close

Вход

Забыли?

вход по аккаунту

?

Psychometric properties of in-situation safety behaviors measures for social anxiety

код для вставкиСкачать
ABSTRACT
PSYCHOMETRIC PROPERTIES OF IN-SITUATION SAFETY BEHAVIORS
FOR SOCIAL ANXIETY
Stephanie L. Gier, M.A.
Department of Psychology
Northern Illinois University. 2010
David P. Valentiner, Director
This study sought to evaluate the psychometric properties of four existing
measures of safety behaviors. A reliable and valid measure of safety behaviors
would be useful to researchers testing social anxiety models. Such a measure would
also be useful to clinicians to assist in designing the appropriate treatment for their
clients and for tracking treatment progress. Currently, there is not an adequately
validated measure of safety behaviors. In the present study, the Social Phobia Safety
Behaviors Scale (SPSBS; Pinto-Gouveia et al., 2003) was determined to be the
superior measure of in-situation safety behaviors for social anxiety. The SPSBS
displayed adequate psychometric properties. Further, a two-component structure of
the SPSBS best defined two-higher order components inferred from correlations
among the weighted factor scores of the four measures of safety behaviors. The
SPSBS components (i.e., Laying Low and Putting On) appear to represent behaviors
intended to keep a low profile and impression-management. Taken as a whole, the
present study provides evidence that the SPSBS has superior psychometric
properties and is recommended for future use.
NORTHERN ILLIONIS UNIVERSITY
DE KALB, ILLINOIS
MAY 2010
PSYCHOMETRIC PROPERTIES OF IN-SITUATION SAFETY BEHAVIORS
FOR SOCIAL ANXIETY
BY
STEPHANIE L. GIER
© 2010 Stephanie L. Gier
A THESIS SUBMITTED TO THE GRADUATE SCHOOL
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE
MASTER OF ARTS
DEPARTMENT OF PSYCHOLOGY
Thesis Director:
David P. Valentiner
UMI Number: 1477027
All rights reserved
INFORMATION TO ALL USERS
The quality of this reproduction is dependent upon the quality of the copy submitted.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.
UMI 1477027
Copyright 2010 by ProQuest LLC.
All rights reserved. This edition of the work is protected against
unauthorized copying under Title 17, United States Code.
ProQuest LLC
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, MI 48106-1346
ii
TABLE OF CONTENTS
Page
LIST OF TABLES .......................................................................................
vi
LIST OF FIGURES ......................................................................................
vii
LIST OF APPENDICES ..............................................................................
viii
Chapter
I.
INTRODUCTION ............................................................................
1
Introduction ..........................................................................
1
Social Anxiety ..........................................................
2
Cognitive Theory for Social Anxiety ........................
4
Therapy for Social Anxiety ......................................
10
Exposure with decreased safety behaviors ....
12
Alternate conceptualizations of exposure
with decreased safety behaviors ...................
13
The Present Study .....................................................
15
Validity Issues of Four Safety Behaviors Measures..
16
Issue 1: Representational Validity ............................
17
Content validity ............................................
17
Construct validity .........................................
21
iii
Chapter
II.
III.
Page
Issue 2: Criterion Validity ........................................
22
METHOD .........................................................................................
24
Sample ..................................................................................
24
Procedures ............................................................................
25
Safety Behaviors Measures ..................................................
26
Liebowitz Social Anxiety Scale - Modified .............
26
Social Behavior Questionnaire – Modified ..............
27
Social Phobia Safety Behaviors Questionnaire ........
27
Presentation-Related Safety Behaviors Scale ...........
28
Measures of Related and Dissimilar Constructs ...................
28
Social Self-Presentation Style Scale .........................
28
Positive and Negative Affect Scale ..........................
29
Brief Fear of Negative Evaluation Scale ..................
30
Center for Epidemiologic Studies Depression Scale
31
Social Anxiety Measures ......................................................
32
Social Interaction Anxiety Scale ..............................
32
Social Phobia Scale ..................................................
32
Other ...................................................................................
33
Demographic Questionnaire .....................................
33
RESULTS .........................................................................................
34
iv
Chapter
IV.
Page
Descriptive Statistics ............................................................
34
Reliability .............................................................................
34
Internal consistency ..................................................
34
Average inter-item correlations ................................
37
Disattenuated correlations ........................................
39
Representational Validity .....................................................
40
Convergent validity ..................................................
41
Divergent validity .....................................................
42
Summary ...................................................................
44
Criterion Validity ..................................................................
45
Correlations with criterion measures ........................
45
Partial correlations ....................................................
45
Component Structures ..........................................................
46
LSAS-M component structure ..................................
48
SBQ-M component structure ....................................
50
SPSBS component structure .....................................
52
PRSBQ component structure ....................................
54
Summary ...................................................................
55
Correlations among Component Scores ...............................
56
DISCUSSION ...................................................................................
58
v
Chapter
Page
Issues of validity ...................................................................
59
Validity of the LSAS-M ...........................................
60
Validity of the SBQ-M .............................................
61
Validity of the SPSBS ..............................................
62
Validity of the PRSBQ .............................................
63
Component Structure ............................................................
64
Component structure of the LSAS-M .......................
64
Component structure of the SBQ-M .........................
65
Component structure of the SPSBS ..........................
67
Component structure of the PRSBQ .........................
68
Cross-measure Correlations among Components ................
68
Component structure of the SBQ-M-UI, SPSBS-LL
and PRSBQ cluster ....................................................
69
Component structure of the SBQ-M-IM and
SPSBS-PO cluster .....................................................
70
Strengths and Limitations of the Present Study ...................
70
Implications for Theory ........................................................
72
Practical Implications ...........................................................
76
Concluding Remarks ............................................................
76
REFERENCES .............................................................................................
78
APPENDICES ..............................................................................................
84
LIST OF TABLES
Tables
1.
2.
3.
4.
5.
6.
Page
Sample Size, Cronbach’s Alpha, Means, Standard Deviations of
and Correlations Between Measures of Safety Behaviors,
Related, Dissimilar, and the Social Anxiety Constructs ...................
35
Correlations Between the Social-Self Presentation Style Scale,
Brief Fear of Negative Evaluation Scale, Positive and Negative
Affect Scale, Center for Epidemiologic Studies Depression Scale,
Social Interaction Anxiety Scale, and the Social Phobia Scale ........
36
Partial Correlations between the Safety Behaviors Measures and
the Social Interaction Anxiety Scale and the Social Phobia Scale
Controlling for the Center for Epidemiologic Studies Depression
Scale .................................................................................................
46
Principal Components Analysis of the Liebowitz Social Anxiety
Scale-Modified: Component Loadings for the Two-Component
Solution..............................................................................................
49
Principal Components Analysis of the Social Behavior
Questionnaire-Modified: Component Loadings for the
Three-Component Solution ...............................................................
51
Principal Components Analysis of the Social Phobia Safety
Behaviors Scale: Component Loadings for the Two-Component
Solution..............................................................................................
53
7.
Principal Components Analysis of the Presentation-Related Safety
Behaviors Questionnaire: Component Loadings for the One-Component
Solution..............................................................................................
55
8.
Correlations among the Component Scores ......................................
57
LIST OF FIGURES
Figures
1.
A model of the processes that occur when a socially anxious
individual enters a feared social situation .........................................
Page
5
LIST OF APPENDICIES
Appendix
Page
A.
Liebowitz Social Anxiety Scale – Modified (LSAS-M) ..................
86
B.
Social Behavior Questionnaire – Modified (SBQ-M) ......................
88
C.
Social Phobia Safety Behaviors Scale (SPSBS) ...............................
90
D.
Presentation-Related Safety Behaviors Questionnaire (PRSBQ) .....
92
E.
Social Self-Presentation Style Scale (SSPSS) ..................................
94
F.
Brief Fear of Negative Evaluation Scale (BFNE) ............................
96
G.
Positive and Negative Affect Scale (PANAS) .................................
98
H.
Social Interaction Anxiety Scale (SIAS) ..........................................
100
I.
Social Phobia Scale (SPS) ................................................................
102
J.
Center for Epidemiologic Studies Depression Scale (CES-D) .........
104
CHAPTER I
INTRODUCTION
Several theorists (Clark, 2001; Clark & Wells, 1995; Salkovskis, 1991) have
proposed a relation between the use of safety behaviors (e.g., avoiding eye contact
or pausing while speaking) and the maintenance of social anxiety (SA) and social
anxiety disorder (SAD). Individuals with SA are thought to believe that safety
behaviors prevent the occurrence of feared social catastrophes. Individuals with SA
who habitually engage in safety behaviors benefit from the addition of a safety
behaviors reduction component to standard exposure treatment (Wells et al., 1995).
Exposure with instructions to decrease safety behaviors is believed to provide
individuals with SA the opportunity to discover that reliance on safety behaviors
actually maintains social anxiety. Anxiety lessens when individuals with SA learn
that safety behaviors do not prevent the feared social catastrophe; the catastrophe
itself is improbable and has less severe social consequences than imagined. A
reliable and valid measure of safety behaviors would be useful to researchers testing
SA models. Such a measure would also be useful to clinicians to assist in designing
the appropriate treatment for their clients and for tracking treatment progress.
2
Currently, there is not an adequately validated measure of safety behaviors.
This study sought to evaluate the psychometric properties of four existing measures
of safety behaviors.
Social Anxiety
Social anxiety disorder is characterized by a marked and persistent fear of
social or performance situations in which embarrassment may occur (American
Psychiatric Association [APA], 2000). Fear of embarrassment or humiliation in
social situations must significantly interfere with daily routine, occupational
performance, or social functioning in order to meet full diagnostic criteria for SAD.
Individuals diagnosed with SA often avoid feared social situations altogether, but, if
necessary, social situations may be tolerated with dread or distress. Fear of
humiliation in social situations may be interpreted by socially anxious individuals as
shyness or weakness that should be endured (Noyes & Hoehn-Saric, 1998).
Social anxiety disorder is the second most prevalent 12-month mental
disorder (6.8%), following specific phobia (8.7%; Kessler, Chiu, Demler, &
Walters, 2005). Although SAD is a prevalent disorder (Wittchen & Fehm, 2003),
only 5.4% of individuals with SAD initially seek treatment for SA as the primary
presenting problem (Noyes & Hoehn-Saric, 1998). Exposure to feared social
situations, a common treatment for SA (Gould, Buckminster, Pollack, Otto, & Yap,
1997), may provoke additional anxiety in socially anxious individuals. Fear of
3
exposure treatment may be a reason why few people with SA seek professional
treatment for the disorder. Although individuals with SA appear to avoid
professional treatment, they nonetheless regularly experience non-clinical social
exposures (e.g., speaking with a bank teller). Repetitive non-occurrence of feared
social catastrophes (e.g., the bank teller is pleasant following a request for a bank
account number) experienced from non-clinical exposures does not appear to benefit
individuals with SA (Wells et al., 1995). It is believed that individuals with SA do
not realize that expected irrational social catastrophes likely will not occur.
Performing safety behaviors is believed to prevent a reevaluation of these
expectations and may thereby maintain social anxiety (Clark & Wells, 1995).
The Diagnostic and Statistical Manual of Mental Disorders –IV-Text
Revised (DSM-IV-TR; APA, 2000) necessitates the specification of generalized
social phobia when fears include most social situations. Another subtype that is not
in the DSM-IV-TR, but found in the research literature, is specific social anxiety.
Specific social anxiety is characterized by fear of circumscribed performance
situations (Hook & Valentiner, 2002; Noyes & Hoehn-Saric, 1998). There appear to
be many differences between the subtypes of SA. Individuals with specific social
anxiety report less neuroticism and more extroversion than individuals with
generalized social anxiety (see Hook & Valentiner for review). Individuals with
generalized social anxiety tend to score higher than individuals with specific social
anxiety on indices of pathology, depression, trait anxiety, and shyness. In addition,
individuals with specific social anxiety respond to feared social situations with an
4
acute heart rate increase. Individuals with generalized social anxiety do not show
this significant heart rate elevation (Noyes & Hoehn-Saric). The many differences
between the SA subtypes raise the possibility that safety behaviors differ by
subtype.
Cognitive Theory for Social Anxiety
Clark and Wells (1995) and Clark (2001) present a cognitive theory for the
maintenance of SA (see Figure 1). According to this theory, some basic cognitive
assumptions are activated when socially anxious individuals enter a social situation.
Cognitive assumptions identified by Clark include: excessively high expectations
for social performance (e.g., “I must always sound intelligent.”), conditional beliefs
about the consequences of behavior (e.g., “If I disagree, they will think I am
stupid.”), and unconditional negative beliefs about the self (e.g., “I am stupid.”).
These assumptions are thought to lead to perceived social danger because socially
anxious individuals assume that social failure is certain. Certainty of failure is
believed to create anxiety and distress, which distracts socially anxious individuals
from cognitive reappraisal of the situation.
5
Social Situation
Activates
Assumptions
Perceived social danger
Safety
Behaviors
Processing of
Self as a
Social Object
Somatic &
Cognitive
symptoms
Figure 1. A model of the processes that occur when a socially anxious individual
enters a feared social situation (adapted from Clark, 2001).
Once social danger is perceived, individuals are believed to process the self
as a social object (Clark, 2001). Processing of the self as a social object is supposed
to involve heightened self-awareness (e.g., listening to yourself as you speak) and
attention to detail (e.g., noticing that others are watching you). Clark proposes that
socially anxious individuals monitor their behavior and social value in detail to
appraise the likelihood of negative evaluation. Following processing of the self,
social danger is thought to be reevaluated. According to Clark and Wells (1995),
6
perceived social danger cognitions may include exaggerated thoughts of loss of
status, loss of worth, or rejection.
Perceived social danger is also believed to be related to somatic and
cognitive symptoms (Clark, 2001). Common somatic symptoms associated with SA
are heart palpitations, blushing, sweating, and trembling (Clark & Wells, 1995). An
example of a cognitive symptom is a mental blank (Clark). For some socially
anxious individuals, it is thought that experience of somatic and cognitive symptoms
indicate social failure. Fear of social failure leads to further processing of the self,
and thus maintenance of the SA cycle.
Safety behaviors are identified by Clark and Wells (1995) as another
maintaining component of SA. Safety behaviors are covert or overt behaviors
performed in threatening social situations to reduce the risk of negative evaluation.
For example, socially anxious individuals may walk with their head down to avoid
social interaction or they may laugh frequently at what others say to create a
positive impression. A well-rehearsed safety behavior is likely believed to be
successful at reducing negative evaluation from others because it is thought to
distract others from engaging in social judgment. Thus, the safety behavior served
its purpose and temporarily decreases anxiety. However, continued use of safety
behaviors is self-defeating, as fear of negative evaluation ultimately remains.
Individuals often engage in many safety behaviors in each feared situation
(Clark, 2001); however, the circumstances of feared social situations appear to
directly affect selection of safety behaviors (e.g., avoiding eye contact during a
7
speech vs. asking a lot of questions during conversation; McManus, Saradura, &
Clark, 2008; Wells et al., 1995). Compared to individuals with low SA, highly
socially anxious individuals tend to use a greater number of safety behaviors, across
a wider range of situations, at a higher frequency (McManus et al.).
Performing safety behaviors is thought to create two main problems for
individuals with SA. First, by performing safety behaviors, individuals with SA only
experience social situations with reliance on safety behaviors and consequently, the
safety behaviors prevent disconfirmation of unrealistic beliefs about the feared
social situation. When socially anxious individuals realize a social catastrophe did
not occur, this outcome is attributed to use of safety behaviors and not to the
situation being less dangerous than previously thought (Clark, 2001; Clark & Wells,
1995).
