acceptance and commitment therapy for generalized social ...612... · acceptance and commitment...
TRANSCRIPT
Acceptance and Commitment Therapy for Generalized Social Anxiety Disorder: A
Pilot Study
A Thesis
Submitted to the Faculty
Of
Drexel University
by
Kristy L. Dalrymple
in partial fulfillment of the
requirements for the degree
of
Doctor of Philosophy
September 2005
ii
ACKNOWLEDGEMENTS
I would like to thank my dissertation chair and graduate advisor, Dr. James Herbert, for
his invaluable guidance and support during this study and my graduate career. I am
particularly indebted to my colleagues who assisted in data collection and served as
therapists for this study: Elizabeth Nolan, LeeAnn Cardaciotto, Meagan Parmley, Heather
Murray, Angela Gorman, Peter Yeomans, and Ethan Moitra. My committee members
were extremely helpful during the course of this study: Dr. Lamia Barakat, Dr. Evan
Forman, Dr. Martin Franklin, and Dr. Pamela Geller. I would also like to thank Dr.
Brandon Gaudiano for his helpful feedback and constant support throughout the course of
this study. I would like to acknowledge those whose work formed the basis for this
study: Dr. Steven Hayes, Dr. Kirk Strosahl, and Dr. Kelly Wilson, who developed
Acceptance and Commitment Therapy (ACT); and Dr. Jennifer Block-Lerner and Dr.
Edelgard Wulfert, who first investigated the efficacy of ACT for Social Anxiety
Disorder. Finally, I am truly grateful for my family and friends, who provided unyielding
support and helped to make this project possible.
iii
TABLE OF CONTENTS
LIST OF TABLES……………………………………………………………………….vii
LIST OF FIGURES……………………………………………………………………..viii
ABSTRACT……………………………………………………………………………...ix
1. INTRODUCTION……………………………………………………………….........1
1.1. Cognitive Behavior Therapies for SAD………………………………………….2
1.2. Pharmacotherapy…………………………………………………………………4
1.3. Combined Pharmacotherapy and CBT…………………………………………...6
1.4. Acceptance and Commitment Therapy……………………………………..........9
1.5. Empirical Evidence for ACT……………………………………………………16
1.6. Efficacy of ACT for Anxiety Disorders………………………………………...21
1.7. Summary and Study Rationale………………………………………………….26
2. METHOD..…………………………………………………………………………..30
2.1. Participants……………………………………………………………………...30
2.2. Measures………………………………………………………………………...31
2.2.1. Structured Clinical Interview for DSM-IV Axis I Disorders
(SCID-I/P)………………………………………………………………….31
2.2.2. Structured Clinical Interview for DSM-IV
Personality Disorders (SCID-II)...………………………………………….31
2.2.3. Anxiety Disorders Interview Schedule for DSM-IV (ADIS-IV)………...32
2.2.4. Social Phobia and Anxiety Inventory (SPAI)……………………………32
2.2.5. Liebowitz Social Anxiety Scale (LSAS)………………………………...32
2.2.6. Beck Depression Inventory-2nd Edition (BDI-II)………………………..33
iv
2.2.7. Fear Questionnaire (FQ)…………………………………………………33
2.2.8. Brief Version of the Fear of Negative Evaluation Scale (Brief FNE)…...33
2.2.9. Sheehan Disability Scale (SDS)………………………………………….34
2.2.10. Quality of Life Inventory (QOLI)………………………………………..34
2.2.11. Acceptance and Action Questionnaire (AAQ)…………………………..35
2.2.12. Valued Living Questionnaire (VLQ)…………………………………….35
2.2.13. Automatic Thoughts Questionnaire (ATQ)……………………………...36
2.2.14. Anxiety Control Questionnaire (ACQ)…………………………………..36
2.2.15. Thought Control Questionnaire (TCQ)…………………………………..37
2.2.16. Willingness Scale (WS)………………………………………………….37
2.2.17. Social Interaction Self-Statement Test (SISST)…………………………37
2.2.18. Demographics Questionnaire…………………………………………….38
2.2.19. Clinical Global Impression Scales (CGI)………………………………..38
2.2.20. Behavioral assessment…………………………………………………...39
2.3. Treatment……………………………………………………………………....40
2.3.1. Acceptance and Commitment Therapy (ACT)…………………………..40
2.4. Procedure………………………………………………………………………..41
2.5. Statistical Analyses……………………………………………………………...43
2.5.1. Statistical Power………………………………………………………….43
2.5.2. Preliminary Analyses…………………………………………………….43
2.5.3. Primary Analyses………………………………………………………...44
2.5.4. Analysis of Clinical Significance………………………………………..45
2.5.5. Secondary Analyses……………………………………………………...46
v
2.6. Design and Data Collection Considerations…………………………………….47
3. RESULTS…………………………………………………………………………....50
3.1. Sample Description……………………………………………………..............50
3.2. Preliminary Analyses……………………………………………………………51
3.2.1. Dropouts…………………...…………………………………………….51
3.2.2. Baseline Period…………………………………………………………..51
3.2.3. Exploratory Demographic Comparisons…………………………………52
3.3. Primary Analyses………………………………………………………………..52
3.3.1. Outcome Measures……...………………………………………………..52
3.3.2. Process Measures……….….…………………………………………….54
3.3.3. Clinician-rated Measures.………………………………………………..56
3.3.4. Behavioral Assessment………….……………………………………….56
3.3.5. Intention-to-Treat Analyses……………………………………………...57
3.4. Analyses of Clinical Significance………………………………………………58
3.5. Secondary Analyses……………………………………………………………..60
3.5.1. Effect Size Comparisons………...……………………………………….60
3.5.2. Correlation Analyses……………………………………………………..61
4. DISCUSSION………………………………………………………………………..64
4.1. Summary of Results……………………………………………………………..64
4.2. Support for Hypotheses…………………………………………………………64
4.2.1. Hypothesis #1……………………………………………………………64
4.2.2. Hypothesis #2…………………………………………………………….65
4.2.3. Hypothesis #3…………………………………………………………….69
vi
4.2.4. Hypothesis #4…………………………………………………………….71
4.2.5. Hypothesis #5…………………………………………………………….72
4.3. Comparison and Contrast to Block (2002)……..………………………….........73
4.4. Limitations………………………………………………………………………75
4.5. Implications and Future Directions……………………………………………..77
LIST OF REFERENCES………………………………………………………………...82
VITA……………………………………………………………………………………111
vii
LIST OF TABLES
1. Assessment Schedule for Current Study………...………………………….………..94 2. Demographic Characteristics of the Sample..……………………………….……….95 3. Means (Standard Deviations), Effect Sizes, and p-Values of Baseline, Pre-, Mid-, and
Post-Treatment Measures for Completers Only and Intention to Treat Analyses….…………………………………………………………………………..96
viii
LIST OF FIGURES
1. Participant Flow Diagram for Study Phases………………………………………..101 2. Baseline, Pre-, Mid-, and Post-Treatment Mean Scores on the SPAI-SP………….102 3. Baseline, Pre-, Mid-, and Post-Treatment Mean Scores on the Brief FNE…..…….103 4. Baseline, Pre-, Mid-, and Post-Treatment Mean Scores on the FQ-SP…………….104 5. Baseline, Pre-, Mid-, and Post-Treatment Mean Scores on the LSAS Fear and
Avoidance Subscales……………………………………………………………….105
ix
ABSTRACT Acceptance and Commitment Therapy for Generalized Social Anxiety Disorder: A Pilot
Study Kristy L. Dalrymple
James D. Herbert, Ph.D.
Despite the demonstrated efficacy of cognitive-behavior therapy for Social Anxiety
Disorder (SAD), many individuals do not respond to treatment or demonstrate residual
symptoms and impairment after treatment. Preliminary evidence indicates that
incorporating mindfulness and acceptance techniques within traditional behavior therapy,
through psychotherapy programs such as Acceptance and Commitment Therapy (ACT),
can be helpful for a variety of disorders. Only one study to date has been conducted on
ACT for public speaking anxiety in a college sample, which showed promising results.
We examined the efficacy of ACT in individuals diagnosed with SAD in a pilot study.
Participants received 12 weekly individual sessions of ACT for SAD. The treatment
incorporated mindfulness and acceptance techniques within a standard exposure-based
intervention protocol for SAD. Multi-modal assessments were conducted using
standardized measures at pre-treatment, mid-treatment, and post-treatment. Self-reported
baseline assessments were also included to control for threats to internal validity; results
showed no change in symptoms from baseline to pre-treatment. Results showed
significant pre- to post-treatment improvement in self-reported and clinician-rated social
anxiety symptoms and observer-rated social skills, as well as significant improvement on
ACT-specific measures of willingness, experiential avoidance, and valued action. Large
effect size gains were found in social anxiety symptoms and quality of life, and were
comparable to those of other studies examining the efficacy of cognitive behavior therapy
x
(CBT) for SAD. Furthermore, 37.5% of participants met criteria for reliable and
clinically significant change, and change in quality of life and experiential avoidance
were significantly associated with treatment outcome. Results from the present study
suggest the potential efficacy of ACT for SAD and highlight the need for future research
utilizing larger samples and directly comparing ACT to CBT.
1
1. INTRODUCTION
Social Anxiety Disorder (SAD), also known as Social Phobia, is an extreme fear
of embarrassment or humiliation in social or performance situations, and is usually
characterized by avoidance in these situations (APA, 1994). According to the current
fourth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-
TR; APA, 2000), the diagnostic criteria for SAD include: Exposure to the feared
situation provokes anxiety, which may take the form of a panic attack; the individual
recognizes that the fear is excessive or unreasonable; the feared situations are avoided or
are endured with significant anxiety; the fear or avoidance interferes significantly with
the person’s normal routine, or occupational or social functioning; the duration of the
anxiety is at least 6 months in individuals under age 18 years; and the fear or avoidance is
not better accounted for by the physiological effects of a substance, a general medical
condition, or another Axis I disorder.
Some estimates indicate that SAD is the third most common psychiatric disorder
in the U.S., following Major Depressive Episode and Alcohol Dependence (Kessler et al.,
1994). Kessler et al. found in the National Comorbidity Survey that SAD has a lifetime
prevalence rate of 13.3 %. Little is known about the etiology of SAD, although some
research has indicated that factors such as traumatic conditioning (Stemberger, Turner,
Beidel, & Calhoun, 1995), behavioral inhibition (Kagan, Reznick, & Snidman, 1988),
and child-rearing practices (Arrindell et al., 1989; Bruch & Heimberg, 1994) may be
implicated (see Herbert & Dalrymple, in press; Morris, 2001 for reviews). Despite this
lack of clear knowledge on etiological factors, many advances have been made in the
treatment of SAD.
2
1.1. Cognitive Behavior Therapies for SAD
Behavioral and cognitive behavioral interventions have been the most studied
psychosocial treatments for SAD. Cognitive Behavior Therapy (CBT) emphasizes the
cognitive factors that maintain SAD, such as exaggerated negative beliefs about one’s
performance in social situations, as well as behavioral factors, such as avoidance of these
situations. CBT targets these maintaining factors via cognitive restructuring to modify
negative beliefs, and in vivo and simulated exposure exercises to decrease avoidance and
to test dysfunctional beliefs (Hope et al., 1995).
Cognitive Behavioral Group Therapy (CBGT; Heimberg, 1991; Heimberg &
Becker, 2002) is the most extensively studied treatment program for SAD. CBGT is
typically delivered over 12 weeks, and includes simulated exposure exercises and
cognitive restructuring. An early study of CBGT (Heimberg, Dodge, et al., 1990)
compared the program to educational-supportive group psychotherapy, and found that
those who received CBGT were rated as less impaired by clinicians and reported less
anxiety during a behavioral assessment task at post-treatment and 6 month follow-up
compared to individuals in the control condition. In addition, both treatments showed an
increase in positive cognitions and a decrease in negative cognitions at post-treatment,
but only the CBGT group maintained these gains at follow-up. Several other studies
have continued to support the efficacy of CBGT (e.g., Gelernter et al., 1991; Heimberg,
Salzman, et al., 1993; Heimberg et al., 1998; Herbert et al., 2005; Hope, Herbert, &
White, 1995). In addition, CBGT is included on the list of empirically supported
treatments developed by the American Psychological Association’s Committee on
Science and Practice (Chambless et al., 1996).
3
Other treatments for SAD that have been shown to be effective are social skills
training (Stravynski, Marks, & Yule, 1982; Wlazlo, Schroeder-Hartwig, Hand, Kaiser, &
Münchau, 1990) and social effectiveness therapy (Turner, Beidel, Cooley, Woody, &
Messer, 1994). Research has suggested that some persons with SAD demonstrate
problems with social skills (Norton & Hope, 2001; Stopa & Clark, 1993), although it is
unclear whether lower performance levels reflect a social skills deficit or inhibition due
to high anxiety (Morris, 2001). Nonetheless, research on the effectiveness of social skills
training, especially when combined with exposure exercises and cognitive restructuring,
has been promising (Herbert et al., 2005; Herbert, Rheingold, & Goldstein, 2002). In
addition, Turner and colleagues’ social effectiveness therapy, which combines exposure
and social skills training but does not directly address cognition via cognitive
restructuring, also has shown promising results (Turner et al., 1994).
In addition, more recently researchers have been examining modified CT for SAD
that places less emphasis on formal cognitive restructuring. For example, Clark (1997)
developed a modified version of CT based on Clark and Wells’s (1995) cognitive model
of the maintenance of SAD, which posits that SAD is maintained by the use of self-
focused attention, misleading internal information to make negative inferences about
appearance, excessive safety behaviors, and negatively biased anticipatory and post-event
processing. Based on this model, Clark’s (1997) treatment emphasizes identifying
problematic anticipatory and post-event processing, decreasing self-focused attention and
use of safety behaviors, and increasing focus of attention to the social situation. Recently
Clark et al. (2003) compared this modified version of CT to fluoxetine plus self-exposure
4
and placebo plus self-exposure. Results showed that on measures of social phobia, CT
was superior to the other two conditions at mid- and post-treatment.
Component analyses and meta-analytic reviews have been conducted to examine
the effectiveness of different components of CBGT. Results from these studies have
been mixed, although few studies have been able to demonstrate that the cognitive
restructuring component adds to the efficacy of exposure alone. For example, a meta-
analysis by Gould et al. (1997) found that exposure interventions produced the largest
effect sizes, either alone or in combination with cognitive restructuring. In addition, a
meta-analysis by Feske and Chambless (1995) found no differential drop out or relapse
rates between exposure (n=9) and cognitive restructuring (n=12) interventions. A
dismantling study by Hope, Heimberg, and Bruch (1995) found that exposure alone was
at least as effective as exposure plus cognitive restructuring. These studies indicate no
clear advantage of cognitive restructuring over exposure, and highlight the importance of
exposure in the treatment of SAD.
1.2. Pharmacotherapy
Many studies have demonstrated the efficacy of pharmacotherapy for SAD. For
instance, four clinical trials have shown that the antidepressant monoamine oxidase
inhibitor (MAOI) phenelzine is efficacious for the treatment of SAD (Gelernter et al.,
1991; Heimberg et al., 1998; Liebowitz et al., 1992; Versiani et al., 1992). A meta-
analysis by Blanco et al. (2003) found that phenelzine produced the largest effect sizes on
measures of social anxiety relative to pill placebo (overall ES = 1.02); however, it did not
perform significantly better than the other medications included in the meta-analysis
(e.g., clonazepam, gabapentin, and brofaromine).
5
Because of the dietary restrictions and adverse effects associated with MAOIs,
research has been conducted on other medications, such as selective serotonin reuptake
inhibitors (SSRIs). For example, a study by Stein et al. (1998) examined the efficacy of
paroxetine in a 12 week placebo-controlled double-blind study, and found that 55% of
those receiving paroxetine were classified as responders based on clinical global
improvement ratings. Paroxetine has been approved by the FDA for the treatment of
SAD, making it the first medication approved for the treatment of SAD in the United
States (Hofmann & Barlow, 2002). More recently, sertraline and venlafaxine also have
been indicated by the FDA for the treatment of SAD (FDA, 2003a,b). Other SSRIs have
shown promising results, such as fluvoxamine (van Vliet, den Boer, & Westenberg,
1994), sertraline (Katzelnick et al., 1995) and fluoxetine (van Ameringen, Mancini, &
Streiner, 1993). For instance, a meta-analysis by Van der Linden et al. (2000) reported
effect sizes ranging from .30 to 2.2 relative to pill placebo for sertraline, fluvoxamine,
and paroxetine.
There is also preliminary support for the use of benzodiazepines in the treatment
of SAD (Davidson, Potts, et al., 1993; Gelernter et al., 1991), although there is concern
about the possibility of physical dependence, which limits the long-term use of this class
of medication. Older tricyclic antidepressants also have been studied, but poor results
have been obtained (e.g., Simpson et al., 1998; Versiani et al., 1988). Finally,
preliminary support has been demonstrated for the use of beta blockers on an as needed
basis for the treatment of discrete social phobia in performance situations such as
speeches (Pohl, Balon, et al., 1998).
6
In summary, some classes of medication such as SSRIs and MAOIs can be
helpful for treating SAD, with a main advantage of pharmacotherapy being the rapid
onset of treatment effects (Preston, O’Neal, & Talaga, 2002). However, many of these
medications also include adverse side effects, dietary restrictions, physical dependence,
or a high risk of relapse, making them a less desirable treatment option for many
individuals.
1.3. Combined Pharmacotherapy and CBT
Some studies have compared the relative efficacy of pharmacotherapy to CBT in
the treatment of SAD, with the general finding of comparable short-term efficacy
between these two treatments (Gould et al., 1997; Heimberg et al., 1998). For example, a
study by Heimberg et al. (1998) compared phenelzine, placebo, CBGT, and educational-
supportive psychotherapy. Results showed that both phenelzine and CBGT were
efficacious; however, results after a 6 month maintenance phase and 6 month follow-up
phase also showed that those treated with phenelzine were more likely to relapse than
those treated with CBGT, indicating that CBGT has better long-term effects (Liebowitz
et al., 1999).
Investigators are beginning to question whether the rapid onset of treatment
effects that is gained from pharmacotherapy can be combined with the maintenance gains
from CBT in order to maximize treatment efficacy. Two studies examined the
incremental benefit of adding fluoxetine or phenelzine to CBGT compared to CBGT
alone (Heimberg, 2002; Foa et al., 2003). Preliminary reports from these studies suggest
no incremental benefits to combined treatment over either monotherapy. In addition, a
study by Haug et al. (2003) compared exposure alone, sertraline alone, exposure plus
7
sertraline, and pill placebo. Results showed that all four groups showed improvement
from baseline to post-treatment; however, individuals in the exposure alone group
continued to improve through follow-up, whereas individuals in the sertraline alone and
exposure plus sertraline groups showed deterioration. These results once again highlight
exposure alone as an important intervention for SAD.
Although standard cognitive behavioral treatments for SAD, either alone or in
combination with medication, have been shown to be efficacious, many individuals
continue to demonstrate residual symptoms and impairment following treatment. In
addition, a percentage of individuals do not even respond to treatment at all (Herbert et
al., 2005; Herbert, Rheingold, Gaudiano, & Myers, 2004). For example, approximately
¼ of completers did not respond to 12 weeks of CBT in some studies (Heimberg et al.,
1998; Herbert et al., 2005), while approximately 1/6 of participants were considered non-
responders at a 6 month follow-up period for other studies (Liebowitz et al., 1999;
Stangier et al., 2003). Response in these studies was defined as a statistically or clinically
significant improvement on self-report or clinician-rated measures, and did not
necessarily mean that participants were symptom-free. Although many participants in
these studies were considered “responders,” their scores did not reach those of non-
clinical populations and they still continued to experience significant symptoms post-
treatment. Therefore, new treatments are needed to enhance the effects of existing
treatments, and to provide treatment that may be helpful for non-responders to standard
CBT.
Given that some researchers (e.g., Hope, Heimberg, and Bruch, 1995) have found
that there is no clear advantage of cognitive restructuring over exposure, and that social
8
effectiveness therapy (Turner et al., 1994), which does not directly target cognitive
restructuring, has shown promising results, this indicates that new treatments for SAD
should include exposure and may not need to directly address cognitions via cognitive
restructuring. Therefore, acceptance-based approaches (such as Acceptance and
Commitment Therapy, or “ACT”) that emphasize the acceptance of negative thoughts
and feelings rather than attempting to change their content, may be particularly helpful
especially when conducted within the context of exposure-based treatments. In addition,
clients with anxiety disorders typically engage in a range of avoidance behavior, and
consequently are cautious to engage in exposure-based treatments that target avoidance
and encourage them to experience fear (Barlow & Craske, 1994). Therefore, acceptance-
based approaches that foster willingness to engage in fearful situations and target
avoidance of the experience of anxiety instead of reducing the anxiety itself may increase
receptiveness to engage in exposure therapy (Eifert & Heffner, 2003).
Finally, Eng, Coles, Heimberg, et al. (2001) suggest the need for interventions
that can help to further improve quality of life. They examined the relationship between
quality of life and treatment outcome after 12 weeks of CBGT. Results showed that
quality of life improved from pre-treatment to post-treatment, but no further gains
occurred from post-treatment to follow-up. Although CBGT improved quality of life,
scores still did not approach those of non-anxious persons. Eng and colleagues
hypothesize that perhaps CBGT can improve quality of life in interpersonal domains, but
not in other ones. ACT, with its emphasis on clarification of personal values across
multiple life domains rather than symptom reduction per se, has the potential to improve
9
quality of life in several domains, rather than focusing solely on amelioration of
symptoms.
1.4. Acceptance and Commitment Therapy
The present study examined a promising new cognitive behavioral treatment,
Acceptance and Commitment Therapy (ACT; Hayes, Strosahl, & Wilson, 1999), which
incorporates mindfulness and acceptance components within standard exposure-based
treatment for SAD. Mindfulness techniques have recently been incorporated into CBT
treatments for several disorders, including Borderline Personality Disorder (using
Dialectical Behavior Therapy (DBT); Linehan, Armstrong, Suarez, & Allmon, 1991;
Linehan, Heard, & Armstrong, 1994), couples discord (using Integrative Behavioral
Couple Therapy (IBCT); Jacobson, Christensen, Prince, Cordova, & Eldridge, 2000), and
relapse prevention in Major Depressive Disorder (using Mindfulness-Based Cognitive
Therapy (MBCT); Teasdale, et al., 2000).
The mindfulness/acceptance based CBT approach that has received the most
attention thus far is Acceptance and Commitment Therapy (ACT; Hayes, Strosahl, &
Wilson, 1999). ACT is based on Relational Frame Theory (RFT; Hayes, Barnes-Holmes,
& Roche, 2001), which describes the nature of human language and cognition, and how
they are related to psychopathology. Simply put, RFT posits that cognitions exert their
effects based not only on their form or frequency, but also based on the context in which
they occur (Hayes, Masuda, Bissett, Luoma, Guerrero, 2004). Problematic contexts are
described by Hayes and colleagues as those that encourage the control of “private”
experiences such as thoughts or emotions. The ACT model explains that many
difficulties arise from attempting to control or avoid private events and from “fusion”
10
with cognitions, or believing that a thought that interprets experience is necessarily
literally true. Efforts to control or eliminate private events can interfere with movement
toward personally-identified valued goals. Therefore, the goal of ACT is not to modify
the content or frequency of private events as in traditional cognitive therapy, but rather to
learn how to fully experience such events in the service of achieving valued goals, thus
altering the problematic contexts of these private events (Herbert, 2002). At a technical
level, ACT borrows strategies not only from “standard” cognitive behavioral
interventions, but incorporates techniques from humanistic and experiential approaches
as well. In particular, liberal use is made of metaphors and experiential exercises to
convey core concepts of the model.
