Journal of Consulting and Clinical Psychology

2016, Vol. 84, No. 9, 783794

2016 American Psychological Association

0022-006X/16/$12.00 http://dx.doi.org/10.1037/ccp0000100

Responsive Management of Early Resistance in CognitiveBehavioral

Therapy for Generalized Anxiety Disorder
Adi Aviram and Henny A. Westra

Michael J. Constantino

York University

University of Massachusetts Amherst

Martin M. Antony

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Ryerson University
Objective: Client resistance limits the efficacy of cognitive behavioral therapy (CBT), rendering it an
important process marker for clinicians. Yet, little is known about how CBT therapists responsively
navigate resistance. This study examined whether differences in therapists behavior during moments of
early resistance in CBT predict distal (posttreatment worry) and proximal (subsequent resistance) therapy
outcomes. We hypothesized that clinicians use of supportive, motivational interviewing (MI)-consistent
strategies versus directive CBT methods during resistance episodes would relate to more favorable
outcomes. Method: Clients (N 30) with severe generalized anxiety disorder (GAD) were treated with
CBT (without MI) in the context of a randomized controlled trial. An adapted version of the Client
Resistance Code was used to identify moments of client disagreement with therapist direction, and these
episodes were rated for therapist behavior using the Motivational Interviewing Treatment Integrity
(MITI). Context-specific variations in therapists MITI ratings during disagreement were compared with
variations in ratings during randomly selected moments to examine whether the timing of therapist use
of theoretically indicated MI skills differentially impacted outcomes. Results: Clients whose therapists
displayed greater MI-consistent responses during disagreement had lower levels of posttreatment worry,
t 2.84, p .009, and subsequent resistance, t 2.96, p .013. Additionally, MI-consistent
behavior during disagreement was substantially related to treatment outcomes, but MI-like behavior
during randomly selected therapy segments was not (t .15, p .886 for posttreatment worry, and
t .09, p .934 for subsequent resistance). Conclusions: Findings support the importance of
context-responsivity to resistance markers, suggesting that systematic incorporation of MI for managing
resistance holds promise for improving CBT for GAD.

What is the public health significance of this article?

This study reinforces the clinical importance of client resistance as a marker for therapist responsiveness. Moreover, the findings suggest that the systematic incorporation of MI relational skills for
managing resistance holds significant promise for improving CBT for GAD. Therapists learning to
provide even modestly more empathic, evocative, and autonomy-supportive responses during moments of client resistance can potentially yield large dividends.

Keywords: cognitive behavioral therapy, resistance, context-responsivity, motivational interviewing,

generalized anxiety disorder
Supplemental materials: http://dx.doi.org/10.1037/ccp0000100.supp

Although cognitive behavioral therapy (CBT) is considered an

evidence-based, front-line intervention (National Institute of Clinical Excellence, 2011), research demonstrates that a substantial

portion of clients receiving CBT do not adequately respond, drop

out prematurely, or relapse at follow-up (e.g., Westen & Morrison,
2001). Investigators in recent years have suggested that treatment

This article was published Online First March 31, 2016.

Adi Aviram and Henny A. Westra, Department of Psychology, York
University; Michael J. Constantino, Department of Psychological and
Brain Sciences, University of Massachusetts Amherst; Martin M. Antony,
Department of Psychology, Ryerson University.
Our sincere thanks to the Canadian Institute of Health Research (CIHR)
for their financial support of this study (MOP114909). We also wish to
acknowledge the hard work and dedication of the people who made this

study possible, including our wonderful coders who gave generously of

their time and effort on this project. We also wish to thank David Flora for
his excellent statistical and data analysis guidance and support. Finally,
sincere thanks go to Melissa Button and Kimberley Hara for their tremendous help during the coding training.
Correspondence concerning this article should be addressed to Adi
Aviram, Department of Psychology, York University, 4700 Keele Street,
Toronto, Ontario, Canada, M3J 1P3. E-mail: aviram@yorku.ca
783

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

784

AVIRAM, WESTRA, CONSTANTINO, AND ANTONY

noncompliance and resistance to change may play a significant

role in limiting response to CBT (Antony, Ledley, & Heimberg,
2005; Westra, 2012). For example, while homework assignments
are widely hypothesized as essential to the efficacy of CBT, client
noncompliance with in-session treatment activities and betweensession homework is a commonly acknowledged problem among
CBT practitioners (e.g., Kazantzis, Lampropoulos, & Deane, 2005;
Schmidt & Woolaway-Bickel, 2000; Wolf & Goldfried, 2014).
Beyond homework noncompliance, Newman (2002) outlined various behaviors that may constitute resistance in CBT, including
high levels of expressed negative emotion toward the therapist,
in-session avoidance such as frequent use of I dont know or
prolonged silence, as well as disagreeing with therapist comments.
In surveys of expert CBT clinicians, the most frequently cited
reasons for insufficient treatment response were lack of engagement in behavioral experiments and noncompliance (Bruce &
Sanderson, 2010). Similarly, in a survey of therapists experiences
conducting CBT for generalized anxiety disorder (GAD), most
respondents identified client resistance (i.e., opposition to therapy
or to the therapists direction) as a barrier to treatment efficacy
(Szkodny, Newman, & Goldfried, 2014).
Research also demonstrates that navigating client resistance is a
challenge for many CBT therapists, leading to decreases in adherence (i.e., therapist use of prescribed techniques and avoidance of
proscribed techniques) and competence (i.e., therapist skillful use
of techniques; see Waltz, Addis, Koerner, & Jacobson, 1993)
ratings. For example, in a study on factors associated with therapist
adherence in CBT for panic disorder, client resistance at midtreatment substantially decreased CBT therapists adherence to the
treatment model, even among the most experienced therapists in
the study (Zickgraf et al., 2015). Relatedly, higher levels of client
anger have been associated with lower therapist adherence and
competence ratings (which deteriorated over the course of therapy)
among highly trained and supervised therapists using a manualized
CBT protocol for panic disorder (Boswell et al., 2013). Such
findings have led researchers to conclude that CBT manuals may
not provide enough guidance concerning how to cope with markers of client hostility, anger, or resistance (Boswell et al., 2013;
Zickgraf et al., 2015).
A growing body of psychotherapy process research consistently
indicates that therapist cultivation of a more supportive and less
directive clinical stance in the presence of resistance is particularly
effective in improving treatment engagement and outcomes (e.g.,
Aviram & Westra, 2011; Bischoff & Tracey, 1995; Miller, Benefield, & Tonigan, 1993; Patterson & Forgatch, 1985). Thus, client
resistance appears to be an important therapy process marker that
indicates the use of supportive rather than directive methods (see
Beutler, Harwood, Michelson, Song, & Holman, 2011, for a review). Flexibility in the presence of resistance seems particularly
important, given that CBT therapists have been shown in some
instances to respond by becoming increasingly directive (i.e.,
amplifying adherence to treatment rationale and techniques),
which in turn has been shown to result in further negative process
and reduced engagement (Ahmed, Westra, & Constantino, 2012),
as well as poorer treatment outcomes (Aspland, Llewelyn, Hardy,
Barkham, & Stiles, 2008; Castonguay, Goldfried, Wiser, Raue, &
Hayes, 1996). These findings suggest that therapists need to be
trained to identify instances in which clients do not react favorably
to a therapy direction, and to respond to such markers in a