Second, use of safety behaviors can actually make the feared catastrophe
more likely to occur (e.g., holding a drink tightly from fear of spilling may actually
cause liquid to spill; Clark & Wells, 1995). Some safety behaviors (e.g., increasing
the distance between yourself and the person you are talking to) may even generate
negative evaluation. Compensating for or hiding physical displays of anxiety often
make socially anxious individuals appear awkward or inappropriate (Noyes &
Hoehn-Saric, 1998). Indeed, socially anxious individuals may appear preoccupied
and uninterested in social situations because of the cognitive effort necessary to
perform safety behaviors (e.g., thinking; “Don’t shake,” during a conversation). In
8
sum, safety behaviors are thought to create greater social discomfort than not using
safety behaviors at all.
Salkovskis and colleagues (1999) expanded the theory of safety behaviors by
organizing safety behaviors into response classes based on function. Avoidance of
social situations (e.g., calling in sick to work the day of a work presentation) is the
first safety behaviors response class. Escape from a current social situation (e.g.,
excusing oneself when introduced to a stranger) is the second safety behaviors
response class. The third safety behaviors response class is in-situation response
(e.g., avoiding eye contact during a conversation). If properly executed, behaviors in
each response class (i.e., avoidance, escape, and in-situation safety behaviors) are
believed to prevent feared catastrophes from occurring (Salkovskis et al.).
Some clinical researchers have proposed that some safety behaviors may
actually be adaptive coping strategies, though this remains a debated issue in the
field. Thwaites and Freeston (2005) attempted to differentiate adaptive coping
strategies from safety behaviors based on function, intention of the individual, and
perceived function for a given context. Safety behaviors are thought to be used
intentionally by socially anxious individuals to reduce anxiety by avoiding or
escaping imagined social catastrophe (e.g., excusing oneself from a dinner
conversation when overwhelmed). Temporary anxiety reduction and evasion of
perceived social catastrophe are believed to be powerful reinforcers for individuals
with SA. If anxiety reduction and evasion are indeed reinforcers behavioral theory
9
would suggest that use of safety behaviors is likely to increase in the future.
Accordingly, Clark and Wells’ (1995) model of SA is maintained.
Temporary anxiety reduction is also achieved by using adaptive coping
strategies (Thwaites & Freeston, 2005). However, when adaptive coping strategies
are used, feared social catastrophes are not avoided or escaped, but endured (e.g.,
only attending a work outing because a friend accompanied the individual for
support, Thwaites & Freeston). It is thought that adaptive coping strategies provide
socially anxious individuals with confidence to face feared social situations.
Confidence may come from the knowledge that coping strategies can be used to
reduce in-situation anxiety (e.g., if the socially anxious individual becomes
overwhelmed at dinner, the supportive friend can continue the conversation). Since
evasion of imagined social catastrophe is not attained even with adaptive coping
strategies, it is believed that individuals with SA are not likely to use coping
strategies in the future. For this reason, adaptive coping strategies are not thought to
maintain SA.
If safety behaviors differ from adaptive coping strategies, it would be
beneficial to distinguish adaptive behaviors (i.e., coping strategies) from
maladaptive behaviors (i.e., safety behaviors) to determine which behaviors should
be reduced during exposure treatment (Thwaites & Freeston, 2005). By having
individuals refrain from performing maladaptive safety behaviors prior to or during
exposure of feared social situations, clinicians would be attempting to provide
disconfirmation of negative beliefs. Adaptive coping strategies do not interfere with
10
disconfirmation, since they do not involve avoidance or permanent escape from
feared social situations. Therefore, adaptive behaviors would remain unchanged
during exposure treatment.
Therapy for Social Anxiety
By the late 1990s, meta-analyses (Feske & Chambless, 1995; Gould et al.,
1997) established several types of cognitive-behavioral therapy as equally effective
treatments for SA. Among the several types of cognitive-behavioral therapy, those
with exposure yielded the largest effects (ES = 0.74; Gould et al.). When exposurebased cognitive-behavioral therapies were compared, those with cognitive
restructuring (ES = 0.80) were not more effective than those with exposure alone
(ES = 0.89). Exposure appears to be the essential element in reducing SA. Cognitive
restructuring does not appear to create any additional benefits over exposure alone.
Around the same time, Clark and Wells (1995) developed a theory-driven
SA treatment. The goal of Clark and Wells’ treatment is to interrupt the vicious SA
cycle. In this model, four main elements are believed to maintain SA: self-focus of
attention; misleading, negative information used to make inferences about how one
appears; covert and overt safety behaviors; and problematic pre- and post-event
processing. For example, self-focus on fear of sweating during social interactions
may evoke a misleading inference that one appears unattractive to others. This
inference may lead to the safety behavior of wearing an unnecessary coat to disguise
11
anticipated sweating. The coat induces sweating, inviting continued self-focus and
negative evaluation from others, and ultimately, the individual may perceive social
failure.
There are many procedures in Clark and Wells’ (1995) treatment such as:
audio/video feedback, mirror exposure, self-focus of attention manipulation, and
dropping safety behaviors during exposure. Although Clark and Wells’ treatment
was shown to be effective at treating SA (Clark et al., 2006; Hofmann &
Scepkowski, 2006; Stangier, Heidenreich, Peitz, Lauter, & Clark, 2003), it remains
unclear which procedures are responsible for treatment change beyond the effects of
exposure. Dismantling studies show that audio/video feedback and mirror exposure
do not appear to maintain treatment effects at follow-up for patients diagnosed with
SAD (Hofmann & Heinrichs, 2003; Smits, Powers, Buxkamper, & Telch, 2006).
Self-focus of attention manipulation demonstrates a modest treatment effect when
compared to exposure only in a non-clinical sample of socially anxious individuals
(Renner & Valentiner, 2008). Several studies (Wells et al., 1995; see also GarciaPalacios & Botella, 2003; Kim, 2005; McManus et al., 2008; Morgan & Raffle,
1999) have found that instructions to decrease safety behaviors, one of the
procedures in the Clark and Wells’ treatment, reduces SA beyond the effects of
exposure only for individuals with SAD.
12
Exposure with Decreased Safety Behaviors
During exposure with decreased safety behaviors, individuals with SA are
encouraged to refrain from engaging in safety behaviors. Therapists prepare their
clients for exposure with decreased safety behaviors by explaining that safety
behaviors maintain SA by preventing disconfirmation of unrealistic beliefs about
feared social situations. Exposure with decreased safety behaviors appears to be
more effective than exposure alone at reducing SA for individuals with SAD
(Garcia-Palacios & Botella, 2003; Kim, 2005; McManus et al., 2008; Morgan &
Raffle, 1999; Wells et al., 1995). In a key study (Wells et al.), eight participants
diagnosed with SAD (according to DSM-III-R criterion; APA, 1987) alternated
trials of exposure alone and exposure with decreased safety behaviors. Results
indicate exposure with decreased safety behaviors was superior to exposure only in
reducing SA and belief in feared social catastrophes (Wells et al., 1995).
Kim (2005) found similar results to Wells and colleagues (1995) using an
experimental design. In addition to replicating the effect, Kim investigated the
importance of providing the rationale for decreasing safety behaviors. Each of 45
participants, who were diagnosed with SAD, were assigned to one of three treatment
groups: exposure with no changes in safety behaviors, exposure with decreased
safety behaviors with a cognitive rationale, or exposure with decreased safety
behaviors with an extinction rationale. Providing participants with a cognitive
13
rationale (i.e., “discover if feared catastrophes can actually happen”) for why safety
behaviors should be decreased appeared to make exposure therapy more effective
than an extinction rationale (i.e., “anxiety is temporarily decreased by using safety
behaviors, but most research and clinical cases have proven that safety behaviors
actually increase anxiety”), or no rationale (i.e., exposure with no changes in safety
behaviors condition; Kim).
Cognitive-behavioral group treatment with decreased safety behaviors also
appears to be an effective treatment for decreasing SA (Morgan & Raffle, 1999).
Thirty individuals diagnosed with SAD participated in 80-hours of cognitivebehavioral group therapy. Half of the participants were instructed to drop safety
behaviors during exposure; the other participants received standard cognitivebehavioral treatment. Results show that instructions to drop safety behaviors was
effective in reducing anxiety, but it is unknown how many participants actually
restrained from engaging in safety behaviors because data collection was based
solely on self-report. Furthermore, if participants interpreted safety behaviors to be
adaptive coping strategies, they may not have dropped their behaviors (Morgan &
Raffle).
Alternate Conceptualizations of Exposure with Decreased Safety Behaviors
A dialectical behavior therapy technique called “opposite action” (Linehan,
1993) may have some of the same characteristics as exposure with decreased safety
14
behaviors. When an unwanted emotion (e.g., fear) is experienced, behaviors
associated with that feeling are identified. Then, instinctual behaviors are replaced
with opposing, incompatible action(s). For example, a woman who feared her
husband was attracted to beautiful, thin women looked at pictures of attractive
females with him (Rizvi & Linehan, 2005). Refraining from instinctual behaviors
lessens unwanted emotions and strengthens constructive emotions (e.g., acceptance)
associated with incompatible behaviors. Exposure treatment sometimes reflects
“opposite action” by having individuals with SA confront feared situations to
experience the possible negative consequences. Exposure and “opposite action” are
believed to teach a similar lesson that anxiety felt during exposure was not as
intense as anticipated.
Existing behavioral conceptualization for obsessive-compulsive disorder
uses exposure and response prevention (ERP) for treatment. During ERP,
individuals remain in feared conditions/situations without using compulsions until
anxiety decreases. For example, individuals who are fearful of contamination would
refrain from washing their hands after touching a public door handle until their
anxiety decreases. Exposure and response prevention can be done in vivo or as an
imaginary exercise (Abramowitz, 2006). Exposure and response prevention is
similar to decreasing safety behaviors for SA because individuals learn that the
feared situation is not actually dangerous and anxiety will decrease naturally with
time.
15
The Present Study
There is a wide array of research on the well-theorized area of safety
behaviors in relation to SA (Clark & Wells, 1995; Kim, 2005; Morgan & Raffle,
1999; Wells & Papageorgiou, 1998). Decreasing the use of safety behaviors during
exposure treatment for SA is superior to exposure only treatment (Wells et al.,
1995). Although studies have replicated this effect and though measures of safety
behaviors are available (Kim, 2005; Morgan & Raffle, 1999; Wells &
Papageorgiou, 1998), a standardized self-report measure for safety behaviors had
yet to be validated (Garcia-Palacios & Botella, 2003). The psychometric properties
of four existing safety behaviors measures were examined to identify whether the
measures would be appropriate for use with socially anxious individuals.
There are several reasons why a rigorously examined safety behaviors
measure will be useful. First, a standardized self-report measure of safety behaviors
can direct treatment planning. Individuals that do not rely on safety behaviors in
social situations, whom could be identified using a safety behaviors measure, will
benefit from exposure only. Individuals that frequently engage in safety behaviors
will benefit from exposure with decreased safety behaviors. Second, a safety
behaviors measure will provide therapists with information as to which specific
safety behaviors their clients should be encouraged to decrease. Third, treatment
compliance and treatment progress can easily be tracked by administering a safety
behaviors measure before and after exposure. A decrease in number of items (i.e.,
16
safety behaviors) endorsed would indicate that safety behaviors are actually
decreased during exposure treatment. Fourth, an established safety behaviors
measure can be compared with anxiety reports. This comparison can teach socially
anxious individuals about the relationship between use of safety behaviors and
anxiety. Lastly, an empirically established self-report measure will allow
researchers to test SA theories and, perhaps, to test hypotheses about other anxiety
disorders that appear to use safety behaviors as coping mechanisms (e.g., obsessivecompulsive disorder, panic disorder, specific phobia; Bennet-Levy, Butler, Fennell,
Hackmann, Mueller, & Westbrook, 2004; Freeman et al., 2007).
The importance of determining a valid and reliable safety behaviors measure
for SA is clear. However, it was not certain if any existing safety behaviors measure
accurately assesses safety behaviors in a way that would be useful to clinicians and
researchers because they had yet to be rigorously examined. This study evaluated
the psychometric properties of four existing safety behaviors measures to determine
if any would be beneficial to the study and treatment of SA.
Validity Issues of Four Safety Behaviors Measures
Four self-report safety behaviors measures have been briefly described in the
current literature: the Liebowitz Social Anxiety Scale - Modified (LSAS-M; see
Appendix A; McManus et al., 2008); the Social Behavior Questionnaire - Modified
(SBQ-M; see Appendix B; McManus et al., 2008); the Social Phobia Safety
17
Behaviors Scale (SPSBS; see Appendix C; Pinto-Gouveia, Cunha, & Salvador,
2003); and the Presentation-Related Safety Behaviors Scale (PRSBQ; see Appendix
D; Kim, 2005). To determine if there is an adequate measure of safety behaviors,
the following issues were evaluated by examining the psychometric properties of
these four safety behaviors measures:
Issue 1: Representational Validity
A good measure of safety behaviors should demonstrate representational
validity (see Foster & Cone, 1995 for review), whether the safety behaviors measure
really assesses safety behaviors for individuals with SA. Establishing
representational validity is important for two reasons. First, representational validity
ensures that the safety behaviors measure properly defines the theoretical construct
of safety behaviors. Second, representational validity ensures the safety behaviors
measure properly interprets, measures, and tests the construct of safety behaviors,
according to SA theory. As subsequently described, representational validity
subsumes content validity and construct validity.
Content validity
According to Haynes, Richard, and Kubany (1995), content validity is the
degree to which items, response format, and measure instructions are relevant to and
18
representative of the construct of safety behaviors. Content validity includes domain
validity and face validity (Foster & Cone, 1995). Domain validity accounts for the
degree to which the safety behaviors measure covers the breadth of the domain of
safety behaviors. Face validity is whether the measure’s items appear to be
semantically related to safety behaviors. The LSAS-M, SBQ-M, and the SPSBS
appear to have adequate content validity, as described below. Content validity of the
PRSBQ may be problematic because of the small number of items.
The LSAS-M is an adaptation of the Liebowitz Social Anxiety Scale (LSAS;
Liebowitz, 1987). The Liebowitz Social Anxiety Scale - Modified (McManus et al.,
2008) instructs individuals to rate 24 social situations (e.g., meeting strangers) for
frequency of engagement in “preventative behaviors” (i.e., safety behaviors) that
keep other people from forming a negative impression. Two types of social
situations are measured, 13 performance anxiety situations (e.g., participating in
small groups) and 11 social anxiety situations (e.g., talking to people in authority).
The LSAS-M contains two scales. The frequency scale instructs individuals to
answer, “How frequently do you make active efforts to prevent other people getting
a negative impression of you in the following situations?” The frequency scale has
four response options from 0 = Never (0%) to 3 = Usually (67-100%). The
frequency scale appears to cover the breadth of social situations in which safety
behaviors may be used. While this scale does not directly assess safety behaviors
use, it was expected to provide a measure of likelihood of safety behaviors use
across various social situations. This scale was considered a safety behaviors
19
measure in this study. The second scale on the LSAS-M is the fear scale.
Participants rate each of the 24 social situations for degree of fear or anxiety felt in
each situation (McManus et al.). Response options range from 0 = None to 3 =
Severe for the fear scale. Consequently, the correlation between the frequency scale
and the fear scale is unavailable as data on the fear scale are not available from this
study.
The Social Behavior Questionnaire - Modified (McManus et al., 2008)
assesses the frequency that individuals engage in social anxiety-related safety
behaviors by rating how often given safety behaviors are performed prior to or
during a feared social situation. The SBQ-M is a modified version of the
unpublished Social Behavior Questionnaire (SBQ) by Clark and colleagues (1995).
The original SBQ lists 28 discrete (e.g., wear clothes or makeup to hide blushing)
and broad safety behaviors (e.g., think positive) related to social anxiety. The SBQ
was modified by McManus and colleagues to include six more safety behaviors,
with the intent to ensure broad application of the measure to most individuals. The
rating scale for the SBQ-M has four frequency options including Never (0),
Sometimes (1), Often (2), and Always (3). The SBQ-M appears to have adequate
face validity and domain validity.