The goals of ACT parallel many concepts that are part of traditional Buddhist
philosophy. Although ACT was not intentionally based on Buddhism, the influence of
Buddhist philosophy can be seen in the underlying theory of ACT, as well as at the level
of applied technology. For example, both ACT and Buddhist philosophy consider human
suffering to be an unalienable part of human existence (Hayes, 2002). According to
Kumar (2002), Buddhism holds that “suffering is generated by the mental tendency
toward essentialism” based on “experiencing thoughts, emotions, behaviors, or self as
discrete and unchanging.” This parallels the concept of cognitive fusion in ACT, in
which suffering is thought to be caused by thoughts or emotions that are perceived as true
and part of the self, thus making them unchangeable (Hayes, 2002).
The concepts of acceptance and mindfulness in ACT also parallel Buddhism, both
at the level of theory and technique. For instance, Buddhist philosophy suggests that the
way to emerge from suffering is to accept the reality of it, identify the source, and detach
11
from the source (Hart, 1987). This idea is similar to processes posited by ACT, which
entails nonjudgmental awareness of thoughts and emotions, identification of cognitive
fusion as the source of suffering, and detachment from thoughts and emotions. At the
level of technique, both ACT and Buddhism utilize mindful meditation as one way to
achieve the process described above.
Finally, both ACT and Buddhism emphasize the importance of valued action. In
Buddhism, the process of seeing suffering, identifying its source, and detaching from the
source are all in the service of performing tasks which allow one to be “living and doing”
(Hayes, 2002). In ACT, acceptance and cognitive defusion are not seen in terms of
outcomes, but processes which can lead to more successful living (Batchelor, 1997). For
example, a student with public speaking anxiety would be taught acceptance and
cognitive defusion in the context of observing anxiety while giving a presentation, thus
fulfilling a course requirement and being able to graduate.
Several techniques are used to illustrate the various components of ACT in
order to meet the goals of seeing suffering, identifying the source, and detaching from the
source. These interventions are used to reduce cognitive fusion, undermine experiential
avoidance, teach acceptance and willingness as an alternative strategy, come in contact
with a transcendent sense of self in order to facilitate acceptance and cognitive defusion,
clarify personal values, and behave in ways consistent with those values (Hayes, 2002).
Sample interventions will be discussed in the context of the various stages of ACT,
described below.
One early component of ACT is called “creative hopelessness.” Past attempts to
alleviate problems are discussed, and the overall failure of those strategies is highlighted.
12
Clients are then asked to consider the possibility that these “solutions” are actually part of
the problem (Hayes, Strosahl, & Wilson, 1999). For example, many individuals with
SAD use avoidance of social situations as a way to decrease or eliminate anxiety.
However, this avoidance typically only makes the anxiety worse, which then makes this
particular strategy part of the problem, not the solution. Therefore, creative hopelessness
is used as a way to set the stage for the client to consider alternative ways of responding,
which may seem counter-intuitive to what are considered “normal” strategies of coping.
Several metaphors can be used to illustrate this point, such as the “Man in the Hole
Metaphor” (Hayes et al., 1999):
The situation you are in seems a bit like this. Imagine that you’re placed in a
field, wearing a blindfold, and you’re given a little tool bag to carry. You’re told
that your job is to run around this field, blindfolded. That is how you are
supposed to live life. And so you do what you are told. Now, unbeknownst to
you, in this field there are a number of widely spaced, fairly deep holes. You
don’t know that at first – you’re naive. So you start running around and sooner or
later you fall into a large hole. You feel around, and sure enough, you can’t climb
out and there are no escape routes you can find. Probably what you would do in
such a predicament is take the tool bag you were given and see what is in there;
maybe there is something you can use to get out of the hole. Now suppose that
the only tool in the bag is a shovel. So you dutifully start digging, but pretty soon
you notice that you’re not out of the hole. So you try digging faster and faster.
But you’re still in the hole. So you try big shovelfuls, or little ones, or throwing
the dirt far away or not. But still you are in the hole. All this effort and all this
13
work, and oddly enough the hole has just gotten bigger and bigger and bigger.
Isn’t that your experience? So you come to see me thinking, “Maybe he has a
really huge shovel – a gold-plated steam shovel.” Well, I don’t. And even if I
did, I wouldn’t use it, because digging is not a way out of the hole – digging is
what makes holes. So maybe the whole agenda is hopeless – you can’t dig your
way out, that just digs you in. (p.101).
Another component of ACT identifies attempts to control disturbing or unpleasant
private events as central to the problem. Therefore, in the case of the client with SAD,
past strategies such as avoidance are described as efforts to control anxiety, and the
futility of attempting to control private events is discussed. A particularly helpful
metaphor used to illustrate this is the “Polygraph Metaphor” (Hayes et al., 1999):
Suppose I had you hooked up to the best polygraph machine that’s ever been
built. This is a perfect machine, the most sensitive ever made. When you are all
wired up to it, there is no way you can be aroused or anxious without the
machine’s knowing it. So I tell you that you have a very simple task here: All
you have to do is stay relaxed. If you get the least bit anxious, however, I will
know it. I know you want to try hard, but I want to give you an extra incentive, so
I also have a .44 Magnum, which I will hold to your head. If you just stay
relaxed, I won’t blow your brains out, but if you get nervous (and I’ll know it
because you’re wired up to this perfect machine), I’m going to have to kill you.
So, just relax!...What do you think would happen? ...Guess what you’d get?...The
tiniest bit of anxiety would be terrifying. You’d naturally be saying, “Oh, my
14
gosh! I’m getting anxious! Here it comes!” BAMM! How could it work
otherwise? (p.123).
This component of ACT also introduces willingness as an alternative to control of
private experiences. Various exercises can be used to describe how control of private
experiences is a function of not being willing to experience thoughts and emotions as
they arise. For example, two types of discomfort are described: “clean” and “dirty.”
Clean discomfort is the discomfort that is derived directly from a stimulus, such as
feeling anxious while speaking with an authority figure. Dirty discomfort is the
discomfort that arises secondarily from feeling discomfort, such as when one is upset
about feeling upset. Therefore, dirty discomfort is derived from an unwillingness to
experience unwanted thoughts or feelings (Hayes et al., 1999).
In another component of ACT, cognitive defusion and acceptance are introduced
to facilitate willingness. During cognitive defusion, the goal is to teach clients how
language is inadequate in describing experience, and how to separate their thoughts from
a core sense of self using cognitive distancing techniques. An example of such a
technique is the “Soldiers in the Parade Exercise” (Hayes et al., 1999, p. 159), in which
clients are asked to picture thoughts, feelings or images on soldiers as they walk by in a
parade. Other variants can be used, such as leaves in a stream or clouds in the sky. This
exercise is used to practice distancing oneself from thoughts and to practice
nonjudgmental awareness (or acceptance) of thoughts and feelings.
The final component of ACT addresses values and choosing to behave in ways
consistent with these personal values. During this component, values and goals are
clarified, and barriers to these values are identified. In addition, willingness is re-
15
introduced as a choice to take valued action even with the barriers that may arise. The
“Bubble in the Road Metaphor” (Hayes et al., 1999) can be helpful in illustrating this
point:
Imagine that you are a soap bubble. Have you ever seen how a big soap bubble
can touch smaller ones and the little ones are simply absorbed into the bigger one?
Well, imagine that you are a soap bubble like that and you are moving along a
path you have chosen. Suddenly, another bubble appears in front of you and says,
“Stop!” You float there for a few moments. When you move to get around, over,
or under that bubble, it moves just as quickly to block your path. Now you have
only two choices. You can stop moving in your valued direction, or you can
touch the other soap bubble and continue on with it inside you. This second move
is what we mean by “willingness.” Your barriers are largely feelings, thoughts,
memories, and the like. They are really inside you, but they seem to be outside.
Willingness is not a feeling or a thought – it is an action that answers the question
the barrier asks: “Will you have me inside you by choice, or will you not?” In
order for you to take a valued direction and stick to it, you must answer yes, but
only you can choose that answer. (p. 230).
In summary, ACT identifies control of private experiences as the problem, and
emphasizes acceptance of the presence of these private experiences while being
committed to valued action. ACT shares some similarities with standard CBT, such as
the element of being aware of one’s own thoughts, or meta-cognitive awareness
(Teasdale et al., 2002). In addition, both ACT and CBT emphasize the role of behavior
(e.g., avoidance) in psychopathology, and utilize similar techniques to create behavior
16
change (e.g., exposure exercises). In fact, most traditional behavior therapy techniques
(e.g., exposure, skills training) can be incorporated into the ACT model. However, ACT
is different from traditional CBT in that it emphasizes acceptance of unpleasant thoughts
and feelings, rather than attempts to decrease or eliminate them. Traditional CBT for
SAD attempts to change the content of unpleasant thoughts, with the goal of decreasing
the discomfort felt as a result of those thoughts. ACT, on the other hand, does not
attempt to change the content of thoughts because the thought itself is not viewed as
problematic; rather, the context of attempting to control thoughts is problematic, and
ACT strives to change the context of control by fostering acceptance of these thoughts
(Hayes et al., 2004). At the level of techniques, ACT differs from standard CBT in that it
relies heavily on metaphors and experiential exercises to illustrate concepts. Although
ACT is comparable to CBT in some ways, it has the potential to offer unique approaches
to the treatment of SAD.
1.5. Empirical Evidence for ACT
Preliminary studies have shown promising results for the efficacy of ACT in a
variety of psychiatric conditions. In general, average post-treatment Cohen’s d effect
sizes for randomized control trials of ACT range from .55 to .99, depending on the
comparison group (e.g., no treatment/treatment as usual, CT/CBT, or another active
treatment), and average follow-up effect sizes range from .55 to .80 (Hayes, 2005).
Below is a brief, but not exhaustive, review of ACT efficacy trials.
The first randomized clinical trials of ACT were conducted on depressed
populations. A study by Zettle and Hayes (1986) randomly assigned 18 depressed
women to 12 sessions of either ACT or 2 variations of CT (cognitive restructuring with
17
or without cognitive distancing). Results showed that both ACT and CT produced
significant decreases in the Hamilton Rating Scale for Depression (HAM-D) at post-
treatment, and that ACT showed a greater decrease in depression than CT at a two-month
follow-up.
A similar study was conducted by Zettle and Rains (1989), which compared three
types of group treatment for 31 depressed women. Participants were randomly assigned
to 12 sessions of either a complete CT package (Beck, Rush, Shaw, & Emery, 1979), a
partial CT package that omitted cognitive distancing, or ACT. All three treatment groups
showed a significant decrease in depression at post-treatment and follow-up, with no
differences between conditions. These studies (Zettle & Hayes, 1986; Zettle & Rains,
1989) indicate that ACT has the potential to be helpful for depression, although further
research needs to be conducted.
Mindfulness and acceptance-based techniques have also been applied to medical
populations, such as chronic pain patients. A study by Geiser (1992) investigated the
efficacy of ACT for chronic pain in a quasi-experimental design. Thirty-three
participants were assigned to either an ACT-based treatment or CBT for 20 sessions.
Both treatments showed a clinically significant improvement at post-treatment, and
maintained treatment gains at a three-month follow-up.
In a study by Hayes and colleagues (1999), the impact of two intervention
rationales on pain tolerance were compared. Thirty-two college students were randomly
assigned to receive either an ACT-based rationale, control-oriented rationale, or an
attention placebo rationale. Participants completed a cold pressor task (i.e., submerge
non-dominant hand in ice water) pre-intervention and post-intervention. Results showed
18
that participants in the ACT-based rationale group kept their hands under water
significantly longer than both the control-oriented rationale group and the placebo group.
In addition, the subjective experience of pain in the cold pressor task did not differ across
rationale groups, even though the control-oriented group targeted this variable. This
study gives some preliminary information on the possible benefits that an acceptance-
based rationale can have on behavior change. However, this study did not account for
other coping strategies that participants may have utilized, and was conducted using a
non-clinical population.
ACT has also been examined in the workplace. A randomized controlled trial by
Bond and Bunce (2000) compared ACT to a behaviorally-oriented intervention and a
wait-list control group for workplace stress management. Ninety workers were randomly
assigned to 9 hours of ACT, Innovation Promotion Program (IPP; an intervention that
taught how to identify and change causes of occupational stress), or wait-list control.
Results found that both interventions produced a decrease in depression and an increase
in propensity to take concrete actions in order to reduce workplace stress. However,
ACT produced significantly greater improvements in stress and psychological health at
post-treatment and follow-up compared to IPP and the waitlist control. This study also
examined possible mediators of change for both interventions. Change in the ACT group
was mediated only by acceptance of undesirable thoughts and feelings, while change in
the IPP group was mediated only by attempts to modify stressors. This study provides
preliminary information on the usefulness of ACT in workplace stress. However, as with
the Hayes et al. (1999) study, this study was not completed on a clinical sample.
19
Bond and Bunce (2003) recently completed another study examining ACT in the
workplace with 412 customer service center workers. They investigated the influence of
acceptance on mental health, job satisfaction, and job performance. Results showed that
acceptance predicted mental health and job performance above and beyond job control,
as well as negative affectivity and locus of control. They also found that job control was
enhanced by higher levels of acceptance.
The efficacy of ACT has also been examined in severely mentally ill populations.
A study by Bach & Hayes (2002) examined the efficacy of ACT in an inpatient
population with psychotic symptoms. Eighty inpatients were randomly assigned to either
treatment as usual (TAU) or 4 sessions of TAU plus ACT, which focused on acceptance
of symptoms and taking action towards goals. Although ACT participants showed
significantly higher symptom reporting than those in the TAU group, they also showed a
significant reduction in rehospitalization rates (50% fewer) over a 4 month follow-up
than the TAU group. Also, ACT participants showed significantly lower levels of
symptom believability at follow-up. Limitations such as use of non-standardized
assessments of psychotic symptoms, a short follow-up period, and additional treatment in
the ACT group limit conclusions that can be drawn. However, this study provides
preliminary support for the efficacy of ACT in severely mentally ill populations. A study
replicating the Bach and Hayes (2002) study has also found similar preliminary results
(Gaudiano & Herbert, in press(a)).
Recent research efforts have also been focused on applying ACT to substance
abuse populations. Hayes and colleagues (2002) conducted a large randomized clinical
trial with polysubstance abusing opiate addicted individuals maintained on methadone.
20
Participants (n = 114) were randomly assigned to 16 weeks (48 sessions) of either
methadone maintenance alone, methadone plus ACT, or methadone plus Intensive
Twelve Step Facilitation (ITSF). There was no difference between active treatments at
post-treatment; however, ACT participants showed a greater decrease in opiate use (as
measured by urinalysis) at 6 month follow-up than those in methadone maintenance
alone. Both active treatment groups also showed lower levels of objectively measured
total drug use than methadone maintenance alone.
Another randomized controlled trial examined the efficacy of ACT compared to
nicotine replacement therapy (NRT) for smoking cessation (Gifford, 2002). Fifty-seven
smokers were randomly assigned to either 12 sessions of ACT or the nicotine
replacement patch. Monitoring of CO levels was used to objectively measure quit rates,
and it was found that both groups had equivalent quit rates at post-treatment. However,
the ACT group maintained gains at a one-year follow-up while the NRT quit rates had
fallen.
Few studies have examined the effectiveness of ACT in “real-world” contexts.
However, one study by Strosahl, Hayes, Bergan and Romano (1998) examined the
effectiveness of ACT within the context of a health maintenance organization setting.
Researchers used a “manipulated training” method to assess impact of clinicians’ work,
provide training, and reassess impact of work post-training as compared to those who did
not receive the training. Participants included 17 masters-level therapists and one
psychologist. Therapists volunteered for training in ACT, which consisted of a two day
didactic workshop, three days of clinical training with the ACT manual, and one year of
three-hour monthly supervision sessions. After training was completed, clients of the
21
ACT-trained clinicians were more likely to complete therapy within 5 months, more
likely to agree with their therapist on the conclusion of therapy, and significantly better in
coping outcomes than clients of the clinicians without ACT training. This suggests that
the ACT training helped clinicians to provide better treatment in a shorter amount of
time. However, factors that limit the conclusions include non-random assignment of
therapist into training, non-random assignment of clients to therapists, and lack of a
control training group.
1.6. Efficacy of ACT for Anxiety Disorders
Researchers have more recently begun to apply ACT to various anxiety disorders.
Case studies have been reported on individuals suffering from various anxiety disorders,
including Obsessive-Compulsive Disorder (Hayes, 1987), Generalized Anxiety Disorder
(Huerta-Romero, Gomez-Martin, Molina-Moreno, & Luciano-Soriano, 1998), and
Agoraphobia with and without panic attacks (Carrascoso Lopez, 2000; Hayes, 1987;
Zaldivar Basurto & Hernandez Lopez, 2001). Many of these case studies have shown
promising results.
Small randomized controlled trials have examined the efficacy of ACT for
various anxiety disorders, including Generalized Anxiety Disorder (GAD), Panic
Disorder, Trichotillomania, and mathematics anxiety. In research with GAD, Roemer &
Orsillo (2002) proposed a conceptualization of GAD that includes: 1) a belief that
worrying will reduce the probability of a future negative event occurring, and 2)
experiential avoidance or worrying about minor matters in order to avoid more global
internal distress. The hypothesized connection between experiential avoidance and GAD
was examined in a preliminary investigation by Roemer & Orsillo (2001). The
22
Acceptance and Action Questionnaire (AAQ; Hayes, Bissett, et al., 2002), a trait measure
of experiential avoidance, was administered along with two other GAD measures to 100
women ages 18 to 49 years old. Results showed that experiential avoidance was
significantly and positively associated with levels of trait worry, levels of distress
associated with GAD symptoms, and interference of GAD symptoms in daily life. These
results give preliminary support to the association between experiential avoidance and
GAD symptoms, suggesting that interventions incorporating mindfulness and acceptance
(such as ACT), which attempt to reduce experiential avoidance of internal experiences,
may be beneficial for treating GAD.
The same researchers are currently investigating an intervention for GAD that
includes standard behavior therapy for GAD (e.g., Borkovec & Roemer, 1994; Borkovec
et al., in press; Craske et al., 1992) integrated with mindfulness and acceptance-based
approaches (e.g., Hayes, Strosahl, & Wilson, 1999; Linehan, 1993). Treatment consists
of a psychoeducation component, a component which combines cognitive-behavioral
monitoring and mindful awareness of anxious responding, a component utilizing
relaxation and mindfulness techniques, and a component emphasizing effective action in
the presence of perceived difficulties, using techniques such as assessing values,
problem-solving, and exposure exercises. Orsillo, Roemer, & Barlow (2001) presented
preliminary results on 4 individuals who underwent 10 weeks of the above treatment
protocol. They found that two of the participants showed a substantial reduction in
anxious and depressive symptoms, a third participant showed modest improvement, and
the fourth participant showed no improvement (although this patient missed several
sessions). Although encouraging, conclusions are limited by the small sample.
23
Preliminary studies have been conducted on the impact of acceptance
interventions on subjective and psychophysiological reactions to aversive interoceptive
stimulation among individuals with Panic Disorder or panic-related symptoms. For
example, Eifert and Heffner (2003) conducted a study on 60 high anxiety sensitive
females who underwent a carbon dioxide challenge. Prior to the challenge, participants
were assigned to one of three groups: instructions to mindfully observe symptoms,
instructions to control symptoms via diaphragmatic breathing, or no instructions. Results
showed that those presented with the acceptance instructions were less behaviorally
avoidant, reported less intense fear, and reported fewer catastrophic thoughts during the
carbon dioxide challenge compared to participants in the other conditions.
A second study has been conducted on the impact of an acceptance intervention
on response to a carbon dioxide challenge, in individuals diagnosed with Panic Disorder
(Levitt, Brown, Orsillo, & Barlow, 2004). Sixty participants were assigned to one of
three conditions (acceptance, suppression of emotion, and control group) prior to
undergoing a carbon dioxide challenge. Results showed that those in the acceptance
condition reported less subjective anxiety and greater willingness to participate in a
second challenge compared to the suppression and control conditions. However, those in
the acceptance group did not differ from the other groups in their self-reported panic
symptoms or physiological measures. Both of these experimental studies show the
potential usefulness of acceptance interventions for those with Panic Disorder or panic-
related symptoms; however, treatment outcome studies need to be conducted in the future
to examine the efficacy of ACT for this particular population.
24
A preliminary study by Twohig & Woods (2004) examined the combination of
ACT and habit reversal in treating 6 adults diagnosed with Trichotillomania. A multiple
baseline across subjects design was used, in which participants began treatment after a
steady rate of hair pulling was established. Treatment consisted of 7 sessions, of which
the first 4 consisted of ACT and the final 3 of habit reversal within the overall ACT
framework. Outcome data consisted of self-report measures and photograph rating data,
in which photographs were taken of the damaged area and raters blind to assessment
point ranked the photographs for each participant from least to most damaged. Results
showed that ACT plus habit reversal produced self-reported decreases in pulling from
pre-treatment to post-treatment in 4 of the 6 participants. In addition, 3 of the 4
participants who completed follow-up maintained their gains. Results from the
photograph ratings also indicated a significant pre- to post-treatment change, but no
significant change from pre-treatment to follow-up.
Secondary analyses in this study showed that 3 of the 6 participants stopped
pulling as a result of the ACT sessions, giving some support for the utility of ACT alone.
However, the order of treatments was not alternated (i.e., some receiving ACT first then
habit reversal, and others receiving habit reversal first then ACT), therefore it is difficult
to determine whether this observation was due to the intervention itself or novelty effects.
The researchers also found that significant changes were seen in the participants’ pulling
without significant changes in levels of anxiety or depression. Furthermore, this study
examined whether ACT produced behavior change by decreasing experiential avoidance,
as measured by the AAQ. The AAQ did not show any changes from pre-treatment to
post-treatment; however, the authors cite limited statistical power resulting from the
25
small sample size as a possible reason for this finding. This study shows promising
results; however, as with many other studies examining the efficacy of ACT, it was
limited by a small sample size and lack of comparison condition.
One small randomized study has been conducted with 24 college students
experiencing mathematics anxiety (Zettle, 2003). Participants were randomly assigned to
6 weeks of ACT or systematic desensitization. Results showed that both groups had
reduced math and test anxiety from pre- to post-treatment, but only those receiving
systematic desensitization showed a significant decrease in trait anxiety from pre- to post-
treatment. In addition, analyses of clinical significance found that a majority of
participants in both groups were categorized as “recovered” or “improved” in their math
anxiety by post-treatment, with no differences between the two groups. Zettle also
examined experiential avoidance based on the AAQ. Results demonstrated that both
groups showed equal reductions in experiential avoidance from pre- to post-treatment.
He also examined the hypothesis that pre-treatment levels of experiential avoidance
would be positively associated with therapeutic change for the ACT group only, and
results revealed that levels of experiential avoidance at pre-treatment were significantly
associated with reductions in math anxiety only for those who received ACT. Therefore,
both groups showed equal reductions in math anxiety and experiential avoidance, but the
systematic desensitization group also showed reductions in trait anxiety, and experiential
avoidance was related to therapeutic change for only the ACT group. This is consistent
with previous findings regarding ACT, such that it does not necessarily decrease levels of
anxiety, but it does decrease experiential avoidance. This study suggests preliminary
support for using ACT to treat mathematics anxiety; however, a convenience sample was
26
used. In addition, outcome measures were only self-report; for a more rigorous study, a
multi-modal assessment approach could be employed, including behavioral and clinician
ratings.
