A good place to start when considering a diagnosis of SAD or GAD is with semi-structured clinical interviews of Axis-I disorders. The Anxiety Disorders Interview Schedule for DSM-IV (ADIS-IV; Brown, Di Nardo, & Barlow, 1994) can be used to assess current and past episodes of anxiety in enough detail to determine DSM-IV diagnosis as well as to gather information useful in a functional analysis of symptoms. The ADIS-IV has good to excellent test–retest reliability (Brown, Di Nardo, Lehman, & Campbell, 2001). The Structured Clinical Interview for DSM-IV (SCID-IV; First, Spitzer, Gibbon, & Williams, 1996) systematically assesses for the presence of current and past Axis-I disorders and has demonstrated good discriminant validity and inter-rater reliability for DSM-IV anxiety disorder diagnoses (Kranzler et al., 1995; Zanarini & Frankenburg, 2001). Both measures contain modules on anxiety, mood, substance use, eating, and somatoform disorders, but the ADIS-IV focuses more in depth on symptoms of anxiety, whereas the SCID focuses more broadly on Axis-I psychopathology.

One of the most widely used measures to assess SAD is the Liebowitz Social Anxiety Scale (LSAS; Liebowitz, 1987). The LSAS assesses 24 situations (13 performances and 11 social interactions) that individuals with SAD may fear or avoid, such as going to a party, meeting strangers, and speaking up at a meeting. For each situation, fear and avoidance are rated separately on a scale of 0–3. Thus, six subscale scores can be derived from the LSAS: fear of performance situations, fear of social interaction situations, total fear, avoidance of performance situations, avoidance of social interaction situations, and total avoidance. The total score can be obtained by adding the fear and avoidance ratings for all items. The LSAS has high reliability, high convergent, and discriminant validity, and its subscales have been found to be normally distributed (Heimberg et al., 1999).

Another clinician-administered measure of social anxiety is the 18-item Brief Social Phobia Scale (BSPS; Davidson et al., 1991, 1997). The BSPS assesses seven situations commonly feared or avoided by individuals with SAD, with fear and avoidance of these situations coded separately. In addition, the BSPS assesses the extent of four physiological reactions to social anxiety, including blushing, heart palpitations, tremors, and sweating. Thus, the BSPS has three subscales: fear, avoidance, and physiological arousal. The BSPS has high inter-rater reliability, as well as test–retest reliability, internal consistency, and construct validity (BSPS; Davidson et al., 1991, 1997).

A self-report version of the LSAS (LSAS-SR) was evaluated and has highly similar psychometric properties compared to the clinician-administered version (Fresco et al., 2001). The LSAS-SR was further found to be sensitive to treatment change (Baker, Heinrichs, Kim, & Hofmann, 2002) and to have the same four-factor structure as the clinician-administered version (Oakman, Van Ameringen, Mancini, & Farvolden, 2003). Finally, the LSAS-SR has successfully discriminated between individuals with and without SAD and between diagnostic subtypes of SAD (Rytwinski et al., 2009).

The Social Phobia and Anxiety Inventory (SPAI; Turner, Beidel, Dancu, & Stanley, 1989) is a widely used, comprehensive measure of SAD. It includes 32 items tapping anxiety and avoidance in social situations, as well as cognitions and somatic responses occurring before and during social situations. The SPAI also includes an agoraphobia subsection (13 items), which is subtracted from the total score. Thus, the SPAI measures social anxiety beyond agoraphobia. The SPAI has good internal consistency and high test–retest reliability (Turner et al., 1989) and demonstrated good concurrent validity with other measures of social anxiety (Herbert, Bellack, & Hope, 1991) and with reports of daily social behavior (Beidel, Borden, Turner, & Jacob, 1989; Beidel, Turner, Stanley, & Dancu, 1989).