The Social Phobia Safety Behaviours Scale (Pinto-Gouveia et al., 2003)
rates frequency of safety behaviors used in social situations. It is a 17-item selfreport questionnaire that instructs participants to rate how frequently they use safety
behaviors in feared social situations (e.g., looking away from or avoiding eye
20
contact with the person with whom you are interacting). Development of the SPSBS
resulted from interviews aimed at gathering information on common safety
behaviors used by individuals with SA. Once the list of commonly used safety
behaviors was formed, item-total correlations were completed for each of the
measure’s items. Following this process, seven items were removed from the
SPSBS, resulting in 15 items. The SPSBS was then administered to several more
participants with two open-response options for inclusion of safety behaviors not
listed. Two items, “trying to disguise your trembling” and “thinking very carefully
about what you are going to say before you speak,” were added to the SPSBS after
administering the measure with the two open-response options. To respond to each
item, participants select one of four frequency options Never (1), Occasionally (2),
Often (3), or Usually (4). Content validity of the SPSBS appears to be adequate.
The Presentation-Related Safety Behaviors Scale (Kim, 2005) measures the
frequency that individuals engage in safety behaviors during exposure treatment.
Participants are asked, “To what degree does each of the following statements
accurately describe how often you engage in these behaviors?” Nine commonly
used SA safety behaviors (e.g., avoided eye contact, focused on my voice, focused
on my enunciation) are listed on the PRSBQ with response scale ranging from 0 =
Not at All to 7 = Very Much. Face validity of the PRSBQ appears to be adequate.
Domain validity for the PRSBQ was questionable due to the small number of items,
which do not appear to cover the breadth of safety behaviors.
21
Construct validity
Construct validity is composed of convergent validity and discriminant
validity (Foster & Cone, 1995). Convergent validity of the self-report safety
behaviors measures would provide evidence that the safety behaviors scales all
measure the same construct: safety behaviors. High correlations between the safety
behaviors measures would demonstrate convergent validity (Campbell & Fiske,
1959). Safety behaviors measures should also display divergent validity 1 (Foster &
Cone, 1995). For example, low to moderately low correlations between the
measures of safety behaviors and measures of positive affect or acquisitiveness
would demonstrate divergence.
Convergent and divergent correlations were compared to establish that the
convergent correlations are larger than the divergent correlations. Convergent
correlations were correlations among the safety behaviors measures (LSAS-M,
SBQ-M, SPSBS, & PRSBQ). There were two types of divergent correlations –
correlations with related constructs and correlations with dissimilar constructs.
Related constructs were protectiveness as measured by the Social Self-Presentation
Styles Scale-Protectiveness (S-SPSS-P; see Appendix E; Meleshko & Alden, 1993),
fear of negative evaluation as measured by the Brief Fear of Negative Evaluation
Scale (BFNE; see Appendix F; Leary, 1983), and negative affect as measured by the
1
“Divergence” is substituted for “discriminant” to reduce confusion with
“discriminative” validity.
22
Positive and Negative Affect Scales-Negative Affect (PANAS-NA; see Appendix
G; Watson, Clark, & Tellegen, 1988). Dissimilar constructs were acquisitiveness as
measured by the Social Self-Presentation Styles Scale-Acquisitiveness (S-SPSS-A;
Meleshko & Alden), positive affect as measured by the Positive and Negative
Affect Scale-Positive Affect (PANAS-PA; Watson et al.), and depression as
measured by the Center for Epidemiological Studies Depression Scale (CES-D;
Radloff, 1977). The hypothesis for construct validity was:
The correlations among the safety behaviors measures are hypothesized to
be larger than the correlations of the safety behaviors measures with
measures of related constructs and measures of dissimilar constructs.
Issue 2: Criterion validity
According to Foster and Cone (1995), criterion validity involves relating
scores on the safety behaviors measures to some criterion of practical value, such as
scores of social anxiety. To establish criterion validity (Foster & Cone, 1995),
scores on the four safety behaviors measures should correlate with scores of social
anxiety as measured on the Social Interaction Anxiety Scale (SIAS; see Appendix
H; Mattick & Clarke, 1998) and the Social Phobia Scale (SPS; see Appendix I;
Mattick & Clarke). Establishing criterion validity was also important to ensure that
the safety behaviors measures did not predict a theoretically unrelated criterion,
such as depressive symptoms, based on scores of the Center for Epidemiologic
23
Studies Depression Scale (CES-D; see Appendix J; Radloff, 1977). The hypotheses
for criterion validity were as follows:
Scores of the safety behaviors measures are expected to moderately correlate
with scores of the Social Phobia Scale and the Social Interaction Anxiety
Scale, as would be expected by Clark & Wells’ (1995) theory of social
anxiety. These associations are hypothesized to be evident even after
controlling for the Center for Epidemiologic Studies Depression Scale.
CHAPTER II
METHOD
Sample
In this study, an undergraduate sample (N = 353) from a large Midwestern
university was used to evaluate the psychometric properties of four existing selfreport questionnaires intended to measure safety behaviors. A non-clinical sample
was used to study SA, as the appraisal processes underlying SA appear similar
between normal and disordered populations (Stopa & Clark, 2001). Though the
sample is considered an analogous sample, 32% of sample (n = 107) scored above
the recognized cut-off of 34 on the SIAS indicating high levels of social anxiety
(Rodebaugh, Woods, Heimberg, Leibowitz, & Schneier, 2006). Participants were
recruited through an introductory psychology course offered at the university.
Partial course credit was earned for participation. Approximately half of the sample
was male (51.1%; 46.4% female; and 2.5% preferred not to respond) and the
majority were Caucasian (68.5%; 14.4% African American; 13% Asian; 9.1%
Other; 0.3% American Indian; and 2.5% preferred not to respond). Average age of
the sample was 19 years-old (SD = 1.51; range 18-28). The majority of the
25
sample was in their first or second year of college (84.7%; 15.3% third year of
college and beyond), was never married (90.1%; 5.5% living with significant other;
1.1% married; and 0.6% preferred not to respond), and had an estimated family
income of $80,000 or less (87%; 13% preferred not to respond).
Procedures
The self-report measures (i.e., safety behaviors, related constructs, dissimilar
constructs, and social anxiety) were administered using internet-based programs
(i.e., SSiWeb and World Cast). SSiWeb is a program designed for creating online
questionnaires. Several questionnaire formats (e.g., open-ended, matrix, and
ranking) and HTML formatting are available to create online questionnaires that
appear to visually match paper-and-pencil versions of the same questionnaires.
Routine data collected using SSiWeb includes user id, completion time, and
responses to all included items. The World Cast program is a mass-emailing system.
World Cast was used to email individualized user identification numbers and
passwords to participants along with the website address that contained the study.
After reading the email describing the study, participants went to the study’s
website and completed the study using a secure computer. Data was routinely
downloaded from SSiWeb’s administration website and imported to SPSS for
analysis.
26
The random-ordered questionnaires were completed on the internet allowing
participants to complete the self-report questionnaires from any comfortable
location with secure internet access. First, the program led participants through an
introduction to the study, then an informed consent form followed by the option to
either agree or decline to participate in the research study. Second, ten self-report
questionnaires were completed and then demographic information was collected.
Each item in the study contained “prefer not respond” as a response option. Third,
debriefing, resources for psychological services, and instructions on how to obtain
the results of the study were provided to participants. If participants became
uncomfortable and wished to quit the study they could exit without penalty (i.e., loss
of credit for participation) by closing the internet browser window. It was possible
for participants to return to the website later to complete the study, if desired. The
entire study took participants an average of 30 minutes to complete.
Safety Behaviors Measures
Liebowitz Social Anxiety Scale – Modified
For the purposes of this study, only the frequency scale (i.e., frequency of
use of safety behaviors) score was administered and scored as it contains
information relevant to safety behaviors use across social situations (McManus et
al., 2008). The LSAS-M accounts for two types of situations: performance situations
27
and social situations. The total frequency score (range of zero to 72) was calculated
by summing scores on both performance situations and social situations. Internal
consistency (Cronbach’s alpha) for the total frequency score on the LSAS-M was
.93 for a high and low social anxiety sample (McManus et al). In the present study,
Cronbach’s alpha for the total frequency score was .95.
Social Behavior Questionnaire – Modified
The SBQ-M was scored by summing the scores for all 34 items (range of
zero to 102). Internal consistency for the SBQ-M was .88 for a high and low
socially anxious sample (McManus et al.). Cronbach’s alpha for the present study
was .92.
Social Phobia Safety Behaviors Questionnaire
The total score of the SPSBS was calculated by summing the scores to all
17 items (range of 0 to 51). The SPSBS has reliably discriminated individuals with
generalized social anxiety from individuals with non-generalized social anxiety,
other anxiety disorders, and from the normal population. Test-retest reliability
across four weeks was .69 and internal consistency was .82, .82, and .87 for the
general population, social anxiety, and other anxiety-disordered patients,
28
respectively (Pinto-Gouveia et al., 2003). For this study, internal consistency of the
SPSBS was .90.
Presentation-Related Safety Behaviors Scale
Total scores on the PRSBQ were calculated by summing the scores of all
nine items (range of zero to 54). Validity of the PRSBQ with a clinical sample of
individuals with SAD has been evaluated using correlations with two related ratings,
anxiety and belief in feared outcomes (Kim, 2005). Correlation of anxiety ratings
and belief in feared outcomes ratings with the PRSBQ were r = .53 and r = .50,
respectively. The correlation between anxiety ratings and belief in feared outcome
ratings was r = .74. Internal consistency was .80 (Kim, 2005). Cronbach’s alpha was
.87 in the present study.
Measures of Related and Dissimilar Constructs
Social Self-Presentation Style Scale
The S-SPSS (Meleshko & Alden, 1993) is an eight-item self-report measure
that assesses the style (i.e., protective or acquisitive) in which individuals present
themselves during social situations. The acquisitive subscale (S-SPSS-A; e.g., Once
I knew what the situation called for, it was easy for me to regulate my behavior) and
29
protective subscale (S-SPSS-P; e.g., I was careful about what I said because I was
afraid that I might say or do something wrong) reflect two opposing motivational
states during social interactions. Participants are instructed to report to what extent
each of the social styles accurately describes feelings and reactions during
conversation. Response options are on a 7-point scale from Completely False (1) to
Completely True (7). Items are equally divided into two subscales (i.e., acquisitive
and protective) and appropriately totaled (range of 4 to 28 for each subscale).
Cronbach’s alphas were .73 for the protective subscale and .70 for the acquisitive
subscale in a sample of high and low socially anxious individuals (Meleshko &
Alden). In this study, internal consistency was .79 for the protective subscale and
.76 for the acquisitive subscale. The correlation of the protective subscale and the
acquisitive subscale was .16, p < .01.
Positive and Negative Affect Scale
The Positive and Negative Affect Scale (PANAS; Watson et al., 1988)
assesses feelings and emotions that individuals have felt during the past month.
There are 20 items and 5 response options on the PANAS. Response options are
Very slightly or not at all, A little, Moderately, Quite a bit, or Extremely. The
PANAS is composed of two subscales, positive affect (PA; e.g., interested) and
negative affect (NA; e.g., upset). Total scores are calculated by totaling the PA
items and the NA items respectively (range of zero to 40 for each subscale). Positive
30
affect refers to how enthusiastic, active, and alert a person feels (Watson). People
that rate high on PA generally are full of energy, can concentrate, and are pleasant.
People who rate low on PA generally are sad and lethargic. Negative affect
measures subjective distress and lack of pleasure that may represent anger,
contempt, disgust, fear, and nervousness. People who score low on NA generally are
calm and serene. According to Watson and colleagues, scoring low on PA and high
on NA is theoretically characteristic of people with depression and anxiety. Internal
consistency was between .86 and .90 for PA and between .84 and .87 for NA for an
analogue sample (Watson et al.). Cronbach’s alpha was .89 for NA and .89 for PA
in the present study. The correlation between NA and PA was .06.
Brief Fear of Negative Evaluation Scale
The BFNE (Leary, 1983) captures self-presentation concerns of socially
anxious individuals. The original Fear of Negative Evaluation measure (Watson &
Friend, 1969) is a 30-item, true-false questionnaire that assesses fear of negative
evaluation from others (e.g., I am afraid that I may look ridiculous and make a fool
of myself or I feel very upset when I commit some social error). The BFNE contains
12 of the original 30 items and uses a five-point response format. Each item is rated
from Not at all characteristic (1) to Extremely characteristic (5). As recommended
by Rodenbaugh and colleagues (2004), the BFNE was scored using only the eight
straight-forwardly worded items (range of 8 to 40). Internal consistency for the
31
BFNE was.89 for a social phobic sample (Weeks et al., 2005). For the present study,
Cronbach’s alpha was .82.
Center for Epidemiologic Studies Depression Scale
The Center for Epidemiologic Studies Depression Scale (CES-D; Radloff,
1977) is a self-report questionnaire that measures depressive symptomology in the
general population. The CES-D items are based on items used in previously
validated scales for clinical diagnosis or evaluation of depression severity. The
directions instruct participants to indicate how they have felt or behaved in the past
week. “Past week” is included in the instructions to orient participants to their
current functioning. The CES-D contains 20 items rated from 0 = Rarely or none of
the time (less than 1 day) to 3 = Most of the time (5- 7 days). Select items are “I was
bothered by things that usually don’t bother me,” “I felt depressed,” and “I had
crying spells.” Total scores are calculated by reverse-scoring four items (e.g., I was
happy) and summing all the items. Possible scores range from zero to 60, with
higher scores indicating more symptoms, weighted by frequency of occurrence.
Internal consistency for the CES-D is .85 in the general sample and .90 in a
clinically depressed sample (Radloff). Cronbach’s alpha for the present study was
.89.
32
Social Anxiety Measures
Social Interaction Anxiety Scale
The SIAS (Mattick & Clarke, 1998) measures typical cognitive, affective, or
behavioral reactions to situations requiring social interaction (e.g., meeting and
talking with other people). The SIAS contains 20 items rated from 0 = Not at all
characteristic or true of me to 4 = Extremely characteristic or true of me. All items
on the SIAS are negatively-worded except for three positively-worded items. Total
scores are obtained by summing all item scores after reversing the three positivelyworded items. Total scores range from zero to 80; high scores represent higher
levels of social interaction anxiety. Internal consistency for the SIASwas .93 for a
social phobic sample and .88 for a college sample (Mattick & Clarke). In the present
study, Cronbach’s alpha was .92.
Social Phobia Scale
The SPS (Mattick & Clarke, 1998) is the companion scale to the SIAS. The
SPS measures social anxiety the authors defined as “anxiety and fear at the prospect
of being observed or watched by other people, in particular, where the individual
expresses distress when undertaking certain activities in the presence of others.”
There are 20 items on the SPS (e.g., I become anxious if I have to write in front of
33
other people). Participants are asked to indicate the degree to which they feel the 20
items are characteristic or true. Responses options are Not at all, Slightly,
Moderately, Very, and Extremely. To score the SPS, all 20 items are totaled (range
of zero to 80). Internal consistency for the SPS was .89 for a social phobic sample
and .80 for a college sample (Mattick & Clarke). Internal consistency was .95 in the
present study. The correlation between the SPS and SIAS was .76, p < .01.
Other
Demographic Questionnaire
The Demographic Questionnaire (DQ) is a self-report questionnaire. The DQ
consists of questions regarding gender, age, ethnic background, annual family
income, education level, and marital status. The DQ was included to describe
participants’ demographic characteristics.