clinically flexible, responsive way (Castonguay, Boswell, Constantino, Goldfried, & Hill, 2010).
Responding to resistance with increased empathy and support is
at the heart of motivational interviewing (MI; Miller & Rollnick,
2002). Developed as an alternative to the traditional directive
therapeutic approaches for treating addictions, MI is based on Carl
Rogers client-centered therapy (Rogers, 1956), with particular
emphasis on enhancing intrinsic motivation for change by exploring and resolving client ambivalence. MI departs from many
directive psychotherapy approaches with respect to its conceptualization and management of resistance. In particular, resistance is
considered to be the product of both client ambivalence to change
and therapist management of that ambivalence (as opposed to
reflecting a negative and oppositional client trait). Moreover, sustained client resistance is considered a clinician skill error. Thus,
the onus is on the therapist to continuously attend to and supportively navigate moments of resistance as they occur throughout
treatment in order to reduce their negative impact (Miller &
Rollnick, 2002). MI therapists faced with resistance are encouraged to shift into a supportive style by rolling with resistance,
which includes responding evocatively (i.e., drawing out the clients ideas about change), using empathic reflections, and supporting client autonomy to hold beliefs and make decisions that run
counter to the goals of the therapist.
Previous studies have typically investigated the differential effects of general counseling style to demonstrate that more supportive and less directive approaches are associated with increased
cooperation and improved therapy outcomes (e.g., Aviram & Westra, 2011; Bischoff & Tracey, 1995; Miller et al., 1993; Patterson
& Forgatch, 1985). However, surprisingly little is known about
therapists responsivity (Stiles, Honos-Webb, & Surko, 1998) during key moments of clients resistant responses and behaviors over
the course of a therapy session. In other words, what are lacking
are investigations of the impact of appropriate responsiveness
(therapist supportive vs. directive behavior) on treatment outcomes
at the precise time(s) when resistance markers emerge. Such
investigations hold promise for increasing our understanding of
factors associated with effective CBT, such as the management of
treatment noncompliance and resistance.
In this vein, the current study examined therapist responsiveness
(i.e., level of therapist support and MI-consistent behavior) during
identified moments of resistance in the context of a recent randomized controlled trial (RCT) that investigated the efficacy of
integrated MI-CBT compared to CBT alone in the treatment of
high severity GAD (Westra, Constantino, & Antony, in press).
Therapists were nested within a treatment group (to control for
allegiance effects) and only CBT alone therapists were included in
the present study for two reasons: (a) MI-CBT therapists were
explicitly trained in supportive MI methods for rolling with resistance, and (b) the MI-CBT group was characterized by substantially lower levels of resistance than the CBT alone group, and had
a very low overall level of resistance (M 0.06, SD 0.07 vs.
CBT alone: M 0.16, SD 0.14; Constantino, Westra, &
Antony, 2015). In particular, we were interested in examining
whether differences in CBT therapists style in the presence of
resistance (specifically, moments of client disagreement with therapist input or direction) predict treatment outcomes.
Responsive management of disagreement was investigated in
two ways. First, variations between CBT therapists supportive

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

MANAGEMENT OF RESISTANCE IN CBT FOR GAD

versus directive behavior during early disagreement episodes were

assessed for their predictive capacity. We hypothesized that higher
ratings of therapist MI-consistent behavior (i.e., higher levels of
support) during disagreement would be associated with better
treatment outcomes, as reflected by (a) lower levels of subsequent
resistance (i.e., resistance coded in the session immediately following the session identified for disagreement), and (b) lower
ratings of posttreatment worry. Moreover, we included a measure
of CBT therapist competence in order to control for variability in
CBT skill. Second, therapists level of MI-consistent behavior was
also examined at several randomly selected points in treatment.
This general MI-consistent response style was then compared to
context-specific levels of MI-consistent behavior (during moments
of disagreement) to determine any differential impact of the timing
of therapist supportive behavior. We predicted that contextspecific therapist MI-consistent behavior (during disagreement)
would exert more pronounced effects on therapy outcomes than
therapist general MI-consistent behavior (i.e., during the randomly
selected therapy moments).

Method
Participants and Selection
Participants were recruited from community advertisements in
the greater Toronto area. Only those individuals who had a principal diagnosis of GAD (via the Structured Clinical Interview for
DSMIV Disorders; First, Spitzer, Gibbon, & Williams, 1996)1
were included. Additionally, participants had to score above the
cutoff for high worry severity, as defined by a baseline Penn State
Worry Questionnaire (PSWQ; described below) score 68 out of
80 points (Meyer, Miller, Metzger, & Borkovec, 1990). The current subsample included 30 adults who were predominantly female
and Caucasian, generally well-educated, and presented with a high
rate of diagnostic comorbidity (see Table 1 for subsample descriptives). A local Institutional Ethics Review Board for research
involving human participants approved all measures and procedures in the larger RCT. Informed consent was obtained at study
intake.

