362 TREATMENT OF SPECIFIC ANXIETY DISORDERS
whereas Brown et al. (1997) note that cutoff scores of 24 (SPS) and 34 (SIAS) may be
useful for screening but not diagnosing social phobia. Copies of the questionnaires can
be found in Orsillo (2001, Appendix B) or Mattick and Clarke (1998), where all items
are reproduced except SIAS item 5.
Social Phobia and Anxiety Inventory
The Social Phobia and Anxiety Inventory (SPAI) is a 45-item empirically derived ques-
tionnaire that assesses physical, cognitive, and behavioral responses to various social
interaction, performance, and observation situations (Turner, Beidel, Dancu, & Stan-
ley, 1989). The administration and scoring of the SPAI are more complex and time-
consuming than other social anxiety questionnaires. The questionnaire includes a
13-item Agoraphobia subscale that was included to provide a better differentiation of
social phobia from agoraphobia (Beidel & Turner, 2007). The SPAI Total score is calcu-
lated by subtracting the total score on the Agoraphobia subscale from the Social Anxi-
ety subscale score, which is intended to be a “purer” measure of social phobia (Turner,
Stanley, Beidel & Bond, 1989).
The SPAI Social Anxiety and Agoraphobia subscales have acceptable internal con-
sistency (alphas range from .83 to .97; D. B. Clark et al., 1994; Osman et al., 1996), and
SPAI Total has a 2-week test– retest reliability of .86 (Turner et al., 1989). In addition
the SPAI correlates highly with other measures of social anxiety and related constructs
(Beidel, Turner, Stanley, & Dancu, 1989; Cox et al., 1998; Osman et al., 1996; Ries et
al., 1998), and discriminates social phobia from other anxiety disorders and nonclinical
controls (D. B. Clark et al., 1994; Peters, 2000; Turner, Beidel et al., 1989). In addition
individuals with generalized social phobia score significantly higher than those with
the circumscribed subtype (Ries et al., 1998). It is also sensitive to treatment effects
(Cox et al., 1998) and factor analysis confirms the existence of separate social anxiety
and agoraphobic dimensions (Osman et al., 1996; Turner, Stanley, et al., 1989). Peters
(2000) recommends a SPAI Total cutoff score of 88 to distinguish social phobia from
panic disorder, whereas the manual recommends a cutoff of 60 for distinguishing social
phobia in treatment- seeking samples (Turner, Beidel, & Dancu, 1996). However, cutoff
scores must be used cautiously and for screening only since 10% of nonclinical indi-
viduals score above the cutoff (Gillis, Haaga, & Ford, 1995). An abbreviated 23-item
SPAI was recently developed that holds promise as a comparable measure to the original
inventory (Robertson-Ny, Strong, Nay, Beidel, & Turner, 2007). The 45–item SPAI can
be purchased from Multi- Health Systems Inc. (Turner et al., 1996).
Clinician Guideline 9.14
Assessment for social phobia should include the ADIS-IV (current or lifetime version) and
one of the specialized symptom questionnaires. Either the SPAI or the SPS and SIAS com-
panion scales will provide comparable clinical information on social anxiety severity as well
as an evaluation of treatment effectiveness.