CHAPTER III
RESULTS
Descriptive Statistics
Means and standard deviations for the safety behaviors, related, dissimilar,
and social anxiety measures are shown in Table 1. The correlations among the
related, dissimilar, and social anxiety measures, although not needed to test the
study’s hypotheses, are presented in Table 2.
Reliability
Internal Consistency
To examine internal consistency reliability, Cronbach’s alphas were
calculated for each of the measures. To be considered adequate, alphas should be
greater than .70. In Table 1, the internal consistency coefficients of the measures
(i.e., safety behaviors, related constructs, dissimilar constructs, and social anxiety)
are displayed. The alpha coefficients range from .87 to .95 for the safety behaviors
35
Table 1: Sample Size, Cronbach’s Alpha, Means, Standard Deviations of and
Correlations Between Measures of Safety Behaviors, Related, Dissimilar, and the
Social Anxiety Constructs
N
Cronbach’s
alpha
Mean (SD)
LSAS-M
LSAS-M
327
.95
45.52 (16.30)
-
SBQ-M
327
.92
41.59 (15.02)
.45a
SPSBS
337
.90
35.37 (8.87)
.36
b
b,c
SBQ-M
SPSBS
PRSBQ
.77c
.69d
-
.56e,f
.63e
.61e
.25c
.59e
.63e
.51f
25.31 (8.54)
.23c
.49a,f
.49f
.41b
.76
17.46 (5.19)
.26c
.18g
.02g
.19g
339
.89
34.50 (7.59)
.17c
.15g
.00g
.14h
CES-D
336
.89
21.14 (8.80)
.06d
.32h
.38b
.26b,g
SIAS
339
.92
26.94 (12.81)
.13
.54
.67
.55
SPS
337
.95
18.58 (15.08)
.18
.60
.72
.55
PRSBQ
341
.87
22.30 (10.58)
.30
S-SPSS-P
340
.79
12.83 (5.12)
.23c
BFNE
PANASNA
338
.82
19.67 (7.51)
340
.89
S-SPSS-A
337
PANAS-PA
.69
d
Note. By the Meng, Rosenthal, and Rubin (1992) method of comparing correlated
coefficients, correlations between measures of monotrait, related, and dissimilar
constructs in the same column with the same superscript do not differ significantly
at p < .05.
LSAS-M = Liebowitz Social Anxiety Scale-Modified; SBQ-M = Social Behavior
Scale-Modified; SPSBS = Social Phobia Safety Behaviours Scale; PRSBQ =
Presentation Related Safety Behaviors Questionnaire; S-SPSS-P = Social SelfPresentations Style Scale- Protective; BFNE = Brief Fear of Negative Evaluation;
PANAS-PA = Positive and Negative Affect Scale-Positive Affect; S-SPSS-A =
Social Self-Presentations Style Scale- Acquisitive; PANAS-PA = Positive and
Negative Affect Scale-Positive Affect; CES-D = Center for Epidemiologic Studies
Depression Scale; SIAS = Social Interaction Anxiety Scale; SPS = Social Phobia
Scale.
Correlations where p < .01, two-tailed are in boldface type; bold italicized
correlations are where p < .05, two-tailed; nonsignificant correlations are in plain
type.
36
S-SPSS-P
340
.79
BFNE
338
.82
.63
PANAS-NA
340
.89
.36
.48
S-SPSS-A
337
.76
.16
-.86
-.02
PANAS-PA
339
.89
-.02
-.14
.06
.30
CES-D
336
.89
.27
.43
.68
-.16
-.24
SIAS
339
.92
.64
.77
.41
-.23
-.15
.38
SPS
337
.95
.60
.71
.44
-.04
-.05
.45
SIAS
CES-D
PANAS-PA
S-SPSS-A
PANASNA
BFNE
S-SPSS-P
N
Cronbach’s
alpha
Table 2: Correlations Between the Social-Self Presentation Style Scale, Brief Fear
of Negative Evaluation Scale, Positive and Negative Affect Scale, Center for
Epidemiologic Studies Depression Scale, Social Interaction Anxiety Scale, and the
Social Phobia Scale
.76
Note. S-SPSS-P = Social Self-Presentations Style Scale- Protective; BFNE = Brief
Fear of Negative Evaluation; PANAS-PA = Positive and Negative Affect ScalePositive Affect; S-SPSS-A = Social Self-Presentations Style Scale- Acquisitive;
PANAS-PA = Positive and Negative Affect Scale-Positive Affect; CES-D = Center
for Epidemiologic Studies Depression Scale; SIAS = Social Interaction Anxiety
Scale; SPS = Social Phobia Scale.
Correlations where p < .01, two-tailed are in boldface type; bold italicized
correlations are where p < .05, two-tailed; nonsignificant correlations are in plain
type.
37
measures, indicating moderate-to-high internal consistency. For the related
constructs, alpha coefficients range from .79 to .89, indicating moderate internal
consistency. The alpha coefficients of the dissimilar constructs ranged from .76 to
.89, each indicating moderate internal consistency. Lastly, the social anxiety
measures had alpha coefficients of .92 (SIAS) and .95 (SPS), indicating high
internal consistency. All preceding reliability coefficients were significant at an
alpha level of .01. In sum, reliability analyses indicate that the measures used in the
present study display adequate internal consistency.
Average Inter-item Correlations
According to Clark and Watson (1995), a narrow construct such as safety
behaviors should have an average inter-item correlation in the .30 to .50 range (or
perhaps higher). To examine this issue, the average inter-item correlation was
calculated for each of the safety behaviors measures. The average inter-item
correlations were .46 for the LSAS-M, .24 for the SBQ-M, .34 for the SPSBS, and
.42 for the PRSBQ.
Next, the range and distribution of the correlations were examined. For
measures thought to assess a single construct (unidimensional measure) the interitem correlation coefficients should be moderate in magnitude (i.e., within the range
.15 to .50) and cluster around the measure’s average inter-item correlation.
Moderate magnitude is preferred to high magnitude to avoid the attenuation paradox
38
(internal consistency beyond a given point may decrease construct validity;
Loevinger, 1954). A more valid safety behaviors measure will have differentiated
items that are moderately correlated, producing more information about the breath
and intensity of safety behavior use.
Inter-item correlations were then compared to the appropriate ranges. The
range of inter-item correlations for the LSAS-M was .31 to .78. The distribution of
the inter-item correlations (SD = .09) shows that the inter-item correlations cluster
around the average inter-item correlation.
On the SBQ-M, the inter-item correlation range was -.03 to .58 with a SD of
.13. Combined, the range, average inter-item correlation, and the distribution of the
SBQ-M items support the notion that the SBQ-M items are measuring at least two
related constructs. The measure appears to adequately represent the breadth and
intensity of safety behavior use.
The SPSBS inter-item correlations ranged from .14 to .58, SD = .10. It is
probable that the SPSBS is measuring two related constructs given the range,
average inter-item correlation, and the distribution of the inter-item correlations.
Lastly, inter-item correlations for the PRSBQ were in the .21 to .79 range
with a SD of .13. The range of inter-item correlations should be interpreted with
caution given that the PRSBQ is a nine-item measure. When examined on an
individual item basis, the inter-item correlations suggest that there are two
underlying constructs that are partially represented by the PRSBQ items.
39
Disattenuated correlations
To examine the true correlation between the conceptual variables (i.e., safety
behaviors construct) rather than the correlation between the observed measures, an
estimate of what the correlation would be after taking random measurement error
into account was calculated using the disattenuated correlations formula (Aiken &
West, 1991). The magnitude of the relationship, represented by a Pearson’s
correlation coefficient, between the safety behaviors constructs was found by
dividing the observed correlation between two safety behaviors measures by the
square root of the product of the reliabilities of such measures. For all calculations,
measurement error was assumed to be random and estimates of the reliabilities were
assumed to be unbiased.
Six disattenuated correlations were calculated to determine the true
correlations between the constructs underlying the safety behaviors measures. The
disattenuated correlations of the LSAS-M were .48, .39, and .33 with the SBQ-M,
SPSBS, and the PRSBQ, respectively. The SBQ-M and the SPSBS disattenuated
correlation was .85. The SBQ-M and the PRSBQ disattenuated correlation was .77.
Lastly, the SPSBS and the PRSBQ disattenuated correlation was .78.
40
Representational Validity
This study included several measures thought to be distinct but theoretically
related to or dissimilar from safety behaviors common to social anxiety believed to
be measured by the safety behaviors measures. Validity issues were evaluated using
Pearson’s correlations (with two-tailed tests). The correlated coefficients
(correlations obtained from a single sample) method by Meng, Rosenthal, and
Rubin (1992) was used to examine whether the differences between correlations are
significant at p < .05. This method compares correlated correlation coefficients that
have a variable in common (e.g., safety behaviors). Table 1 displays the correlations
between the safety behaviors measures and measures of related constructs,
dissimilar constructs, and social anxiety. Table 2 displays the correlations between
the measures of related constructs, dissimilar constructs, and social anxiety
measures.
To test representational validity, convergent and divergent correlations were
compared. The correlations among the safety behaviors measures were hypothesized
to be larger than the correlations of the safety behaviors measures with measures of
related constructs and measures of dissimilar constructs.
41
Convergent Validity
Convergence was examined by calculating Pearson’s correlations of the four
safety behaviors measures (LSAS-M, SBQ-M, SPSBS, and PRSBQ). As seen in
Table 1, all the convergent correlations were significant at an alpha level of .01
(two-tailed). Although all the correlations were significantly different from zero,
visual inspection of the convergent correlations indicated that the LSAS-M might
not converge with the other measures. Additional analyses were conducted to
determine if the magnitude of the correlation coefficients with the LSAS-M were
significantly different from the correlation coefficients among the other safety
behavior measures. Specifically, the correlation between the LSAS-M and the SBQM was .45, which was significantly different from all other convergent correlations
(z = 9.60, p < .01). Correlations of the LSAS with the SPSBS (r = .36) and PRSBQ
(r = .30) were not statistically different from each other (z = 1.47, p > .05), but the
correlation coefficients were statistically different from all other convergent
correlations (all p’s < .05). These correlations indicate that the LSAS-M shows
significantly lower convergent validity with the other measures of safety behaviors.
Additional analyses were conducted to examine relationships among the
remaining convergent correlations. The correlation coefficient between the SBQ-M
and the SPSBS (r = .77) was significantly different from all other safety behaviors
correlation coefficients (p < .05). The correlation coefficient of the SBQ-M and
PRSBQ (r = .69) did not differ significantly from the correlation coefficient of the
42
SPSBS and PRSBQ (r = .69; z = 0.00, p > .05). In sum, anticipated convergence of
the remaining safety behavior measures was confirmed by correlated coefficients
analysis.
Divergent Validity
Next, divergent validity was tested. The Meng and colleagues (1992) method
of examining correlated coefficients was again used to determine whether the
correlations significantly differed as proposed. The magnitude of the correlations
among safety behaviors correlations were compared to the magnitude of the
correlations between safety behaviors measures and measures of related and
dissimilar constructs (see Table 1).
To evaluate the divergent validity of the LSAS-M, its convergent
correlations (range = .30 to .45, see Table 1) were compared to its divergent
correlations (range = .06 to .25). The lowest convergent correlation of the LSAS-M
(r = .30 with the PRSBQ) did not differ significantly from the three highest
divergent correlations r = .25 with the BFNE; r = .23 with the PANAS-NA; and r =
.23 with the S-SPSS-P (z’s = 0.96, -1.04, and 0.07, respectively, all p’s > .05). The
lowest convergent correlation of LSAS-M did not significantly differ from the SSPSS-A (r = .26; z = 0.51, p > .05). Lastly, the lowest convergent correlation of
LSAS-M did differ significantly from the correlation of the LSAS-M and the CESD (r = .06, p > .05; z = 3.10, p < .01).
43
To evaluate the divergent validity of the SBQ-M, its convergent correlations
(range = .45 to .77, see Table 1) were compared to its divergent correlations (range
= .15 to .59). The lowest convergent correlation of the SBQ-M (r = .45 with the
LSAS-M) was significantly greater than three of the divergent correlations of the
SBQ-M (r = .15 with the PANAS-PA; r = .18 with the S-SPSS-A; and r = .32 with
the CES-D; z’s = 3.84, -3.68, and -1.65, respectively, p < .05), but was smaller in
magnitude than the three remaining divergent correlations (r = .49 with the PANASNA (z = 0.59, p > .05); r = .56 with the S-SPSS-P (z = 1.70, p < .05); and r = .59
with the BFNE (z = 2.22, p < .05). The other convergent correlations of the SBQ-M
(r = .69 with the PRSBQ and r = .77 with the SPSBS) were significantly greater
than all of the divergent correlations of the SBQ-M (all p’s < .05).
To evaluate the divergent validity of the SPSBS, its convergent correlations
(range = .36 to .77, see Table 1) were compared to its divergent correlations (range
= -.00 to .63). The lowest convergent correlation of the SPSBS (r = .36 with the
LSAS-M) was significantly greater than two of the divergent correlations of the
SPSBS (r = -.00 with the PANAS-PA and r = .02 with the S-SPSS-A; z’s = 4.40,
4.41, respectively, both p’s < .01), but was not significantly different from one
divergent correlation (r = .38 with the CES-D; z = 0.25, p > .05). Further, the three
remaining divergent correlations (r = .49 with the PANAS-NA; z = 1.86, p < .05; r
= .63 with the S-SPSS-P, z = 4.16, p < .01; and r = .63 with the BFNE; z = 4.21, p <
.01) were significantly greater than the convergent correlation of the SPSBS and the
LSAS-M. The other convergent correlations of the SPSBS (r = .77 with the SBQ-M
44
and r = .69 with the PRSBQ) were significantly greater than all of the divergent
correlations of the SPSBS (all p’s < .05).
To evaluate the divergent validity of the PRSBQ, its convergent correlations
(range = .30 to .69, see Table 1) were compared to its divergent correlations (range
= .14 to .61). The lowest convergent correlation of the PRSBQ (r = .30 with the
LSAS-M) was significantly different from the PANAS-PA (r = .14; z = 1.97, p <
.05) but did not differ significantly from two other divergent correlations of the
PRSBQ (r = .19 with the S-SPSS-A; and r = .26 with the CES-D; z’s = 1.45 and
0.48, respectively, both p’s > .05). The correlation of the PRSBQ with the LSAS-M
was not significantly lower the correlation of the PRSBQ and the PANAS-NA (r =
.41; z = 1.50, p > .05). The correlation of the PRSBQ with the LSAS-M was
significantly lower than the two remaining divergent correlations (r = .61 with the
S-SPSS-P; and r = .51 with the BFNE; z’s = 4.60 and 3.00, respectively, both p’s <
.01). The other convergent correlations of the PRSBQ (r = .69 with the SBQ-M and
r = .69 with the SPSBS) were significantly greater than all of the divergent
correlations of the PRSBQ (all p’s < .05).
Summary
In sum, the LSAS-M did not show adequate convergent and divergent
validity. The SBQ-M, SPSBS, and the PRSBQ did show adequate convergent and
divergent validity.
45
Criterion Validity
Correlations with Criterion Measures
To test criterion validity, Pearson’s correlations between the safety behaviors
measures and the social anxiety measures (SIAS & SPS) were examined. Table 1
presents these correlations. All correlations were significant (two-tailed) at an alpha
level of .01 with the exception of the correlation between the LSAS-M and the
SIAS, which was significant at an alpha level of .05.
Partial Correlations
Partial correlations were calculated after controlling for depression (using
the CES-D) to examine the degree of association between the safety behaviors
measures and the social anxiety measures. Partial correlations of the LSAS-M with
SIAS and the SPS were .11 and .17, correspondingly (both p’s < .05). The partial
correlations of the SIAS and the remaining safety behaviors measures were R = .48
with the SBQ-M; R = .63 with the SPSBS; and R = .51 with the PRSBQ (all p’s <
.01). Partial correlations of the SPS and the three safety behaviors measures were R
= .54 with the SBQ-M; R = .67 with the SPSBS; and R = .50 with the PRSBQ (all
p’s < .01). In sum, all partial correlations among the safety behaviors measures and
46
the criterion measures were significant after controlling for symptoms of depression
(see Table 3).