Therapists and Therapist Training

CBT therapists (N 13, all females) included 12 doctoral
graduate students in clinical psychology and one postdoctoral
fellow. All CBT therapists in the parent RCT had at least one case
in the current sample (i.e., had at least one disagreement episode
from the range of session numbers under consideration, as described below). Therapists ages ranged from 26 to 41 years old,
with a mean age of 29.15 years (SD 4.10 years). Eleven of the
therapists identified their primary orientation as cognitive
behavioral and two identified themselves as integrative. All therapists self-selected into the CBT arm of the parent trial, and none
had formal training in MI. Training consisted of four daylong
workshops, including discussion and role play. All therapists saw
at least one practice case with feedback until they became proficient. Therapists received ongoing case supervision during the
trial, which consisted of videotape review and weekly individual
meetings with a senior CBT postdoctoral fellow, under the supervision of a highly experienced CBT expert.

785

Treatment
CBT in this trial, which consisted of 15 weekly, 1-hr individual
sessions, was constructed from a number of evidence-based protocols (e.g., Craske & Barlow, 2006; Zinbarg, Craske, & Barlow,
2006) and included progressive muscle relaxation, cognitive restructuring, and behavioral strategies. Therapists were instructed to
implement treatment in a specific order, but the length of time
spent on each treatment component was left to the judgment of the
therapist, as indicated by the needs and response of clients to each
treatment element. In addition, in order to establish consistency in
the management of homework noncompliance, procedures for
CBT-consistent management were extracted from the literature
(e.g., Beck, 2005; Kazantzis & Shinkfield, 2007) and included in
the manual.

Measures
Worry. To assess worry, patients completed the PSWQ
(Meyer et al., 1990), a widely used 16-item instrument assessing
trait worry. The PSWQ possesses high temporal stability and
internal consistency, as reflected by a Cronbachs alpha of .93 for
all anxiety disorders, and .86 for GAD, as well as good convergent
and discriminant validity (Brown, Antony, & Barlow, 1992; Meyer
et al., 1990). It also differentiates individuals with GAD from those
with other anxiety disorders (Brown et al., 1992). PSWQ Scores
range from 16 to 80, with higher scores indicating greater worry.
For the current study, Cronbachs alpha was .68 at baseline and .96
at posttreatment.
Resistance. Given that there is no published measure of
therapist-rated resistance, Westra et al. (in press) constructed one
for the parent trial. Namely, therapists completed three visual
analogue scales (VASs) where they rated clients on the dimensions
of Passive-Active, Defensive-Receptive, and Rigid-Flexible. The
average Cronbachs alpha for these items over the first seven
sessions was .79, indicating good internal consistency. A higher
score on each dimension indicated more positive therapist ratings
(i.e., less resistance).
Observer-rated client resistance was assessed with the Client
Resistance Code (CRC; Chamberlain, Patterson, Reid, Kavanagh,
& Forgatch, 1984). Modified for use in studies of individual
therapy for generalized anxiety (Westra, Aviram, Kertes, Ahmed,
& Connors, 2009), resistance in the CRC is defined as any behavior that opposes, blocks, diverts, or impedes the direction set by the
therapist. Rather than being considered a characteristic of clients
(typically pejoratively), resistance is viewed as inextricably embedded in the interpersonal process between client and therapist
and is thus considered a measure of interpersonal process. The
CRC has been shown to possess good construct and predictive
validity (Chamberlain et al., 1984; Patterson & Forgatch, 1985;
Tracey & Ray, 1984), as well as face and content validity
(Bischoff & Tracey, 1995). Although the CRC defines 11 catego1
Criteria for GAD were mostly unchanged from DSMIV to DSM5
(American Psychiatric Association, 2000, 2013), with the exception of the
requirement in DSMIV that the GAD symptoms not occur exclusively
during the course of a mood disorder, psychotic disorder, pervasive developmental disorder, or posttraumatic stress disorder (this requirement was
removed in DSM5). All participants in the current study would have
received a diagnosis of GAD, based on current DSM5 criteria.

AVIRAM, WESTRA, CONSTANTINO, AND ANTONY

786
Table 1
Sample Characteristics
Measure
Gender
Age in years
Ethnicity

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Marital status

Highest level of education

Average family income

Employment/education status
Concurrent antidepressant medication use
Comorbidity

ries of resistance (e.g., challenging, disagreeing, sidetracking),

only an overall rating of the level of resistance was coded in the
present study, given that the presence of resistance in any form was
of primary interest, rather than the particular subtypes. Coders
watch an entire therapy session from beginning to end, making
individual ratings at 30-s intervals. Each time bin is rated for the
presence of resistance on a rating scale of 0 (absence of resistance), 1 (minimal/qualified resistance), 2 (clear resistance), or 3
(hostile/confrontational resistance).
Motivational interviewing integrity. To assess therapist
level of supportive behavior in the context of resistance, an MI
adherence measure was used. In particular, independent observers
rated sessions with the Motivational Interviewing Treatment Integrity (MITI) scale, version 3.1.1 (Moyers, Martin, Manuel,
Miller, & Ernst, 2010). The MITI scale has been used in numerous
studies to assess clinician competence and fidelity to the principles
of MI (e.g., Jensen et al., 2011; Moyers, Martin, Manuel, Hendrickson, & Miller, 2005). It has demonstrated sound reliability
and sensitivity, and has been found to predict treatment outcome
across a wide range of behavioral domains (Moyers, Martin,
Houck, Christopher, & Tonigan, 2009).
The MITI includes five global dimensions known as global
scores, which require coders to assign a single number from a
5-point Likert scale to characterize the entire interaction, thus
capturing the raters global impression or overall judgment about
the dimension. The five global dimensions include therapist Empathy, Evocation, Collaboration, respect for client autonomy (i.e.,
Autonomy/Support), and Direction (i.e., focus) of the session
toward the target behavior.2,3
CBT competence. Therapist competence in conducting CBT
was assessed using the Cognitive Therapy Scale (CTS; Young &

Participants
28 Female
2 Male
M 34.80, SD 12.72, min 21, max 63
23 White/European
3 Asian (e.g., South Asian, East Asian, Southeast Asian)
2 Hispanic/Latin American
1 Biracial/Multiracial
1 Other
15 Single
12 Married/Cohabitating
1 Divorced/Separated
2 No data
1 Some high school
1 Completed high school
7 Some postsecondary education
21 Completed postsecondary degree or diploma
13 less than $50,000
12 $50,001$100,000
4 $100,001$150,000
1 $150,001$175,000
11 Unemployed/Temporarily unable to go work/school
15 Employed currently
4 In school currently
11 Yes
19 No
26 (87%) Other anxiety disorder
22 (73%) Major depressive disorder/Dysthymic disorder

Beck, 1980). Ratings are made on 11 different dimensions, including General Skills (e.g., collaboration, understanding) and Specific
Cognitive Therapy Skills (e.g., guided discovery, focus on key
cognitions). Scores range from 0 to 66, with higher scores reflecting higher levels of overall CBT competence. Five undergraduate
psychology students were trained to criterion over a period of 6
months by the postdoctoral fellow specializing in CBT. The raters
worked independently, met regularly to reduce rater drift, and
resolved any disagreements through discussion to achieve consensus. Double coding a subset of 25% of independently coded tapes
to assess rater reliability yielded an intraclass correlation coefficient (ICC) of .84.

