (Furmark, 2002, Kessler et al., 2005a, Kessler et al., 2005b). Given the relative propensity for SAD in the general population, increased understanding and robust measurement of SA is crucial.

The latent structure of SA has been debated theoretically (Hofmann, 2007, Rapee and Heimberg, 1997) and implicitly as a function of psychometric investigations exploring the factor structure of SA-related measures (Carleton et al., 2009, Carleton et al., 2007a). Specifically, researchers have explored whether SA is taxonic (i.e., a qualitative boundary delineates pathology) or dimensional (i.e., a quantitative boundary delineates pathology) (Kollman et al., 2006, Weeks et al., 2009) as part of investigating SA cognitive models. There have also been a variety of measures implying important differences in symptom presentation—such as social interaction anxiety relative to social phobia or evaluation fears (Carleton et al., 2009). Although there is a general consensus that SA comprises a fear of negative evaluation and avoidance of social situations (Hofmann, 2007, Leary, 1983), the nature of those dimensions and the extent to which they sufficiently represent the SA construct for research and clinical utility remains uncertain (Antony and Swinson, 2000, Hofmann, 2007).

The Social Phobia Inventory (SPIN; Connor et al., 2000) was designed to assess three dimensions of SA, specifically fear, avoidance, and physiological arousal. An implicit assumption is that researchers expected delineation between fear and avoidance as described by anxiety disorder theorists (Barlow, 2002) and diagnostic protocols (American Psychiatric Association, 2000). In the original study (Connor et al., 2000), an exploratory factor analysis (EFA) was intended to evaluate and ratify the three dimensions as being factorially coherent. The EFA was conducted on data from 148 people with SAD using principal components analysis with an unspecified method of factor rotation. Eigenvalues greater than one were used to determine the number of factors retained, and in contrast to the author's expectations, a 5-factor structure emerged. The factors were conceptualized as fears of: (1) talking to strangers in social situations, (2) criticism and embarrassment, (3) physiological changes, (4) people in authority, and (5) being the center of attention and public speaking. Researchers have since used the SPIN in clinical and nonclinical samples (Ashbaugh, Antony, McCabe, Schmidt, & Swinson, 2005), and the measure has been translated from English into French (Radomsky et al., 2006).

Confirmatory factor analyses (CFAs) have been used to evaluate the 3-factor structure originally suggested by Connor et al. (2000), using data from English and French speaking undergraduate participants (Radomsky et al., 2006). Results supported the 3-factor structure with excellent fit indices; however, the authors cautioned that subsequent evaluations of the SPIN factor structure using different samples are warranted. To our knowledge to date, there have been no published CFAs evaluating the 5-factor structure found by Connor et al. when they performed their principal components analyses.

The psychometric properties of the SPIN have also been assessed in a diagnostically diverse clinical sample (Antony, Coons, McCabe, Ashbaugh, & Swinson, 2006). Results from the investigation provided support for excellent internal consistency, test–retest reliability, convergent validity, and discriminant validity. The SPIN distinguished well between people with SAD and other anxiety disorders and was sensitive to changes in SA subsequent to group cognitive–behavioral therapy. Despite the thorough nature of the analyses, the study did not include an evaluation of the factor structure of the SPIN.

Subsequent to the creation of the SPIN, a 3-item version called the Mini-SPIN was found to have good sensitivity and specificity for generalized SAD (Connor, Kobak, Churchill, Katzelnick, & Davidson, 2001). Interestingly, the Mini-SPIN has also demonstrated sound psychometric properties in a treatment-seeking sample (Weeks, Spokas, & Heimberg, 2007). Relative success of the Mini-SPIN implies the possibility that several items in the SPIN may be redundant; however, such redundancy would not necessarily detract from the utility of the SPIN.

The purpose of the present study was to assess the factor structure of the SPIN using EFA and CFA with data from both an undergraduate and a clinical sample. Changes to recommended methods for conducting EFA (Osborne, 2008) may yield increasingly robust results relative to previous methods, such as those used in the original SPIN study (Connor et al., 2000) and the translation study (Radomsky et al., 2006). The factor structure identified in the present analysis as providing the best fit to the data will also be compared to the original 3- and 5-factor structures suggested by Connor et al. (2000). Finally, sensitivity and specificity of each solution (i.e., original, Mini-SPIN, and current) will be compared using a Receiver Operating Characteristic (ROC) analysis across the undergraduate and clinical samples.

Section snippets

Participants

Two samples of participants who completed larger batteries of self-report questionnaires were obtained. The first was an undergraduate sample (n = 227) from the University of Regina [50 men, 18–29 years (Mage = 20.3; SD = 2.6) and 177 women, 18–45 years (Mage = 20.1; SD = 3.3)] all of whom scored less than 22 (M = 9.11; SD = 6.08; Mode = 4), suggesting against clinically significant levels of SA, on the Social Interaction Phobia Scale (Carleton et al., 2009)—a merged scale containing subsets of the Social

Descriptive statistics

The item means and standard deviations for the SPIN from the clinical and undergraduate samples are presented in Table 1. None of the indices of univariate skewness and kurtosis in the clinical sample was sufficiently out of range to preclude the EFA (i.e., had positive standardized skewness values that exceeded 2 or positive standardized kurtosis values that exceeded 7; see Curran et al., 1996, Tabachnick and Fidell, 2001). In the undergraduate sample, only item 8 had a positive standardized