Table 3: Partial Correlations between the Safety Behaviors Measures and the Social
Interaction Anxiety Scale and the Social Phobia Scale Controlling for the Center for
Epidemiologic Studies Depression Scale.
N
SIAS
N
SPS
LSAS-M
310
.11
309
.17
SBQ-M
312
.48
311
.54
SPSBS
321
.63
320
.67
PRSBQ
323
.51
322
.50
Note. LSAS-M = Liebowitz Social Anxiety Scale-Modified; SBQ-M = Social
Behavior Scale-Modified; SPSBS = Social Phobia Safety Behaviours Scale;
PRSBQ = Presentation Related Safety Behaviors Questionnaire; SIAS = Social
Interaction Anxiety Scale; SPS = Social Phobia Scale.
Correlations where p < .01, two-tailed are in boldface type; bold italicized
correlations are where p < .05, two-tailed.
Component Structures
To examine the component structure underlying each of the safety behaviors
measures exploratory component analyses were conducted using principal
component analysis (PCA) with oblique (Oblimin) rotations. The variance-focused
47
approach of PCA was used to examine component structure because the objective
was data reduction (Bryant & Yarnold, 1995). Although the correlation-focused
principal axis factoring (PAF) approach (a.k.a., principal factor analysis; PFA) was
considered as a method to examine factor structure, PAF was not reported as it is
intended for theory confirmation and causal modeling; which were not objectives of
this study. Results of the PFA were substantially similar to the results of the PCA
with the following exceptions: going to a party and expressing a disagreement or
disapproval to people you don’t know very well no longer had salient loadings on
either component of the LSAS-M (both had salient loadings on the Verbal
Situations component in the PCA) and stopping what you were doing while being
observed no longer had a salient loading on either component of the SPSBS (loaded
saliently on Putting On component in the PCA).
As recommended by Zwick and Velicer (1986), parallel analyses (Horn,
1965; Longman, Cota, Holden, & Fekken, 1989) were conducted to determine how
many components could be reliably extracted and interpreted from the PCA. Parallel
analyses were conducted at both the mean and 95th percentile eigenvalues.
Component loadings greater than |.40| were considered salient item loadings
(Thurstone, 1947).
48
LSAS-M Component Structure
The first four eigenvalues and corresponding variance estimates on the
LSAS-M were 11.58 (48.2%), 1.68 (7.0%), 1.12 (4.7%), and 0.82 (3.4%). Three
eigenvalues were above 1.00 explaining 59.9% of the variance; however, parallel
analyses, for both mean and 95th percentile, indicated that two components could
reliably be extracted and interpreted. Two components were extracted from the
principal components analysis of the LSAS-M using the oblique extraction method
(see Table 4 for component loadings and communalities). A substantial amount of
the total variance, 55.2%, was explained by two components.
The LSAS-M two-component solution had good simple structure based on
the criteria that there were no items with loadings on multiple components, each
item had a salient loading, and each component had an adequate number of items
with salient loadings (Thurstone, 1947). Of the 24 items on the LSAS-M, 15 items
had salient loadings on the first component; component loadings ranged from .40 to
.93. The first component was labeled “Verbal Situations (VS)” as the items concern
verbal situations in which negative evaluation is possible (e.g., acting, performing
or giving a talk in front of an audience; talking to people in authority). The
remaining 9 items of the LSAS-M had salient loadings on the second component of
the LSAS-M. Component loadings on the second component ranged from .55 to .88.
The second component was labeled “Nonverbal Situations (NS)” as the items
49
Table 4: Principal Components Analysis of the Liebowitz Social Anxiety ScaleModified: Component Loadings for the Two-Component Solution
LSAS-M item
1.
2.
3.
4.
5.
6.
Telephone in public
Participating in small groups
Eating in public places
Drinking with others in public places
Talking to people in authority
Acting, performing or giving a talk in front of an
audience
7. Going to a party
8. Working while being observed
9. Writing while being observed
10. Calling someone you don’t know very well
11. Talking with people you don’t know very well
12. Meeting strangers
13. Urinating in a public bathroom
14. Entering a room when others are already seated
15. Being the center of attention
16. Speaking up at a meeting
17. Taking a test
18. Expressing disagreement or disapproval to people
you don’t know very well
19. Looking at people you don’t know very well in
the eyes
20. Giving a report to a group
21. Trying to pick up someone
22. Returning goods to a store
23. Giving a party
24. Resisting a high pressure salesperson
LSAS-M Component
VS
NS
h2
.02
.40
.62
.29
.46
.46
.14
.49
.61
.18
.46
.55
-.14
.59
.85
.93
-.20
.67
.43
.83
.70
.84
.81
.75
-.22
.17
.49
.63
.05
.28
-.03
.08
-.01
.02
.07
.88
.63
.29
.19
.70
.41
.65
.57
.69
.68
.63
.59
.56
.50
.58
.53
.40
.37
.47
.24
.60
.60
.64
.52
.23
.51
-.04
.20
.29
.57
.30
.72
.61
.53
.54
.54
.48
Note. LSAS-M = Liebowitz Social Anxiety Scale-Modified; VS = Verbal
Situations; NS = Nonverbal Situations; h2 = communality.
Component loadings > |.40| are listed in boldface type.
50
concern nonverbal or performance situations in which negative evaluation is
possible (e.g., urinating in a public restroom; taking a test).
SBQ-M Component Structure
The first nine eigenvalues and corresponding variance estimates on the SBQM were 9.33 (27.5%), 3.19 (9.4%), 1.97 (5.8%), 1.32 (3.9%), 1.26 (3.7%), 1.19
(3.5%), 1.10 (3.2%), 1.01 (3.2%), and 0.92 (2.7%). The principal components
analysis of the SBQ-M revealed eight eigenvalues above 1.0 which represented
59.9% of the total variance. Parallel analyses for both mean and 95th percentile
indicated that three components were clearly present and interpretable. Three
components were extracted from the principal components analysis of the SBQ-M
using the oblique extraction method (see Table 5 for component loadings and
communalities). Three components explained 42.6% of the total variance.
Of the 34 items on the SBQ-M, 29 items had salient loadings. Fifteen items
had salient loadings on the first component with component loadings ranging from
.42 to .77. The first component was labeled “Unobtrusive Intent (UI)” as the items
consist of purposeful behaviors to maintain inconspicuousness (e.g., hide your face;
avoid eye contact). Eleven items of the SBQ-M had salient component loadings
ranging from .45 to .78 on the second component. The second component was
labeled “Impression Management (IM)” as the items concern self-monitoring of
social behavior and resultant behavioral adjustments meant to create or maintain a
51
Table 5: Principal Components Analysis of the Social Behavior QuestionnaireModified: Component Loadings for the Three- Component Solution
SBQ-M item
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
Use alcohol to manage anxiety
Try not to attract attention
Make an effort to get your words right
Check that you are coming across well
Avoid eye contact
Talk less
Avoid asking questions
Try to picture how you appear to others
Grip cups or glasses tightly
Position yourself so as not to be noticed
Try to control shaking
Choose clothes that will prevent or conceal
sweating
Wear clothes or makeup to hide blushing
Rehearse sentences in your mind
Censor what you are going to say
Blank out or switch off mentally
Avoid talking about yourself
Keep still
Ask lots of questions
Think positive
Stay on the edge of groups
Avoid pauses in speech
Hide your face
Try to think about other things
Talk more
Try to act normal
Try to keep tight control of your behavior
Make an effort to come across well
Try to hide any signs of anxiety
Try to think of entertaining things to say
Focus on yourself
Try to appear clever or knowledgeable
Try to look at ease
Try to appear witty
UI
.28
.52
.12
.12
.74
.64
.66
.23
.50
.73
.45
SBQ-M Component
IM
CB
h2
-.04
.37
.23
.27
-.33
.46
-.16
.42
.62
-.04
.42
.51
-.01
-.11
.53
.19
-.37
.57
.04
-.22
.46
.08
.35
.45
.01
.29
.37
.04
-.05
.55
.10
.26
.36
.38
.10
.30
.32
.54
.45
.44
.68
.52
.36
.03
-.38
.42
.24
.77
.57
-.05
.00
.20
-.11
.31
-.11
.07
-.05
-.05
-.03
-.15
.20
.34
.04
.11
.12
.04
.27
.19
.16
-.19
.10
.07
.70
.58
.78
.57
.69
.14
.70
.77
.56
.36
.11
.09
.15
-.03
.05
.59
.29
.20
.31
.15
.10
.65
-.13
-.04
-.01
-.05
.20
.54
.17
.05
.27
.41
.34
.45
.53
.44
.18
.37
.25
.35
.26
.57
.40
.44
.45
.44
.56
.52
.54
.37
.55
.60
.45
Note. SBQ-M = Social Behavior Questionnaire-Modified; UI = Unobtrusive Intent;
IM = Impression Management; CB = Compensatory Behavior; h2 = communality.
Component loadings > |.40| are listed in boldface type.
52
positive impression (e.g., make an effort to come across well; try to look at ease).
Three items of the SBQ-M had salient loadings ranging from .54 to .59 on the third
component. The third component was labeled “Compensatory Behavior (CB)” as
the items concern overcompensating for perceived social failure (e.g., talk more).
The remaining five items (i.e., use alcohol to manage anxiety; choose clothes that
will prevent or conceal sweating; keep still; think positive; and avoid pauses in
speech) did not load saliently on any of the components (i.e., component loadings
less than |.40|).
SPSBS Component Structure
The first four eigenvalues and corresponding variance estimates on the
SPSBS were 6.47 (38.1%), 1.66 (9.7%), 1.04 (6.1%), and 0.91 (5.3%). Three
eigenvalues were above 1.0 explaining 54.0% of the total variance; however,
parallel analyses, for both mean and 95th percentile, indicated that the first two
components could reliably be extracted and interpreted. Two components were
extracted from the principal components analysis of the SPSBS using the oblique
extraction method (see Table 6 for component loadings and communalities). Two
components explained 42.6% of the total variance.
The SPSBS two-component solution had good simple structure based on the
criteria that there were no items with loadings on multiple components, one item
with no salient component loading, and each component had an adequate number of
53
Table 6: Principal Components Analysis of the Social Phobia Safety Behaviors
Scale: Component Loadings for the Two-Component Solution
SPSBS item
1. Looking away from or avoiding eye
contact with the person with whom you are
interacting.
2. Speeding up your speech, talking quickly
and without pauses.
3. Shortening your speech, drastically
reducing what you have to say.
4. Avoiding attracting attention to yourself.
SPSBS Component
Laying
Putting
h2
Low
On
.67
-.12
.38
.54
.11
.36
.69
.06
.52
.77
.00
.60
5. Getting a seat as hidden as you can.
.83
-.09
.61
6. Pretending you are not interested or you
are distant from what is happening.
7. Limiting yourself to being a passive
spectator of a situation.
8. Pretending you did not see someone.
.71
-.04
.48
.80
-.02
.62
.55
.13
.40
9. Walking with your head down.
.69
-.01
.47
10. Putting your hands in your pockets.
.32
.27
.26
11. Stopping doing what you were doing (e.g.,
writing, drinking, etc.) while being
observed.
12. Trying to look at ease.
.29
.42
.38
-.02
.69
.47
-.06
.81
.61
-.08
.84
.64
.48
.28
.44
.19
.54
.43
.01
.68
.46
13. Laughing to hide the fact that you are
nervous.
14. Constantly checking if you are presentable.
15. Increasing the distance between yourself
and the person you are talking to.
16. Trying to disguise your trembling.
17. Thinking very carefully about what you are
going to say before you speak.
Note. SPSBS = Social Phobia Safety Behaviors Scale; h2 = communality.
Component loadings > |.40| are listed in boldface type.
54
items with salient loadings (Thurstone, 1947).Of the 17 items on the SPSBS, 10
items had salient loadings on the first component with component loadings ranging
from .48 to .83. The first component was labeled “Laying Low (LL)” as the items
consist of deliberate behaviors that produce physical, mental, or emotional distance
in order to minimize the likelihood of unwanted social attention (e.g., getting a seat
as hidden as you can; limiting yourself to being a passive spectator of a situation).
Six items of the SPSBS had salient loadings (ranging from .42 to .84) on the second
component which was labeled “Putting On (PO).” “Putting On” items assess the
desire to present oneself as physically, mentally, and emotionally composed via
suppression or disguise of emotional distress (e.g., constantly checking if you are
presentable; laughing to hide the fact that you are nervous). The remaining item,
putting your hands in your pockets, did not load saliently on either component (i.e.,
component loading less than |.40|).
PRSBQ Component Structure
The first three eigenvalues and corresponding variance estimates on the
PRSBQ were 4.20 (49.1%), 1.11 (12.3%), and 0.90 (10.0%).The principal
components analysis of the PRSBQ revealed two eigenvalues above 1.0 which
represented 61.4% of the total variance. Parallel analyses for both mean and 95th
percentile indicated that one component was clearly present and interpretable. One
component was extracted from the principal components analysis of the PRSBQ
55
(see Table 7 for component loadings and communalities). Principal component
extraction revealed that all 9 items of the PRSBQ had salient loadings on the first
component; loadings ranged from .54 to .81. One component explained 49.1% of
the total variance.
Table 7: Principal Components Analysis of the Presentation-Related Safety
Behaviors Questionnaire: Component Loadings for the One-Component Solution
PRSBQ item
1.
2.
3.
4.
5.
6.
7.
8.
9.
Avoided eye-contact
Focused on my voice
Focused on my enunciation
Monitored continuously what I said
Avoided stop talking
Thought ahead about what to say
Talked as little as possible
Drew as little attention to self as possible
Took deep breaths
PRSBQ Component
Factor I
h2
.29
.54
.57
.76
.56
.75
.65
.81
.36
.60
.49
.70
.53
.73
.57
.76
.41
.64
Note. PRSBQ = Presentation-Related Safety Behaviors Questionnaire; h2 =
communality.
Component loadings > |.40| are listed in boldface type.
Summary
In sum, principal component analysis indicated that the pattern of the
relationship among respective variables is well-explained by a LSAS-M twocomponent structure, a SPSBS two-component structure, and a PRSBQ one-
56
component structure. Principal component analysis of a SBQ-M three-component
structure indicated that the component structure sufficiently explains the
relationship among the SBQ-M variables.
Correlations among Component Scores
Correlations among the weighted component scores for all measures are
presented in Table 8. In all, convergent correlations among the SBQ-M-UI, SPSBSLL and the PRSBQ component scores suggest a strong, positive relationship among
the variables. Correlations among the SBQ-M-UI, SPSBS-LL and the PRSBQ
component scores and the other measures’ component scores suggest convergence
with theoretically related constructs and nonconvergence with theoretically
dissimilar constructs. The SBQ-M-IM component score and the SPSBS-PO
component score were highly correlated (r = .65, p < .01). Correlations of the SBQM-IM and the SPSBS-PO component scores suggest adequate convergence and
divergence. Overall, the LSAS-M-VS, LSAS-M-NS, and the SBQ-M-CB
component scores did not display convergence.