Procedure
Figure 1 contains a flowchart of the sampling and coding
procedures in the present study. Therapists completed VAS ratings
of client defensiveness, rigidity, and passivity following every
2
In the present study, the global score Direction was not included as part
of the MITI codes differentiating therapist directive vs. supportive behavior, given that the degree of therapist focus on the target behavior during
moments of resistance was not deemed as necessarily reflecting therapist
directive (i.e., coercive) vs. supportive (i.e., collaborative and autonomypreserving) behavior.
3
The MITI also includes a second component known as behavioral
counts, which requires coders to tally frequencies of specific MI-consistent
and inconsistent therapist behaviors without regard to how well they fit into
the overall impression of therapist use of MI. Therefore, the coder is not
required to judge the quality or overall adequacy of the therapists behavior
as with global scores, but simply to count it. Given that no significant
relationships were found among the MITI behavioral counts and the study
outcome measures, these were not included in the present study.

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

MANAGEMENT OF RESISTANCE IN CBT FOR GAD

Figure 1.

787

Flow diagram of coding procedures.

CBT session. The PSWQ was administered at baseline, immediately following every CBT session, and post-CBT. For the purposes of the present study, the CBT competence (CTS) ratings for
midtreatment (Session 6) were included given that this was within
the time frame of the disagreement session selection (see below).
This midtreatment session was coded for each of the therapist
client dyads.
Sample selection. Given our interest in examining therapist
responsiveness to resistance, CBT sessions containing resistance
were identified using a two-stage process. First, therapist VAS
ratings of client defensiveness for early CBT sessions (i.e., Sessions 2 through 6) were reviewed in order to identify sessions in
which therapists rated clients as highly defensive.4 Sessions that
were rated as highest on client defensiveness were selected for

resistance coding. A total of 17 sessions were selected using this

identification method. In cases where therapists did not rate their
clients as defensive for any of the sessions under review, a second
step was taken to identify sessions in which resistance may have
been present. Namely, four undergraduate coders who were trained
in the identification of resistance reviewed Sessions 2 through 6 of
the remaining dyads, and selected the session which they deemed
as containing the highest level of observed resistance. A total of 13

4
Session 1 was not included in this sample selection given that it was of
interest to preserve an early therapy session (i.e., a session prior to the
identification of resistant disagreements) to include as a baseline measure
for resistance.

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

788

AVIRAM, WESTRA, CONSTANTINO, AND ANTONY

additional potential sessions were selected using this identification

method. In summary, out of the 44 participants in the CBT alone
group who had completed treatment, only those who were identified as having a sufficient level of resistance were included in the
final sample for the current study (N 30).
Resistance coding. Following session selection, the videorecordings for the identified sessions were coded in their entirety
for resistance using the adapted CRC (Westra et al., 2009). Given
previous research indicating that clear and hostile resistance accounted for the highest variance in the prediction of treatment
outcomes (Aviram, Westra, & Eastwood, 2011), the present study
only considered the frequency of clear and hostile resistance in
observer coder ratings. That is, each time bin could receive a code
of 1, 2, and/or 3, and only those time bins receiving a code of 2
and/or 3 (i.e., clear or hostile resistance) were considered in the
present study. The frequency of clear and hostile resistance was
calculated by dividing the number of 30-s time bins containing a
code of 2 or 3 by the total number of time bins in the session.
In addition, given our interest in examining whether differences
in therapist style in the presence of client disagreement predict
subsequent levels of resistance, resistance coding was also conducted for the session immediately following the target session
containing disagreement. Finally, a subsample of 15 early sessions
(i.e., sessions that took place prior to the target session identified
for the presence of disagreement) were also coded for resistance to
provide a baseline measure of client opposition to therapist direction.
The team of resistance coders consisted of three graduate students in clinical psychology (2 doctoral, 1 Masters level), and one
PhD psychologist. Training consisted of two day-long workshops.
Samples of publicly available therapy sessions, followed by therapy session video-recordings from clients who were not included
in the present study were reviewed and discussed by the group
extensively at weekly meetings. Coders independently coded practice sessions, meeting regularly to review discrepancies in coding,
until adequate interrater reliability as assessed by 85% observer
agreement was achieved. Coders were kept blind to client outcome
status throughout the coding process. Reliability was examined
continuously throughout the coding process. Interrater reliability
was calculated by double-coding 25% of all tapes. ICCs ranged
from .74 to .98 with a mean of .88, indicating good to excellent
agreement.
Disagreement episode identification. In designing the current study, we chose to focus on interpersonal resistance that
involves client disagreements with therapist direction, suggestion,
and/or input. The selection of instances representing client disagreements was guided by both practical and empirical reasons.
Practically, while many forms of resistance are transient and occur
only momentarily (e.g., interrupting, sidetracking), disagreements
are often episodic in nature, and are thus more amenable to getting
a large enough sample of therapist behavior to enable valid coding.
Moreover, disagreements often represent instances of clear, explicit client opposition to therapist direction, and such episodes can
therefore be easily and reliably identified (see Appendix A in the
online supplemental materials for examples). Finally, episodes of
client disagreement were deemed especially relevant to examining
CBT therapist management of resistance, given that in these instances, therapist responses of coercion versus preservation of
client autonomy are thought to be especially apparent. That is,