The Social Phobia Inventory (SPIN; Connor et al., 2000) was developed in an attempt to create a comprehensive measure of SAD that would also be short and easy to administer. Based on the BSPS clinician-administered measure, the SPIN includes 17 items assessing SAD-related fear, avoidance, and physiological symptoms. The SPIN has good internal consistency, test–retest reliability, and convergent and divergent validity (Antony, Coons, McCabe, Ashbaugh, & Swinson, 2006; Connor et al., 2000). It successfully discriminates between individuals with and without SAD (Connor et al., 2000) and is sensitive to both pharmacological (Connor et al., 2000) and psychological treatments (Antony et al., 2006).

The Social Phobia Scale (SPS; Mattick & Clarke, 1989) includes 20 items that pertain to situations in which one is observed by others (e.g., speaking to a group, writing in public). The SPS has good internal consistency as well as convergent and discriminant validity (Heimberg, Mueller, Holt, Hope, & Liebowitz, 1992). The SPS ­successfully differentiates between individuals with and without SAD (Brown et al., 1997; Heimberg et al., 1992) but not between diagnostic subtypes of the disorder (Heimberg et al., 1992). In addition, the SPS could not discriminate between individuals with speech phobia and individuals with SAD (Ries et al., 1998) or between individuals with panic disorder and individuals with SAD (Brown et al., 1997; Peters, 2000). The SPS was sensitive to treatment effects (Ries et al., 1998) and predicted anxious response to a stressful social challenge (Gore, Carter, & Parker, 2002).

The Social Interaction and Anxiety Scale (SIAS; Mattick & Clarke, 1989) is comprised of 20 items measuring anxiety in social interactions. The SIAS has good internal consistency as well as convergent and discriminant validity (Heimberg et al., 1992). It discriminates between individuals with and without SAD (Brown et al., 1997; Heimberg et al., 1992), between individuals with SAD and individuals with other anxiety disorders, and between diagnostic subtypes of the disorder (Heimberg et al., 1992; Ries et al., 1998). The SIAS was sensitive to the effects of treatment (Ries et al., 1998) and predicted anxious response to a stressful social challenge (Gore et al., 2002). The SIAS has been shown to have a single interaction-anxiety factor, thus supporting construct validity of the measure (Safren, Turk, & Heimberg, 1998).

The Brief Fear of Negative Evaluation Scale (BFNE; Leary, 1983) was developed using 12 of the 30 original items included in the Fear of Negative Evaluation scale (Watson & Friend, 1969). Using item response theory, Rodebaugh et al. (2004) found the BFNE to be superior to the full-length version, as it better discriminated between a wider range of severity levels of fear of negative evaluation. The BFNE has been found to have excellent internal consistency, test–retest reliability, convergent and discriminant validity and to successfully discriminate between individuals with SAD, individuals with panic disorder, and community controls (Collins, Westra, Dozois, & Stewart, 2005). Moreover, the BFNE was found to be sensitive to the effects of psychological treatment (Collins et al., 2005).

Linked to the concept of fear of negative evaluation, fear of positive evaluation has also been found among individuals with SAD (e.g., Alden, Taylor, Mellings, & Laposa, 2008; Wallace & Alden, 1997). The Fear of Positive Evaluation Scale (FPES; Weeks, Heimberg, & Rodebaugh, 2008) was developed to assess this construct. The FPES includes ten items that focus solely on fears of positive evaluation, and respondents are required to indicate the degree to which the statement is true for them on a 0–9 scale. Of the ten items, only the eight straightforwardly worded items comprise the final score. The FPES has good internal consistency and test–retest reliability and good convergent and discriminant validity (Weeks, Heimberg, & Rodebaugh, 2008) Weeks, Heimberg, Rodebaugh, & Norton, 2008). The NPES has also been found to have a single factor which was related but distinct from fear of negative evaluation (Weeks, Heimberg, & Rodebaugh, 2008).