Only one study to date has examined the efficacy of ACT for social anxiety
symptoms. Block (2002) semi-randomly assigned (due to scheduling constraints) 39
college students with public speaking anxiety to 6 weeks of either ACT, CBGT, or
waitlist control. Results showed that scores on social anxiety measures decreased for
both of the treatment groups relative to the control condition, and willingness to engage
in public speaking situations increased for both of the treatment conditions relative to the
control group. However, only the ACT group showed significant decreases in behavioral
avoidance. Although these results are promising, there are several limitations to this
study, such as use of a non-clinical population, a small sample size, lack of a structured
interview to establish a diagnosis, lack of true random assignment to treatment
conditions, short duration of treatment, and lack of an independent evaluator. The
present study examined ACT in a more internally and externally valid way than Block
(2002) by using a clinical sample of adults diagnosed with generalized SAD,
administering an ACT protocol of longer duration, and using independent evaluators to
assess treatment effects.
1.7. Summary and Study Rationale
Although CBT has been shown to be an effective treatment for SAD, many
individuals experience residual symptoms and impairment after treatment, and some
individuals fail to respond to treatment altogether. Pharmacotherapy can also be
effective, especially SSRIs such as paroxetine, which is FDA approved for the treatment
27
of SAD. However, studies have shown high relapse rates following discontinuation of
medication and some individuals are affected by negative side effects. Results on
combination treatments including CBT and medication have been disappointing.
Furthermore, research has indicated that cognitive restructuring, a component thought to
be of key importance in CBT, appears to be no more effective than exposure alone. The
current study therefore combined exposure interventions with ACT, an intervention that
emphasizes acceptance of thoughts and emotions rather than efforts to modify them.
Results from Eng, Coles, Heimberg, et al. (2001) indicate the need for
interventions that can improve quality of life across multiple domains of functioning.
One goal of the current study was to target improvement in quality life. For example, the
values clarification component of ACT helps clients to clarify personally-relevant values
in social relationships and other domains, such as family, spirituality, citizenship, and
work, and to establish specific goals consistent with those values.
The current study extended the findings of the Block (2002) study by
incorporating mindfulness and acceptance techniques with standard exposure to treat
individuals with SAD. First, the Block study used a non-clinical, homogenous sample
with discrete social anxiety (i.e., public speaking fears). The current study used a clinical
sample of individuals diagnosed with generalized SAD based on structured clinical
interviews. In addition, efforts were made to collect a diverse sample (e.g., ranging in
age from 18-60) rather than the college sample utilized by Block (2002). Second, Block
administered a brief (6 week) protocol of ACT. The current study administered a more
comprehensive 12 session protocol (consistent with Hayes, Strosahl, & Wilson (1999),
and other treatment manuals for anxiety disorders, such as Twohig, Hayes, & Masuda, in
28
press; Hayes, Wilson, Afari, & McCurry, 2003), which also included exposure exercises
to emphasize engaging in valued behaviors despite anxiety. Third, the Block study did
not utilize independent evaluators to assess treatment effects. The current study used
more stringent methodological procedures, including independent assessors, multiple
assessment points (including a baseline assessment to control for regression to the mean
and natural recovery), treatment integrity checks, and a multi-modal assessment
approach.
The specific aims of the current study were as follows: 1) to conduct a pilot study
to evaluate the efficacy of a novel psychosocial treatment for Social Anxiety Disorder; 2)
to measure treatment outcome in a multi-modal fashion, including self-report, clinician
ratings, and objective indices, including behavioral assessment tasks; 3) to examine
theoretically derived psychological factors (e.g., experiential avoidance, believability in
negative cognitions) to determine their association with treatment outcome; and 4) to
compare the outcomes (via effect sizes) obtained from the current study to existing data
on the efficacy of standard CBT for SAD, including both data collected through Drexel
University’s Anxiety Treatment and Research Program as well as data collected
elsewhere. This method was chosen given that the current study is not a randomized
clinical trial; therefore, using effect sizes is a more standardized way of comparing results
from the current study to other studies using CBT, to determine whether or not further
research on ACT for SAD is warranted.
In the current study, participants were administered a trial of ACT adapted for
treating SAD consisting of 12 sessions. ACT included psychoeducation, goal setting,
mindfulness and acceptance-based techniques, and simulated and in vivo exposure
29
exercises. The current study was conducted through Drexel University’s Anxiety
Treatment and Research Program, which specializes in the assessment and treatment of
SAD.
Specific hypotheses for the current study included: 1) Participants would not
demonstrate a change in symptoms from baseline to pre-treatment; 2) Participants would
demonstrate significant improvements in outcomes (e.g., symptomatology, impairment,
etc.) from pre-treatment to post-treatment; 3) Experiential avoidance and believability in
negative cognitions would be lower at post-treatment, and would be associated with
treatment outcome; 4) Effect sizes obtained from the current study would be comparable
to effect sizes of other studies examining the efficacy of CBT for SAD; and 5)
Participants would demonstrate a significant improvement in quality of life from pre- to
post-treatment, and quality of life would be associated with treatment outcome.
30
2. METHOD
2.1. Participants
Information on participant flow and demographic characteristics are presented in
the Results section. Participants were recruited via community media through the
Anxiety Treatment and Research Program at Drexel University. Participants met DSM-
IV (APA, 1994) criteria for Social Anxiety Disorder (SAD), generalized subtype, based
on a standard structured clinical interview. For the purposes of this study, the generalized
subtype was operationally defined as fear and avoidance in three or more distinct social
situations. Because epidemiological data have indicated high rates of other Axis I
comorbidity with SAD, participants with comorbid diagnoses were included in the study.
However, the diagnosis of SAD was judged to be clearly primary to and of greater
severity to the secondary diagnoses in order for inclusion.
The inclusion criteria for the study included:
1. Adults ages 18 and over with a primary diagnosis of Social Anxiety Disorder,
generalized subtype;
2. Fluency in English;
3. Consent to participate.
The exclusion criteria included:
1. A primary diagnosis of any disorder other than SAD;
2. Diagnosis of Mental Retardation or a Pervasive Developmental Disorder;
3. Diagnosis of a psychiatric disorder due to a medical condition;
4. Current diagnosis of substance dependence (within the past 6 months);
5. Acute suicide potential;
31
6. A general medical condition that would contraindicate treatment;
7. Previous trial of behavior or cognitive-behavior therapy for SAD.
2.2. Measures
2.2.1. Structured Clinical Interview for DSM-IV Axis I Disorders (SCID-I/P).
The SCID-I/P (First, Spitzer, Gibbon, & Williams, 1996) is a widely used diagnostic
structured clinical interview for the major Axis I disorders, based on DSM-IV (1994)
criteria. Results from several studies have found that the SCID-I/P has moderate to high
inter-rater reliability for most of the major mental disorders (Williams et al., 1992;
Riskind, Beck, Berchick, Brown, & Steer, 1987; also see Segal, Hersen, & Van Hasselt,
1994 for a review of the literature on inter-rater reliability of the SCID-I/P).
2.2.2. Structured Clinical Interview for DSM-IV Personality Disorders (SCID-II).
The SCID-II (First, Spitzer, Gibbon, Williams, & Benjamin, 1994) is a diagnostic
structured clinical interview for Axis II personality disorders, based on DSM-IV (1994)
criteria. Only the Avoidant Personality Disorder (APD) section was used because of the
high comorbidity between SAD and APD (Herbert, in press; Herbert, Hope, & Bellack,
1992). The SCID-II has been found to have adequate inter-rater reliability (First, Spitzer,
Gibbon, Williams, Davies, et al., 1995; Rennenberg, Chambless, and Gracely, 1992).
Studies of validity have compared SCID-II diagnoses to those generated by Spitzer’s
(1983) LEAD standard (i.e., a longitudinal, expert, evaluation using all data) and found
the overall diagnostic power to be good (kappa = .70) for 8 of 12 disorders (Segal, 1997;
Skodol, Rosnick, Kellman, Oldham, & Hyler, 1988). In addition, studies comparing
SCID-II diagnoses to Personality Disorder Examination (PDE) diagnoses have found
32
agreement to be moderate (O’Boyle & Self, 1990; Skodol, Oldham, Rosnick, Kellman, &
Hyler, 1991).
2.2.3. Anxiety Disorders Interview Schedule for DSM-IV (ADIS-IV). The ADIS-
IV (Brown, Di Nardo, & Barlow, 1994) is a diagnostic clinical interview designed
primarily for the assessment of anxiety disorders. For this study, only the SAD section of
the ADIS-IV was used to improve diagnostic reliability. The ADIS-IV has demonstrated
adequate reliability in the diagnosis of SAD (DiNardo, Brown, Lawton, & Barlow, 1995).
In general, the ADIS-IV possesses good reliability as a structured clinical interview for a
variety of Axis I disorders (DiNardo, Moras, Barlow, Rapee, & Brown, 1993).
2.2.4. Social Phobia and Anxiety Inventory (SPAI). The SPAI (Turner, Beidel,
Dancu, & Stanley, 1989) is a 45-item self-report measure that assesses clinical symptoms
of SAD. In the proposed study, the 32-item Social Phobia subscale (SPAI-SP) was used
in analyses because it is a better index of social anxiety symptoms than the difference
subscale score (Herbert, Bellack, & Hope, 1991). The SPAI is an empirically validated
measure of SAD, with psychometric research indicating that it has good test-retest
reliability, internal consistency, and discriminant, concurrent, and external validity
(Beidel, Bordon, Turner, & Jacob, 1989; Beidel, Turner, Stanley, & Dancu, 1989;
Herbert et al., 1991; Turner et al., 1989).
2.2.5. Liebowitz Social Anxiety Scale (LSAS). The LSAS (Liebowitz, 1987) is a
24-item inventory assessing fear and avoidance of several social situations (e.g., going to
a party or returning goods to a store). Participants are asked to rate their fear and
avoidance of these situations on a 4-point Likert scale, ranging from 0 (no
fear/avoidance) to 3 (severe fear/usually avoid). The LSAS has high internal consistency,
33
good discriminant and convergent validity, and has demonstrated treatment sensitivity
(Baldwin et al., 1999; Bouwer & Stein, 1998; Heimberg et al., 1999; Lott et al., 1997).
2.2.6. Beck Depression Inventory-II (BDI-II). The BDI-II (Beck, Steer, & Brown,
1996) is a 21-item inventory assessing symptoms of depression. The BDI-II has been
used extensively, and has been shown to possess good reliability and validity (Beck,
Steer, Ball, & Ranieri, 1996). It is based largely on the first edition of the BDI (Beck &
Steer, 1987), which numerous studies have indicated possesses good reliability and
validity in use with clinical and nonclinical samples (see Beck & Steer, 1988 for a
review).
2.2.7. Fear Questionnaire (FQ). The FQ (Marks & Mathews, 1979) is a 15-item
measure assessing avoidance behaviors commonly related to social situations,
agoraphobia, and blood/injury phobia. Of these 3 subscales, only the social phobia
subscale was used in data analysis for this study. Participants were asked to rate severity
of avoidance for 15 specific situations, their main phobia, and any other situations on a 9-
point Likert scale. The FQ has high test-retest reliability, good internal consistency, and
good discriminant validity (Cox, Parker, & Swinson, 1996; Cox Swinson, & Parker,
1993; Cox, Swinson, & Shaw, 1991; Michelson & Mavissakalian, 1983; Oei, Moylan, &
Evans, 1991; Van Zuuren, 1988).
2.2.8. Brief Version of the Fear of Negative Evaluation Scale (Brief FNE). The
Brief FNE (Leary, 1983) is a 12-item measure assessing concerns of negative evaluation
by others. The Brief FNE is based upon the original FNE, which contains 30 items. The
Brief FNE correlates highly with the FNE, and therefore was used in this proposed study.
Participants were asked to rate how each item is characteristic of them on a 5-point Likert
34
scale. The Brief FNE has good test-retest reliability and internal consistency (Leary,
1983), as well as good concurrent validity with other measures of social anxiety (Saluck,
Herbert, Rheingold, & Harwell, 2000).
2.2.9. Sheehan Disability Scale (SDS). The SDS (Leon, Olfson, Portera, Farber,
& Sheehan, 1997) is a self-report measure assessing impairment of symptoms related to a
psychiatric illness. The SDS assesses impairment in work, social/leisure activities, and
family/home life on a 10-point Likert scale. This measure has adequate internal
consistency, construct validity, and criterion-related validity (Leon, Shear, Portera, &
Klerman, 1992).
2.2.10. Quality of Life Inventory (QOLI). The QOLI (Frisch, 1994) is a 32-item
measure assessing importance and satisfaction in several domains, such as health,
friendships, and work. Participants were instructed to rate the personal importance of
these domains on a Likert scale ranging from 0 (not important) to 2 (extremely
important). Participants were then asked to rate satisfaction with these domains on a
Likert scale ranging from –3 (very dissatisfied) to +3 (very satisfied). The QOLI has
been validated on clinical samples, and has good internal consistency and test-retest
reliability (Frisch, Cornell, Villanueva, & Retzlaff, 1992). In addition, the QOLI
possesses good convergent, discriminant, and criterion-related validity (Frisch et al.,
1992). Treatment sensitivity from pre- to post-treatment has also been demonstrated with
the QOLI following 12 weeks of CBGT for SAD (Eng, Coles, Heimberg, & Safren,
2001).
35
2.2.11. Acceptance and Action Questionnaire (AAQ). The AAQ (Hayes et al.,
2002) is a 9-item measure assessing emotional avoidance and inaction, related to anxiety.
Sixteen-item and 22-item versions also have been created; however, most of the
psychometric research has been conducted on the 9-item version, which was used for the
proposed study. Items include statements such as “Anxiety is bad” and “I rarely worry
about getting my anxieties, worries, and feelings under control.” Items are rated on a 10-
point Likert scale ranging from “never true” to “always true.” Preliminary evidence
indicates that this measure possesses good internal consistency, as well as good
concurrent, convergent, and construct validity (Hayes et al., 2002).
2.2.12. Valued Living Questionnaire (VLQ). The VLQ (Wilson & Groom, 2002)
is a 10-item measure assessing the importance and consistency of personal values in
several domains, such as work, family, and recreation/fun. Participants first rated the
importance of these values in their life and then how consistent their actions are with
them. Items for the importance scale are rated on a 10-point Likert scale ranging from
“not at all important” to “extremely important.” Items on the consistency scale are also
rated on a 10-point Likert scale, ranging from “not at all consistent with my value” to
“completely consistent with my value.” A total discrepancy score was calculated to
determine the discrepancy between stated values and consistent action. Preliminary
research on the psychometric properties of the VLQ indicated that this measure possesses
good test-retest reliability (Groom & Wilson, 2003). Validity data on the VLQ are
currently being collected and are not yet available (Wilson & Murrell, 2004).
36
2.2.13. Automatic Thoughts Questionnaire (ATQ). The ATQ (Hollon & Kendall,
1981) is a 30-item measure assessing the frequency of typical negative automatic
thoughts (e.g., “No one understands me” and “My future is bleak”). Participants were
asked to rate the frequency of the 30 automatic thoughts on a 7-point Likert scale,
ranging from “never” to “always.” The ATQ was adapted by another researcher (Zettle
& Hayes, 1986) to include believability ratings of each of the automatic thoughts.
Participants were asked to rate the believability of these thoughts on a 7-point Likert
scale, ranging from “not at all believable” to “completely believable.” The adapted
version, which was used in the proposed study, has been used as an outcome measure in
several ACT studies (e.g., Bach & Hayes, 2002). The ATQ has demonstrated good split-
half reliability and internal consistency, as well as good convergent and discriminant
validity (Deardorff, Hopkins, & Finch, 1984; Harrell & Ryon, 1983).
2.2.14. Anxiety Control Questionnaire (ACQ). The ACQ (Rapee, Craske, Brown,
& Barlow, 1996) is a 30-item measure assessing perception of control over emotional
reactions and external events and situations. The ACQ consists of two subscales, events
and reactions. Participants were asked to rate items on a 6-point Likert scale and indicate
the degree to which they agree or disagree with each statement. Higher scores on the
ACQ reflect higher levels of perceived control. The ACQ possesses good internal
consistency and good test-retest reliability (Rapee et al., 1996). In addition, the ACQ
possesses good convergent validity, specificity to individuals with anxiety disorders, and
treatment sensitivity (Rapee et al., 1996).
37
2.2.15. Thought Control Questionnaire (TCQ). The TCQ (Wells & Davies, 1994)
is a 30-item measure assessing the use of different strategies for controlling unwanted
and unpleasant thoughts. The TCQ consists of five subscales: distraction, social control,
worry, punishment, and reappraisal. Participants were asked to rate how often they use
each strategy on a 4-point Likert scale. The TCQ possesses adequate internal consistency
(Reynolds & Wells, 1999) and good test-retest reliability (Wells & Davies, 1994). The
TCQ also has good convergent validity, construct validity, and treatment sensitivity
(Reynolds & Wells, 1999).
2.2.16. Willingness Scale (WS). The WS (Block & Wulfert, 2000) is an 8-item
measure assessing willingness to engage in public speaking situations. The WS includes
situations such as “Raising your hand in a small seminar class to ask a question or make a
comment” and “Giving a presentation in a large classroom setting.” Because these items
are aimed at a college-aged population, the items were adapted for a more general clinical
population for the proposed study. Therefore, the adapted version included items such as
“Approaching a professor/boss to speak with him/her personally in the office.”
Participants were asked to rate their willingness to engage in these activities based on a
10-point Likert scale, ranging from “completely unwilling” to “completely willing.”
Participants receiving ACT and CBGT in the Block (2002) study showed an increase in
willingness from pre- to post-treatment compared to the waitlist control group, suggesting
that this measure is treatment sensitive.
2.2.17. Social Interaction Self-Statement Test (SISST). The SISST (Glass,
Merluzzi, Biever, & Larsen, 1982) is a 30-item scale designed to measure the frequency
of positive and negative self-statements that arise before, during, or after a social
38
interaction. The SISST consists of two subscales (positive and negative self-statements),
and items are rated on a 5-point Likert scale ranging from 1(hardly ever) to 5 (very
often). The SISST possesses good internal consistency (Osman, Markway, & Osman,
1992; Zweig & Brown, 1985), split-half reliability (Glass et al., 1982), and test-retest
reliability (Zweig & Brown, 1985). In addition, the SISST has demonstrated good
discriminant validity (Dodge et al., 1988; Glass et al., 1982; Zweig & Brown, 1985),
convergent validity (Glass et al., 1982; Osman et al., 1992), and treatment sensitivity
(Turner, Beidel, & Jacob, 1994; Heimberg et al., 1990). The SISST was administered
after completion of the video taped behavioral assessments at pre- and post-treatment.
2.2.18. Demographics Questionnaire. A demographics questionnaire was created
which collected demographic information such as age, ethnicity, level of education,
marital status, and occupation. In addition, similar questions were asked of spouses, if
applicable.
2.2.19. Clinical Global Impression Scale (CGI). The CGI (National Institutes of
Mental Health, 1985) is a clinician global rating of severity and improvement, on a 7-
point Likert scale. Severity ratings were completed by assessors at pre-treatment, mid-
treatment, post-treatment, and follow-up. Improvement ratings were also completed by
assessors at mid-treatment, post-treatment and follow-up, based on comparisons of the
current evaluation to the pre-treatment assessment results. The CGI scales have been
used extensively in clinical trials, and they have demonstrated good interrater reliability
(Lipsitz, Mannuzza, Klein, Ross, & Fyer, 1999). In addition, self-reported improvement
ratings correlated significantly and highly with independent evaluator and therapist
ratings of improvement on the CGI Improvement scale (Lipsitz et al., 1999). A recent
39
study by Zaider, Heimberg, Fresco, Schneier, & Liebowitz (2003) examined the
psychometric properties of both CGI scales adapted for social anxiety disorder. Results
found that the CGI Severity possesses good convergent validity with measures of social
anxiety, depression, impairment, and quality of life, supporting its use as a global index
of severity. In addition, results showed that the CGI Improvement possesses good
convergent validity only with change in social anxiety symptoms, supporting its use as a
symptom-specific measure of improvement for individuals with SAD.
2.2.20. Behavioral assessment. Three standardized behavioral role play tasks
were administered to assess behavioral performance. These tasks included: (a) a dyadic
role play simulating an interaction with a confederate; (b) a triadic role play simulating a
conversation with two confederates; (c) an impromptu speech. Role play tasks are
frequently used for behavioral assessment of social anxiety (Herbert, Rheingold, &
Brandsma, 2001; McNeil, Ries, & Turk, 1995; Turner, Beidel, & Larkin, 1986). Ratings
of skill and anxiety were obtained from participant self-report and observer ratings
conducted by assessors. There is sufficient support for the reliability and validity of
social skills ratings in behavioral assessment tasks (Herbert et al., 2003). For the current
study, the role play tasks were video-taped and viewed by observers blind to assessment
time point. The observers rated participants’ quality of social skills on a 5-point Likert
scale ranging from 1 (poor) to 5 (excellent), on the following dimensions: verbal content,
non-verbal content, paralinguistic features, and overall social skills. In addition,
observers rated participants’ observed level of anxiety based on the Subjective Units of
Discomfort (SUDS) scale (Wolpe & Lazarus, 1966), which ranges from 0-100.
Observers used anchors developed from previous studies in the Anxiety Treatment and
40
Research Program (Herbert et al., 2003, 2004, 2005) and were trained to a reliability of
.80. Previous agreement between observers on these ratings for other studies in the
Anxiety Treatment and Research Program has been high (intraclass correlation α = .96;
Herbert et al., 2003), and agreement for the current study was also high (intraclass
correlation α = .87).
2.3. Treatment
2.3.1. Acceptance and Commitment Therapy (ACT). The cognitive-behavior
therapy used in the current study was delivered in an individual format using a modified
treatment manual based on the work of Hayes et al. (1999) and Block (2002).
Participants received 12 one-hour sessions of ACT through the Anxiety Treatment and
Research Program at Drexel University.
Four major concepts of ACT were presented in treatment, the first of which is
termed “creative hopelessness.” The primary purpose of this stage is to help participants
examine the futility of past attempts to control unwanted levels of social anxiety. The
next phase introduced acceptance or “willingness” as an alternative to controlling
unwanted private events. This stage consists of allowing oneself to have unwanted
thoughts or feelings while engaging in goal-directed behavior (e.g., attending a party,
initiating a conversation). Mindfulness and other techniques were then introduced in the
next stage to facilitate nonjudgmental awareness of unwanted private events and
willingness to experience them without analyzing their veracity. This exercise of
separating oneself from internal experiences has been termed “cognitive defusion,” or
“deliteralization.” Although values and goals were elicited in the beginning of treatment,
the final stage consisted of clarifying participants’ values and facilitating their ability to
41
engage in valued actions (e.g., engaging in social interactions which will lead to more
meaningful social relationships) despite perceived obstacles. These key concepts were
explained via metaphors and mindfulness exercises, as the experiential aspect of this
treatment is theoretically important. As in standard behavior therapy for social anxiety,
role play exercises with confederates, in-vivo exposure exercises assigned as homework,
and social skills training were incorporated into treatment. Each session ended with a
brief review, suggested exercises to practice between sessions, and specific homework
assignments.
2.4. Procedure
Potential participants underwent an initial 20-minute telephone screening
interview, in which the purpose of the study was discussed and a brief description of
presenting problems was determined. Those individuals still interested in participating
were invited to the anxiety clinic for an evaluation by a trained diagnostician using the
SCID/IP. At that time, informed consent was obtained. In addition, the SAD section of
the ADIS and the Avoidant Personality Disorder section of the SCID-II were
administered to increase the accuracy of diagnosis and to obtain further information on
participants. Diagnosis was primarily determined by the SCID-I/P. Diagnosticians were
advanced doctoral clinical psychology students. Diagnosticians were trained to
proficiency, and the assessments were presented in weekly supervision meetings.