CBT therapists ability to roll with resistance seems especially

transparent when clients explicitly voice disagreement with the
therapist.
Disagreement episodes consisted of a section of the video that
began with an instance of clear client disagreement with therapist
direction, input, or suggestion, and this would often center on
disputes concerning the tasks of treatment (e.g., clients objecting to
specific CBT tasks). With respect to the length of the extracted
disagreement episode, this was defined as beginning in the first
instance of client disagreement, and as ending once the client and
therapist had shifted into discussing another topic. Accordingly,
the end of a disagreement episode was not dependent on its
resolution, but on the end of the topic. A total of 67 disagreement
episodes were identified, which varied in length from 1.14 to 34.23
mins (M 9.34 mins, SD 7.64 mins). In addition, the number
of disagreement episodes per session varied, ranging from one to
five episodes (M 2.23, SD 1.04).
MITI coding. Therapists level of supportive behavior was
rated using the MITI (Moyers et al., 2010) for all identified
disagreement episodes, in order to provide context-specific
(i.e., during disagreement) MI-consistent behavior ratings. In
addition, therapists general MI-consistent behavior ratings
were derived for randomly selected 20-min segments from
sessions representing early, middle, and late phases of therapy
(i.e., Sessions 1, 6, and 11, respectively). The team of MITI
coders consisted of six advanced undergraduate students in
psychology. Coders participated in two day-long workshops,
which included didactic presentations, readings, as well as
coding and discussion of publicly available videotapes of MI
and CBT, followed by therapy sessions from clients who were
not included in the present study. In turn, only those coders who
had maintained 85% observed agreement against criterion
scores were included in the final group of coders. To prevent
drift, coders continued to independently code test materials
against criterion scores throughout the coding process. Coders
were kept blind to client outcome status and to the study
hypotheses. Interrater reliability as assessed by ICCs was calculated by double coding 25% of all tapes. ICCs ranged from
.71 to 1.0 with a mean of .89, indicating good to excellent
agreement.

Results
Calculating MITI Indices
Consistent with the MITI manual, and because the three
global scores reflecting MI Spirit (i.e., Evocation, Collaboration, and Autonomy/Support) were highly correlated (ranging
from .82 to .93 during disagreement episodes, and .92 to .94 for
randomly selected therapy segments), they were averaged to
yield a total MI Spirit score. The total MI Spirit scores were
also highly correlated with the Empathy global scores (.91
during disagreement episodes, and .89 during randomly selected therapy segments). Thus, an overarching global dimension of MI-consistent behavior was created by aggregating
Empathy and MI Spirit ratings. The MI-consistent behavior
index for disagreement episodes will be referred to as
Disagreement-Specific MI Behavior, and the MI-consistent behavior index for randomly selected therapy segments will be

MANAGEMENT OF RESISTANCE IN CBT FOR GAD

referred to as General MI Behavior. For Disagreement-Specific

MI Behavior, in cases where there was more than one disagreement episode for the session, the MITI ratings were averaged
across the episodes.5 In addition, for General MI Behavior,
MITI scores for the three randomly selected therapy segments
(from Sessions 1, 6, and 11) were averaged for each therapist to
yield an overall index of general MI-consistent response style.

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Preliminary Analyses
The means and standard deviations for all study variables are
presented in Table 2. No substantial violation of normality was
uncovered that would jeopardize the assumptions of the analyses.
No outliers were identified. Table 3 presents the intercorrelations
among all study measures. Initial worry severity was not significantly related to any of the variables. Ratings of General MI
Behavior and Disagreement-Specific MI Behavior were not significantly related. Moreover, ratings of General MI Behavior were
not significantly related to posttreatment worry, or to levels of
resistance in the session following the identified disagreement
episode session. In contrast, ratings of Disagreement-Specific MI
Behavior were significantly negatively correlated with clients
posttreatment worry scores, as well as with observed levels of
resistance following the disagreement episode session. Higher
midtreatment CBT competence was significantly associated with
higher Disagreement-Specific MI behavior, but not significantly
correlated with outcomes. Finally, early resistance levels were
positively related to subsequent resistance, but not significantly
related to posttreatment worry.6 Subsequent resistance (i.e., resistance measured immediately after the disagreement episode session) was significantly positively correlated with posttreatment
worry.

Regression Analyses
Hypothesis 1: Higher ratings of therapist MI-consistent behavior in the context of disagreement will be associated with
lower ratings of posttreatment worry.
We used multilevel modeling (MLM) to examine the relationship between therapist MI-like behavior during moments of disagreement and posttreatment PSWQ. Because clients were nested
within therapists, we first explored the amount of variability in
outcome that occurred at the therapist level by calculating the ICCs
from 2-level unconditional models using restricted maximum likelihood (REML). The intraclass correlation was .171, suggesting
that 17.1% of the total variance in posttreatment worry was accounted for by differences between therapists. A two-level
random-intercepts model was thus examined for the regression of
posttreatment worry on the Level 1 variable DisagreementSpecific MI Behavior, while also accounting for clients baseline
worry and therapist midtreatment CBT competence ratings. Regression diagnostics were assessed for the random-intercepts
model, and there was no evidence to suggest significant departures
from model assumptions. When the random-intercepts model was
fit to the data using REML, the estimated fixed Level 1 slope was
10 12.04, indicating that, while accounting for client baseline
worry severity and therapist midtreatment CBT competence ratings, for each 1-point increase in MI-consistent responding during