Frequently employed types of standardized behavioral assessment tasks (BATs) include conversation with a same-gender stranger, conversation with an opposite-gender stranger, and an impromptu speech given to a small audience. Other situations sometimes include solving simple math problems on a chalkboard in front of an audience or discussing controversial topics with strangers (e.g., Beidel, Borden, Turner, & Jacob, 1989; Hofmann, 2000; Hofmann et al., 2004). Subjective anxiety ratings are typically obtained by using a subjective unit of discomfort scale (SUDS). Patients are then asked to give SUDS ratings before and after the BAT, and sometimes at regular intervals during the BAT. In addition, behavioral indicators (such as length of an impromptu speech) and physiological indicators (such as heart rate) are assessed before, during, and after the test. Ideally, it is recommended to choose a multimodal assessment strategy in conjunction with a well-controlled behavioral test (i.e., participants are asked to speak about one or more specific topics for a specified amount of time). However, practical limitations often need to be considered. For example, whereas heart rate data can be easily and inexpensively recorded, electrodermal activity is considerably more complicated to record ambulatorily in the context of a behavioral test. It should also be noted that different autonomic indicators measure different aspects of psychophysiological arousal; electrodermal activity is primarily influenced by the sympathetic nervous system, heart rate frequency by both the sympathetic and the parasympathetic nervous system, and a certain frequency band of heart rate variability primarily by parasympathetic arousal (vagal tone). In all cases, the clinician or researcher needs to be aware that the act of speaking leads to heightened physiological arousal. Furthermore, the posture (standing vs. sitting) is an important factor to consider. Greater arousal will be obtained if the subject is asked to stand.

The Social Performance Rating Scale (SPRS; Fydrich, Chambless, Perry, Buergener, & Beazley, 1998) was developed to assess structured, videotaped interactions, in terms of eye gaze, vocal quality, length, discomfort, and conversation flow. Each facet of social performance is rated on a 5-point scale with specific behavioral anchors for each rating. Other ratings are similarly anchored in behavioral descriptions. The SPRS was found to have excellent inter-rater reliability and good internal consistency, as well as divergent and convergent validity (Fydrich et al., 1998). The SPRS successfully differentiated between individuals with SAD, individuals with other anxiety disorders and individuals without a psychiatric disorder (Fydrich et al., 1998).

The Social Behavior and Anxious Appearance Rating Scale (SBA; Voncken & Bögels, 2008) was based on earlier behavioral rating scales of Rapee and Lim (1992) and Bögels, Rijsemus, and De Jong (2002) and was developed to rate behaviors and appearance in social interactions. The scale consists of 11 items assessing anxious appearance (e.g., blushing, fidgeting, laughing nervously) and 16 items measuring social behavior (e.g., making eye contact, completing of sentences, coherence, silences). Each item of the SBA scale is rated on a 9-point Likert scale, with a high score indicating a more anxious appearance and better social behavior. The SBA has excellent internal consistency and good inter-rater reliability (Voncken & Bögels, 2008) and has been applied to both social interactions and speech tasks (Voncken, Alden, Bögels, & Roelofs, 2008; Voncken, Dijk, de Jong, & Roelofs, 2010).

The Trier Social Stress Test (TSST; Kirschbaum, Pirke, & Hellhammer, 1993) is a structued procedure for eliciting social-evaluative stress. It consists of a 5-min anticipatory period, a 5-min public speaking task, and a 5-min mental-arithmatic task, all performed in front of an audience. Most studies use endocrine and cardiovascular measures of stress measured before, during, and 1 h following the test, but use of psychological outcome measures is also possible (Williams, Hagerty, & Brooks, 2004). The TSST has been found to result in significant increases in heart rate and cortisol (both serum and saliva) and is not affected by personality traits (Kirschbaum et al., 1993). It is important to note that many factors (e.g., smoking, pregnancy, time of day) can influence cortisol levels and thus may have an impact on the results of the TSST (Williams et al., 2004).