Weekly supervision was conducted by the director of the clinic, who is a licensed clinical
psychologist with extensive experience in the assessment and treatment of SAD and the
use of ACT in this population. If questions arose as to the diagnostic status of a
participant, the case was discussed and a decision was made via group consensus.
42
At the time of the diagnostic interview, demographic information and baseline
measures were obtained from self-report questionnaires. If participants met criteria based
on the diagnostic interviews, they completed the video-taped role play tasks prior to
beginning treatment. All participants meeting criteria for the study underwent a standard
baseline waiting period of 4 weeks between the diagnostic interview and the video-taped
role play tasks. At the time of the role play tasks, participants were given a second
questionnaire packet to complete and bring to the first session. Also, participants were
given a fear hierarchy form to complete for the first session.
Once the pre-treatment assessments were completed, participants received 12 one-
hour, weekly individual sessions of ACT. Therapists consisted of doctoral clinical
psychology students, who underwent protocol training in ACT by the director of the
clinic. Weekly supervision meetings were held by the director to provide ongoing
supervision. Treatment sessions were audiotaped with participants’ consent, and 10% of
treatment tapes were randomly selected and assessed using a treatment integrity form to
determine adherence to the manual. Results of this review showed 100% adherence to
the manual, with no errors of commission or omission.
Mid-way through treatment (after 6 sessions) participants completed the same
assessment self-report measures. In addition, participants were administered the SAD
section of the SCID-I/P, as well as the SAD section of the ADIS, the APD section of the
SCID-II, and the CGI Severity and Improvement scales. At post-treatment participants
completed the same assessment self-report measures, and assessors administered the
same abbreviated structured clinical interviews and completed the CGI Severity and
Improvement scales. Participants also completed the video-taped behavioral assessment
43
tasks at post-treatment. At 3-month follow-up, assessment data was collected by
contacting participants via telephone. The assessor interviewed the participant by
telephone using the same abbreviated structured clinical interviews. Once this
assessment was completed, the assessor completed the CGI Severity and Improvement
scales based on information obtained from the interview. Participants also completed a
follow-up questionnaire packet via mail. Because collection of follow-up information is
ongoing, the results from this information will not be reported in the current study.
However, these results will be presented in a subsequent manuscript that will be
submitted for publication. See Table 1 for an overview of the assessment procedures.
2.5. Statistical Analyses
2.5.1. Statistical power. Power was calculated using the computer program G-
Power (Faul & Erdfelder, 1992) for repeated measures analysis of variance with an alpha
set at .05 and a medium-to-large effect size (f = .35). This effect size was chosen because
ACT is a relatively new treatment, but is a form of behavior therapy that includes
exposure. Studies based on exposure treatment for SAD have demonstrated large pre-to-
post effect sizes (Gould et al., 1997; Taylor, 1996). A sample of 28 participants would
yield an estimated power of .80, which is acceptable for behavioral research (Cohen,
1988). A post-hoc power analysis was conducted based on 17 completers, a large effect
size (f = .40), and an alpha level of .05, resulting in power of .62.
2.5.2. Preliminary analyses. Baseline scores (consisting of a questionnaire packet
administered 4 weeks prior to beginning treatment) were compared to pre-treatment
scores to determine if symptoms were likely to change over time without treatment. No
changes in symptoms were expected from baseline to pre-treatment, based on previous
44
research showing no change over time in waitlist control conditions compared to
treatment conditions (Hope, Heimberg, & Bruch, 1995; Mattick, Peters, & Clarke, 1989).
T-tests were used to compare means on the outcome measures between baseline and pre-
treatment measures. As the number of treatment drop outs was small (n = 2), the pre-
treatment scores and demographics of these participants are not analyzed statistically and
instead are descriptively presented in order to compare them to the remainder of the
sample.
2.5.3. Primary analyses. To test the hypothesis that ACT would result in
improved outcome from pre-treatment to post-treatment, continuous measures were
analyzed using multivariate analysis of variance and appropriate post hoc tests. A one-
way repeated measures MANOVA (levels: pre-treatment, mid-treatment, and post-
treatment) was conducted on the following measures of social anxiety symptoms: the
SPAI-SP, FQ-SP, Brief FNE, and the fear and avoidance total subscale scores of the
LSAS. Significant results were followed up by univariate ANOVAs and Bonferroni post
hoc tests. In addition, repeated measures MANOVAs were conducted on the five
subscales of the TCQ, the two subscales of the ACQ and ATQ, and the three subscales of
the SDS. Separate repeated measures ANOVAs (pre-treatment, mid-treatment, and post-
treatment) were conducted on all other measures. Separate ANOVAs as opposed to
MANOVAs were used for the ACT questionnaires because they were conceptualized as
process measures. Significant results from the ANOVAs were followed up using
Bonferroni post hoc tests. To account for attrition, separate treatment completer and
intent-to-treat analyses (carrying the last set of data obtained forward) were conducted for
the analyses described above.
45
Paired samples t-tests were used to examine pre- to post-treatment changes on the
participant and observer ratings from the behavioral assessment task. For the observer-
rated social skills analysis, ratings in the 4 dimensions were averaged across all three role
play situations and then compared using paired samples t-tests. Participant self-
performance ratings were averaged across the three role play situations, and participant
and observer SUDS ratings were also averaged across the three role play situations and
compared pre- to post-treatment via paired samples t-tests. This method has been used in
previous research studies examining the efficacy of CBT for SAD (Herbert et al., 2004,
2005).
2.5.4. Analysis of clinical significance. Analyses were conducted to determine the
proportion of participants achieving clinically significant improvement using the reliable
change index (Jacobson & Truax, 1991). Treatment responders were defined, based on
Jacobson and Truax, as those whose post-treatment scores on the SPAI-SP fell closer to
the mean of the functional, rather than dysfunctional, population. In addition, the
percentage of participants meeting criteria for both reliable change and clinically
significant improvement was calculated. As few other studies examining CBT for SAD
have calculated clinical significance based on this definition, treatment response was also
calculated based on definitions used by other studies (e.g., Heimberg et al., 1998 and
Herbert et al., 2005). This included calculating percentage of treatment responders based
on a one standard deviation improvement from pre- to post-treatment, and based on those
who were rated by an independent assessor as a 1 or 2 (markedly or moderately
improved) on the CGI Improvement scale at post-treatment. Demographic characteristics
of responders vs. non-responders were then descriptively examined to determine any
46
trends suggesting demographic differences related to treatment response. Finally, the
proportion of those who changed SAD diagnostic status at post-treatment based on the
SCID/IP was computed.
2.5.5. Secondary analyses. Due to the lack of a comparison group, effect sizes
derived from the current study were compared to effect sizes obtained from existing data
on the efficacy of standard CBT for SAD, both from Drexel’s Anxiety Treatment and
Research Program, as well as from other researchers elsewhere (Clark et al. 2003;
Davidson et al. 2004; Heimberg et al. 1998; Herbert et al. 2005; Herbert et al. 2004).
Effect sizes were converted to Pearson r coefficients, using a method described by
Rosnow and Rosenthal (1996), and the coefficients were then compared statistically
using Fisher’s Z test.
Pearson correlation analyses were conducted to examine the hypothesis that a
change in ACT-related process measures would be associated with a change in treatment
outcome. Pearson correlations were conducted between the pre- to post-treatment change
score on the SPAI-SP and pre- to post-treatment change scores on ACT-related measures
such as the AAQ and ATQ believability and frequency scales. The same analyses were
also conducted using the QOLI as an outcome measure. To better assess the timing of
changes, correlations were also conducted between pre- to mid-treatment changes on the
AAQ and ATQ and mid- to post-treatment changes on the SPAI-SP and QOLI. Several
other studies have examined the relationship between experiential avoidance (using the
AAQ) and treatment outcome or symptom severity (e.g., Roemer & Orsillo, 2002;
Twohig & Woods, 2004; Zettle, 2003). In addition, some studies have examined the
relationship between believability and frequency of symptoms or negative thoughts and
47
treatment outcome (Bach & Hayes, 2002; Gaudiano & Herbert, in press(b); Zettle &
Hayes, 1986). Although many studies examining these variables used regression
analyses to conduct formal mediational analyses, the current study used Pearson
correlations due to the small sample size. Block (2002) also examined the relationship
between change in negative and positive thoughts (based on the SISST) and treatment
outcome. In order to replicate this analysis, the current study also conducted Pearson
correlations between the change in the SISST positive and negative subscales and the
change in the SPAI-SP.
Finally, a Pearson correlation analysis was conducted to examine the hypothesis
that a change in quality of life would be associated with treatment outcome, using the
SPAI-SP. Previous research by Eng et al. (2001) examined the relationship between
quality of life and measures of social anxiety and depression, and found that quality of
life significantly correlated with depression scores only at post-treatment. Therefore, the
current study attempted to replicate these analyses by examining the relationship between
changes in quality of life and social anxiety and depression scores. As above, the
relationship between pre- to mid-treatment changes in quality of life and mid- to post
changes in social anxiety and depression was examined, in addition to correlations
utilizing pre- to post-treatment change scores.
2.6. Design and Data Collection Considerations
Attempts were made to anticipate any difficulties that might arise when
conducting the study and to minimize their effects. For example, drop out rate and
difficulty collecting follow-up are common problems in treatment outcome studies. To
safeguard against a high attrition rate, a baseline assessment period was used as opposed
48
to having a waitlist control group. To address difficulty with obtaining follow-up
information, clients were reminded of the follow-up assessment at post-treatment, and
phone calls were made by the therapist to the client if the assessor was unable to reach
the client. If the client could not be reached by telephone, a letter was sent to the client
via post. Finally, both completer and intent to treat analyses were conducted (see
Primary analyses section).
Alternative designs to the current study were also considered. One alternative
design considered was administering treatment in a group format, similar to the
preliminary study completed by Block (2002). However, individual treatment provides a
more practical alternative to group treatment, given the difficulty of scheduling a cohort
of patients in the community (Herbert et al., 2004). In addition, individual treatment can
provide an opportunity to develop a more detailed and personalized assessment and
formulation of values and goals, a major component of ACT. Previous studies using
CBT for SAD have suggested comparable efficacy between group and individual
treatment (Gould et al., 1997; Lucas, 1994). Although no direct comparisons between
individual and group ACT have been conducted, ACT has been delivered successfully in
both individual and group formats (Orsillo, Roemer, Block, LeJeune, & Herbert, 2004;
Walser & Pistorello, 2004). However, preliminary evidence on an indirect comparison
between individual and group ACT treatment for depression suggests that the efficacy of
ACT may be diminished when delivered in a group format (Zettle & Rains, 1989).
Another alternative design considered was one using a waitlist control group.
However, having a waitlist control group was difficult due to the resources and time
available for this study. In addition, the use of waitlist control groups can increase the
49
rate of attrition (Kazdin, 1998, p.138). Instead, baseline measures were obtained and
analyzed, and effect sizes from this study were compared to effect sizes from other
studies examining the efficacy of CBT for SAD. Effect sizes also were compared to
other studies conducted in the Anxiety Treatment and Research Program, which
examined the efficacy of CBT for SAD. These data were conducted in the same clinic as
the current study, and utilized the same assessment procedures, independent evaluators,
and therapists, thus providing an appropriate comparison to the current study.
Research on the efficacy of ACT, especially related to anxiety disorders, is in its
infancy. One goal of preliminary research on ACT should be to examine its efficacy for
specific forms of psychopathology. The current study attempted to examine the efficacy
of ACT as an intervention for SAD. Although the design of the current study did not
provide information on the specific mechanisms of action of ACT, it represents a
necessary first step in the investigation of its efficacy in order to determine whether ACT
is worthy of the time and resources needed for further investigation.
50
3. RESULTS
3.1. Sample Description
Figure 1 depicts a diagram showing participant flow throughout the study phases.
A total of 86 participants completed a telephone screening procedure, and of those, 47
passed the telephone screening and were scheduled for a SCID assessment. Thirty-nine
participants did not meet inclusion/exclusion criteria during the telephone screening: 24
did not meet criteria for generalized SAD or SAD was not primary; 4 met criteria for
current psychotic symptoms; 2 were over 60 years of age; 2 were already in treatment
studies elsewhere; 6 had a previous trial of CBT; and 1 was unreachable due to
disconnected phone service. Of the 47 participants who completed the SCID, 30 met
inclusion/exclusion criteria and were provided with informed consent. Ten participants
were no longer interested in participating in the study after the SCID, and 7 participants
did not meet study criteria upon completion of the SCID. Of the 30 participants that
passed the SCID assessment, 6 dropped out of the study during the 4-week baseline
period (therefore not completing the behavioral assessment) and 4 completed the
behavioral assessment but dropped out before beginning treatment. Twenty participants
began treatment; 2 dropped out before the 6-week assessment period, and 1 participant
was withdrawn after beginning treatment as it became apparent that SAD was not
primary. Therefore, 17 participants completed treatment, and 16 were included in the
completer analyses (1 participant completed treatment but refused to complete post-
treatment assessments).
Average age of participants was 31 years (SD = 10), and 52.8% of participants
were female. A majority of the sample was Caucasian (63.9%), single (80.6%), and
51
employed full-time (54.3%). Educational attainment of the sample was relatively high,
with 22.2% having a graduate/professional school education, 38.9% having a college
degree, and 27.8% having some college education. Almost half (48.6%) of participants
carried at least one comorbid Axis I disorder; 29.7% had a comorbid depressive disorder,
and 24.3% had a comorbid anxiety disorder. In addition, 59.5% of participants met
criteria for Avoidant Personality Disorder. Finally, approximately 16% of participants
were taking at least one psychotropic medication. Two participants were taking one or
more antidepressants, 1 was taking anxiolytics, and 2 participants were taking both
antidepressants and anxiolytics; 2 participants were taking psychostimulants. See Table 2
for details of demographic characteristics.
3.2. Preliminary Analyses
3.2.1. Drop outs. Because there were so few treatment dropouts (n=2) in relation
to treatment completers (n=17), statistical analyses could not be conducted to compare
dropouts to completers on variables. One drop out was female, and 1 was male. Both
drop outs were single or separated, had a college degree, and were employed full time.
Only one of the drop outs was taking a psychotropic medication (a PRN anxiolytic). One
participant dropped out of treatment due to lack of belief in the treatment rationale, and
the other dropped out because of the time commitment involved. Overall, the 2
participants who dropped out appeared representative of the larger sample, and did not
appear to differ along any demographic dimension.
3.2.2. Baseline period. Baseline scores were compared to pre-treatment scores
using paired samples t-tests, to determine if symptoms changed over the 4 week baseline
period. Results showed no significant differences between baseline and pre-treatment
52
scores for all self-report measures, including social anxiety and depressive symptoms,
control over anxiety and thoughts, willingness, experiential avoidance, quality of life,
impairment, discrepancy between stated values and consistent action, and frequency and
believability of automatic thoughts.
3.2.3. Exploratory Demographic Comparisons. Although formal statistical
analyses could not be conducted on pre-treatment SPAI-SP scores between demographic
variables due to small sample size, these data were descriptively examined. Pre-
treatment SPAI-SP scores appeared to be greater for females (M = 146.14, SD = 26.28)
than males (M = 121.82, SD = 31.12), non-Caucasians (M = 148.33, SD = 30.71) than
Caucasians (M = 133.03, SD = 33.23), married/divorced/separated (M = 134.97, SD =
25.22) than single (M = 132.90, SD = 32.59) participants, and those with a high school
degree or some college education (M = 141.02, SD = 31.03) than a college degree or
graduate/professional education (M = 130.04, SD = 31.35).
3.3. Primary Analyses
Of the 17 participants who completed treatment, one participant completed
treatment but refused to complete the post-treatment assessments. In addition, missing
data on various questionnaires resulted in different sample sizes for analyses. See Table
3 for raw score means and standard deviations of outcome and process measures.
3.3.1. Outcome measures. A one-way repeated measures (pre-, mid-, and post-
treatment) MANOVA was conducted on the measures of social phobia symptoms (i.e.,
the SPAI-SP, FQ-SP, Brief FNE, and the total fear and avoidance subscales of the LSAS)
with results showing a significant difference (F2,46 = 2.88, p = .015). Separate one-way
repeated measures ANOVAs yielded significant differences on all of the questionnaires
53
(ps: SPAI-SP < .001; FQ-SP = .002; Brief FNE = .006; LSAS Fear = .001; LSAS
Avoidance < .001). On the SPAI-SP, Bonferroni post hoc tests showed significant
decreases in severity from pre- to mid-treatment (p = .02) and pre- to post-treatment (p =
.003), but not from mid- to post-treatment (p = .089). Post hoc tests also showed
significant decreases on the FQ-SP from pre- to mid-treatment (p = .057) and pre- to
post-treatment (p = .026), but not from mid- to post-treatment (p = .18). Follow-up
comparisons on the Brief FNE showed a significant improvement from pre- to post-
treatment (p = .028) but not pre- to mid-treatment or mid- to post-treatment (ps = .142
and .333, respectively). Finally, Bonferroni post hoc comparisons on the LSAS total fear
subscale yielded significant decreases from mid- to post-treatment, and pre- to post-
treatment (ps = .037 and .016, respectively), but not from pre- to mid-treatment (p =
.412). In addition, follow-up tests on the LSAS total avoidance subscale showed
significant decreases in avoidance from pre- to mid-treatment (p = .024), mid- to post-
treatment (p = .003), and pre- to post-treatment (p = .001). These finding suggest that
avoidance changed before fear, given that there was no difference on the LSAS fear
subscale from pre- to mid-treatment.
The repeated measures ANOVA on the BDI-II yielded significant results (F2,30 =
5.8, p = .007), with Bonferroni follow-up tests showing marginally significant
differences from mid- to post-treatment (p = .068) and pre- to post-treatment (p = .071),
but not pre- to mid-treatment (p = .367).
A repeated measures ANOVA on the QOLI also showed significant results (F2,30
= 6.46, p = .005), with post hoc tests revealing greater perceived quality of life from pre-
to post-treatment (p = .01), but not pre- to mid-treatment or mid- to post-treatment (ps =
54
.185 and .280, respectively). In addition, a repeated measures MANOVA was conducted
on the three subscales of the SDS (work, social, family); results were significant (F2,58 =
3.97, p < .001). Separate follow-up ANOVAs were conducted, and revealed significant
differences for the work (F2,30 = 11.12, p < .001), social (F2,30 = 9.11, p = .001), and
family (F2,30 = 7.24, p = .003) subscales. Bonferroni post hoc tests showed that none of
the subscales were significant from pre- to mid-treatment (ps: work = .276; social = .130;
family = 1.00), but all three subscales were significant from mid- to post-treatment (ps:
work = .022; social = .047; family = .055), showing a decrease in self-reported
impairment in work, social, and family domains. In addition, all three subscales were
significant from pre- to post-treatment (ps: work = .001; social = .013; family = .01),
indicating less self-reported impairment in work, social, and family domains at post-
treatment.
3.3.2. Process measures. Separate repeated measures MANOVAs were also
conducted on the two maladaptive subscales (worry and punishment) and three adaptive
subscales (social control, distraction, and reappraisal) of the TCQ. Results on the
maladaptive subscales MANOVA were not significant (p = .625). However, results from
the adaptive subscales MANOVA were significant (F2,46 = 2.29, p = .05). Follow-up
univariate ANOVAs showed a significant difference on the social control subscale only
(F2,24 = 6.70, p = .005), with Bonferroni’s post hoc tests showing a marginal difference
from mid- to post-treatment (p = .057), and a significant difference from pre- to post-
treatment (p = .043), but no difference from pre- to mid-treatment (p = 1.00). Therefore,
participants reported greater use of social control strategies from pre- to post-treatment
and mid- to post-treatment.
55
A repeated measures MANOVA also was conducted on the two ACQ subscales,
reactions and events. This MANOVA was significant (F2,48 = 2.62, p = .047). Follow-
up univariate ANOVAs showed a significant difference on the reactions subscale (F2,24 =
5.89, p = .008), with follow-up Bonferroni’s tests showing a significant difference from
pre- to post-treatment (p = .026), but not from pre- to mid-treatment or mid- to post-
treatment (ps = .228 and .322, respectively). More specifically, participants reported
greater perceived control over emotional reactions at post-treatment compared to pre-
treatment. The follow-up ANOVA on the ACQ events subscale was marginally
significant (F2,24 = 3.13, p = .062); however, post hoc tests showed no significant
differences across the three time points (ps = .488, .984, and .119, respectively).
Finally, separate one-way repeated measures MANOVAs and ANOVAs were
conducted on the ACT-specific measures. The MANOVA on the ATQ believability and
frequency subscales was significant (F2,44 = 2.87, p = .034). Results from the follow-up
univariate ANOVAs showed a significant difference on the ATQ believability subscale
(F2,22 = 7.3, p = .004), with Bonferroni tests showing less believability in automatic
thoughts from pre- to post-treatment (p = .011), but not from pre- to mid-treatment and
mid- to post-treatment (ps = .623 and .108, respectively). Results also showed a
significant difference on the ATQ frequency subscale (F2,22 = 5.17, p = .014), with
Bonferroni tests showing less frequency of automatic thoughts from pre- to post-
treatment (p = .046) but not pre- to mid-treatment or mid- to post-treatment (ps = .555
and.22, respectively).
The ANOVA on the AAQ showed significant results (F2,26 = 7.09, p = .003), with
post hoc tests revealing less experiential avoidance from pre- to post-treatment (p = .025)
56
but not from pre- to mid-treatment or mid- to post-treatment (ps = .182 and .119,
respectively). Furthermore, the ANOVA on the WS yielded a significant difference (F2,26
= 25.4, p < .001), with post hoc tests showing that willingness to engage in public
speaking situations increased significantly from pre- to mid-treatment, mid- to post-
treatment, and pre- to post-treatment (ps = .007, .002, and < .001, respectively). The
ANOVA on the VLQ also showed a significant difference (F2,26 = 3.47, p = .046), with
participants reporting significantly less discrepancy between stated values and consistent
action from pre- to post-treatment (p = .031) but not pre- to mid-treatment or mid- to
post-treatment (ps = 1.00 and .372, respectively).
3.3.3. Clinician-rated measures. A one-way repeated measures ANOVA was
conducted on CGI Severity ratings at pre-, mid-, and post-treatment. Results revealed a
significant difference (F2,28 = 19.99, p < .001), with post hoc tests showing significantly
decreased severity from pre- to mid-treatment, mid- to post-treatment, and pre- to post-
treatment (ps = .011, .009, and < .001, respectively). In addition, a repeated measures
ANOVA on CGI Improvement ratings at mid- and post-treatment showed a significant
difference (F1,14 = 6.67, p = .022), with greater improvement at post-treatment compared
to mid-treatment.
3.3.4. Behavioral assessment. Separate paired samples t-tests were conducted on
the average self-ratings of performance and average SUDS ratings across the three role
play situations at pre- and post-treatment. The t-test on the average self-ratings of
performance revealed a significant difference (t10 = -4.68, p = .001), with self-rated
performance greater at post-treatment. The t-test on the average SUDS ratings was also
significant (t10 = 5.13, p < .001), with lower SUDS ratings at post-treatment.
57
Separate paired samples t-tests were also conducted on observer ratings of social
skills and anxiety (using the SUDS scale), averaged across the three role play situations.
The t-test on observer ratings of social skills revealed a significant difference (t15 = -6.84,
p < .001), with observers rating participants’ social skills significantly higher at post-
treatment. In addition, the t-test on observed anxiety was also significant (t15 = 5.95, p <
.001), with observers rating participants’ anxiety lower at post-treatment.