789

disagreement, posttreatment worry scores were predicted to decrease by 12.04. This effect was significant, t(25.47) 2.84,
p .009, 95% CI [12.09, 11.98], with Disagreement-Specific
MI Behavior accounting for 14.5% of the total variance in posttreatment worry.
Hypothesis 2: Higher ratings of therapist MI-consistent behavior in the context of disagreement will be associated with
lower levels of subsequent resistance.
When the random-intercepts model was estimated to account for
the nonindependent observations in this data set, a statistically
inadmissible solution was obtained. That is, the variability in
observed levels of resistance in the session immediately following
the disagreement episode session that was due to nesting of clients
within therapists was found to be negligible, thus leading to an
improper solution. This implies that the ICC is very close to zero.
A simplified, ordinary fixed-effects regression model was thus
estimated, using Disagreement-Specific MI Behavior in the prediction of subsequent resistance. Regression analyses indicated
that, while accounting for therapist midtreatment CBT competence
ratings, Disagreement-Specific MI Behavior was a significant predictor of subsequent resistance, t 2.87, p .013, 95% CI
[.14, .02]. Specifically, higher levels of observed MI-consistent
responding during disagreement were significantly and substantively related to lower levels of observed resistance in the following session, accounting for 18.7% of the variance in subsequent
resistance.
A second hierarchical linear regression was conducted in which
observed levels of resistance measured in a session prior to the
disagreement session were first entered into the equation, in order
to control for the impact of clients baseline tendency to oppose the
therapist on the observed findings. Regression analyses indicated
that, even while accounting for early levels of resistance,
Disagreement-Specific MI Behavior was a significant predictor of
subsequent resistance, t 2.96, p .013, 95% CI [.18, .03];
namely, higher levels of therapist MI-like behavior at the time of
disagreement were significantly and substantively related to lower
levels of observed resistance in the following session, accounting
for 31.2% of the variance in subsequent resistance.
5
It was also of interest to examine the level of consistency in therapists
management of disagreement in cases where more than one disagreement
episode was identified within the session. Here, standard deviations for
Disagreement-Specific MI Behavior ratings were calculated for each session with two or more disagreement episodes (n 21). These standard
deviations ranged from 0 to 1.44 (M .55), indicating that therapists could
be inconsistent in their management of resistance within a given session.
Namely, whereas some sessions were characterized by virtually no variation in ratings of therapist MI Spirit in response to resistance (SD 0),
other sessions were characterized by relatively high variability in therapist
behavior (SD 1.44) for different disagreement episodes.
6
It is possible that early levels of resistance were not significantly
related to posttreatment worry given the sampling method used in this
study to identify resistant disagreements. That is, given that disagreement
episode sessions were chosen to reflect the first instance in each dyad
where there was evidence for clear disagreement, this may have inadvertently led to a limited range for estimates of early resistance. Nevertheless,
given that higher levels of resistance that took place before the disagreement episode session were positively associated with higher levels of
resistance immediately following the disagreement episode session, early
resistance was arguably a solid index of resistance that can be used to
provide a baseline measure of client opposition to therapist direction.

AVIRAM, WESTRA, CONSTANTINO, AND ANTONY

790

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Table 2
Means and Standard Deviations for Study Variables
Measure

SD

Range

Pre-CBT PSWQ
Post-CBT PSWQ
Midtreatment CTRS
Resistance Pretarget Disagreement Session (n 15)
Resistance Posttarget Disagreement Session
Disagreement-Specific MI Behavior
General MI Behavior

75.62
44.43
43.79
.09
.18
2.37
1.99

3.27
17.30
10.73
.09
.13
.80
.64

69.0080.00
16.0080.00
19.0061.00
.01.30
.00.41
1.003.84
1.063.34

Note. CBT CognitiveBehavioral Therapy; PSWQ Penn State Worry Questionnaire; CTRS Cognitive
Therapy Rating Scale; MI Motivational Interviewing; Disagreement-Specific and General MI Behavior were
both rated on a Likert scale ranging from 1 to 5; The overall rating of the CBT sessions for adherence was good,
with total scores on the CTRS averaging 45.54 (SD 5.28). This compares favorably with the average score
of 41.28 (SD 4.24) on the CTRS in the CBT arm of the Treatment of Depression Collaborative Research
Program (Shaw et al., 1999).

Hypothesis 3: Context-specific ratings of therapist MIconsistent behavior (during disagreement) would exert more
pronounced effects on therapy outcomes than ratings of therapist general MI-consistent behavior (during the randomly
selected therapy moments).
To examine this hypothesis, the abovementioned set of regression analyses were repeated using the General MI Behavior scores.
Namely, a random-intercepts model was first examined for the
regression of post-CBT PSWQ on the Level 1 variable General MI
Behavior, while accounting for client baseline worry scores and
therapist midtreatment CBT competence ratings. Regression diagnostics were assessed for the random-intercepts model, and there
was no evidence to suggest significant departures from model
assumptions. When the random-intercepts model was fitted to the
data using REML, the estimated fixed Level 1 slope was
10 .91, indicating that, for each 1-point increase in General
MI Behavior, post-CBT PSWQ scores were predicted to decrease
by .91. This effect was not significant, t(17.36) .15, p .886,
95% CI [.98, .83]. That is, higher levels of General MI
Behavior were not significantly related to posttreatment worry.
Similarly, an ordinary fixed effects regression model was
estimated using General MI Behavior in the prediction of subsequent resistance.7 General MI Behavior was not found to
predict subsequent resistance, t .68, p .503, 95% CI
[.10, .05]. A second hierarchical linear regression was also
conducted, in which observed levels of resistance measured in
a session prior to the disagreement episode session were first
entered into the equation in order to control for the impact of
clients tendency to oppose the therapist on the level of resistance following the disagreement session. Regression analyses
again indicated that, while accounting for early levels of resistance, General MI Behavior was not a significant predictor of
subsequent resistance, t .09, p .934, 95% CI [.09, .09].
A summary of the main findings is presented in Table 4.
Examples of therapist MI-consistent (i.e., supportive) and MIinconsistent (i.e., directive) management of disagreement are
presented in Appendix B in the online supplemental materials.

Discussion
The findings of the present study are highly consistent with
previous research in supporting the relationship between cultivat-

ing a supportive relational stance to resistance and improved

treatment outcomes (e.g., Beutler et al., 2011; Miller et al., 1993).
By identifying precise moments of disagreement and examining
therapist behavior at those specific times, the present study not
only supported the importance of appropriate responsiveness in the
moment, but also provided estimates of the magnitude and the
relative impact of supportive versus directive management of
resistance. In particular, higher levels of MI-like behavior (i.e.,
empathy, evocation, collaboration, and support of client autonomy) in response to early client disagreement were associated with
substantially lower subsequent resistance and posttreatment worry,
while non-context-specific (i.e., random or general) variations in
ratings of therapists MI-consistent responses were not. In terms of
magnitude, context-specific MI-consistent behavior had over 10
times the impact on posttreatment worry outcomes compared to
general MI-consistent behavior. A similar pattern was observed
with respect to proximal therapy outcomes of subsequent resistance.
Collectively, these findings suggest that client disagreement
and opposition to therapist direction may represent a significant
event or key moment in the therapy process, such that how the
therapist responds in this key moment has particular and
marked significance. In other words, doing the right thing at
the right time seems to be much more impactful than doing
that same thing at any time (or at a randomly selected time).
While empathy is associated with positive treatment outcomes,
there may be times when empathy is especially indicated, and
even critical. Indeed, a growing body of research supports that
upon noticing opposition, therapists should become increasingly empathic and supportive, encouraging clients to express
their concerns rather than persisting with technical interventions (e.g., Aspland et al., 2008; Aviram & Westra, 2011;
Beutler et al., 2011; Bischoff & Tracey, 1995; Castonguay et
al., 1996; Miller et al., 1993; Patterson & Forgatch, 1985).
More broadly, the delineation of contextual markers that indi7
Given that subsequent resistance was predominantly measured at an
earlier time point than were therapist MI-consistent behavior estimates
taken in the middle (i.e., Session 6) and late (i.e., Session 11) phases of
therapy, only early therapist MI-consistent behavior ratings (i.e., Session 1)
were used for the purposes of the regression analyses examining General
MI Behavior in the prediction of subsequent resistance.