Table 8: Correlations among the Component Scores
.34
.12
.76
.51
.62
.59
.63
.47
-.04
-.07
.43
.70
.70
IM
.26
.35
.65
.54
.39
.41
.30
.29
.21
-.11
.30
.29
CB
.09
.28
.14
.01
.03
.27
.16
.21
.18
.00
.16
LL
.51
.61
.60
.60
.43
-.09
-.07
.39
.71
.71
PO
.58
.48
.48
.42
.16
.08
.26
.47
.51
.60
.51
.40
.18
.14
.27
.57
.53
9
10
11
12
13
14
15
SIAS
PRSBQ
S-SPSS-P
BFNE
PANAS-NA
S-SPSS-A
PANAS-PA
CES-D
SIAS
SPS
.26
.41
.15
.27
.36
.28
.22
.25
.25
.19
.15
.14
-.17
.23
UI
SPSBS
8
CES-D
SPSBS
NS
7
PANAS-PA
SBQ-M
VS
.61
.18
.50
.08
.19
.36
.26
.19
.21
.16
.25
.15
.02
.09
.10
SBQ-M
6
S-SPSS-A
LSAS-M
VS
NS
UI
IM
CB
LL
PO
5
PANAS-NA
LSAS-M
4
BFNE
3
S-SPSS-P
2
PRSBQ
1
.63
.36
.15
-.02
.28
.65
.60
.49
-.09
-.14
.44
.77
.71
-.02
.05
.69
.42
.43
.31
-.13
-.19
-.06
-.20
-.13
-.06
.38
.46
.77
57
Note. LSAS-M = Liebowitz Social Anxiety Scale-Modified, VS = Verbal Situations, NS = Nonverbal Situations; SBQ-M = Social Behavior ScaleModified, UI = Unobtrusive Intent, IM = Impression Management, CB = Compensatory Behavior; SPSBS = Social Phobia Safety Behaviours Scale, LL
= Laying Low, PO = Putting On; PRSBQ = Presentation Related Safety Behaviors Questionnaire; S-SPSS-P = Social Self-Presentations Style ScaleProtective; BFNE = Brief Fear of Negative Evaluation; PANAS-PA = Positive and Negative Affect Scales-Positive Affect; S-SPSS-A = Social SelfPresentations Style Scale- Acquisitive; PANAS-PA = Positive and Negative Affect Scales-Positive Affect; CES-D = Center for Epidemiologic Studies
Depression Scale; SIAS = Social Interaction Anxiety Scale; SPS = Social Phobia Scale.
Correlations where p < .01, two-tailed are in boldface type; bold italicized correlations are where p < .05, two-tailed; nonsignificant correlations are in
plain type.
CHAPTER III
DISCUSSION
A literature review indicated that it was unclear whether any existing safety
behaviors measures accurately assesses safety behaviors for social anxiety in a way
that will be useful to clinicians and researchers because they had yet to be rigorously
examined. In the present study, psychometric properties, representational validity
(i.e., convergence and divergence), and criterion validity of four existing safety
behaviors measures were examined using a large sample of undergraduate students.
The LSAS-M, SBQ-M, SPSBS, and PRSBQ demonstrated adequate internal
consistency. As an indication of representational validity, correlations among the
safety behaviors measures were hypothesized to be larger than correlations of the
safety behaviors measures with measures of related constructs and measures of
dissimilar constructs. Analyses indicated that the LSAS-M did not show adequate
convergent and divergent validity. The SBQ-M, SPSBS, and the PRSBQ did
demonstrate adequate convergent and divergent validity.
To substantiate adequate criterion validity, scores of the safety behaviors
measures were expected to moderately correlate with scores of social anxiety even
after controlling for scores of depression. As predicted, safety behaviors scores
59
uniquely predicted social anxiety scores after controlling for depression scores.
Collectively, the pattern of results suggests the SBQ-M, SPSBS, and the PRSBQ
adequately assess safety behaviors in relation to social anxiety, even when
controlling for depression. Furthermore, results imply promising application of
these measures to research and clinical practice. Results of the LSAS-M do not
support its utility as a safety behaviors measure for social anxiety.
Issues of Validity
Each of the safety behaviors measures will be reviewed individually for
evidence of validity. To examine representational validity, convergent and divergent
correlations were compared to establish that the expected convergent correlations
were larger than the expected divergent correlations. Related constructs were
protectiveness, fear of negative evaluation, and negative affect. Dissimilar
constructs were acquisitiveness, positive affect, and depression. To establish
criterion validity, scores on the four safety behaviors measures were correlated with
scores of social anxiety. To examine incremental validity, partial correlations were
conducted for safety behaviors scores controlling for depression scores.
60
Validity of the LSAS-M
Overall, analyses demonstrated inadequate convergence of the LSAS-M and
the other safety behaviors measures. Bear in mind that the LSAS-M does not assess
safety behaviors use directly, but was expected to provide a measure of likelihood of
safety behaviors use across various social situations. Lack of convergence among
the safety behaviors measures and the LSAS-M is likely related to content validity.
The LSAS-M did not demonstrate significant divergence as was expected. In fact,
likelihood of use of safety behaviors is weakly related to feelings of protectiveness
and fear of negative evaluation, which is not in agreement with cognitive theory of
SA (Clark & Wells, 1995). Further, acquisitiveness and positive affect were
similarly related to use of safety behaviors in social situations. There appears to be
little or no relationship between likelihood of safety behaviors use and symptoms of
depression which is unexpected given the generally moderate relationship between
anxiety and depression. The weak relationship between likelihood of safety
behaviors use and social anxiety indicates minimal value for the LSAS-M for
predicting symptoms of social anxiety. Further analyses suggested that likelihood of
safety behaviors use accounts for minimal value in predicting social anxiety
symptoms after removing the influence of depressive symptoms. In sum, the LSASM did not demonstrate adequate psychometric properties.
61
Validity of the SBQ-M
The SBQ-M score was found to significantly correlate with the SPSBS score
and the PRSBQ score indicating convergence. Since the LSAS-M was determined
not to demonstrate sufficient convergent or divergent validity, the other convergent
correlations of the SBQ-M (i.e., correlations with the SPSBS and PRSBQ)
substituted as the criteria to examine representational validity. Based on the findings
and in concordance with cognitive theory, frequent use of safety behaviors is related
to higher levels of protectiveness, fear of negative evaluation, and negative affect.
Further, frequent use of safety behaviors is weakly related to acquisitiveness,
positive affect, and depression. The SBQ-M demonstrated moderate criterion
validity indicating that use of safety behaviors demonstrates significant value in
predicting symptoms of social anxiety. In addition, the SBQ-M score showed
incremental validity over the depression score in predicting social anxiety symptoms
suggesting that safety behaviors use accounts for some of the variance in predicting
social anxiety symptoms after removing the influence of depressive symptoms. In
sum, the SBQ-M demonstrates acceptable psychometric properties and is
functioning as hypothesized.
62
Validity of the SPSBS
The SPSBS score was found to highly correlate with the SBQ-M and the
PRSBQ scores indicating measurement of the same underlying construct of safety
behaviors for social anxiety. Because the LSAS-M was determined not to
demonstrate sufficient convergent or divergent validity, the other convergent
correlations of the SPSBS (i.e., correlations with the SBQ-M and PRSBQ) were
used instead as the criteria to examine representational validity. Under these
conditions, the results indicate that frequent use of safety behaviors in social
situations is strongly related to higher levels of protectiveness, fear of negative
evaluation, and negative affect. Frequent use of safety behaviors, as measured by
the SPSBS, does not appear to be related to acquisitiveness, positive affect, or
depression. Safety behaviors use in social situations demonstrates considerable
value in predicting symptoms of social anxiety. Furthermore, use of safety
behaviors showed incremental validity over depressive symptoms score in
predicting symptoms of social anxiety. In sum, the SPSBS demonstrated adequate
psychometric properties and is the pattern of results is in concordance with
cognitive theory of social anxiety.
63
Validity of the PRSBQ
As mentioned earlier, the PRSBQ score was found to highly correlate with
the SBQ-M and the SPSBS scores indicating convergence in measuring the
construct of safety behaviors. Because the LSAS-M was determined not to
demonstrate sufficient convergent or divergent validity, the other convergent
correlations of the PRSBQ (i.e., correlations with the SBQ-M and SPSBS) were
used instead as the criteria to examine representational validity. Under these
conditions, the PRSBQ indicated that a greater degree of engagement in safety
behaviors is related to higher levels of protectiveness, fear of negative evaluation,
and negative affect. Degree of engagement in safety behaviors does not appear to
have a relationship with acquisitiveness or positive affect as cognitive theory would
predict. The relationship between engagement in safety behaviors and depression is
minimal. Engagement in safety behaviors demonstrates substantial value in
predicting symptoms of social anxiety. Furthermore, engagement in safety
behaviors showed incremental validity over depressive symptoms in predicting
social anxiety symptoms. As hypothesized, the PRSBQ demonstrated adequate
validity and functioning.
64
Component Structure
Factor analysis with parallel analysis was conducted for each of the safety
behaviors measures to determine lower-order component structure and how many
components could be reliably extracted and interpreted. Correlations among the
component scores and criterion indices were then examined.
Component Structure of the LSAS-M
Analyses of the LSAS-M resulted in a two-component solution, in which
each item of the LSAS-M loaded on one of two components representing Verbal
Situations (VS) or Nonverbal Situations (NS). Communalities suggested that the
two components performed well in explaining the underlying constructs of the
LSAS-M. As mentioned earlier, the LSAS-M has previously been thought to
represent two constructs of performance situations and social anxiety situations. The
component-structure found in this study did not adequately match onto the construct
distinction made by others. Specifically, only 14 of the 24 items on the LSAS-M
were found to load onto a VS or NS component and performance situations or social
anxiety situations. Results indicate that the VS and NS component structure better
explains the underlying constructs measured by the LSAS-M. In other words, the
LSAS-M appears to better measure likelihood of safety behaviors use in verbal and
nonverbal situations. Examination of the two-component structure in relation to
65
theoretically related constructs demonstrated poor convergent validity suggesting
that likelihood of safety behaviors use in verbal and nonverbal situations is not
related to feelings of protectiveness, fear of negative evaluation and negative affect
as cognitive theory would suggest. Likelihood of safety behaviors use in verbal and
nonverbal situations was also poorly related to acquisitiveness, positive affect, and
depressive symptoms.
Examination of the two-component solution of the LSAS-M in regards to
social anxiety demonstrated that likelihood of safety behaviors use in verbal and
nonverbal situations poorly correlates with symptoms of social anxiety. In fact, the
relationship between likelihood of safety behaviors use in verbal situations and
symptoms of social anxiety was nonsignificant. Not unexpectedly, the LSAS-M
two-component solution demonstrated poor criterion validity.
Component Structure of the SBQ-M
Analyses of the SBQ-M resulted in a three-component solution, in which 29
of the 34 items of the SBQ-M loaded on one of three components representing
Unobtrusive Intent (UI), Impression Management (IM), or Compensatory Behavior
(CB). Communalities suggested that the three components perform well in
explaining the underlying constructs of the SBQ-M.
Examination of the component structure in relation to related constructs
demonstrated that the UI and IM components correlated moderately with the
66
theoretically related constructs. This finding suggests that higher levels of
unobtrusive intent and impression management are related to higher levels of
protectiveness, fear of negative evaluation, and negative affect. The third
component, CB, was not significantly correlated with protectiveness and fear of
negative evaluation suggesting that compensatory behavior was not responsible for
feelings of protectiveness and fear of negative evaluation. Compensatory behavior
was weakly correlated with negative affect suggesting a minimal relationship. The
overall pattern of correlations suggests adequate convergent validity of the UI and
IM components and poor convergent validity of the CB component. With the
exception of the moderate relationship of impression management and depression,
the relationship of unobtrusive intent and compensatory behavior with
acquisitiveness, positive affect, and depression were weak. The overall component
structure suggests adequate divergent validity. The moderate relationship between
unobtrusive intent and depression indicates that behaviors intended to keep a lowprofile are positively related to symptoms of depression.
Examination of the three-component solution of the SBQ-M in regards to
social anxiety resulted in discrepant findings. The UI component highly correlated
with the criterion constructs suggesting that unobtrusive intent is highly related to
symptoms of social anxiety. Weak correlations were found among the IM
component and the criterion constructs suggesting a modest relationship between
impression management and social anxiety symptoms. The correlation between the
CB component and the SIAS was nonsignificant, which suggests no or little
67
relationship between compensatory behavior and social interaction anxiety. The
correlation of between the CB component and the SPS was weak but significant
which indicates that compensatory behavior is weakly related to social anxiety
symptoms as measured by the SPS.
Component Structure of the SPSBS
Analyses of the SPSBS resulted in a two-component solution, in which 16 of
the 17 items of the SPSBS loaded on to one of two components representing Laying
Low (LL) or Putting On (PO). Communalities suggested that the two components
performed well in explaining the underlying constructs of the SPSBS.
Examination of the component structure in relation to protectiveness, fear of
negative evaluation, and negative affect demonstrated that laying low and putting on
are moderately correlated with the theoretically related constructs. In other words,
greater attempts to lay low or to put on a show result in higher levels of
protectiveness fear of negative evaluation, and negative affect. Correlations among
laying low and putting on and theoretically dissimilar constructs of acquisitiveness
and positive affect were weak suggesting no to little relationship between the
constructs and acquisitiveness or positive affect. Moderate correlations were found
among laying low and putting on and symptoms of depression suggesting that
attempts to lay low or put on a show influence symptoms of depression.
68
Examination of the two-component solution of the SPSBS in regards to
social anxiety found that laying low was highly correlated with the criterion
constructs, suggesting that attempts to lay low are strongly related to greater levels
of social anxiety. Moderate correlations were found among putting on and the
criterion constructs suggesting a reasonable relationship between greater efforts to
put on a show and greater levels of social anxiety.
Component structure of the PRSBQ
Analyses of the PRSBQ resulted in a one-component solution, in which each
item of the PRSBQ loaded on to a single component. The component appears to
describe safety behaviors engagement. Communalities suggested that one
component performed well in explaining the underlying construct of the PRSBQ.
The component pattern was the same as the scale pattern in regards to convergence,
divergence, and criterion validity.
Cross-measure Correlations among Components
Three components converged with each other: Unobtrusive Intent, Laying
Low, and safety behaviors engagement as measured by the PRSBQ. Two other
components also showed convergence: Impression Management and Putting On.
69
The remaining three components, Compensatory Behavior, Verbal Situations, and
Nonverbal Situations did not converge with any other components.
Component Structure of the SBQ-M-UI, SPSBS-LL and PRSBQ Cluster
High convergent correlations among the SBQ-M-UI, SPSBS-LL and the
PRSBQ components suggest a strong, positive relationship among unobtrusive
intent, laying low, and safety behaviors engagement. The underlying higher-order
component could be described as “behaviors intended to keep a low profile.” High
correlations of this cluster with theoretically related constructs suggest that
behaviors intended keep a low profile are robustly related to protectiveness, fear of
negative evaluation, and negative affect. Low correlations among the cluster and
theoretically dissimilar constructs suggest that behaviors intended to keep a low
profile are weakly associated with acquisitiveness and positive affect. Moderate
correlations were found among cluster and depression suggesting a likely
relationship between keeping a low profile and depression. Strong correlations
among the SBQ-M-UI, SPSBS-LL and the PRSBQ cluster and the criterion
constructs suggest that behaviors intended keep a low profile are strongly related to
symptoms of social anxiety. All findings are in concordance with what would be
expected by cognitive theory for social anxiety.
70
Component Structure of the SBQ-M-IM and SPSBS-PO Cluster
A moderately high convergent correlation between the SBQ-M-IM and the
SPSBS-PO component suggests a reasonable relationship between impression
management and putting on. This higher-order component could be described as
“impression-management.” Examination of the component score cluster in relation
to theoretically related constructs suggests greater levels of impression-management
are related to greater levels of protectiveness, fear of negative evaluation, and
negative affect. Correlations among the cluster and theoretically dissimilar
constructs were weak suggesting that impression-management is weakly associated
with acquisitiveness, positive affect, and depression. Examination of the criterion
validity of the SBQ-M-IM and the SPSBS-PO cluster suggested a reasonable
relationship between impression-management and symptoms of social anxiety.