Discussion

The current study assessed the SPIN factor structure (Connor et al., 2000) using current recommendations for EFA (Osborne, 2008) and CFA with data from clinical and undergraduate samples. Based on Osborne's recommendations and our previous experience performing factor analyses with those recommendations (Carleton, Norton, & Asmundson, 2007), we expect the current EFA results to be robust; moreover, the current study provides the first comparative examination of potential SPIN factors using CFA

Acknowledgements

Dr. Asmundson is supported by a Canadian Institutes of Health Research (CIHR) Investigator's Award (FRN: 63186), R.N. Carleton by a CIHR Canada Graduate Scholarship Doctoral Research Award (FRN: 80455), and K.C. Collimore is supported by a CIHR Doctoral Research Award (FRN: 85321).

References (48)

  • M.M. Antony et al.

    The shyness and social anxiety workbook: Proven techniques for overcoming your fears

    (2000)
  • J.L. Arbuckle

    Amos 6.0.0

    (2005)
  • A.R. Ashbaugh et al.

    Self-evaluative biases in social anxiety

    Cognitive Therapy and Research

    (2005)
  • D.H. Barlow

    Unraveling the mysteries of anxiety and its disorders from the perspective of emotion theory

    American Psychologist

    (2000)
  • D.H. Barlow

    Anxiety and its disorders

    (2002)
  • J.G. Beck et al.

    When the heat is on: Romantic partner responses influence distress in socially anxious women

    Behaviour Research and Therapy

    (2006)
  • M.W. Browne et al.

    Single sample cross-validation indices for covariance structures

    Multivariate Behavioral Research

    (1989)
  • M.W. Browne et al.

    Alternative ways of assessing model fit

  • B. Byrne

    Structural equation modeling with AMOS: Basic concepts, applications, and programming

    (2001)
  • R.N. Carleton et al.

    Refining and validating the Social Interaction Anxiety Scale and the Social Phobia Scale

    Depression and Anxiety

    (2009)
  • R.N. Carleton et al.

    The Illness/Injury Sensitivity Index: An examination of construct validity

    Depression and Anxiety

    (2006)
  • J. Cohen

    Statistical power analysis for the behavioral sciences

    (1988)
  • K.M. Connor et al.

    Psychometric properties of the Social Phobia Inventory (SPIN): New self-rating scale

    British Journal of Psychiatry

    (2000)
  • K.M. Connor et al.

    Mini-SPIN: A brief screening assessment for generalized social anxiety disorder

    Depression and Anxiety

    (2001)
  • Cited by (26)

    • Social anxiety questionnaire (SAQ): Development and preliminary validation

      2018, Journal of Affective Disorders
      Citation Excerpt :

      With the prevalence of SAD (see Heimberg et al., 2000; Kessler et al., 2005; Stein et al., 2017), and both theoretical and empirical interest in social anxiety growing, there is a need for measures of social anxiety that are brief and can be used more practically while assessing the construct comprehensively. A review of currently available and used measures of social anxiety, for instance, the Liebowitz Social Anxiety Scale (LSAS; Liebowitz, 1987), the Social Phobia Weekly Summary Scale (SPWSS; Clark et al., 2003), the Social Interaction Phobia Scale (Carleton et al., 2009; Reilly et al., 2012), the Social Interaction Anxiety Scale (SIAS) and the Social Phobia Scale (SPS; Mattick and Clarke, 1998), the Social Cognitions Questionnaire (SCQ; Wells et al., 1993), and the Social Phobia Inventory (SPIN; Connor et al., 2000; Carleton et al., 2010), revealed some issues to be resolved. First and foremost, there exists no Polish language reliable and valid self-report measures of social anxiety.

    • The aetiology and maintenance of social anxiety disorder: A synthesis of complimentary theoretical models and formulation of a new integrated model

      2016, Journal of Affective Disorders
      Citation Excerpt :

      Again, we consider these behavioural processes to be ‘primary’ given they are part of the first set of processes that develop in response to increases in the threat value represented in the SET principle.3 Our review of the social anxiety literature has yielded cross-sectional studies that have shown that avoidance and escape are consistently associated with high levels of social anxiety and SAD (e.g., Carleton et al., 2010; Heuer et al., 2007; Ottenbreit et al., 2014; Whiteside et al., 2013; Wong and Moulds, 2011a; see also Table 4). These findings provide support for our proposal that avoidance and escape play a role in the elimination of social-evaluative threat.

    • Latent dimensions of social anxiety disorder: A re-evaluation of the Social Phobia Inventory (SPIN)

      2015, Journal of Anxiety Disorders
      Citation Excerpt :

      CFA allows for detection and modeling of covariance among items that is attributable to sources other than a shared latent factor; in this case, the possibility of a method effect due to similar item wording should be considered (Brown, 2015; Campbell-Sills & Brown, 2005). Second, although Carleton et al. (2010b) pointed out that redundant items may be a problem for the SPIN, the SPIN-R retained several redundant items and instead dropped items with unique content (e.g., those related to physical manifestations of anxiety; public speaking). Moreover, the removal of the items assessing distress related to physical symptoms may diminish the construct validity and clinical utility of the scale.

    View all citing articles on Scopus
    View full text