MANAGEMENT OF RESISTANCE IN CBT FOR GAD

791

Table 3
Intercorrelations Among Study Variables
Disagreement-Specific
MI Behavior

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Variable
General MI Behavior
Disagreement-Specific MI
Behavior
Midtreatment CTRS
Pre-CBT PSWQ
Post-CBT PSWQ
Resistance Pretarget
Disagreement Session
(n 15)

.30 (p .113)

Midtreatment
CTRS

Pre-CBT
PSWQ

.24 (p .186) .13 (p .500)

Post-CBT
PSWQ

Resistance Pretarget
Disagreement Session Resistance Posttarget
(n 15)
Disagreement Session

.02 (p .921)

.33 (p .226)

.18 (p .351)

.41 (p .025) .09 (p .639) .42 (p .023)

.13 (p .511) .08 (p .670)
.06 (p .760)

.46 (p .083)
.17 (p .552)
.01 (p .971)
.34 (p .218)

.52 (p .003)
.31 (p .095)
.14 (p .462)
.61 (p .001)
.51 (p .053)

Note. MI Motivational Interviewing; CTRS Cognitive Therapy Rating Scale; CBT CognitiveBehavioral Therapy; PSWQ Penn State Worry
Questionnaire.
p .05, two-tailed. p .01, two-tailed.

cate the use of particular therapist responses on a moment-tomoment basis should emerge as a priority for psychotherapy
research (Constantino, Boswell, Bernecker, & Castonguay,
2013; Stiles et al., 1998).
Considering the magnitude of the impact of responsive navigation of resistance on therapy process and outcomes, and to the
extent that the current findings are replicated in future studies,
these results suggest that it is essential that CBT therapists resist
tendencies to be directive when navigating client opposition and
noncompliance in favor of supportive methods such as those
derived from MI. In fact, a growing number of CBT therapists are
recommending the integration of MI to increase engagement in
CBT (e.g., Federici, Rowa, & Antony, 2010; Flynn, 2011; Westra,
2012). As noted by Zickgraf et al. (2015), investigators need to
consider building in evidence-based modules such as MI for addressing therapy processes such as resistance that often derail CBT
therapists.
Notably, the current findings suggest that one may not need to
be an extremely proficient MI therapist in order to effectively
respond to resistance in therapy. Namely, the bandwidth of MIconsistent responding among CBT therapists during moments of
disagreement in the present study was relatively narrow and reflective of poor-to-moderate MI adherence overall. Nevertheless, it

Table 4
Random-Intercepts Models and Regression Analyses Examining
Disagreement-Specific MI Behavior and General MI Behavior in
the Prediction of Worry Reduction and Subsequent Resistance

Post-CBT PSWQ

Resistance Posttarget
Disagreement
Session

Disagreement-Specific
MI Behavior

General MI
Behavior

Coefficient 12.04
p .009
se 4.24
t Ratio 2.84
R2 .312
t 2.96
p .013
Beta .76

Coefficient .91
p .886
se 6.22
t Ratio .15
R2 .000
t .09
p .934
Beta .02

Note. MI Motivational Interviewing; CBT CognitiveBehavioral

Therapy; PSWQ Penn State Worry Questionnaire.

is quite striking that within this relatively limited range, even a

modest increase of 1 point on the 5-point MI adherence scale was
associated with a large effect. In other words, learning to cultivate
even modestly more empathic, evocative, and autonomysupportive responses during moments of client opposition can
potentially yield large dividends.
This finding should be encouraging to CBT practitioners who
are contemplating integrating MI skills for more effectively managing resistance in therapy. Nevertheless, it is important to note
that true embodiment of the MI relational style (and indeed the
effective use of MI) requires a fundamental shift in frame of
reference that is not always easily accomplished and requires
considerable practice. That is, given that the conceptual framework
of MI and its underlying spirit are fundamentally humanistic rather
than behaviorist (Miller & Rollnick, 2009), this implies that the
incorporation of humanistic principles to the training and practice
of CBT therapists may be required for the optimal integration of
these two methods. This observation is further strengthened by the
present findings demonstrating that therapist MI-consistent behavior was significantly related to distal and proximal therapy outcomes, even when accounting for ratings of therapist CBT competence. That is, the current findings cannot be explained by
therapist competence in interpersonal effectiveness, understanding, or collaboration as defined and assessed by CBT adherence
measures.
It is also noteworthy that CBT therapists in the present study
who displayed higher levels of MI relational conditions in the
presence of disagreement were in fact not formally trained in MI.
That is, while some CBT therapists were at times demonstrating
appropriate responsivity in the presence of resistance, this was
likely based on natural therapist inclinations, as opposed to a clear
and consistent set of guidelines concerning strategies for effectively negotiating opposition in therapy. Moreover, even within
therapists in the present study, there was often inconsistency in
responding to disagreement. Indeed, CBT investigators have recently noted that the cognitive behavioral model provides little to
no guidance in dealing with patient hostility, anger, or resistance
(Boswell et al., 2013; Zickgraf et al., 2015). Accordingly, the
current findings suggest that it may be important to systematically
train CBT therapists in appropriate responsivity during moments