Although findings for “impression-management” were not as robust as findings for
“behaviors intended to keep a low profile,” all findings are in agreement with
cognitive theory for social anxiety.
Strengths and Limitations of the Present Study
Strengths of the present study include: large sample size; use of
disattenuated correlations to directly compare the true correlations between the
safety behaviors variables after removing measurement error; principal component
71
analysis with oblique rotations as the investigatory strategy to examine component
structure; use of parallel analysis to determine the true number of components; and
the thorough investigation of representational and criterion validity at both measure
and component levels.
Weaknesses of the present study should be noted. First, because a nonclinical sample of young adult undergraduates was used to examine the processes
underlying social anxiety, generality to a clinical population is unknown. Future
studies are encouraged to examine the psychometric properties of in-situation safety
behaviors measures in a clinical sample to address concerns of external validity.
Second, use of only self-report measures likely inflated the observed
correlations due to common method variance. Ideally, future studies will use
multiple measurement methods (e.g., behavioral assessment, direct observation,
other-report, or performance appraisal) to explicate convergence between observed
and true relationships among constructs.
Third, subsequent to data collection for this study, two new measures of
safety behaviors for social anxiety have been developed. The Social Coping Scale
(SCS; Dobson, Hertzsprung, & Dobson, unpublished) is 22-item instrument
designed to measure the use and frequency of internal and external safety behaviors
related to social phobia. The Subtle Avoidance Frequency Examination (SAFE;
Cuming, Rapee, Kemp, Abbott, Peters, & Gaston, 2009) is a 32-item instrument
designed to measure active safety behaviors, as well as the restriction of behaviors,
72
associated with social anxiety. Future studies of psychometric properties of safety
behaviors measures would be wise to include the SCS and SAFE.
Fourth, there was a univariate pattern of missing data in this study. Because
the questionnaires were designed to be administered randomly, the missing data
appears arbitrarily. Pairwise analyses were used to address the issue of missing data
in this study; however, listwise deletion or multiple imputation may have been more
powerful missing data techniques to address the missing data (Schafer, 1999;
Schafer & Graham, 2002).
A fifth weakness of the present study was the decision not to collect data on
the LSAS-M fear scale. The fear scale is generally administered along with the
frequency scale. Though not definite, it is possible that the pattern of results for the
LSAS-M frequency scale may have been different had participants completed the
fear scale and the frequency scale.
Implications for Theory
Findings of the present study point to several theoretical implications for
understanding safety behaviors. First, nonconvergence among the safety behaviors
measures and the measures of social anxiety suggest construct uniqueness and nonredundancy of social anxiety content. Accordingly, the dimensions of safety
behaviors were examined. The results of the present study suggest that there are two
higher-order components that underlie the construct of safety behaviors: behaviors
73
intended to keep a low profile and impression-management. The suggestion that low
profile and impression-management behaviors are core dimensions underlying the
construct of safety behaviors should be taken with caution as these dimensions can
only be a result of the range of items on the original measures used in this study.
Nonetheless, face validity of the higher-order components, development of the
original measure items (i.e., based on theory and clinical interviews), and high
correlations with measures of social anxiety suggest that low profile and
impression-management behaviors are important dimensions to consider.
Second, social anxiety theory can be further tested using a psychometricallysound safety behaviors measure. For example, indices of the two higher-order
components do not appear to differentially correlate with the SIAS and SPS
suggesting that low profile and impression-management behaviors do not
differentially associate with the subtypes. In future studies, these indices can be used
to examine other social anxiety dimensions (e.g., neuroticism, loneliness, social
satisfaction, educational or career attainment) or social anxiety subtypes (i.e.,
generalized social phobia and specific social phobia). For example, Davies and
colleagues (1998) found that greater levels of neuroticism were related to lower
levels of self-deceptive enhancement (a protective component against antisocial
thinking and behavior) in a student sample. It is speculated that use of impressionmanagement behaviors and high levels of self-deceptive enhancement would be
highly correlated. Additionally, a strong relationship is thought to exist between
neuroticism and impression-management behaviors due to hypothesized use of
74
impression-management behaviors for the dual purpose of coping with negative
state and to function in a socially desirable way.
Third, the relationship of safety behaviors to other domains of the Clark and
Wells’ (1995) model of social anxiety could be examined using a psychometricallysound safety behaviors measure. Future research could examine the relationship
between safety behaviors and the domains of somatic symptoms, cognitive
processes (i.e., internal self-focus of attention vs. external self-focus of attention),
assumptions about the self (e.g., negative self-statements), and perception of social
danger. It is hypothesized that behaviors intended to lay low are strongly related to
conditional beliefs about the self and unconditional beliefs about the self because of
the avoidant tendencies these thoughts evoke, whereas impression-management
behaviors are thought to be more closely related to excessively high expectations for
the self because individuals with high expectations of themselves may be more
likely to engage with others to prove themselves a worthy member of society.
Internal self-focus of attention is hypothesized to be strongly related with both
laying low and impression-management behaviors as self-focus plays a vital role in
maintenance of SA. Somatic symptoms may be more strongly related to impressionmanagement behaviors than laying low behaviors due to the greater physical
contribution to impression management (e.g., shaking hands with strangers)
compared to the non-physical nature of laying low behaviors (e.g., avoiding
strangers altogether).
75
Fourth, a psychometrically-sound safety behaviors measure could be used to
test hypotheses about other anxiety disorders. Use of safety behaviors appears to be
present in other disorders such as obsessive-compulsive disorder, panic disorder,
and specific phobia (Bennet-Levy et al., 2004; Freeman et al., 2007). Future
researchers could use the safety behaviors constructs of laying low and impressionmanagement for research or they could adapt the social anxiety safety behaviors
measures to examine the issue of alternative conceptualizations of safety behaviors
in other disorders. For example, it is hypothesized that there is a negative
relationship between laying low and frequency of panic symptoms because the
physical distance from social situations, obtained from laying low, may make an
anxious person less likely to experience symptoms of panic. As another example of
application of safety behaviors to other disorders, the relationship between
impression-management behaviors and symptoms of obsessive compulsive disorder
are hypothesized to be strong because of an underlying desire to present oneself as
socially acceptable and non-disorderly.
Fifth, the issue of whether safety behaviors are distinct from adaptive coping
mechanisms can be examined. As mentioned prior, some clinical researchers have
proposed that safety behaviors may actually be adaptive in some contexts. Though
not directly addressed in this study, visual inspection of item content suggests that
safety behaviors are used intentionally by socially anxious individuals to
temporarily reduce anxiety or to avoid negative consequences. If safety behaviors
do differ from adaptive coping strategies, it would be beneficial to distinguish the
76
behaviors to determine which behaviors should be targeted during exposure
treatment.
Practical Implications
There are several reasons why a rigorously examined safety behaviors
measure is useful. First, a standardized self-report measure of safety behaviors can
direct treatment planning. Individuals who do not rely on safety behaviors in social
situations, as identified by a safety behaviors measure, will benefit from exposure
only. Individuals who frequently engage in safety behaviors will benefit from
exposure with decreased safety behaviors. Second, a safety behaviors measure will
provide therapists with information on which specific safety behaviors their clients
should be encouraged to decrease. Third, treatment progress can easily be tracked
by administering a safety behaviors measure before and after exposure.
Concluding Remarks
In conclusion, results point toward the SPSBS as the superior measure of
safety behaviors. Results of the LSAS-M preclude its utility as a safety behaviors
measure for social anxiety. Though the SBQ-M and the PRSBQ displayed adequate
internal consistency and psychometric properties, the pattern of results for the
SPSBS showed the strongest evidence of internal consistency, convergent and
77
divergent validity, and criterion validity. Further, the SPSBS two-component
structure best defined the two-higher order components (Laying Low representing
behaviors intended to keep a low profile and Putting On as impressionmanagement). Other benefits of the SPSBS include brevity (17-items), ease of
administration, self-report format, and ease of scoring. Taken as a whole, the present
study provides evidence that the Social Phobia Safety Behaviors Scale has superior
psychometric properties and is recommended for future use.
REFERENCES
Abramowitz, J. (2006). The psychological treatment of obsessive-compulsive
disorder. The Canadian Journal of Psychiatry – In Review, 51, 407-416.
Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting
interactions. Thousand Oaks: Sage.
American Psychiatric Association. (1987). Anxiety Disorders in Diagnostic and
statistical manual of mental disorders (3rd ed. R). Washington, DC: Author.
American Psychiatric Association. (2000). Anxiety Disorders in Diagnostic and
statistical manual of mental disorders: DSM-IV-TR (4th ed.). Washington,
DC: Author.
Bennet-Levy, J., Butler, G., Fennell, M., Hackmann, A., Mueller, M., & Westbrook,
D. (2004). Oxford Guide to Behavioural Experiments in Cognitive Therapy.
New York: Oxford University Press.
Bryant, P. G., & Yarnold, P. R. (1995). Principal components analysis and
exploratory and confirmatory factor analysis. In Grimm and Arnold,
Reading and understanding multivariate analysis. American Psychological
Association Books.
Campbell, D. T., & Fiske, D. W. (1959). Convergent and discriminant validation by
the multitrait-multimethod matrix. Psychological Bulletin, 56, 81-105.
Clark, D. M. (2001). A cognitive perspective on social anxiety. New York: John
Wiley & Sons.
Clark, D. M., Butler, G., Fennell, M., Hackman, A., McManus, F., & Wells, A.
(1995). Social Behaviour Questionnaire. Unpublished manuscript. Oxford
Department of Psychiatry, Oxford University.
79
Clark, D., Ehlers, A., Hackmann, A., McManus, F., Fennell, M., Grey, N., et al.
(2006). Cognitive therapy versus exposure therapy and applied relaxation in
social anxiety: A randomized controlled trial. Journal of Counseling and
Clinical Psychology, 74, 568-578.
Clark, L. A., & Watson, D. (1995). Constructing validity: Basic issues in objective
scale development. Psychological Assessment, 7, 309-319.
Clark, D. M., & Wells, A. (1995). A cognitive model of social anxiety. New York:
Guilford Press.
Davies, M. F., French, C. C.., & Keogh, E. (1998). Self-deceptive enhancement and
impression management correlated of EPQ-R dimensions. The Journal of
Psychology: Interdisciplinary and Applied, 132, 401-406.
Feske, U., & Chambless, D. ( 1995). Cognitive behavioral versus exposure only
treatment for social phobia: A meta-analysis. Behavior Therapy, 26, 695720.
Foster, S., & Cone, J. (1995). Validity issues in clinical assessment. Psychological
Assessment, 7, 248-260.
Freeman, D., Garety, P. A., Kuipers, E., Fowler, D., Bebbington, P. E., & Dunn, G.
(2007). Acting on persecutory delusions: The importance of safety seeking.
Behaviour Research and Therapy, 45, 89-99.
Garcia-Palacios, A., & Botella, C. (2003). The effects of dropping in-situation
safety behaviors in the treatment of social anxiety. Behavioral Interventions,
18, 23-33.
Gould, R. A., Buckminster, S., Pollack, M. H., Otto, M. W., & Yap, L. (1997).
Cognitive-behavioral and pharmacological treatment for social phobia: A
meta-analysis. Clinical Psychology: Science and Practice, 4, 291-306.
Haynes, S. N., Richard, D. C. S., & Kubany, E. S. (1995). Content validity in
psychological assessment: A functional approach to concepts and methods.
Psychological Assessment, 7, 238-247.
Hofmann, S. G., & Heinrichs, N. (2003). Differential effect of mirror manipulation
on self-perception in social phobia subtypes. Cognitive Therapy and
Research, 27, 131-142.
80
Hofmann, S. G., & Scepkowski, L. A. (2006). Social self-reappraisal therapy for
social phobia: Perliminary findings. Journal of Cognitive Psychotherapy, 20,
45-57.
Hook, J. N., & Valentiner, D. P. (2002). Are specific and generalized social anxiety
qualitatively distinct? Clinical Psychology: Science and Practice, 9, 393409.
Horn, J. L. (1965). A rationale and test for the number of factors in factor analysis.
Psychometrika, 30, 179-185.
Kessler, R. C., Chiu, W. T., Demler, O., & Walters, E. E. (2005). Prevalence,
severity, and comorbidity of 12-month DSM-IV disorders in the National
Comorbidity Survey Replication. Archives of General Psychiatry, 62, 617627.
Kim, E. (2005). The effect of the decreased safety behaviors on anxiety and
negative thoughts in social phobics. Journal of Anxiety Disorders, 19, 69-86.
Leary, M. R. (1983). A brief version of the Fear of Negative Evaluation Scale.
Personality and Social Psychology Bulletin, 9, 371-375.
Liebowitz, M. R. (1987). Social anxiety. Modern Problems of Pharmacopsychiatry,
22, 141-173.
Linehan, M. M. (1993). Skills training manual for treating borderline personality
disorder. New York: Guilford Press.
Longman, R. S., Cota, A. A., Holden, R. R., & Fekken, G. C. (1989). A regression
equation for the parallel analysis criterion in principal components analysis:
mean and 95th percentile eigenvalues. Multivariate Behavior Research, 24,
59-69.
Loevinger, J. (1954). The attenuation paradox in test theory. Psychological Bulletin,
51, 493-504.
Mattick, R. P., & Clark, J. C. (1998). Development and validation of measures of
social anxiety scrutiny fear and social interaction anxiety. Behaviour
Research and Therapy, 36, 455-470.
McManus, F., Saradura, C., & Clark, D. M. (2008). Why social anxiety persists: An
experimental investigation of the role of safety behaviours as a maintaining
factor. Journal of Behavior Therapy and Experimental Psychiatry.
81
Meleshko, K. G., & Alden, L. E. (1993). Anxiety and self-disclosure: Toward a
motivational model. Journal of Personality and Social Psychology, 64,
1000-1009.
Meng, X. L., Rosenthal, R., & Rubin, D. B. (1992). Comparing correlated
correlation coefficients. Psychological Bulletin, 111, 172-175.
Morgan, H., & Raffle, C. (1999). Does reducing safety behaviors improve treatment
response in patients with social anxiety? Australian and New Zealand
Journal of Psychiatry, 33, 503-510.
Noyes, R., & Hoehn-Saric, R. (1998). The Anxiety Disorders. New York:
Cambridge University Press.
Pinto-Gouveia, J., Cunha, M. I., & Salvador, M-C. (2003). Assessment of social
anxiety by self-report questionnaires: The Social Interaction and
Performance Anxiety and Avoidance Scale and the Social Anxiety Safety
Behaviours Scale. Behavioural and Cognitive Psychotherapy, 31, 291-311.
Radloff, L. S. (1977). The CES-D Scale: A self-report depression scale for research
in the general population. Applied Psychological Measurement, 3, 385-401.
Renner, K. A., & Valentiner, D. P. (in submission). The effects of the self-focused
attention behavioral experiment on social anxiety. Manuscript in preparation.
Rizvi, S. L., & Linehan, M. M. (2005). The treatment of maladaptive shame in
borderline personality disorder: A pilot study of “Opposite Action.”
Cognitive and Behavioral Practice, 12, 437-447.
Rodenbaugh, T. L., Woods, C. M., Heimberg, R. G., Leibowitz, M. R., & Schneier,
F. R. (2006). The factor structure and screening utility of the Social
Interaction Anxiety Scale. Psychological Assessment, 18, 231-237.
Rodenbaugh, T. L., Woods, C. M., Thissen, D. M., Heimberg, R. G., Chambless, D.
L., & Rapee, R. M. (2004). More information from fewer questions: The
factor structure and item properties of the original and Brief Fear of
Negative Evaluation Scale. Psychological Assessment, 16, 169-181.