In addition to the comprehensive diagnostic interviews described at the beginning of the chapter, clinicians and researchers may also consider clinician-administered assessments for the evaluation of generalized anxiety symptoms, such as the Hamilton Anxiety Rating Scale (HARS or HAM-A; Hamilton, 1959, 1969). The HARS was originally designed as a fairly unstructured interview that assessed the severity of 14 symptom clusters (anxiety, tension, sleep, concentration, etc.) that are frequently experienced in those with high levels of anxiety, particularly, generalized anxiety. In order to standardize its administration and to provide clear anchor points, the Structured Interview Guide for the Hamilton Anxiety Scale (SIGH-A; Shear et al., 2001) was developed to provide a structured guide to the assessment with a focus on the frequency and intensity of the various symptoms. The SIGH-A has demonstrated high inter-rater and test–retest reliability (Shear et al., 2001).

Although they are quite different disorders, symptoms of GAD and obsessive–compulsive disorder (OCD) can oftentimes be described similarly by patients, as they share a core feature of intrusive, difficult-to-control thoughts. One key difference, however, is that the content of intrusive thoughts (or obsessions) in OCD is not simply of excessive worries about real-life problems, as in GAD. Specifically, patients with OCD have intrusive thoughts that surround their core fear (contamination, harm, inappropriate or unacceptable behavior, etc.), while patients with GAD worry about a number of events and activities, spanning from work to health to relationships. Furthermore, while patients with GAD find the worries difficult to control, patients with OCD will actively try to ignore, suppress, or neutralize such thoughts (or images, impulses) with other thoughts or actions (i.e., compulsions).

The Generalized Anxiety Disorder Question­naire-IV (GADQ-IV; Newman et al., 2002) was designed to assess the presence of worry and its excessiveness and uncontrollability, duration, presence of the six associated symptoms, as well as the degree of interference and distress. Each item is assessed on a 9-point Likert-type scale. The GADQ-IV has demonstrated good test–retest reliability as well as convergent and discriminant validity. Additionally, it has been shown to have excellent specificity and sensitivity (89% and 83%, respectively; Newman et al., 2002).

The Intolerance of Uncertainty Scale (IUS; Freeston, Rhéaume, Letarte, Dugas, & Ladouceur, 1994) is a 27-item self-report measure that is designed to assess intolerance of uncertainty using a five-point Likert-type scale. Intolerance of uncertainty, or a “dispositional characteristic that results from a set of negative beliefs about uncertainty and its implications” (Dugas & Robichaud, 2007, p. 24), is often considered a cognitive vulnerability factor for the development of chronic worry (Koerner & Dugas, 2008). In addition to GAD, intolerance of uncertainty is also considered a key factor in the development and maintenance of OCD (Tolin, Abramowitz, Brigidi, & Foa, 2003), and individuals with both disorders report higher levels of intolerance of uncertainty than individuals with other anxiety disorders (Dugas, Gagnon, Ladouceur, & Freeston, 1998; Steketee, Frost, & Cohen, 1998). The IUS has been shown to have excellent internal consistency (α  =  0.95) and convergent validity in a sample of students (Buhr & Dugas, 2002). Likewise, the IUS discriminated between a clinical sample of individuals diagnosed with GAD from a non-­anxious control sample (Dugas et al., 1998).

The Penn State Worry Questionnaire (PSWQ; Meyer, Miller, Metzger, & Borkovec, 1990) is a 16-item self-report questionnaire designed to measure the intensity and excessiveness of worry without assessing the particular content of worry. This widely used measure has good to very good internal consistency across clinical and college student samples (α from 0.86 to 0.93, Molina & Borkovec, 1994). Likewise, it has been shown to be related to other measures of worry (Davey, 1993; Meyer et al., 1990). PSWQ scores have also been shown to be higher for clients diagnosed with GAD compared to clients diagnosed with other anxiety disorders, notably OCD (Brown, Antony, & Barlow, 1992).