Finally, separate paired samples t-tests (pre- to post-treatment) were conducted on
the SISST negative and positive thoughts subscales, which were administered
immediately following the behavioral assessment task. Both subscales were significant
(positive: t13 = -2.99, p = .011; negative: t13 = 2.35, p = .035), showing greater frequency
of positive thoughts and lower frequency of negative thoughts from pre- to post-
treatment.
3.3.5. Intention-to-treat (ITT) analyses. Missing data from completers and
dropouts (including mid-treatment dropouts) was replaced using the last observation
carried forward (LOF) method, resulting in a sample size of 19 participants for the
intention-to-treat analyses. Results from the ITT MANOVA on the SPAI-SP, FQ-SP,
Brief FNE, and LSAS were the same as the completer MANOVA, with the exception that
the SPAI-SP significantly decreased across all time points (ps < .05). The ITT
MANOVAs of the TCQ were nearly identical to the completer analyses, with the social
control subscale only marginally significant from pre- to post-treatment (p = .062); for
the ITT MANOVA of the ACQ, the reactions subscale was only marginally significant
from pre- to post-treatment (p = .062). The ITT MANOVA of the SDS was identical to
the completer analysis. The separate ANOVAs on all other questionnaires were similar
58
to the completer analyses, with one exception: the VLQ was only marginally significant
from pre- to post-treatment (p = .064). Results from the ANOVAs on the CGI Severity
and Improvement scales were identical to the completer analyses. Finally, the paired
samples t-tests on the self and observer ratings of the behavioral assessment, as well as
the SISST, were also similar to the completer analyses.
3.4. Analyses of Clinical Significance
Jacobson and Truax (1991) define clinically significant improvement as the post-
treatment score of a particular individual falling closer to the mean of the functional,
rather than the dysfunctional, population. Clinical significance analyses were also
conducted based on this definition, by comparing the current sample to a non-clinical
college undergraduate sample (Osman et al., 1995). Results showed that 56.3% of
participants met criteria for clinically significant improvement.
Jacobson and Truax (1991) also recommend calculating a reliable change index to
account for measurement error. Therefore, reliable change was calculated on the SPAI-
SP from pre- to post-treatment, using the test-retest reliability (r = .86) obtained from
Turner et al. (1989). Results showed that 62.5% of participants achieved reliable change
above and beyond measurement error. Furthermore, it was calculated how many
participants met criteria for both reliable and clinically significant change; results showed
that 37.5% of participants met criteria for both. Pre-treatment SPAI-SP scores for
participants meeting criteria for clinically significant and reliable change were in the
clinical range. Finally, forty-four percent of participants no longer met DSM-IV criteria
for generalized SAD at post-treatment as determined by the SCID.
59
Responder status was also calculated based on definitions used in other studies
examining traditional CBT for SAD. For example, Herbert et al. (2005) defined
responders as those demonstrating a pre- to post-treatment improvement of at least one
standard deviation. Based on this definition, 62.5% of participants in the current study
were considered responders, compared to 79% in the Herbert et al. (2005) study. In
addition, responder status was calculated based on similar methods used by Heimberg et
al. (1998), in which responders were defined as receiving a rating of 1 or 2 (markedly or
moderately improved) on the CGI-I. Based on this definition, 37.5% of participants in
the current study were considered responders, compared to 75% in Heimberg et al.
(1998).
Finally, demographic characteristics of responders vs. non-responders were
descriptively examined using both a more liberal (1 standard deviation improvement) and
more conservative (Jacobsen & Truax, 1991) definition of response. When examining a
1 standard deviation improvement, there appeared to be fewer female non-responders
than responders (33% vs. 60%), fewer single non-responders than responders (67% vs.
90%), and greater non-responders than responders taking psychotropic medication (83%
vs. 60%). Using the Jacobsen & Truax (1991) definition, there appeared to be fewer
Caucasian non-responders than responders (60% vs. 83%), greater highly educated non-
responders than responders (80% vs. 67%), fewer non-responders than responders taking
psychotropic medications (10% vs. 67%), and greater non-responders than responders
meeting criteria for Avoidant Personality Disorder (60% vs. 33%).
60
3.5. Secondary Analyses
3.5.1. Effect size comparisons. Effect sizes (pre- to post-treatment) were
calculated for the current study on the SPAI-SP and Brief FNE, and compared to those
calculated from other studies conducted within the Drexel University Anxiety Treatment
and Research Program, as well as from studies conducted elsewhere. Effect sizes for the
SPAI-SP and Brief FNE in the current study were large, based on Cohen’s definitions
(Cohen, 1988). First, the current study’s SPAI-SP effect size (d = 1.01, r = .45) was
compared to two previous studies conducted at Drexel University (Herbert et al., 2004;
Herbert et al., 2005). Herbert et al. (2004) compared an individual 12 session protocol of
CBT administered in 12 weeks versus 18 weeks; the effect size on the SPAI-SP for the
12-week group (d = 1.42; r = .58) was compared to the current study. As discussed
previously, effect sizes were converted into Pearson r coefficients and then compared
using the Fisher Z test. Results showed no significant difference between the two
coefficients (p > .05). Next, the SPAI-SP effect size was compared to Herbert et al.
(2005), in which CBGT with social skills training was compared to CBGT without social
skills training; the current study was compared to only the CBGT + social skills training
group (d = 1.94; r = .70), as the current study also included social skills training. Results
from the Fisher Z test showed no significant difference between the coefficients (p > .05).
The SPAI-SP effect size of the current study was then compared to the
comprehensive cognitive behavioral group therapy (CCBT; d = 1.15; r = .50) condition
of Davidson et al. (2004), which administered 14 sessions of cognitive restructuring,
exposure, and social skills training in a group format. The Fisher Z test resulted in no
significant difference between the coefficients (p > .05). Finally, the current study was
61
compared to Cox et al. (1998), which examined the effects of CBGT on several outcome
measures (SPAI-SP d = .56; r = .26). The Fisher Z test between these two coefficients
was also not significant (p > .05).
Next, the Brief FNE effect size of the current study (d = 1.04; r = .46) was
compared to those of Herbert et al. (2005) (d = 1.31; r = .55), the CBGT condition of
Heimberg et al. (1998) (d = .79; r = .36), and the cognitive therapy condition of Clark et
al. (2003) (d = 1.35; r = .56). There were no significant differences between the current
study and these other studies (ps > .05).
Given that the current study is a partial replication and extension of Block (2002),
effect sizes were also calculated from that study and compared to the current study on the
Brief FNE, WS, and QOLI. Effect sizes on these measures for the current study were
1.04, 1.34, and .80, respectively. For the Block (2002) study, these effect sizes were .52,
1.03, and .16. However, when these effect sizes were converted to r coefficients and
compared via the Fisher Z test, these effect sizes were not significantly different from
each other (all ps > .05).
Finally, to address the hypothesis regarding quality of life, an effect size was
calculated on the QOLI from the current study (d = .80; r = .37) and compared to the
QOLI effect size from Eng et al. (2001) (d = .49; r = .24). The Fisher Z test showed no
significant difference between these two coefficients (p > .05).
3.5.2. Correlation analyses. Pearson correlation analyses were conducted to
examine the relationship between a change in ACT-related process measures and
treatment outcome. Results showed that a pre- to post-treatment change in the SPAI-SP
was significantly correlated with a pre- to post-treatment change in the AAQ (r = .60, p <
62
.05). Therefore, an improvement in social anxiety symptoms from pre- to post-treatment
also was associated with a decrease in experiential avoidance. There was no significant
association between a pre- to mid-treatment change on the AAQ and a mid- to post-
treatment change on the SPAI-SP (r = -.22, p = .43). In addition, mid- to post-treatment
and pre- to post-treatment changes in the SPAI-SP were not associated with pre- to mid-
treatment or pre- to post-treatment changes in believability or frequency of negative
cognitions (ps > .05). Pearson correlations were also conducted to examine the
relationship between a change in the same ACT-related process measures and quality of
life. The correlation between the pre- to post-treatment change on the AAQ and the pre-
to post-treatment change on the QOLI was marginally significant (r = -.50, p = .069), but
the correlation between the pre- to mid-treatment change on the AAQ and mid- to post-
treatment change on the QOLI was not significant (r = .43, p = .10). Pre- to mid-
treatment and pre- to post-treatment changes in believability and frequency of negative
thoughts were not related to mid- to post-treatment or pre- to post-treatment changes in
quality of life (ps > .05).
Block (2002) found that a decrease in the SISST negative subscale and an
increase in the SISST positive subscale (to a lesser extent) were associated with
improvement. To replicate these findings, change in the SISST positive and negative
subscales were compared to change in the SPAI-SP using Pearson correlations. Results
showed that a decrease on the SISST negative subscale was significantly correlated with
a decrease on the SPAI-SP (r = .60, p < .05), while an increase on the SISST positive
subscale was not (r = -.32, p > .05).
63
Finally, Pearson correlations were conducted to compare pre- to mid-treatment
and pre- to post-treatment changes on the QOLI with mid- to post-treatment and pre- to
post-treatment changes on the BDI-II and SPAI-SP, in order to replicate the correlation
analyses from Eng et al. (2001). Results showed that a pre- to post-treatment change in
the QOLI was significantly correlated with a pre- to post-treatment change in the SPAI-
SP (r = -.82 p < .01), indicating that an increase in quality of life was associated with a
decrease in social anxiety symptoms. However, a pre- to mid-treatment change on the
QOLI was not associated with a mid- to post-treatment change on the SPAI-SP (r = .05, p
= .86). Pre- to mid-treatment and pre- to post-treatment changes on the QOLI were not
associated with mid- to post-treatment or pre- to post-treatment changes on the BDI-II (r
= .26, p = .33; r = -.29, p = .28, respectively).
64
4. DISCUSSION
4.1. Summary of Results
The current study was a pilot trial to develop and investigate the preliminary
efficacy of ACT delivered in an individual format in a sample of adults with SAD.
Results showed significant improvement from pre- to post-treatment on self-report
measures of social anxiety, depression, quality of life, impairment, believability and
frequency of automatic thoughts, control over anxiety, and ACT-specific process
measures such as willingness to engage in social situations and experiential avoidance.
In addition, there were significant improvements from pre- to post-treatment on self-rated
performance and anxiety during the behavioral assessment tasks. Clinician-rated severity
and impairment also improved significantly from pre- to post-treatment, as did observer-
rated social skills and anxiety on the behavioral assessment tasks. Furthermore, results
showed large effect size gains from pre- to post-treatment on measures of social anxiety
and quality of life. These effect sizes were comparable to other recent studies that have
examined the efficacy of CBT for SAD. Clinical significance analyses indicated that
37.5% of participants met criteria for both reliable change and clinically significant
improvement compared to a normative sample. Finally, results showed that a change in
quality of life and experiential avoidance were significantly associated with treatment
outcome, but believability and frequency of negative cognitions were not.
4.2. Support for Hypotheses
4.2.1. Hypothesis #1. It was hypothesized that participants would not
demonstrate a change in self-reported symptoms from baseline to pre-treatment. Support
for this hypothesis was found on all except one self-report measure. These results are
65
consistent with previous findings that SAD does not remit without treatment (Davidson,
Hughes, George, & Blazer, 1994; Schneier, Johnson, Hornig, Liebowitz, & Weissman,
1992). The only significant difference from baseline to pre-treatment was found on the
VLQ. More specifically, participants reported less discrepancy between stated values
and consistent action at pre-treatment compared to baseline. This result could be due to
measurement error, as this measure has not yet been well validated or studied
psychometrically. Alternatively, this result could be due to the fact that since participants
were in the process of seeking treatment, they perceived this as being more consistent
with their stated values and thus reported less discrepancy between actions and stated
values at pre-treatment compared to baseline.
4.2.2. Hypothesis #2. It was hypothesized that participants would demonstrate
significant improvements in outcomes (e.g., symptomatology, impairment, etc.) and
changes in process measures from pre- to post-treatment. Significant results were found
in almost all domains and across all assessment modalities from pre- to post-treatment,
lending support to this hypothesis. As mentioned previously, there was a significant
decrease in social anxiety and depressive symptoms from pre- to post-treatment, a
significant increase in willingness to engage in social situations, and a significant
decrease in self-reported avoidance. These results are comparable to those obtained by
Block (2002). The current study also showed significant improvement in clinician-rated
measures of severity and impairment, in addition to observer-rated measures of social
skills and anxiety obtained from the behavioral assessment task, thus extending Block’s
findings.
66
The only domains in which significant results were not obtained were on the TCQ
and the ACQ events subscale. For example, the MANOVA on the two maladaptive
subscales (worry and punishment) of the TCQ was not significant, indicating that
participants’ use of maladaptive thought control strategies did not change over the course
of treatment. In addition, the MANOVA on the adaptive subscales (social control,
reappraisal, and distraction) of the TCQ was significant, but only the social control
subscale showed a significant difference. This finding could indicate that participants
increased their frequency of expressing unpleasant thoughts to others. Based on the ACT
model, which focuses on increasing behaviors consistent with one’s stated values and
goals rather than using control strategies to alter thoughts, one might expect that thought
control strategies would significantly decrease over time. Instead, these results showed
that most thought control strategies did not change over time, and one strategy increased
(i.e., social control). However, the ACT model does not necessarily state that there is not
any utility in control strategies, as the approach adopts a highly pragmatic stance with
regard to the utility of cognitive control strategies. Therefore, there is some utility in
control strategies if they “work” for the individual in some contexts. Furthermore, a
cognitive-based intervention (e.g., reappraisal or distraction) would be expected to
increase adaptive control strategies; but the current ACT intervention, which
deemphasizes controlling cognitions, showed no changes on these scales. The one
adaptive control strategy that did increase seems to be more consistent with exposure-
based treatments, including ACT. For example, the social control subscale included
items such as: “I talk to a friend about the thought” and “I ask my friends if they have
similar thoughts.” Finally, another possible explanation for the equivocal findings on the
67
TCQ could be related to the known psychometric problems with this measure, as its
construct validity is unclear and internal consistency was only marginal to adequate
(Reynolds & Wells, 1999).
In addition, results showed that the reactions subscale of the ACQ was clearly
significant, indicating that participants had greater perceived control over emotional
reactions to events. However, the events subscale was only marginally significant,
indicating a trend toward increasing perceived control over external events at post-
treatment. This is consistent with results obtained by Block (2002), who found greater
perceived control over reactions to events and greater perceived control over external
events, but to a lesser extent. The trend towards increased perceived control over
external events is consistent with ACT and exposure therapies in general, given the
emphasis on decreasing avoidance of situations and engaging in valued actions.
On initial consideration results from the ACQ reactions to internal experiences
subscale appear to be contradictory to the focus and proposed mechanisms of ACT. For
example, metaphors in ACT, such as the Polygraph Metaphor (Hayes et al., 1999, p.
123), emphasize the futility of attempts to control internal experiences and suggest that
control of anxiety is the problem, not the solution. The current study found that
participants reported an increase in perceived control over internal experiences from pre-
to post-treatment, which seemingly contradicts the focus of ACT. One possible
explanation for this finding is measurement error, as the ACQ is a relatively new measure
and few studies have been conducted on its psychometric properties. For example, a
study by Zebb and Moore (1999) was unable to replicate the two-factor structure of this
measure, and instead found three factors: internal sense of control, lack of helplessness
68
over internal events, and lack of helplessness over external events. Therefore, the change
obtained in the present study could be more reflective of a decrease in helplessness over
internal events rather than an increase in control over anxiety per se. This would be
consistent with the ACT model because participants may experience a decrease in
helplessness over internal events as their focus shifts from attempting to change internal
events to attempting to change external events (i.e., engaging in valued action without
attempting to control anxiety first).
It is important to note that the ACQ may actually be assessing coping more
generally. Upon closer examination of ACQ reaction subscale items, many of these
items reflect the degree to which an individual generally copes with anxiety symptoms,
and not necessarily the degree to which one “controls” them. For example, one item is:
“I am unconcerned if I become anxious in a difficult situation, because I am confident in
my ability to cope with my symptoms.” Therefore, the significant results obtained from
this subscale could also be consistent with ACT as participants could have been
describing acceptance-based coping on the ACQ. However, it is impossible to discern
whether this is the case based on the item composition of the ACQ, as it does not directly
assess specific coping strategies for anxiety. Finally, as noted with the TCQ, the ACQ
has demonstrated some problems with construct validity (Zebb & Moore, 1999), which
could serve as another possible explanation for these findings.
Finally, the clinical significance results lend limited support to Hypothesis #1, in
that approximately half of the participants achieved clinically significant change on the
SPAI-SP; these results are comparable to other studies that have investigated traditional
CBT for SAD. For example, Heimberg et al. (1990) found that 65% of participants
69
showed clinically significant change (defined as 2 or more standard deviation
improvement) at 6-month follow-up. In addition, Herbert et al. (2005) found that 79% of
participants responded to 12 weeks of CBGT plus social skills training (defined as at least
one standard deviation improvement from pre- to post-treatment). Unlike these two
studies, the current study defined clinical significance as the post-treatment score of a
particular individual falling closer to the mean of the functional population, which is a
more conservative estimate (Jacobson & Truax, 1991). However, when reliable change
was taken into consideration in the present study, only 37.5% of the 16 completers
included in this analysis met criteria for both clinical significance and reliable change.
Despite this, there is a trend towards clinically significant change comparable to studies
examining traditional CBT for SAD, suggesting that the treatment delivered in the
current study is potentially efficacious for generalized SAD and deserves further
investigation.
4.2.3. Hypothesis #3. It was also predicted that experiential avoidance and
believability in negative cognitions would be significantly decreased at post-treatment,
and that these two variables would be associated with treatment outcome. Some support
for this hypothesis was found, in that both experiential avoidance (measured by the AAQ)
and believability in negative cognitions (as measured by the ATQ) decreased
significantly from pre- to post-treatment. However, results from the correlation analyses
showed that a decrease in experiential avoidance was associated with a decrease in social
anxiety symptoms, but a decrease in believability and frequency of negative cognitions
were not. In addition, pre- to post-treatment change in experiential avoidance was
70
associated with a pre- to post-treatment change in quality of life, although changes in
believability and frequency of negative thoughts were not.
These results partially replicate findings from other studies that have examined
experiential avoidance and believability and frequency of negative cognitions. For
example, Zettle (2003) also found a significant pre- to post-treatment reduction in
experiential avoidance for individuals with mathematics anxiety; in addition, it was found
that pre-treatment levels of experiential avoidance were related to therapeutic change for
the ACT treatment group. Regarding believability and frequency of negative cognitions,
results from the current study are similar to results from a study comparing ACT and CT
for depression (Zettle & Hayes, 1986), which found that participants in the ACT group
showed significant reductions in both believability and frequency of negative thoughts.
These results are also similar to previous studies using ACT with seriously mentally ill
patients, which found a decrease in believability and frequency of hallucinations over
time within the ACT treatment group (Bach & Hayes, 2002; Gaudiano & Herbert, in
press(a)). Based on results from mediational analyses conducted by Gaudiano & Herbert
(in press(b)), it was also expected in the current study that change in believability, but not
frequency, would predict change in treatment outcome; this was not the case. However,
in descriptively examining the magnitude of change in believability and frequency, it was
found that the magnitude of change in frequency of negative thoughts was less than that
of believability across all three time points. This trend is consistent with results from
Gaudiano & Herbert, and should continue to be examined in future research.
Taken together, the findings from the current study showed that the ACT-related
process of experiential avoidance changed during the course of treatment and was
71
associated with treatment outcome, while the symptom-focused variable of frequency of
negative thoughts was not associated with treatment outcome. Although it is unknown
whether these changes were related to ACT specifically or exposure therapy in general,
these findings indicate that this treatment is worthy of further investigation, especially to
examine mechanisms of action compared to traditional CBT.
4.2.4. Hypothesis #4. Although a comparison group was not included in the
current study, effect sizes were used to compare the results to those of other studies that
have utilized CBT for SAD. Effect sizes for the SPAI-SP, Brief FNE, and QOLI in the
current study were very large. It was found that the effect sizes obtained from the current
study were not significantly different from effect sizes obtained in other recent studies
that have used state-of-the-art CBT programs to treat SAD. This suggests that ACT has
the potential to be at least as efficacious as CBT for SAD, and that the treatment used in
the current study is worth further investigation.
Based on the design of the current study, it is not possible to determine which
components of the treatment accounted for the treatment effects. For example, the effects
obtained in the current study could have been a result of the ACT components
(acceptance, defusion, values clarification), exposure, nonspecific effects (novelty effect,
effort justification), or some combination of the above. ACT is conceptualized as a
behavioral/cognitive-behavioral treatment, and has been termed a “third-wave behavior
therapy” along with other approaches such as Dialectical Behavior Therapy and
Functional Analytic Psychotherapy (Hayes, 2004). Even though ACT does not utilize
formal cognitive restructuring, effect sizes were comparable to other studies of traditional
CBT. These findings are consistent with those from previous meta-analyses (Gould et
72
al., 1997; Taylor, 1996), which found no clear advantage in effect sizes of cognitive
therapy plus exposure compared to exposure alone. In addition, the dismantling study by
Hope, Heimberg, and Bruch (1995) found no clear advantage of cognitive restructuring
plus exposure over exposure alone. Therefore, although results from the behavioral
treatment used in the current study are consistent with those obtained from other studies,
future studies should compare ACT, CBT, and exposure therapy alone directly to
determine their relative effects and mechanisms of action.
4.2.5. Hypothesis #5. Finally, Eng et al. (2001) found that although quality of life
improved after 12 weeks of treatment, scores still generally did not approach those of the
functional population, indicating a need for interventions that can further improve quality
of life. In the current study, it was hypothesized that there would be a significant
improvement in quality of life from pre- to post-treatment, and that quality of life would
be associated with treatment outcome. Results showed that quality of life increased
significantly from pre- to post-treatment, and that an increase in quality of life was
significantly associated with a decrease in social anxiety severity, therefore lending
support to the hypothesis. These findings replicate results from Eng et al. (2001), who
also found a significant improvement in quality of life from pre- to post-treatment. The
present study also extended findings from Eng et al. (2001), as it found that an increase in
quality of life was associated with a decrease in social anxiety severity; Eng et al. did not
find an association between quality of life and social anxiety severity at post-treatment.
These results from the QOLI suggest the potential for ACT to increase quality of life, and
perhaps better than in traditional CBT. This is consistent with the values component of
73
ACT, which emphasizes increasing behaviors consistent with chosen values in several
domains, not just social ones (e.g., spirituality, citizenship, work, education).
4.3. Comparison and Contrast to Block (2002)
The current study attempted to extend findings of a previous study conducted by
Block (2002). Both studies were pilot investigations into the potential efficacy of ACT
for social anxiety. However, Block’s investigation was based on public speaking anxiety
in an analogue sample of college undergraduates, whereas the current study investigated
ACT for individuals diagnosed with generalized SAD recruited from the community.
More rigorous assessment procedures were used in the current study, including telephone
screens and diagnostic clinical interviews. Block’s study used self-report questionnaires
to screen participants. Another design difference was that the Block study compared 3
groups (ACT, CBGT, and wait-list control), while the current study examined ACT only
with all participants undergoing a 4-week baseline period. Within the ACT condition of
the Block study, 6 weekly sessions, 1½ hours each, were administered in group format.
Both studies used a treatment that consisted of components described by Hayes et al.