Salkovskis, P. M. (1991). The importance of behavior in the maintenance of anxiety
and panic: A cognitive account. Behavioural Psychotherapy, 19, 6-19.
82
Salkovskis, P. M., Clark, D. M., Hackmann, A., Wells, A., & Gelder, M. G. (1999).
An experimental investigation of the role of safety-seeking behaviours in the
maintenance of panic disorder with agoraphobia. Behaviour Research and
Therapy, 37, 559-574.
Schafer, J. L. (1999). Multiple imputation: A primer. Statistical Methods in Medical
Research, 8, 3-15.
Schafer, J. L., & Graham, J. W. (2002). Missing data: Our view of the state of the
art. Psychological Methods, 7, 147-177.
Smits, A. J., Powers, M. B., Buxkamper, R., & Telch, M. J. (2006). The efficacy of
videotape feedback for enhancing the effects of exposure-based treatment
for social anxiety disorder: A controlled investigation. Behaviour Research
and Therapy, 44, 1773-1785.
Stangier, U., Heidenreich, T., Peitz., M., Lauterbach., W., & Clark, D. M. (2003).
Cognitive therapy for social phobia: Individual versus group treatment.
Behaviour Research and Therapy, 41, 991-1007.
Stopa, L., & Clark, D. M. (2001). Social phobia: Comments on the viability and
validity of an analogue research strategy and British norms for the Fear of
Negative Evaluation Questionnaire. Behavioural and Cognitive
Psychotherapy, 29, 423-430.
Thurstone, L. L. (1947). Multiple factor analysis. Chicago, IL: University of
Chicago Press.
Thwaites, R., & Freeston, M. H. (2005). Safety-seeking behaviours: Fact or
function? How can we clinically differentiate between safety behaviours and
adaptive coping strategies across anxiety disorders? Behavioural and
Cognitive Psychotherapy, 33, 177-188.
Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of
brief measures of positive and negative affect: The PANAS scales. Journal
of Personality and Social Psychology, 54, 1063-1070.
Watson, D., & Friend, R. (1969). Measurement of social-evaluative anxiety. Journal
of Consulting and Clinical Psychology, 33, 448-457.
83
Weeks, J. W., Heimberg, R. G., Fresco, D. M., Hart, T. A., Turk, C. L., Schneier, F.
R., et al. (2005). Empirical validation and psychometric evaluation of the
Brief Fear of Negative Evaluation Scale in patients with social anxiety
disorder. Psychological Assessment, 17, 179-190.
Wells, A., Clark, D. M., Salkovskis, P., Ludgate, J., Hackman, A., & Gelder, M.
(1995). Social anxiety: The role of in-situation safety behaviors in
maintaining anxiety and negative beliefs. Behavior Therapy, 26, 153-161.
Wells, A., & Papageorgiou, C. (1998). Social anxiety: Effects of external attention
on anxiety, negative beliefs, and perspective taking. Behavior Therapy, 29,
357-370.
Wittchen, H-U., & Fehm, L. (2003), Epidemiology and natural course of social fears
and social phobia. Acta Psychiatricia Scandinavica, 108, 4-18.
Zwick, W. R., & Velicer, W. F. (1986). Comparisons of five rules for determining
the number of components to retain. Psychological Bulletin, 99, 432-442.
APPENDICIES
APPENDIX A
LIEBOWITZ SOCIAL ANXIETY SCALE - MODIFIED (LSAS-M)
86
Liebowitz Social Anxiety Scale – Modified (LSAS – M)
Instructions: How frequently do you make active efforts to prevent other people
from getting a negative impression of you in the following situations?
1.
2.
3.
4.
5.
6.
Telephone in public
Participating in small groups
Eating in public places
Drinking with others in public places
Talking to people in authority
Acting, performing or giving a talk in front of
an audience
7. Going to a party
8. Working while being observed
9. Writing while being observed
10. Calling someone you don’t know very well
11. Talking with people you don’t know very well
12. Meeting strangers
13. Urinating in a public bathroom
14. Entering a room when others are already
seated
15. Being the center of attention
16. Speaking up at a meeting
17. Taking a test
18. Expressing a disagreement or disapproval to
people you don’t know very well
19. Looking at people you don’t know very well in
the eyes
20. Giving a report to a group
21. Trying to pick up someone
22. Returning goods to a store
23. Giving a party
24. Resisting a high pressure salesperson
Prevention
0= Never (0%)
1= Occasionally (1-33%)
2= Often (34-67%)
3= Usually (68-100%)
APPENDIX B
SOCIAL BEHAVIOR QUESTIONNAIRE – MODIFIED (SBQ-M)
88
Social Behavior Questionnaire – Modified (SBQ – M)
Directions: Please indicate which word best describes how often you do the
following things when you are anxious in or before a social situation.
0 = Never
1 = Sometimes
_____
_____
_____
_____
_____
_____
_____
_____
_____
_____
_____
_____
_____
_____
_____
_____
_____
_____
_____
_____
_____
_____
_____
_____
_____
_____
_____
_____
_____
_____
_____
_____
_____
_____
Use alcohol to manage anxiety
Try not to attract attention
Make an effort to get your words right
Check that you are coming across well
Avoid eye contact
Talk less
Avoid asking questions
Try to picture how you appear to others
Grip cups or glasses tightly
Position yourself so as not to be noticed
Try to control shaking
Choose clothes that will prevent or conceal sweating
Wear clothes or makeup to hide blushing
Rehearse sentences in your mind
Censor what you are going to say
Blank out or switch off mentally
Avoid talking about yourself
Keep still
Ask lots of questions
Think positive
Stay on the edge of groups
Avoid pauses in speech
Hide your face
Try to think about other things
Talk more
Try to act normal
Try to keep tight control of your behaviour
Make an effort to come across well
Try to hid any signs of anxiety
Try to think of entertaining things to say
Focus on yourself
Try to appear clever or knowledgeable
Try to look at ease
Try to appear witty
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
2 = Often
3 = Always
APPENDIX C
SOCIAL PHOBIA SAFETY BEHAVIORS SCALE (SPSBS)
90
Social Phobia Safety Behaviors Scale (SPSBS)
Instructions: Please indicate a response for each item to indicate how
frequently you used each of the following behaviors in social situations.
0 = Never
1 = Occasionally
2 = Often
3 = Always
1. Looking away from or avoiding eye contact with the person with
whom you are interacting.
2. Speeding up your speech, talking quickly and without pauses.
3. Shortening your speech, drastically reducing what you have to say.
4. Avoiding attracting attention to yourself.
5. Getting a seat as hidden as you can.
6. Pretending you are not interested or you are distant from what is
happening.
7. Limiting yourself to being a passive spectator of a situation.
8. Pretending you did not see someone.
9. Walking with your head down.
10. Putting your hands in your pockets.
11. Stopping doing what you were doing (e.g., writing, drinking, etc.)
while being observed.
12. Trying to look at ease.
13. Laughing to hide the fact that you are nervous.
14. Constantly checking if you are presentable.
15. Increasing the distance between yourself and the person you are
talking to.
16. Trying to disguise your trembling.
17. Thinking very carefully about what you are going to say before you
speak.
APPENDIX D
PRESENTATION-RELATED SAFETY BEHAVIORS SCALE (PRSBQ)
92
Presentation Related Safety Behaviors Questionnaire (PRSBQ)
Instructions: To what degree does each of the following statements
accurately describe how often you engage in these behaviors?
1.
Avoided eye-contact
Not at all
2.
Very much
0
1
2
3
4
5
6
Very much
0
1
2
3
4
5
6
Very much
0
1
2
3
4
5
6
Very much
0
1
2
3
4
5
6
Very much
0
1
2
3
4
5
6
Very much
0
1
2
3
4
5
6
Very much
Drew as little attention to self as possible
Not at all
9.
6
Talked as little as possible
Not at all
8.
5
Thought ahead about what to say
Not at all
7.
4
Avoided stop talking
Not at all
6.
3
Monitored continuously what I said
Not at all
5.
2
Focused on my enunciation
Not at all
4.
1
Focused on my voice
Not at all
3.
0
0
1
2
3
4
5
6
Very much
2
3
4
5
6
Very much
Took deep breaths
Not at all
0
1
APPENDIX E
SOCIAL SELF-PRESENTATION STYLE SCALE (SSPSS)
94
Social Self-Presentations Style Scale (SSPSS)
Instructions: To what extent does each of the following statements accurately
describe your feelings and reactions during conversations?
1. Once I knew what the situation called for, it was easy for me to
regulate my behavior.
completely false 1 2 3 4 5 6 7
completely true
2. I was careful about what I said because I was afraid that I might say
or do something wrong.
completely false 1 2 3 4 5 6 7
completely true
3. I was able to control the way I came across to my partner so that I
gave the impression I wanted to give.
completely false 1 2 3 4 5 6 7
completely true
4. I talked about the same things my partner did because I didn’t want
to appear foolish.
completely false 1 2 3 4 5 6 7
completely true
5. If I felt that I wasn’t giving the impression I wanted to give, I feel I
could have easily changed it.
completely false 1 2 3 4 5 6 7
completely true
6. During the conversation, I tried to behave in such a way that I
wouldn’t draw attention to myself.
completely false 1 2 3 4 5 6 7
completely true
7. I had no difficulty making a good impression during the
conversation because I felt it was to my advantage to do so.
completely false 1 2 3 4 5 6 7
completely true
8. I didn’t talk about the topics I wanted to because I was afraid my
partner would disapprove of them.
completely false 1 2 3 4 5 6 7
completely true
APPENDIX F
BRIEF FEAR OF NEGATIVE EVALUATION SCALE (BFNE)
96
Brief Fear of Negative Evaluation (BFNE)
Instructions: Indicate the degree to which you feel the statement is characteristic
or true of you.
0 = Not at all
1 = Slightly
2 = Moderately
3 = Very
4 = Extremely
_____ 1.
_____ 2.
_____ 3.
_____ 4.
_____
_____
_____
_____
5.
6.
7.
8.
_____ 9.
_____ 10.
_____ 11.
_____ 12.
I worry about what people will think of me even when I know it
doesn’t make any difference.
I am unconcerned even if I know people are forming an
unfavorable impression of me.
I am frequently afraid of other people noticing my
shortcomings.
I rarely worry about what kind of impression I am making on
someone.
I am afraid that others will not approve of me.
I am afraid that people will find fault with me.
Other people’s opinions of me do not bother me.
When I am talking to someone, I worry about what they may be
thinking about me.
I am usually worried about what kind of impression I make.
If I know someone is judging me, it has little effect on me.
Sometimes I think I am too concerned with what other people
think of me.
I often worry that I will say or do the wrong things.
APPENDIX G
POSITIVE AND NEGATIVE AFFECT SCALE (PANAS)
98
Positive and Negative Affect Scale (PANAS)
Instructions: This scale consists of a number of words that describe different
feelings and emotions. Read each item and then indicate the phrase that
describes how you have felt in the past month.
1. Interested
2. Distressed
3. Excited
4. Upset
5. Strong
6. Guilty
7. Scared
8. Hostile
9. Enthusiastic
10. Proud
11. Irritable
12. Alert
13. Ashamed
14. Inspired
15. Nervous
16. Determined
17. Attentive
18. Jittery
19. Active
20. Afraid
Very slightly or
not at all
A little
Moderately
Quite a bit
Extremely
APPENDIX H
SOCIAL INTERACTION ANXIETY SCALE (SIAS)
100
Social Interaction Anxiety Scale (SIAS)
Instructions: Indicate the degree to which you feel the statement is
characteristic or true of you.
0 = Not at all 1 = Slightly 2 = Moderately 3 = Very 4 = Extremely
_____ 1. I get nervous if I have to speak with someone in authority
(teacher, boss, etc.)
_____ 2. I have difficulty making eye-contact with others.
_____ 3. I become tense if I have to talk about myself or my feelings.
_____ 4. I find difficulty mixing comfortably with the people I work with.
_____ 5. I find it easy to make friends of my own age.
_____ 6. I tense-up if I meet an acquaintance in the street.
_____ 7. When mixing socially I am uncomfortable.
_____ 8. I feel tense if I am alone with just one other person.
_____ 9. I am at ease meeting people at parties, etc.
_____ 10. I have difficulty talking with other people.
_____ 11. 11 I find it easy to think of things to talk about.
_____ 12. I worry about expressing myself in case I appear awkward.
_____ 13. I find it difficult to disagree with another’s point of view.
_____ 14. I have difficulty talking to attractive persons of the opposite sex.
_____ 15. I find myself worrying that I won’t know what to say in social
situations.
_____ 16. I am nervous mixing with people I don’t know well.
_____ 17. I feel I’ll say something embarrassing when talking.
_____ 18. When mixing in a group I find myself worrying I will be ignored.
_____ 19. I am tense mixing in a group.
_____ 20. I am unsure whether to greet someone I know only slightly.
APPENDIX I
SOCIAL PHOBIA SCALE (SPS)
102
Social Phobia Scale (SPS)
Instructions: Indicate the degree to which you feel the statement is
characteristic or true of you.
0 = Not at all 1 = Slightly 2 = Moderately 3 = Very 4 = Extremely
_____ 1. I become anxious if I have to write in front of other people.
_____ 2. I become self-conscious when using public toilets.
_____ 3. I can suddenly become aware of my own voice and of others listening to
me.
_____ 4. I get nervous that people are staring at me as I walk down the street.
_____ 5. I fear I may blush when I am with others.
_____ 6. I feel self-conscious if I have to enter a room where others are already
seated.
_____ 7. I worry about shaking or trembling when I am watched by other people.
_____ 8. I would get tense if I had to sit facing other people on a bus or a train.
_____ 9. I get panicky that others might see me to be faint, sick, or ill.
_____ 10. I would find it difficult to drink something if in a group of people.
_____ 11. It would make me feel self-conscious to eat in front of a stranger at a
restaurant.
_____ 12. I am worried people will think my behavior odd.
_____ 13. I would get tense if I had to carry a tray across a crowded cafeteria.
_____ 14. I worry I’ll lose control of myself in front of other people.
_____ 15. I worry I might do something to attract the attention of others.
_____ 16. When in an elevator, I am tense if people look at me.
_____ 17. I can feel conspicuous standing in a line.
_____ 18. I get tense when I speak in front of other people.
_____ 19. I worry my head will shake or nod in front of others.
_____ 20. I feel awkward and tense if I know people are watching me.
APPENDIX J
CENTER FOR EPIDEMIOLOGIC STUDIES DEPRESSION SCALE (CES-D)
104
Center for Epidemiologic Studies Depression Scale (CES-D)
Below is a list of the ways you might have felt or behaved during the past
week. Please rate how often you have felt this way (during the past week).
1 = rarely or none of the time (less than 1 day)
2 = some or a little of the time (1-2 days)
3 = occasionally or a moderate amount of time (3-4 days)
4 = most or all of the time (5-7 days)
During the past week...
_____ 1. I was bothered by things that usually don't bother me
_____ 2. I did not feel like eating my appetite was poor
_____ 3. I felt that I could not shake off the blues even with help from
my family or friends
_____ 4. I felt that I was just as good as other people
_____ 5. I had trouble keeping my mind on what I was doing
_____ 6. I felt depressed
_____ 7. I felt that everything I did was an effort
_____ 8. I felt hopeful about the future
_____ 9. I thought my life had been a failure
_____ 10. I felt fearful
_____ 11. My sleep was restless
_____ 12. I was happy
_____ 13. I talked less than usual
_____ 14. I felt lonely
_____ 15. People were unfriendly
_____ 16. I enjoyed life
_____ 17. I had crying spells
_____ 18. I felt sad
_____ 19. I felt that people disliked me
_____ 20. I could not get "going”
Документ
Категория
Без категории
Просмотров
1
Размер файла
2 646 Кб
Теги
sdewsdweddes
1/--страниц
Пожаловаться на содержимое документа