Other self-report measures for GAD include the Why Worry Scale-II (Freeston et al., 1994), a 25-item measure that assesses the perceived positive consequences of worry, and the Worry Domains Questionnaire (Tallis, Eysenck, & Mathews, 1992), a 25-item measure that was designed to measure nonpathological worry. Thought suppression, which in the context of GAD, pertains to experiential avoidance and the suppression of negative affect, may be assessed using the White Bear Suppression Inventory (Wegner & Zanakos, 1994). This measure contains 15 statements with which respondents rate their agreement on a 5-point Likert scale (e.g., “I wish I could stop thinking about certain things”). For older patients, clinicians may consider utilizing the Worry Scale for Older Adults (Wisocki, 1988), a 35-item measure designed to measure the extent and frequency of which older adults worry about events commonly associated with aging.

A number of exposure exercises have been developed for the treatment and study of GAD. For example, the worry exposure is a method developed by Craske, Barlow, and O’Leary (1992) as an intervention for pathological worry. In this variant of imaginal exposure, individuals are instructed to mentally expose themselves to worry at set times, for a prolonged period, by thinking about the feared events. The exposure takes place by conjuring up an image of the most feared expectation and focusing on this for a period of about 25 min. Individuals are then instructed to brainstorm alternative explanations or outcomes of the feared event and to evaluate these options.

Hofmann et al. (2005) induced worry by asking participants to listen to a script about having to give a presentation in front of a large class. This script contained a number of ruminative self-statements (e.g., “You will feel overwhelmed by negative thoughts as you are facing the audience”) adapted from the Self-Statement During Public Speaking Scale (Hofmann & DiBartolo, 2000). Participants were then asked to worry about this particular situation for 30 s. Psychophysiological data, including heart rate and skin conductance, were collected during the worry period, after which participants were asked to rate their distress on a scale from 0 (no distress) to 100 (very distressed).

Hofmann, Schulz, Heering, Muench, and Bufka (2010) also designed a worry induction as part of a study examining the physiological correlates of GAD and Major Depressive Disorder (MDD). Following a 5-min baseline resting period, participants were asked to engage in a worry period or relaxation period for 5 min, the task order for which was randomly assigned. During the worry period, participants were instructed to worry about their most worriesome topic (e.g., money, health, relationships, etc.) as identified in a diagnostic interview and to redirect their attention to worrying about this topic if they found that their attention had wandered. During the worry and relaxation periods, heart rate, skin conductance level and other psychophysiological data were collected. After each task, participants rated their average level of anxiety and their worry level during the task on a scale from 0  =  not at all to 10  =  extremely.

Major depressive disorder (MDD), and mood disorders in general, are commonly comorbid with anxiety disorders (see also Chap. 23). In a study by Brown, Campbell, Lehman, Grisham, & Mancill (2001), MDD was present in 14% of those diagnosed with SAD and 26% of those diagnosed with GAD. Furthermore, the presence of comorbid depression is associated with greater symptom severity in patients with SAD (Erwin, Heimberg, Juster, & Mindlin, 2002) and in patients with GAD (Newman, Przeworski, Fisher, & Borkovec, 2010). Among patients with SAD, those who also have comorbid depression have greater anxiety-related cognitions and higher levels of negative evaluations of social performances than those patients without comorbid depression (Ball, Otto, Pollack, Uccello, & Rosenbaum, 1995; Wilson & Rapee, 2005).

To assess for depression, the mood disorder modules in the larger overall assessment of Axis I disorders may be used, such as the SCID-I (First et al., 1996) or ADIS-IV (Brown et al., 1994). Clinician-administered interviews designed solely for assessment of depression include the Montgomery-Asberg Depression Rating Scale (Montgomery & Åsberg, 1979) and the Hamilton Depression Rating Scale (Hamilton, 1960). Validated and reliable self-report measures of depression include the Beck Depression Inventory (Beck, Ward, Mendelson, Mock, & Erbaugh, 1961), and the Quick Inventory of Depressive Symptomatology – Self-Report (Rush et al., 2003).