(1999), which included exposure exercises. However, in the current study treatment was
administered for 1 hour weekly, for 12 sessions, and in an individual format. Both
studies included assessments at pre- and post-treatment, but the current study also
included a mid-treatment assessment (after 6 sessions of treatment) to better identify the
timing of changes. Furthermore, assessments in the Block study included self-report
measures and a behavioral performance task, while the current study included self-report
measures, clinician-rated measures (completed by independent evaluators), behavioral
assessments, and observer-rated measures obtained from the behavioral assessments.
74
The decision to deviate from the design of the Block (2002) study was based on
the current study’s aims to provide preliminary data on ACT for generalized SAD, to
determine if it is comparable to other studies on CBT for SAD, and to determine if it
deserves further investigation. In order to accomplish these aims, the internal validity
was increased by using diagnostic clinical interviews and multi-modal assessments, as
well as by assessing treatment integrity. In addition, the format of the treatment in the
current study consisted of 12 sessions to make it comparable to other CBT for SAD
studies, thereby affording a direct comparison of effect sizes. Results from the current
study showed that many of the measures did not change from pre- to mid-treatment (6
weeks), indicating the importance of lengthening the treatment to 12 weeks, compared to
the 6-week protocol of the Block study.
Results from the current study were comparable to those from Block (2002). In
addition, the current study extended findings from Block by showing a significant
decrease in other self-report outcome and process measures, such as depression,
experiential avoidance, impairment, and discrepancy between stated values and
consistent action. Unlike the Block study, the current study also showed significant
increases in quality of life and observer-rated social skills. Finally, findings were
extended in the present study by conducting clinical significance analyses and comparing
effect sizes between the current study and other studies using CBT for SAD. The effect
sizes were also compared to those of Block (2002).
The current study also replicated analyses from Block (2002) on the relationship
between the SISST and treatment outcome, and found that a decrease in negative self-
statements, but not an increase in positive self-statements, was associated with a decrease
75
in social anxiety symptoms. This is similar to results from Block, who found that a
decrease in negative self-statements was related to improvement, as well as an increase in
positive self-statements, but to a lesser extent.
Finally, the current study extended findings from Block (2002) by assessing the
relationship between treatment outcome and experiential avoidance, quality of life, and
believability and frequency of negative thoughts. As described previously, a decrease in
experiential avoidance and an increase in quality of life were associated with a decrease
in social anxiety severity, but decreases in believability and frequency of negative
thoughts were not associated with treatment outcome.
4.4. Limitations
The current study possessed several strengths, such as multi-modal assessments,
independent evaluators, and treatment integrity checks. However, potential limitations of
the current study to consider when interpreting the results include small sample size, lack
of a comparison or control group, composition of the sample, and non-blind assessors.
Despite recruitment efforts, sample size of the current study was small, therefore
limiting the generalizability of the results. Although a priori power analysis indicated
that a greater number of participants than actually obtained would be needed for large
effect sizes from pre- to post-treatment, significant results were obtained on almost every
measure with the more modest sample. In addition, results from the current study were
comparable to other studies using CBT for SAD that had larger sample sizes. Therefore,
while the small sample size limits the extent to which the results can be generalized, the
current sample was large enough to detect differences from pre- to post-treatment that
were comparable to effects found in other studies.
76
Another potential limitation of the current study was the lack of a comparison or
wait-list control group. Regarding the lack of a wait-list control group, participants
underwent a 4-week baseline period before beginning treatment. Results showed no
significant differences on all but one measure from baseline to pretreatment, indicating
that spontaneous recovery is an unlikely explanation for improvement, consistent with
previous findings (e.g., Davidson et al., 1994). In addition to lack of a waitlist condition,
the current study did not include a comparison condition as in Block (2002). Therefore,
it is not known whether the results from the current study can be attributed specifically to
ACT, nor how these results would compare to traditional CBT protocols in a head-to-
head comparative trial. However, to partially account for the lack of a comparison
condition, effect sizes from the current study were compared to other studies utilizing
CBT for SAD. Results showed that the effect sizes from the current study were not
significantly different from other studies, indicating that this treatment is worthy of
further investigation.
Although the sample from the current study was drawn from the community
rather than a college sample, the composition of the sample is worth considering when
interpreting the results. For example, a majority of the sample was Caucasian, single,
employed full-time, and highly educated (at least some college education or higher).
This differs somewhat from the demographic composition of epidemiological samples, in
which social phobia has been associated with lower education, single persons,
unemployed persons, and students (Magee et al., 1996). Therefore, this may limit the
external validity of the study. However, the sample composition of the current study is
77
similar to that of treatment-seeking samples of previous studies (Heimberg et al., 1998;
Herbert et al., 2004 & 2005).
Finally, a potential limitation of the current study is that the independent
evaluators who completed the clinician-rated measures were not blind to assessment time
point. However, the present study used several modes of assessment, including self-
report measures, clinician-rated measures, and behavioral assessments. Results across
these different assessment strategies consistently showed improvement over the course of
treatment, thereby ruling out rater bias as a likely explanation for the improvement on the
clinician-ratings.
4.5. Implications and Future Directions
The current study replicated results from Block (2002) in a sample of adults with
generalized SAD. In both studies, participants reported a decrease in social anxiety
symptoms and behavioral avoidance, as well as an increase in willingness to engage in
social situations. Results from the present study were also comparable to effects from
other studies on CBT for SAD. However, the current study was not designed to
determine whether ACT is specifically efficacious for generalized SAD as no comparison
group was included. Therefore, non-specific factors (e.g., support, novelty effects)
cannot be ruled out as an explanation for improvement. Block (2002) provided some
preliminary support for the benefit of ACT with exposure for public speaking anxiety,
and the current study extended these results by providing preliminary support for the
benefit of ACT plus exposure for generalized SAD. Due to the small sample size and
lack of comparison group, future trials should be conducted in larger samples to
78
systematically compare CBT, ACT, and exposure alone to determine the relative
contribution of each.
It is possible that CBT and ACT produce similar outcomes and work through the
same mechanisms of action. Although CT alone does result in significant improvement
in SAD, it is still unclear what mechanism is accounting for the effect. Cognitive therapy
proposes thought modification as the mechanism of change (Beck, Emery, & Greenberg,
1985), but perhaps other variables, such as “metacognitive awareness,” are responsible
for treatment effects. Metacognitive awareness has been defined as experiencing
thoughts and feelings as transient mental events, rather than as part of the self (Teasdale
et al., 2002). Therefore, the act of recording one’s thoughts in cognitive restructuring
may aid in separating oneself from one’s thoughts, thus leading to decreased believability
in negative thoughts. There has been some preliminary support for metacognitive
awareness as a mechanism of change in the prevention of relapse in depression. For
example, Teasdale et al. (2002) found that both CT and a mindfulness-based CT resulted
in increased metacognitive awareness. As ACT also attempts to achieve greater
metacognitive awareness, it is possible that similar results obtained by ACT and CBT
might be due to the same mechanism of action (e.g., metacognitive awareness). Future
research needs to be conducted in this area to determine the degree to which
metacognitive awareness mediates treatment outcome both in ACT and traditional CBT.
However, based on their differing proposed mechanisms of action, it is also
possible that a comparative study on CBT and ACT would yield similar results, but the
interventions may work through different mechanisms of action. There has been some
preliminary support for this in previous research on ACT. For example, Block (2002)
79
found that the ACT group showed a greater decrease in behavioral avoidance compared
to the CBT group. In addition, the study by Zettle (2003) on ACT versus systematic
desensitization for mathematics anxiety found that experiential avoidance was related to
therapeutic change for the ACT group but not for the systematic desensitization group.
Finally, a study comparing an acceptance-based versus cognitive-control based
intervention for pain tolerance found that the ACT participants showed significantly
higher tolerance to pain compared to the cognitive control group, and the cognitive
control group showed greater reductions in self-reported pain compared to the ACT
participants (Gutierrez, Luciano, Rodriguez, & Fink, 2004).
In the current study, participants reported a significant reduction in social anxiety
fears (e.g., on the SPAI-SP) and frequency of negative thoughts, similar to individuals
who receive CBT for SAD. Therefore, although many of the results from the current
study were comparable to other studies that have examined CBT, some results from the
present study are particularly consistent with ACT and suggest a need for further research
to examine mechanisms of action between ACT and CBT. For example, results on the
LSAS suggested a higher avoidance subscale effect size (although not significantly
larger) than the fear subscale (.98 vs. .59). Furthermore, change on the LSAS avoidance
subscale was achieved earlier than on the LSAS fear subscale. This is consistent with
ACT given that it does not directly target fear reduction itself, but avoidance behaviors
instead, including those related to both internal experiences and external events. In
addition, results from the TCQ support this hypothesis, as the only significant change in
use of thought control strategies was the strategy most consistent with exposure therapies
(i.e., the degree to which one expresses unpleasant thoughts to others). This also is
80
consistent with the ACT model because of its emphasis on control of internal experiences
(e.g., thoughts, feelings, memories) as part of the problem rather than the solution (Hayes
et al., 1999). Although these preliminary findings seem consistent with ACT, the current
study did not specifically examine mediating effects of ACT-relevant variables.
Therefore, future studies are needed to compare CBT and ACT for generalized SAD and
to examine potential mediators of change for each intervention.
There also exist alternative explanations for the treatment effects obtained in the
present study. For example, results could be novelty and expectancy effects, which have
been shown to be significant predictors of treatment outcome in CBT for SAD
(Chambless, Tran, & Glass, 1997; Safren, Heimberg, & Juster, 1997). It also is unclear
whether ACT facilitates exposure better than simply exposure alone. Eifert & Heffnor
(2003) hypothesize that an increase in acceptance of fear and willingness to engage in
social situations despite the fear itself may increase receptiveness to exposure treatment
better than cognitive restructuring. Alternatively, cognitive restructuring attempts to alter
the content of thoughts to decrease anxiety to facilitate social engagement in order to
dispute dysfunctional beliefs (Beck, Emery, & Greenberg, 1985). ACT may have the
ability to better facilitate exposure via its unique use of values clarification and linking
behavior to personally-identified values. Although other researchers are beginning to
utilize traditional cognitive restructuring less and instead emphasize decreasing self-
focused attention and increasing attention towards the social situation (Clark et al., 2003),
these other approaches still do not link the desired behavior to the process of engaging in
valued action; rather, the behavioral change is emphasized in the context of
anxiety/symptom reduction. Therefore, ACT, with its emphasis on engaging in behavior
81
change in order to be consistent with values rather than to reduce symptoms, has the
potential to result in greater functional improvement in the long-term.
The above questions of whether ACT would produce different outcomes from
CBT, whether it would work through different mechanisms of action, and whether it
would better facilitate exposure and improve functional outcome all highlight the need
for future research directly comparing these interventions. For example, future research
should directly compare ACT plus exposure, CT plus exposure, and exposure alone to
better determine the relative treatment effects of each. In addition, several process
variables should be measured to examine potential mechanisms of change, such as
metacognitive awareness, experiential avoidance, believability and frequency of negative
thoughts, and anxiety/thought control strategies.
Cognitive behavior therapy is deemed an empirically supported treatment for
generalized SAD; however, a significant percentage of participants still do not respond to
treatment (Heimberg et al., 1998; Herbert et al., 2005). Therefore, there is a need for
other interventions that can decrease social anxiety symptoms and increase quality of life,
especially for those who may not respond to traditional CBT. The current study was a
pilot study that supported the efficacy of ACT for generalized SAD. However, the
current study did not address the specific efficacy of ACT for SAD relative to established
treatments. The benefits obtained by participants from the present study suggest that
ACT is worthy of further investigation and should be compared directly to traditional
CBT for SAD.
82
LIST OF REFERENCES
American Psychiatric Association (2000). Diagnostic and Statistical Manual of
Mental Disorders Text Revision (4th ed.). Washington, DC: Author. Arrindell, W.A., Kwee, M.G.T., Methorst, G.J., Van der Ende, J., Pol, E., & Moritz,
B.J.M. (1989). Perceived parenting styles of agoraphobic and socially phobic in-patients. British Journal of Psychiatry, 155, 526-535.
Bach, P. & Hayes, S.C. (2002). The use of Acceptance and Commitment Therapy to
prevent the rehospitalization of psychotic patients: A randomized controlled trial. Journal of Consulting and Clinical Psychology, 70, 1129-1139.
Baldwin, D., Bobes, J., Stein, D., Scharwaechter, I., & Faure, M. (1999). Paroxetine in
social phobia/social anxiety disorder: Randomized, double-blind, placebo-controlled study. British Journal of Psychiatry, 175, 120-126.
Batchelor, S. (1997). Buddhism without beliefs: A contemporary guide to awakening.
New York: Riverhead Books. Beck, A.T., Emery, G., & Greenberg, R.L. (1985). Anxiety Disorders and Phobias: A
Cognitive Perspective. New York: Basic Books. Beck, A.T., Rush, A.J., Shaw, B.F., & Emery, G. (1979). Cognitive Therapy of
Depression. New York: Guildford Press. Beck, A.T., Steer, R.A., Ball, R., & Ranieri, W.F. (1996). Comparison of Beck
Depression Inventories IA and II in psychiatric outpatients. Journal of Personality Assessment, 67, 588-597.
Beck, A.T., Steer, R.A., & Brown, G.K. (1996). Manual for the Beck Depression
Inventory (2nd edition). San Antonio, TX: Psychological Corporation. Beck, A. T., Steer, R. A., & Garbin, R. A. (1988). Psychometric properties of the
Beck Depression Inventory: Twenty-five years of evaluation. Clinical Psychology Review, 8, 77-100.
Beck, A. T., & Steer, R. A. (1987). Manual for the Beck Depression Inventory.
San Antonio, TX: Psychological Corporation. Beidel, D. C., Borden, J. W., Turner, S. M., & Jacob, R. G. (1989). The Social
Phobia and Anxiety Inventory: Concurrent validity with a clinic sample. Behaviour Research and Therapy, 27, 573-576.
83
Beidel, D. C., Turner, S. M., Stanley, M. A., & Dancu, C. V. (1989). The Social
Phobia and Anxiety Inventory: Concurrent and external validity. Behavior Therapy, 20, 417-427.
Blanco, C., Schneier, F.R., Schmidt, A., Blanco-Jerez, C.R., Marshall, R.D.,
Sanchez-Lacay, A., et al. (2003). Pharmacological treatment of social anxiety disorder: A meta-analysis. Depression and Anxiety, 18, 29-40.
Block, J.A. (2002). Acceptance or change of private experiences: A comparative
analysis in college students with public speaking anxiety. Doctoral dissertation. University at Albany, State University of New York.
Block, J.A., & Wulfert, E. (2000). Acceptance and change: Treating socially
anxious college students with ACT or CBGT. Behavior Analyst Today, 1, 3-10. Bond, F.W. & Bunce, D. (2000). Mediators of change in emotion-focused and
problem-focused worksite stress management interventions. Journal of Occupational Health Psychology, 5, 156-163.
Bond, F.W. & Bunce, D. (2003). The role of acceptance and job control in mental health,
job satisfaction, and work performance. Journal of Applied Psychology, 88, 1057-1067.
Borkovec, T.D., Alcaine, O., & Behar, E. (in press). Avoidance theory of worry
and generalized anxiety disorder. In R.G. Heimberg, C.L. Turk, & D.S. Mennin (Eds.), Generalized Anxiety Disorder: Advances in Research and Practice. New York: Guilford.
Borkovec, T.D., & Roemer, L. (1994). Generalized anxiety disorder. In R.T.
Ammerman & M. Hersen (Eds.), Handbook of Prescriptive Treatments for Adults (pp. 261-281). New York: Plenum.
Bouwer, C., & Stein, D.J. (1998). Use of the selective serotonin reuptake inhibitor
citalopram in the treatment of generalized social phobia. Journal of Affective Disorders, 49, 79-82.
Brown, T. A., Di Nardo, P. A., & Barlow, D. H. (1994). Anxiety Disorders
Interview Schedule for DSM-IV (ADIS-IV). Albany, NY: Graywind. Bruch, M.A., & Heimberg, R.G. (1994). Differences in perceptions of parental and
personal characteristics between generalized and nongeneralized social phobics. Journal of Anxiety Disorders, 8(2), 155-168.
Carrascoso Lopez, F.J. (2000). Acceptance and commitment therapy (ACT) in
panic disorder with agoraphobia: A case study. Psychology in Spain, 4, 120-128.
84
Chambless, D.L., Sanderson, W.C., Shoham, V., Johnson, S.B., Pope, K.S.,
Crits-Christoph, P., Baker, M., Johnson, B., Woody, S., Sue, S., Beutler, L., Williams, D.A., & McCurry, S. (1996). An update on empirically validated treatments. The Clinical Psychologist, 49, 5-18.
Chambless, D.L., Tran, G.Q., & Glass, C.R. (1997). Predictors of response to cognitive-
behavioral group therapy for social phobia. Journal of Anxiety Disorders, 11, 221-240.
Clark, D.M., Ehlers, A., McManus, F., Hackmann, A., Fennell, M., Campbell, H.,
Flower, T., Davenport, C., & Louis, B. (2003). Cognitive therapy versus fluoxetine in generalized social phobia: A randomized placebo-controlled trial. Journal of Consulting and Clinical Psychology, 71, 1058-1067.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.).
Hillsdale, NJ: Erlbaum. Cox, B. J., Parker, J. D. A., & Swinson, R. P. (1996). Confirmatory factor
analysis of the Fear Questionnaire with social phobia patients. British Journal of Psychiatry, 168, 497-499.
Cox, B.J., Ross, L., Swinson, R.P., & Direnfeld, D.M. (1998). A comparison of social
phobia outcome measures in cognitive-behavioral group therapy. Behavior Modification, 22, 285-297.
Cox, B. J., Swinson, R. P., & Parker, J. D. A. (1993). Confirmatory factor
analysis of the Fear Questionnaire in panic disorder with agoraphobia. Psychological Assessment, 5, 235-237.
Cox, B. J., Swinson, R. P., & Shaw, B. F. (1991). Value of the Fear
Questionnaire in differentiating agoraphobia and social phobia. British Journal of Psychiatry, 159, 842-845.
Craske, M.G., Barlow, D.H., & O’Leary, T. (1992). Mastery of Your Anxiety and
Worry. Albany, NY: Graywind. Davidson, J.R.T., Foa, E.B., Huppert, J.D., Keefe, F.J., Franklin, M.E., Compton, J.S.,
Zhao, N., Connor, K.M., Lynch, T.R., & Gadde, K.M. (2004). Fluoxetine, comprehensive cognitive behavioral therapy, and placebo in generalized social phobia. Archives of General Psychiatry, 61, 1005-1013.
Davidson, R., Hughes, D., George, L., & Blazer, D. (1994). The boundary of social
phobia: Exploring the threshold. Archives of General Psychiatry, 51, 975-983.
85
Davidson, J.R.T., Potts, N.L.S., Richichi, E.A., Krishnan, R.R., Ford, S.M., Smith, R., & Wilson, W.H. (1993). Treatment of social phobia with clonazepam and placebo. Journal of Clinical Psychopharmacology, 13, 423-428.
Deardorff, P.A., Hopkins, L.R., & Finch, A.J. (1984). Automatic Thoughts
Questionnaire: A reliability and validity study. Psychological Reports, 55, 708-710.
DiNardo, P.A., Brown, T.A., Lawton, J.K., & Barlow, D.H. (1995, November). The
Anxiety Disorders Interview Schedule for DSM-IV Lifetime Version: Description and initial evidence for diagnostic reliability. Paper presented at the 29th Annual Convention of the Association for the Advancement of Behavior Therapy, Washington, DC.
DiNardo, P.A., Moras, K., Barlow, D.H., Rapee, R.M., & Brown, T.A. (1993).
Reliability of DSM-III-R anxiety disorder categories using the Anxiety Disorders Interview Schedule-Revised (ADIS-R). Archives of General Psychiatry, 50, 251-256.
Dodge, C.S., Hope, D.A., Heimberg, R.G., & Becker, R.E. (1988). Evaluation of the
Social Interaction Self-Statement Test with a social phobic population. Cognitive Therapy and Research, 12, 211-222.
Eng, W., Coles, M.E., Heimberg, R.G., & Safren, S.A. (2001). Quality of life
following cognitive behavioral treatment for social anxiety disorder: Preliminary findings. Depression and Anxiety,13, 192-193.
Eifert, G.H., & Heffner, M. (2003). The effects of acceptance versus control contexts on
avoidance of panic-related symptoms. Journal of Behavior Therapy and Experimental Psychiatry, 34, 293-312.
Faul, F., & Erdfelder, E. (1992). GPOWER: A priori, post hoc, and compromise
power analyses for MS-DOS [Computer Program]. Bonn, Germany: Bonn University, Department of Psychology.
Federal Drug Administration (2003a). FDA approved labeling of Zoloft for the treatment
of Social Anxiety Disorder. Retrieved September 4, 2005, from http://www.fda.gov/cder/foi/label/2003/19839se1-045_Zoloft_lbl.pdf
Federal Drug Administration (2003b). FDA approved labeling of Effexor XR. Retrieved
September 4, 2005, from http://www.fda.gov/cder/foi/label/2003/20699se1-022_Effexor_lbl.pdf
Feske, U., & Chambless, D.L. (1995). Cognitive behavioral versus exposure only
treatment for social phobia: A meta-analysis. Behavior Therapy, 26, 695-720.
86
First, M.B., Spitzer, R.L., Gibbon, M., Williams, J.B.W., & Benjamin, L. (1994). Structured Clinical Interview for DSM-IV Axis II Personality Disorders (SCID-II, Ver. 2.0). New York: New York State Psychiatric Institute, Biometrics Research Department.
First, M. B., Spitzer, R. L., Gibbon, M., & Williams, J. B. W. (1996). Structured
Clinical Interview for DSM-IV Axis I Disorders - Patient Edition (SCID-I/P, Version 2.0). New York: Biometrics Research Department.
First, M.B., Spitzer, R.L., Gibbon, M., Williams, J.B.W., Davies, M., Borus, J.,
Howes, M.J., Kane, J., Pope, H.G., & Rounsaville, B. (1995). The Structured Clinical Interview for DSM-III-R Personality Disorders (SCID-II): Part II Multisite test-retest reliability study. Journal of Personality Disorders, 9, 92-104.
Foa, E.B., Davidson, J.R.T., Huppert, J.D., Franklin, M.E., Roth, D.A., & Keefe,
F.J. (2003, November). Comprehensive CBT, fluoxetine, and their combination: A randomized, placebo-controlled trial. In D.A. Roth (Ed.), Cognitive behavior therapy for social phobia: Recent findings and future directions. Symposium conducted at the meeting of the Association for the Advancement of Behavior Therapy, Boston.
Frisch, M.B. (1994). Quality of Life Inventory: Manual and treatment guide.
Minneapolis, MN: National Computer Systems. Frisch, M.B., Cornell, J., Villanueva, M., & Retzlaff, P.J. (1992). Clinical
validation of the Quality of Life Inventory: A measure of life satisfaction for use in treatment planning and outcome assessment. Psychological Assessment, 4, 92-101.
Gaudiano, B.A., & Herbert, J.D. (in press(a)). Acute treatment of inpatients with
psychotic symptoms using Acceptance and Commitment Therapy: Pilot results. Behaviour Research and Therapy.
Gaudiano, B.A., & Herbert, J.D. (in press(b)). Believability of hallucinations as a
potential mediator of the relationship between their frequency and associated distress in psychotic inpatients. Behavioural and Cognitive Psychotherapy.
Geiser, D.S. (1992). A comparison of acceptance-focused and control-focused
psychological treatments in a chronic pain treatment center. Doctoral dissertation. University of Nevada, Reno. Dissertation Abstracts International-B 54/02, p. 1096, Aug. 1993.