Substance use disorders are highly comorbid with both SAD and GAD (see also Chap. 22). In a recent epidemiological survey of comorbid substance use and anxiety disorders, 21.3% of patients with SAD and 15.9% of patients with GAD who sought treatment in the past 12 months for these disorders also had comorbid substance use disorders (Grant et al., 2006). In general, both SAD and GAD are characterized by both cognitive and behavioral avoidance of anxiety-provoking experiences due to the physiological and emotional distress that the anxiety can cause. As is described in Morris, Stewart, and Ham (2005), alcohol use may reduce both this physiological and cognitive arousal, providing negative reinforcement for alcohol consumption.

Assessing substance use is particularly important prior to starting therapy, especially exposure-based therapy, since individuals may experience increased distress at the beginning of therapy. This is particularly important if an individual’s habitual coping response to increased stress is to increase substance use. Likewise, substance use prior to engaging in a potential anxiety-provoking situation could function as a safety behavior, potentially minimizing the effect of an intervention. For this reason, a thorough assessment of substance use should also include an evaluation of prescription and nonprescription medication, such as benzodiazepine use.

Although often overlooked in initial assessment and treatment planning, personality disorders can significantly interfere with effective treatment, and patients with SAD and GAD have been found to have higher rates of personality disorders as compared to other anxiety disorders (Reich et al., 1994) (see Chap. 15). Notably, in individuals with SAD, Avoidant Personality Disorder (APD) is highly comorbid (Brooks, Baltazar, & Munjack, 1989; Jansen, Arntz, Merckelbach, & Mersch, 1994). However, the APD diagnosis may simply reflect a higher symptom severity level of SAD (Holt, Heimberg, & Hope, 1992). In general, comorbid personality disorders are associated with poorer treatment response (Pollack, Otto, Rosenbaum, & Sachs, 1992), premature termination of treatment (Sanderson, Beck, & McGinn, 1994), and increased risk of self-injurious thoughts and behaviors (Corbitt, Malone, Haas, & Mann, 1996).

Current and lifetime prevalence rates of cognitive impairment are estimated to be 1.2% and 2.5–13.6%, respectively, for those with SAD, 2.3% and 1.4–6.0% for patients with GAD (Castaneda, Tuulio-Henriksson, Marttunen, Suvisaari, & Lönnqvist, 2008). If impairment is suspected or reported, it is vital to assess cognitive functioning at the outset of treatment, as such impairment may interfere with successful treatment and thus be relevant for treatment recommendations and for treatment planning.

There are a number of measures designed to assess specific areas of cognitive functioning, depending on the area of interest. Intelligence may be measured using the Wechsler Adult Intelligence Scale (Wechsler, 1981) and the Wide Range Achievement Test (Wilkinson, 1993). Memory problems may be tested using the Wechsler Memory Scale (Wechsler, 1997), which contains subscales measuring several aspects of memory. Furthermore, various components of cognition may be measured using the Luria-Nebraska Neuropsychological Battery (Golden, Hammeke, & Purisch, 1980) and the California Verbal Learning Test-II (Delis, Kramer, Kaplan, & Ober, 2000).

It is common for individuals with SAD to perceive their social performances more negatively than those without SAD (e.g., Glasgow & Arkowitz, 1975; Norton & Hope, 2001). However, it is often less clear whether these perceived social skills deficits reflect actual deficits or are biased interpretations influenced by social anxiety. Some studies have shown that there are no differences in social performances between those with higher and lower levels of social anxiety (e.g., Cartwright-Hatton, Tschernitz, & Gomersall, 2005; Rapee & Lim, 1992; Strahan & Conger, 1998), whereas others have demonstrated that those without social anxiety perform better socially than those with social anxiety (e.g., Norton & Hope, 2001; Stopa & Clark, 1993). It is therefore likely that some individuals with SAD will present with some social skill deficits and others will not, making a careful appraisal of social skills an important part of a comprehensive assessment.