Gelernter, C.S., Uhde, T.W., Cimbolic, P., Arnkoff, D.B., Vittone, B.J., Tancer,
M.E., & Bartko, J.J. (1991). Cognitive-behavioral and pharmacological treatment for social phobia: A controlled study. Archives of General Psychiatry, 48, 938-945.
87
Gifford, E. (2002). Acceptance and Commitment Therapy versus Nicotine
Replacement Therapy as methods of smoking cessation. Unpublished doctoral dissertation, University of Nevada, Reno.
Gillis, M.M., Haaga, D.A., & Ford, G.T. (1995). Normative values for the Beck Anxiety
Inventory, Fear Questionnaire, Penn State Worry Questionnaire, and Social Phobia and Anxiety Inventory. Psychological Assessment, 7, 450-455.
Glass, C.R., Merluzzi, T.V., Biever, J.L., & Larsen, K.H. (1982). Cognitive assessment
of social anxiety: Development and validation of a self-statement questionnaire. Cognitive Therapy and Research, 6, 37-55.
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.
Groom J.M., & Wilson, K.G. (2003, May). Examination of the psychometric properties
of the Valued Living Questionnaire (VLQ): A tool of Acceptance and Commitment Therapy (ACT). Paper presented at the annual meeting of the Association for Behavior Analysis, San Francisco.
Gutierrez, O., Luciano, C., Rodriguez, M., & Fink, B.C. (2004). Comparison between an
acceptance-based and a cognitive-control-based protocol for coping with pain. Behavior Therapy, 35, 767-783.
Harrell, T.H., & Ryon, N.B. (1983). Cognitive-behavioral assessment of
depression: Clinical validation of the Automatic Thoughts Questionnaire. Journal of Consulting and Clinical Psychology, 51, 721-725.
Hart, W. (1987). The art of living: Vipassana meditation as taught by S.N.
Goenka. San Francisco: Harper. Haug, T. T., Blomhoff, S., Hellstrøm, K., Holme, I., Humble, M., Madsbu, H. P., &
Wold, J. E. (2003). Exposure therapy and sertraline in social phobia: 1-year follow-up of a randomized controlled trial. British Journal of Psychiatry, 182, 312-318.
Hayes, S.C. (2005). ACT state of the evidence: Creating a psychology more adequate to
the challenge of the human condition. Symposium presented at the ACT Summer Institute II, Philadelphia, PA.
Hayes, S.C. (2004). Acceptance and commitment therapy, relational frame theory, and
the third wave of behavioral and cognitive therapies. Behavior Therapy, 35, 639-665.
88
Hayes, S.C. (2002). Buddhism and acceptance and commitment therapy. Cognitive and Behavioral Practice, 9, 58-66.
Hayes, S.C. (1987). A contextual approach to therapeutic change. In N. Jacobson (Ed.),
Psychotherapists in Clinical Practice: Cognitive and Behavioral Perspectives (pp.327-387). New York: Guilford Press.
Hayes, S.C., Barnes-Holmes, D., & Roche, B. (2001). Relational Frame Theory: A
Post-Skinnerian Account of Human Language and Cognition. New York, NY: Kluwer.
Hayes, S.C., Bissett, R.T., Korn, Z., Zettle, R.D., Rosenfarb, I.S., Cooper, L.D., &
Grundt, A.M. (1999). The impact of acceptance versus control rationales on pain tolerance. The Psychological Record, 49, 33-47.
Hayes, S.C., Bissett, R.T., Strosahl, K., Wilson, K., Pistorello, J., Toarmino, D.,
Batten, S.V., Dykstra, T.A., Stewart, S.H., Zvolensky, M.J., Eifert, G.H., Bond, F.W., & Bergan, J. (2002). Psychometric properties of the Acceptance and Action Questionnaire (AAQ). Manuscript submitted for publication.
Hayes, S.C., Masuda, A., Bissett, R., Luoma, J., & Guerrero, L.F. (2004). DBT,
FAP, and ACT: How empirically oriented are the new behavior therapy technologies? Behavior Therapy, 35, 35-54.
Hayes, S. C., Strosahl, K. D., & Wilson, K. G. (1999). Acceptance and
commitment therapy: An experiential approach to behavior change. New York: Guilford.
Hayes, S.C., Wilson, K.G., Afari, N., & McCurry, S. (2003, November). The use of
Acceptance and Commitment Therapy in the treatment of agoraphobia. Paper presented at the meeting of the Association for the Advancement of Behavior Therapy, Boston, MA.
Hayes, S.C., Wilson, K.G., Gifford, E., Bissett, R., Batten, S., Piasecki, M., Byrd,
M., & Gregg, J. (May 2002). The use of Acceptance and Commitment Therapy and 12-step facilitation in the treatment of polysubstance abusing heroin addicts on methadone maintenance: A randomized controlled trial. Paper presented at the meeting of the Association for Behavior Analysis, Toronto.
Heimberg, R.G. (2002, November). The understanding and treatment of social
anxiety: What a long strange trip it’s been (and will be). Paper presented at the 36th Annual Convention of the Association for the Advancement of Behavior Therapy, Reno, NV.
89
Heimberg, R. G. (1991). Cognitive-behavioral group therapy for social phobia: A treatment manual. Unpublished manuscript, The University at Albany, State University of New York.
Heimberg, R. G., & Becker, R. E. (2002). Cognitive-behavioral group therapy for
social phobia: Basic mechanisms and clinical strategies. New York: Guilford. Heimberg, R.G., Dodge, C.S., Hope, D.A., Kennedy, C.R., Zollo, L.J., & Becker,
R.E. (1990). Cognitive behavioral treatment for social phobia: Comparison with a credible placebo control. Cognitive Therapy and Research, 14, 1-23.
Heimberg, R.G., Horner, K.J., Juster, H.R., Safren, S.A., Brown, E.J., Schneier, F.R., &
Liebowitz, M.R. (1999). Psychometric properties of the Liebowitz Social Anxiety Scale. Psychological Medicine, 29, 199-212.
Heimberg, R. G., & Juster, H. R. (1995). Cognitive-behavioral treatments:
Literature review. In R. G. Heimberg, M. R. Liebowitz, D. A. Hope, & F. R. Schneier (Eds.), Social phobia: Diagnosis, assessment, and treatment (pp. 261-309). New York: Guilford.
Heimberg, R.G., Liebowitz, M.R., Hope, D.A., Schneier, F.R., Holt, C.S.,
Welkowitz, L.A., Juster, H.R., Campeas, R., Bruch, M.A., Cloitre, M., Falloon, B., & Klein, D.F. (1998). Cognitive behavioral group therapy vs. phenelzine therapy for social phobia. Archives of General Psychiatry, 55, 1133-1141.
Heimberg, R.G., Salzman, D.G., Holt, C.S., & Blendell, K.A. (1993). Cognitive
behavioral group treatment for social phobia: Effectiveness at five-year follow-up. Cognitive Therapy and Research, 17, 325-339.
Herbert, J. D. (in press). Avoidant Personality Disorder. In W. O’Donohue, K. A.
Fowler, & S. O. Lilienfeld (Eds.), Handbook of personality disorders. Sage. Herbert, J. D. (2002). Review of S. C. Hayes, K. D. Strosahl, & K. G. Wilson,
Acceptance and Commitment Therapy: An experiential approach to behavior change. Cognitive and Beahvioral Practice, 9, 164-166.
Herbert, J. D., Bellack, A. S., & Hope, D. A. (1991). Concurrent validity of the
Social Phobia and Anxiety Inventory. Journal of Psychopathology and Behavioral Assessment, 14, 357-368.
Herbert, J.D., Hope, D.A., & Bellack, A.S. (1992). Validity of the distinction
between generalized social phobia and avoidant personality disorder. Journal of Abnormal Psychology, 104, 332-339.
Herbert, J. D., & Dalrymple, K. (in press). Social anxiety disorder. Encyclopedia
of Cognitive Behavior Therapy. Norwell, MA: Kluwer.
90
Herbert, J. D., Gaudiano, B. A., Rheingold, A. A., Myers, V. H., Dalrymple, K., &
Nolan, B. M. (2005). Social skills training augments the effectiveness of cognitive behavioral group therapy for social anxiety disorder. Behavior Therapy, 36, 125-138.
Herbert, J. D., Rheingold, A. A., & Brandsma, L. L. (2001). Assessment of social
anxiety and social phobia. In S. G. Hofmann & P. M. DiBartolo (Eds.), From social anxiety to social phobia: Multiple perspectives. Allyn & Bacon.
Herbert, J.D., Rheingold, A.A., Gaudiano, B.A., & Myers, V.H. (2004). Standard
versus extended cognitive behavior therapy for social anxiety disorder: A randomized-controlled trial. Behavioural and Cognitive Psychotherapy, 32, 131-147.
Herbert, J. D., Rheingold, A. A., & Goldstein, S. G. (2002). Brief cognitive
behavioral group therapy for social anxiety disorder. Cognitive and Behavioral Practice, 9, 1-8.
Hofmann, S. G., & Barlow, D. H. (2002). Social phobia (social anxiety disorder).
In D. H. Barlow, Anxiety and its disorders (pp. 454-476). New York: Guilford. Hollon, S. D., & Kendall, P. C. (1981). Cognitive self-statements in depression:
Development of an Automatic Thoughts Questionnaire. Cognitive Therapy and Research, 4, 383-395.
Holt, C.S., & Heimberg, R.G. (1990). The Reaction to Treatment Questionnaire:
Measuring treatment credibility and outcome expectancies. The Behavior Therapist, 13, 213-214, 222.
Hope, D.A., Heimberg, R.G., & Bruch, M.A. (1995). Dismantling cognitive-
behavioral group therapy for social phobia. Behavior Research and Therapy, 33, 637-650.
Hope, D.A., Herbert, J.D., & White, C. (1995). Diagnostic subtype, avoidant
personality disorder, and efficacy of cognitive-behavioral group therapy for social phobia. Cognitive Therapy and Research, 19, 399-417.
Huerto Romero, F., Gomez Martin, S., Molina Moreno, A.M., & Luciano Soriano, M.C.
(1998). Ansiedad generalizada: Un estudio de caso [Generalized anxiety: A case study]. Analisis y Modificacion de Conducta, 24, 751-766.
Jacobson, N. S., Christensen, A., Prince, S. E., Cordova, J., & Eldridge, K.
(2000). Integrative behavioral couple therapy: An acceptance-based, promising new treatment for couple disorder. Journal of Consulting and Clinical Psychology, 68, 351-355.
91
Jacobson, N. S., & Truax, P. (1991). Clinical significance: A statistical approach
to defining meaningful change in psychotherapy research. Journal of Consulting and Clinical Psychology, 59, 12-19.
Kagan, J., Reznick, J.S., & Snidman, N. (1988). Biological basis of childhood shyness.
Science, 240, 167-171. Katzelnick, D.J., Kobak, K.A., Greist, J.H., Jefferson, J.W., Mantle, J.M., & Serlin,
R.C. (1995). Sertraline for social phobia: A double-blind, placebo-controlled crossover study. American Journal of Psychiatry, 152, 1368-1371.
Kazdin, A.E. (1998). Research Design in Clinical Psychology (3rd ed.). Needham
Heights, MA: Allyn & Bacon. Kessler, R.C., McGonagle, K.A., Zhao, S., Nelson, C.B., Hughes, M., Eshleman,
S., Wittchen, H.-U., & Kendler, K.S. (1994). Lifetime and 12-month prevalence of DSM-III-R psychiatric disorders in the United States: Results from the National Comorbidity Survey. Archives of General Psychiatry, 51, 8-19.
Kumar, S.M. (2002). An introduction to Buddhism for the cognitive-behavioral
therapist. Cognitive and Behavioral Practice, 9, 40-43. Leary, M.R. (1983). A brief version of the Fear of Negative Evaluation Scale.
Personality and Social Psychology Bulletin, 9, 37-375. Leon, A. C., Olfson, M., Portera, L. Farbert, L., & Sheehan, D. V. (1997).
Assessing psychiatric impairment in primary care with the Sheehan Disability Scale. International Journal of Psychiatry in Medicine, 27, 93-105.
Leon, A.C., Shear, M.K., Portera, L., & Klerman, G.L. (1992). Assessing
impairment in patients with panic disorder: the Sheehan Disability Scale. Social Psychiatry and Psychiatric Epidemiology, 27, 78-82.
Levitt, J.T., Brown, T.A., Orsillo, S.M., & Barlow, D.H. (2004). The effects of
acceptance versus suppression of emotion on subjective and psychophysiological response to carbon dioxide challenge in patients with panic disorder. Behavior Therapy, 35, 747-766.
Liebowitz, M.R. (1987). Social phobia. Modern Problems in Pharmacopsychiatry, 22,
141-173.
Liebowitz, M.R., Heimberg, R.G., Schneier, F.R., Hope, D.A., Davies, S., Holt, C.S., et al. (1999). Cognitive-behavioral group therapy versus phenelzine in social phobia: Long-term outcome. Depression and Anxiety, 10, 89-98.
92
Liebowitz, M.R., Schneier, F., Campeas, R., Hollander, E., Hatterer, J., Fyer, A., Gorman, J., Papp, L., Davies, S., Gully, R., & Klein, D.F. (1992). Phenelzine vs. atenolol in social phobia: A placebo-controlled comparison. Archives of General Psychiatry, 49, 290-300.
Linehan, M.M. (1993). Skills training manual for cognitive behavioral treatment of
borderline personality disorder. New York: Guilford. Linehan, M. M., Armstrong, H. E., Suarez, A., & Allmon, D. (1991). Cognitive
behavioral treatment of chronically parasuicidal borderline patients. Archives of General Psychiatry, 48, 1060-1065.
Linehan, M. M., Heard, H. L., & Armstrong, H. E. (1994). Naturalistsic follow-up
of a behavioral treatment for chronically parasuicidal borderline patients. Archives of General Psychiatry, 51, 22.
Lipsitz, J.D., Mannuzza, S., Klein, D.F., Ross, D.C., & Fyer, A.J. (1999). Specific
phobia 10-16 years after treatment. Depression and Anxiety, 10, 105-111. Lott, M., Greist, J.H., Jefferson, J.W., Kobak, K.A., Katzelnick, D.J., Katz, R.J., &
Schaettle, S.C. (1997). Brofaromine for social phobia: A multicenter, placebo-controlled, double-blind study. Journal of Clinical Psychopharmacology, 17, 255-260.
Lucas, R. A. (1994). A comparative treatment outcome study of social phobia.
Dissertation Abstracts International, 54 (12-B), 6465. Magee, W.J., Eaton, W.W., Wittchen, H-U, McGonagle, K.A., & Kessler, R.C. (1996).
Agoraphobia, simple phobia, and social phobia in the National Comorbidity Survey. Archives of General Psychiatry 53, 159-168.
Marks, I. M., & Mathews, A. M. (1979). Brief standard rating for phobic patients.
Behaviour Research and Therapy, 17, 263-267. Mattick, R., Peters, L., & Clarke, J. (1989). Exposure and cognitive restructuring for
social phobia: A controlled study. Behavior Therapy, 20, 3-23. McNeil, D. W., Ries, B. J., & Turk, C. L. (1995). Behavioral assessment: Self-
report, physiology, and overt behavior. In R. G. Heimberg, M. R. Liebowitz, D. A. Hope, & F. R. Schneier (Eds.), Social phobia: Diagnosis, assessment, and treatment (pp. 202-231). New York: Guilford.
Michelson, L., & Mavissakalian, M. (1983). Temporal stability of self report
measures in agorophobia research. Behaviour Research and Therapy, 21, 695-698.
93
Miller, J.J., Fletcher, K., & Kabat-Zinn, J. (1995). Three-year follow-up and clinical implications of a mindfulness meditation-based stress reduction intervention in the treatment of anxiety disorders. General Hospital Psychiatry, 17, 192-200.
Morris, T.L. (2001). Social phobia. In M.W. Vasey and M.R. Dadds (Eds.), The
Developmental Psychopathology of Anxiety (pp. 435-458). New York: Oxford University Press.
National Institutes of Mental Health (1985). Clinical Global Impressions Scale.
Psychopharmacology Bulletin, 21, 839-843. Norton, P. J., & Hope, D. A. (2001). Kernels of truth or distorted perceptions:
Self and observer ratings of social anxiety and performance. Behavior Therapy, 32, 765-786.
O’Boyle, M., & Self, D. (1990). A comparison of two interviews for DSM-III-R
personality disorders. Psychiatry Research, 32, 85-92. Oei, T. P. S., Moylan, A., & Evans, L. (1991). Validity and clinical utility of the
Fear Questionnaire for anxiety-disorder patients. Psychological Assessment, 3, 391-397.
Orsillo, S.M., Roemer, L., & Barlow, D.H. (2003). Integrating acceptance and
mindfulness into existing cognitive behavioral treatment for GAD: A case study. Cognitive and Behavioral Practice, 10, 223-230.
Orsillo, S., Roemer, L., Block, J., LeJeune, C., & Herbert, J.D. (2004). ACT
with anxiety disorders. In S.C. Hayes & K. Strosahl (Eds.), A Practical Guide to Acceptance and Commitment Therapy (pp. 103-132). Norwell, MA: Plenum Press.
Osman, A., Barrios, F.X., Aukes, D., & Osman, J.R. (1995). Psychometric evaluation of
the Social Phobia and Anxiety Inventory in college students. Journal of Clinical Psychology, 51, 235-243.
Osman, A., Markway, K., & Osman, J.R. (1992). Psychometric properties of the Social
Interaction Self-Statement Test in a college sample. Psychological Reports, 71, 1171-1177.
Pohl, R.B., Balon, R., Chapman, P., & McBride, J. (1998). A new patient diary to
study performance anxiety. Poster presented at the 151st Annual Convention of the American Psychiatric Association, Toronto.
Preston, J.D., O’Neal, J.H., & Talaga, M.C. (2002). Handbook of Clinical
Psychopharmacology for Therapists, 3rd ed. Oakland, CA: New Harbinger Publications.
94
Rapee, R.M., Craske, M.G., Brown, T.A., & Barlow, D.H. (1996). Measurement
of perceived control over anxiety-related events. Behavior Therapy, 27, 279-293. Rennenberg, B., Chambless, D.L., & Gracely, E.J. (1992). Prevalence of SCID-
diagnosed personality disorders in agoraphobic outpatients. Journal of Anxiety Disorders, 6, 111-118.
Reynolds, M., & Wells, A. (1999). The Thought Control Questionnaire – Psychometric
properties in a clinic sample and relationships with PTSD and depression. Psychological Medicine, 29, 1089-1099.
Riskind, J.H., Beck, A.T., Berchick, R.J., Brown, G., & Steer, R.A. (1987). Reliability of
DSM-III-R diagnoses for major depression and generalized anxiety disorder using the Structured Clinical Interview for DSM-III-R. Archives of General Psychiatry, 44, 817-820.
Roemer, L., & Orsillo, S.M. (2002). Expanding our conceptualization of and treatment
for generalized anxiety disorder: Integrating mindfulness/acceptance-based approaches with existing cognitive-behavioral models. Clinical Psychology: Science and Practice, 9, 54-68.
Roemer, L., & Orsillo, S.M. (2001, July). The role of experiential avoidance in
generalized anxiety disorder. Paper presented at the annual convention of the World Congress of Behavioral and Cognitive Therapy.
Rosnow, R.L., & Rosenthal, R. (1996). Computing contrasts, effect sizes, and
counternulls on other people’s published data: General procedures for research consumers. Psychological Methods, 1, 331-340.
Safren, S.A., Heimberg, R.G., & Juster, H.R. (1997). Clients’ expectancies and their
relationship to pretreatment symptomatology and outcome of cognitive-behavioral group treatment for social phobia. Journal of Consulting and Clinical Psychology, 65, 694-698.
Saluck, R. G., Herbert, J. D., Rheingold, A. A., & Harwell, V. (2000, November).
Validity of the brief and full versions of the FNE scales. Paper presented at the meeting of the Association for Advancement of Behavior Therapy, New Orleans.
Schneier, F., Johnson, J., Hornig, C., Liebowitz, M., & Weissman, M. (1992). Social
phobia: Comorbidity and morbidity in an epidemiologic sample. Archives of General Psychiatry, 49, 282-288.
Segal, D.L. (1997). Structured interviewing and DSM classification. In S.M. Turner &
M. Hersen (Eds.), Adult Psychopathology and Diagnosis, 3rd edition. New York: John Wiley & Sons.
95
Segal, D.L., Hersen, M., & Van Hasselt, V.B. (1994). Reliability of the Structured
Clinical Interview for DSM-III-R: An evaluative review. Comprehensive Psychiatry, 35, 316-327.
Simpson, H.B., Schneier, F.R., Campeas, R.B., Marshall, R.D., Fallon, B.A., Davies, S.,
Klein, D.F., & Liebowitz, M.R. (1998). Imipramine in the treatment of social phobia. Journal of Clinical Psychopharmacology, 18, 132-135.
Skodol, A.E., Oldham, J.M., Rosnick, L., Kellman, H.D., & Hyler, S.E. (1991).
Diagnosis of DSM-III-R personality disorders: A comparison of two structured interviews. International Journal of Methods in Psychiatric Research, 1, 13-26.
Skodol, A.E., Rosnick, L., Kellman, H.D., Oldham, J.M., & Hyler, S.E. (1988).
Validating structured DSM-III-R personality disorder assessments with longitudinal data. American Journal of Psychiatry, 145, 1297-1299.
Spitzer, R.L. (1983). Psychiatric diagnosis: Are clinicians still necessary?
Comprehensive Psychiatry, 24, 399-411. 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.
Stein, M.B., Liebowitz, M.R., Lydiard, R.B., Pitts, C.D., Bushnell, W., & Gergel, I.
(1998). Paroxetine treatment of generalized social phobia (social anxiety disorder): A randomized clinical trial. Journal of the American Medical Association, 280, 708-713.
Stemberger, R.T., Turner, S.M., Beidel, D.C., & Calhoun, K.S. (1995). Social phobia:
An analysis of possible developmental factors. Journal of Abnormal Psychology, 104, 526-531.
Stopa, L., & Clark, D. M. (1993). Cognitive processes in social phobia. Behaviour Research and Therapy, 31, 255-267. Stravynski, A., Marks, I., Yule, W. (1982). Social skills problems in neurotic outpatients:
Social skills training with and without cognitive modification. Archives of General Psychiatry, 39, 1378-1385.
Strosahl, K.D., Hayes, S.C., Bergan, J., & Romano, P. (1998). Assessing the field
effectiveness of acceptance and commitment therapy: An example of the manipulated training research method. Behavior Therapy, 29, 35-64.
Taylor, S. (1996). Meta-analysis of cognitive-behavioral treatments for social phobia.
Journal of Behaviour Therapy and Experimental Psychiatry, 27, 1-9.
96
Teasdale, J. D., Moore, R. G., Hayhurst, H., Pope, M., Williams, S., & Segal, Z. V.
(2002). Metacognitive awareness and prevention of relapse in depression: Empirical evidence. Journal of Consulting and Clinical Psychology, 70, 275-287.
Teasdale, J. D., Segal, Z. V., Williams, J. M. G., Ridgeway, V. A., Soulsby, J. M., & Lau,
M. A. (2000). Prevention of relapse/recurrence in major depression by mindfulness-based cognitive therapy. Journal of Consulting and Clinical Psychology, 68, 615-623.
Turner, S. M., Beidel, D. C., Cooley, M. R., Woody, S. R., & Messer, S. C. (1994). A
multi-component behavioral treatment for social phobia: Social effectiveness therapy. Behaviour Research and Therapy, 32, 381-390.
Turner, S. M., Beidel, D. C., Dancu, C. V., & Stanley, M. A. (1989). An empirically derived inventory to measure social fears and anxiety: The social phobia and anxiety inventory. Psychological Assessment: A Journal of Consulting and Clinical Psychology, 1, 35-40. Turner, S.M., Beidel, D.C., & Jacob, R.G. (1994). Social phobia: A comparison of
behavior therapy and atenolol. Journal of Consulting and Clinical Psychology, 62, 350-358.
Turner, S. M., Beidel, D. C., & Larkin, K. T. (1986). Situational determinants of social
anxiety in clinic and non clinic samples: Physiological and cognitive correlates. Journal of Consulting and Clinical Psychology, 54, 523-527.
Twohig, M.P., Hayes, S.C., & Masuda, A. (in press). Increasing willingness to
experience obsessions: Acceptance and Commitment Therapy as a treatment for obsessive-compulsive disorder. Behavior Therapy.
Twohig, M.P., & Woods, D.W. (2004). A preliminary investigation of acceptance and
commitment therapy and habit reversal as a treatment for trichotillomania. Behavior Therapy, 35, 803-820.
van Ameringen, M., Mancini, C., & Streiner, D.L. (1993). Fluoxetine efficacy in social
phobia. Journal of Clinical Psychiatry, 54, 27-32. Van der Linden, GJH, Stein, DJ, & van Balkom, AJLM (2000). The efficacy of the
selective serotonin reuptake inhibitors for social anxiety disorder (social phobia): A meta-analysis of randomized controlled trials. International Clinical Psychopharmacology, 15, S15-S23.
van Vliet, I.M., den Boer, J.A., & Westenberg, H.G.M. (1994). Psychopharmacological
treatment of social phobia: A double-blind, placebo-controlled study with fluvoxamine. Psychopharmacology, 115, 128-134.
97
van Zuuren, F.J. (1988). The Fear Questionnaire: Some data on validity, reliability, and
layout. British Journal of Psychiatry, 153, 659-662. Versiani, M., Mundim, F.D., Nardi, A.E., & Liebowitz, M.R. (1988).
Tranylcypromine in social phobia. Journal of Clinical Psychopharmacology, 8, 279-283.
Versiani, M., Nardi, A.E., Mundim, F.D., Alves, A.B., Liebowitz, M.R., & Amrein, R.
(1992). Pharmacotherapy of social phobia: A controlled study with moclobemide and phenelzine. British Journal of Psychiatry, 161, 353-360.
Walser, R.D., & Pistorello, J. (2004). ACT in group format. In S.C. Hayes & K.D.
Strosahl (Eds.), A Practical Guide to Acceptance and Commitment Therapy (pp. 347-372). Norwell, MA: Plenum Press.
Wells, A., & Davies, M. (1994). The Thought Control Questionnaire: A measure of
individual differences in the control of unwanted thoughts. Behavior Research and Therapy, 32, 899-913.
Williams, J.B.W., Gibbon, M., First, M.B., Spitzer, R.L., Davies, M., Borus, J., Howes,
M.J., Kane, J., Pope, H.G., Rounsaville, B., & Wittchen, H. (1992). The Structured Clinical Interview for DSM-III-R (SCID): Multisite test-retest reliability. Archives of General Psychiatry, 49, 630-636.
Wilson, K.G., & Groom, J. (2002). The Valued Living Questionnaire. Available from
the first author at Department of Psychology, University of Mississippi. Wilson, K.G., & Murrell, A.R. (2004). Values work in Acceptance and Commitment
Therapy: Setting a course for behavioral treatment. In S.C. Hayes, V.M. Follette, & M.M. Linehan (Eds.), Mindfulness and Acceptance: Expanding the Cognitive-Behavioral Tradition (pp. 120-151. New York, NY: Guilford Press.
Wlazlo, Z., Schroeder-Hartwig, K., Hand, I., Kaiser, G., & Münchau, N. (1990).
Exposure in vivo vs. social skills training for social phobia: Long-term outcome and differential effects. Behaviour Research and Therapy, 28, 181-193.
Wolpe, J., & Lazarus, A. A. (1966). Behavior therapy techniques. New York: Pergamon. Zaider, T.I., Heimberg, R.G., Fresco, D.M., Schneier, F.R., & Liebowitz, M.R. (2003).
Evaluation of the Clinical Global Impression Scale among individuals with social anxiety disorder. Psychological Medicine, 33, 611-622.
98
Zaldivar Basurto, F., & Hernandez Lopez, M. (2001). Terapia de aceptacion y compromiso: Aplicacion a un trastorno de evitacion experiencial con topografia agorafobica [Acceptance and Commitment Therapy (ACT): Application to an experiential avoidance with agoraphobic form]. Analisis y Modificacion de Conducta, 27, 425-454.
Zebb, B.J., & Moore, M.C. (1999). Another look at the psychometric properties of the
Anxiety Control Questionnaire. Behaviour Research and Therapy, 37, 1091-1103.
Zettle, R.D. (2003). Acceptance and commitment therapy (ACT) vs. systematic
desensitization in treatment of mathematics anxiety. Psychological Record, 53, 197-215.
Zettle, R.D., & Hayes, S.C. (1986). Dysfunctional control by client verbal behavior:
The context of reason-giving. The Analysis of Verbal Behavior, 4, 30-38. Zettle, R.D., & Rains, J.C. (1989). Group cognitive and contextual therapies in treatment
of depression. Journal of Clinical Psychology, 45, 438-445. Zweig, D.R., & Brown, S.D. (1985). Psychometric evaluation of a written stimulus
presentation format for the Social Interaction Self-Statement Test. Cognitive Therapy and Research, 9, 285-295.
99
Table 1: Assessment Schedule
Measure
Baseline
Pre-Treatment
Mid-Treatment
Post-Treatment
SCID-I/P
X
SAD section of SCID-I/P
X
X
SAD section of ADIS
X
X
X
Questionnaire Packet
X
X
X
X
CGI Severity Improvement
X
X X
X X
Behavioral Assessment
X
X
Note. SCID-I/P = Structured Clinical Interview for DSM-IV Axis I Disorders; SAD = Social Anxiety Disorder; ADIS = Anxiety Disorders Interview Schedule; CGI = Clinical Global Impression Scales.
100
Table 2: Demographic Characteristics of the Sample
Characteristic
Total Sample
(n=20)
Characteristic
Total Sample
(n=20)
Gender Male Female
55% (11) 45% (9)
Avoidant PD Yes No
60% (12) 40% (8)
Race African American Caucasian Hispanic Asian
5% (1) 65% (13) 10% (2) 20% (4)
Comorbid Axis I†
Mood Anxiety Other None
40% (8) 30% (6) 10% (2) 35% (7)
Education Some H.S. H.S. diploma Some college College degree Graduate or Professional GED
0% (0) 15% (3) 10% (2) 45% (9) 30% (6)
0% (0)
Relationships Single Married Divorced Separated Widowed
80% (16) 5% (1) 5% (1) 10% (2) 0% (0)
Employment Unemployed Part-time Full-time Student Missing
15% (3) 5% (1)
55% (11) 20% (4) 5% (1)
Medication One Two or more None
10% (2) 20% (4) 70% (14)
Note. Avoidant PD = Avoidant Personality Disorder, as diagnosed by the SCID-II. †Percentages in this cell sum to greater than 100% due to some participants meeting criteria for both a comorbid mood and anxiety disorder. Comorbid mood disorders consisted of Major Depressive Disorder (n = 7) and Depressive Disorder NOS (n = 1). Comorbid anxiety disorders included: Generalized Anxiety Disorder (n = 2), Obsessive-Compulsive Disorder (n = 1), Panic Disorder without Agoraphobia (n = 1), Specific Phobia (n = 1), and Anxiety Disorder NOS (n = 1). Other comorbid disorders included Alcohol Abuse (n = 1) and Learning Disability (n = 1).
101
Table 3: Means (Standard Deviations), Effect Sizes, and p-Values of Baseline, Pre-,
Mid-, and Post-Treatment Measures for Completers Only and Intention to Treat Analyses
Measures
Completers Only
ES
p
Intention to
Treat
ES
p
SPAI-SP
n Baseline Pre Mid Post
14 132.25 (29.48) 131.11 (30.23)a
116.30 (28)b
101.81 (33.13)b
.04
.51
.47
.92
1.00 .02 .089 .003
19 129.15 (28.91) 130.81 (31.26)a
118.56 (29.40)b
98.46 (33.34)c
.06
.40
.64 1.00
1.00 .006 .030
.002 FQ-SP
n Baseline Pre Mid Post
14 21.62 (5.59) 22 (6.48)a
17.43 (5.35)b
14.36 (7.34)b
.06
.77
.48 1.10
1.00 .057 .180 .026
19 21.42 (5.92) 21.47 (6.96)a 16.16 (6.78)b
13.89 (7.61)b
.01
.77
.31 1.04
1.00 .009 .096 .003
Brief FNE
n Baseline Pre Mid Post
14 48.69 (5.91) 50.21 (6.67)a 44.93 (7.35)a,b 40.79 (8.57)b
.24
.75
.52 1.23
1.00 .142 .333 .028
19 48.63 (7.07) 49.95 (7.12)a 46.32 (8.21)a,b 41.11 (9.54)b
.19
.47
.59 1.05
1.00 .199 .129 .003
LSAS-Fear
n Baseline Pre Mid Post
14 40.15 (11.36) 40.07 (10.75)a 37.80 (13.14)a 31.71 (12.26)b
.01
.19
.48
.73
1.00 .412 .037 .016
19 39.79 (10.20) 40.70 (11.30)a 38.32 (12.16)a 34.00 (12.19)b
.08
.20
.35
.57
1.00 .163 .041 .007
LSAS-Avoidance n Baseline Pre Mid Post
14 38.00 (11.34) 38.34 (11.75)a
29.99 (14.12)b
23.14 (13.05)c
.03
.64
.50 1.22
1.00 .024 .003 .001
19 37.16 (11.51) 38.14 (12.90)a
30.36 (13.14)b
24.32 (13.12)c
.08
.60
.46 1.06
1.00 .003
.001 <.001
102
Table 3 continued BDI-II n Baseline Pre Mid Post
16 16.75 (13.43) 16.69 (13.26)a 14.23 (11.62)a
9.30 (11.16)a
.004.20 .43 .60
.959
.367
.068
.071
19 16.47 (13.04) 17.16 (12.65)a
14.82 (11.29)a,b 10.67 (11.29)b
.05
.20
.37
.54
.556
.254
.071
.057 QOLI n Baseline Pre Mid Post
16 .23 (1.77) .12 (1.78)a .66 (2.14)a,b 1.54 (1.36)b
.06
.27
.49
.90
.516
.185
.280
.010
19 .12 (1.94)
-.17 (2.05)a .32 (2.39)a,b 1.07 (2.02)b
.03
.07
.34
.44
.114
.135
.280
.008 SDS-Work n Baseline Pre Mid Post
16 6.24 (2.63) 6.5 (2.22)a
5.37 (2.75)a,b 3.69 (2.55)b
.11
.45
.63 1.16
.579
.276
.022
.001
19 6.37 (2.65) 6.32 (2.56)a 5.32 (2.95)a
3.89 (2.83)b
.02
.36
.49
.90
.884
.226
.025
.001 SDS-Social n Baseline Pre Mid Post
16 7.35 (1.94) 7.13 (2.45)a 6.25 (2.49)a
4.50 (2.71)b
.10
.36
.67 1.02
.605
.130
.047
.013
19 7.53 (1.93) 7.16 (2.46)a 6.37 (2.50)a
4.89 (2.83)b
.17
.32
.55
.86
.320
.094
.050 .012
SDS-Family n Baseline Pre Mid Post
16 4.53 (2.60) 4.88 (2.22)a 4.50 (3.03)a
2.81 (2.97)b
.14
.14
.56
.79
.361 1.00 .055 .010
19 4.74 (2.79) 5.21 (2.39)a
4.84 (3.06)a,b
3.42 (3.24)b
.18
.13
.45
.63
.305 1.00 .058 .009
TCQ-Distraction n Baseline Pre Mid Post
13 13.07 (3.91) 12.00 (2.24)a 12.85 (2.73)a 12.69 (1.89)a
.34
.34
.07
.33
.373
.761 1.00 .207
19 13.39 (3.65) 12.55 (2.31)a 12.79 (2.76)a 12.95 (2.44)a
.28
.09
.06
.17
.179 1.00 1.00 1.00
103
Table 3 continued TCQ-Social Control n Baseline Pre Mid Post
13 10.29 (3.29) 9.92 (3.57)a
10.31 (2.87)a,b 12.69 (3.52)b
.11 .12 .74 .78
.928 1.00 .057 .043
19 10.11 (2.91) 10.16 (3.27)a 10.53 (2.70)a 12.05 (3.26)a
.02
.12
.51
.58
.996 1.00 .107 .062
TCQ-Worry n Baseline Pre Mid Post
13 10.89 (4.41) 11.31 (3.71)a 11.15 (3.18)a 10.31 (4.07)a
.10
.05
.35
.26
.648 1.00 .795 .770
19 11.19 (3.95) 11.56 (3.17)a 10.95 (3.26)a 10.63 (3.70)a
.10
.19
.09
.27
.558 1.00 1.00 .421
TCQ-Punishment n Baseline Pre Mid Post
13 9.79 (2.67) 9.69 (2.32)a 9.15 (1.82)a 8.62 (2.36)a
.04
.26
.25
.46
.828 1.00 1.00 .767
19 10.22 (2.80) 10.10 (2.11)a 9.58 (1.95)a 9.32 (2.43)a
.05
.26
.12
.34
.921
.911 1.00 .763
TCQ-Reappraisal n Baseline Pre Mid Post
13 12.43 (3.41) 12.92 (3.95)a 13.54 (2.76)a 13.75 (3.18)a
.13
.18
.07
.23
.302 1.00 1.00 1.00
19 13.50 (3.70) 13.33 (3.42)a 13.58 (3.13)a 13.78 (3.40)a
.05
.08
.06
.13
.789 1.00 1.00 1.00
ACQ-Reactions n Baseline Pre Mid Post
13 30.01 (8.81) 28.62 (8.96)a 32.85 (9.29)a,b 38.00 (10.35)b
.16
.46
.52
.97
1.00 .228 .322 .026
19 32.27 (8.56) 30.00 (8.43)a
33.21 (8.70)a,b 36.21 (9.95)b
.27
.37
.32
.67
.487
.194
.563
.062 ACQ-Events n Baseline Pre Mid Post
13 40.92 (9.46)
39.69 (10.21)a 43.77 (8.42)a 45.85 (8.56)a
.12
.44
.24
.65
1.00 .488 .984 .119
19 40.68 (9.59)
39.32 (10.97)a 41.58 (10.12)a 44.47 (9.01)a
.13
.21
.30
.51
1.00 .854 .477 .145
ATQ-Believability n
12
19
104
Table 3 continued ATQ-Believability Baseline Pre Mid Post
91.50 (35.84) 91.33 (38.20)a 81.25 (47.19)a,b 60.75 (22.21)b
.004.23 .56 .98
.364
.623
.108
.011
94.92 (39.81) 97.37 (42.49)a
88.42 (46.05)a,b 75.47 (36.89)b
.06
.20
.34
.55
.309
.265
.117
.006 ATQ-Frequency n Baseline Pre Mid Post
12 103.45 (34.79) 103.75 (33.85)a 92.67 (43.14)a,b 76.08 (26.44)b
.01
.29
.46
.91
.273
.555
.220
.046
19 106.64 (36.35) 114.63 (39.50)a
103.32 (40.80)a,b 92.84 (35.46)b
.21
.28
.27
.58
.132
.212
.224
.020 AAQ n Baseline Pre Mid Post
14 40.64 (6.97) 41.36 (8.35)a 37.14 (6.63)a,b 33.14 (10.09)b
.09
.56
.47
.89
1.00 .182 .119 .025
19 40.26 (6.82) 41.21 (7.74)a
38.21 (6.42)a,b 35.26 (9.64)b
.13
.42
.36
.68
.435
.229
.125
.038 WS n Baseline Pre Mid Post
14 27.86 (8.70) 26.86 (9.21)a
40.14 (13.07)b
51.29 (15.59)c
.11 1.17.78 1.91
1.00 .007 .002
<.001
19 30.21 (11.24) 29.74 (12.26)a
42.00 (11.83)b
50.21 (13.61)c
.04 1.02.64 1.58
.792
.002
.003 <.001
VLQ n Baseline Pre Mid Post
14 23.29 (20.57) 18.91 (13.24)a 15.99 (24.27)a,b 6.29 (17.51)b
.23
.15
.46
.81
.323 1.00 .372 .031
19 26.16 (20.02) 21.20 (17.56)a 19.52 (26.69)a 12.37 (24.10)a
.26
.07
.28
.42
.110 1.00 .370 .064
CGI-Severity n Pre Mid Post
15 4.67 (.49)a
4.20 (.41)b
3.47 (.74)c
1.041.221.91
.011
.009 <.001
19 4.79 (.54)a
4.42 (.61)b
3.84 (.90)c
.64
.75 1.28
.014
.022
.001 CGI-Improvement n Mid Post
15 3.33 (.62)a
2.60 (1.06)b
.84
.022
17 3.21 (.63)a
2.67 (1.01)b
.64
.038
105
Table 3 continued SISST-Positive n Pre Post
14 28.50 (8.13)a 38.64 (9.04)b
1.18
.011
19 30.58 (8.04)a 39.16 (8.02)b
1.07
.005 SISST-Negative n Pre Post
14 48.79 (13.64)a 40.21 (11.38)b
.68
.035
19 48.04 (12.60)a 41.79 (11.15)b
.53
.039 Self-Rating of Performance n Pre Post
11 2.45 (.75)a 3.76 (.86)b
1.62
.001
19 2.47 (1.01)a 3.23 (1.23)b
.68
.003 Self-Rating of SUDS n Pre Post
11 59.14 (17.09)a 39.65 (16.73)b
1.15
<.001
19 55.05 (19.20)a 43.77 (19.03)b
.59
.002 Social Skills Ratings n Pre Post
16 2.18 (.51)a 3.12 (.81)b
1.39
<.001
18 2.17 (.48)a 3.00 (.83)b
1.22
<.001 Observed SUDS Ratings n Pre Post
16 57.50 (13.81)a 39.06 (14.59)b
1.30
<.001
18 57.59 (13.09)a 41.20 (15.15)b
1.16
<.001 Note. Means with different subscripts differ significantly. Sample sizes vary due to missing data. ES = Cohen’s d effect size; SPAI-SP = Social Phobia and Anxiety Inventory-Social Phobia subscale; FQ-SP = Fear Questionnaire-Social Phobia subscale; Brief FNE = Brief Version of the Fear of Negative Evaluation Scale; LSAS = Liebowitz Social Anxiety Scale; BDI-II = Beck Depression Inventory 2nd Edition; QOLI = Quality of Life Inventory; SDS = Sheehan Disability Scale; TCQ = Thought Control Questionnaire; ACQ = Anxiety Control Questionnaire; ATQ = Automatic Thoughts Questionnaire; AAQ = Acceptance and Action Questionnaire; WS = Willingness Scale; VLQ = Valued Living Questionnaire; CGI-S = Clinical Global Impression Severity Scale; CGI-I = Clinical Global Impression Improvement Scale; SISST = Social Interaction Self-Statement Test; SUDS = Subjective Units of Discomfort scale.
106
Figure 1: Participant Flow Diagram for Study Phases
Diagnostic Assessment (n = 47)
Excluded (n = 39) No longer interested (n = 0) Did not meet criteria (n = 39)
Excluded (n = 17) No longer interested (n = 10) Did not meet criteria (n = 7)
Telephone Screening (n = 86)
Baseline Assessment (n = 30)
Behavioral Assessment (n = 24)
Assigned to Treatment (n = 20) Didn’t start (n = 4) Drop out (n = 1)
Mid-Treatment Assessment (n = 18)
Post-Treatment Assessment (n = 17)
Drop out (n = 1) Withdrawn (n = 1)
Drop out (n = 0) Refused assessment (n = 1)
Drop out (n = 6)
107
50
60
70
80
90
100
110
120
130
140
baseline pre mid post
spai-sp
Figure 2: Baseline, Pre-, Mid-, and Post-Treatment Mean Scores on the SPAI-SP
108
20
25
30
35
40
45
50
55
baseline pre mid post
brief fne
Figure 3: Baseline, Pre-, Mid-, and Post-Treatment Mean Scores on the Brief FNE
109
5
7
9
11
13
15
17
19
21
23
25
baseline pre mid post
fear-sp
Figure 4: Baseline, Pre-, Mid-, and Post-Treatment Mean Scores on the FQ-SP
110
10
15
20
25
30
35
40
45
baseline pre mid post
lsas-fearlsas-avoid
Figure 5: Baseline, Pre-, Mid-, and Post-Treatment Mean Scores on the LSAS Fear
and Avoidance Subscales
111
VITA
KRISTY L. DALRYMPLE
EDUCATION: • Drexel University (formerly MCP Hahnemann University until 2001), 2000 - 2005,
GPA: 4.0/4.0, M.S. in Clinical Psychology, May 2002, Ph.D. in Clinical Psychology, (APA accredited), October 2005.
• SUNY Upstate Medical University, Department of Psychiatry, September 2004 - September 2005, Predoctoral Internship in Clinical Psychology (APA accredited).
• Hope College, Holland, MI, 1996 – 2000, GPA: 3.86/4.0, B.A. in Psychology with a Minor in Spanish, May 2000.
PUBLICATIONS: • Herbert, J. D., & Dalrymple, K. (in press). Social anxiety disorder. In A. Freeman &
S. Felgoise (Eds.), Encyclopedia of Cognitive Behavior Therapy. Norwell, MA: Kluwer.
• Herbert, J. D., Gaudiano, B. A., Rheingold, A.A., Myers, V. H., Dalrymple, K., & Nolan, B. M. (2005). Social skills training augments the effectiveness of cognitive behavioral group therapy for social anxiety disorder. Behavior Therapy, 36, 125-138.
• Gaudiano, B.A., & Dalrymple, K.L. (2005). EMDR variants, pseudoscience, and the demise of empirically supported treatments. [Review of the book Psychotherapeutic Interventions for Emotion Regulation: EMDR and Bilateral Stimulation for Affect Management]. PsycCRITIQUES Contemporary Psychology: APA Review of Books.
• Herbert, J.D., Crittenden, K.B., & Dalrymple, K. (2004). Knowledge of social anxiety disorder relative to attention deficit hyperactivity disorder among educational professionals. Journal of Clinical Child and Adolescent Psychology, 33, 366-372.
• Gaudiano, B.A. & Dalrymple, K. (2002). Reconsidering prescription privileges for psychologists. American Psychological Association for Graduate Students Newsletter, 4(1), 39-40.
• Dalrymple, K., & Motiff, J.P. (2000). [Review of the book Understanding Sleep: The Evaluation and Treatment of Sleep Disorders]. Cognitive and Behavioral Practice, 7, 485.
HONORS AND AWARDS: MCP Hahnemann University Honors Distinction in Clinical Applications and Research Methods (2002), Sigma Xi Research Award (2000), Arthur John Ter Keurst Psychology Scholarship (1999), Hope College Endowed Scholarship (1996 – 2000), Anna R. Pipp Foundation Scholarship (1996-2000), Honor Societies: Phi Beta Kappa (2000), Psi Chi (1998-2000), Hope College Pew Society (1998-2000). PROFESSIONAL MEMBERSHIPS: American Psychological Association, American Psychological Society, Association for Behavioral and Cognitive Therapies, Society for a Science of Clinical Psychology.