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

792

AVIRAM, WESTRA, CONSTANTINO, AND ANTONY

of resistance, rather than allowing these relational skills to freely

vary. Furthermore, given the magnitude of effect, the current
findings suggest that such targeted training in rolling with resistance could be well worth the investment in improving CBT
outcomes.
Accordingly, the responsive management of resistance in psychotherapy requires that therapists learn to become good observers
of important clinical markers indicating opposition and disharmony in the therapy process. Such training in the observation and
identification of in-session cues of client opposition is especially
indicated given that these cues are often subtle and difficult to
identify. For example, Hara et al. (2015) reported that trained
observers ratings of resistance were strongly associated with
treatment outcomes, while CBT therapists own ratings of resistance were not. As such, training should involve systematic guided
practice in learning to identify cues of opposition and resistance,
together with cultivating facility in responding effectively to such
markers.
The present study had a number of limitations. First, there
was not enough power to account for the nesting of disagreement episodes within therapists. It will be important for future
studies with larger samples to account for the influence of the
clustering of disagreement episodes within therapists. In addition, findings are correlational rather than causal, and future
studies should experimentally manipulate therapist management of disagreement in order to examine causal connections.
However, it should be noted that the findings are consistent
with experimental investigations in this area manipulating therapist directive and supportive counseling styles (e.g., Miller et
al., 1993; Patterson & Forgatch, 1985). Moreover, the findings
of the present study are also consistent with findings of the
larger parent clinical trial, where differential levels of resistance were found to strongly mediate between-groups effects
favoring MI-CBT over CBT alone even up to 1 year posttreatment (Constantino et al., 2015). That is, this within-group
study on CBT echoes the between-groups findings in the larger
trial, converging on the importance of managing resistance in
an MI-style. Furthermore with respect to study limitations,
while serving only as a screening measure to locate sessions in
which the phenomenon of resistance was likely to be present,
therapist defensiveness coding would have benefitted from independent assessment. Future research should also include a
more fine-grained transactional analysis that would allow for
the evaluation of temporal relationships between therapist behaviors during disagreement and client outcomes, such as the
Sequential Code for Observing Process Exchanges (SCOPE;
Martin, Moyers, Houck, Christopher, & Miller, 2005). Moreover, future studies should also investigate resistance management across sessions in order to determine how fluctuations in
being more MI-like in response to disagreement influence outcomes. Finally, the present study only included individuals with
high severity GAD who underwent CBT. It would be important
to examine whether the present findings extend to other populations and therapeutic approaches.
Continued research on context-responsivity to resistance
markers is clearly warranted based on the present findings. The
findings also warrant serious consideration of systematically
integrating training in the recognition of resistance markers and
the cultivation of the client-centered relational conditions ad-

vanced in MI for the responsive management of resistance in

CBT. Such an approach appears to hold significant potential for
optimizing CBT outcomes.

References
Ahmed, M., Westra, H. A., & Constantino, M. J. (2012). Early therapy
interpersonal process differentiating clients high and low in outcome
expectations. Psychotherapy Research, 22, 731745. http://dx.doi.org/
10.1080/10503307.2012.724538
American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed., text rev.). Arlington, VA: Author.
American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Arlington, VA: Author.
Antony, M. M., Ledley, D. R., & Heimberg, R. G. (2005). Improving
outcomes and preventing relapse in cognitive-behavioral therapy. New
York, NY: Guilford Press.
Aspland, H., Llewelyn, S., Hardy, G. E., Barkham, M., & Stiles, W.
(2008). Alliance ruptures and rupture resolution in cognitive behavior
therapy: A preliminary task analysis. Psychotherapy Research, 18, 699
710. http://dx.doi.org/10.1080/10503300802291463
Aviram, A., & Westra, H. A. (2011). The impact of motivational interviewing on resistance in cognitive behavioural therapy for generalized
anxiety disorder. Psychotherapy Research, 21, 698 708. http://dx.doi
.org/10.1080/10503307.2011.610832
Aviram, A., Westra, H. A., & Eastwood, J. (2011, September). Patterns of
early resistance and client outcome in cognitive behavioral therapy.
Paper presented at the meeting of the North American Society for
Psychotherapy Research, Banff, AB, Canada.
Beck, J. S. (2005). Cognitive therapy for challenging problems: What to do
when the basics dont work. New York, NY: Guilford Press.
Beutler, L. E., Harwood, T. M., Michelson, A., Song, X., & Holman, J.
(2011). Resistance/reactance level. Journal of Clinical Psychology, 67,
133142. http://dx.doi.org/10.1002/jclp.20753
Bischoff, M. M., & Tracey, T. J. G. (1995). Client resistance as predicted
by therapist behavior: A study of sequential dependence. Journal of
Counseling Psychology, 42, 487 495. http://dx.doi.org/10.1037/00220167.42.4.487
Boswell, J. F., Gallagher, M. W., Sauer-Zavala, S. E., Bullis, J., Gorman,
J. M., Shear, M. K., . . . Barlow, D. H. (2013). Patient characteristics and
variability in adherence and competence in cognitive-behavioral therapy
for panic disorder. Journal of Consulting and Clinical Psychology, 81,
443 454. http://dx.doi.org/10.1037/a0031437
Brown, T. A., Antony, M. M., & Barlow, D. H. (1992). Psychometric
properties of the Penn State Worry Questionnaire in a clinical anxiety
disorders sample. Behaviour Research and Therapy, 30, 3337. http://
dx.doi.org/10.1016/0005-7967(92)90093-V
Bruce, T. J., & Sanderson, W. C. (2010). Understanding and managing
treatment-resistant panic disorder: Perspectives from the clinical experience of several expert therapists. In D. Sookman & R. L. Leahy (Eds.),
Treatment resistant anxiety disorders (pp. 105134). New York, NY:
Routledge.
Castonguay, L. G., Boswell, J. F., Constantino, M. J., Goldfried, M. R., &
Hill, C. E. (2010). Training implications of harmful effects of psychological treatments. American Psychologist, 65, 34 49. http://dx.doi.org/
10.1037/a0017330
Castonguay, L. G., Goldfried, M. R., Wiser, S., Raue, P. J., & Hayes, A. M.
(1996). Predicting the effect of cognitive therapy for depression: A study
of unique and common factors. Journal of Consulting and Clinical
Psychology, 64, 497504. http://dx.doi.org/10.1037/0022-006X.64.3
.497
Chamberlain, P., Patterson, G. R., Reid, J. B., Kavanagh, K., & Forgatch,
M. S. (1984). Observation of client resistance. Behavior Therapy, 15,
144 155. http://dx.doi.org/10.1016/S0005-7894(84)80016-7

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

MANAGEMENT OF RESISTANCE IN CBT FOR GAD

Constantino, M. J., Boswell, J. F., Bernecker, S. L., & Castonguay, L. G.
(2013). Context-responsive psychotherapy integration as a framework
for a unified clinical science: Conceptual and empirical considerations.
Journal of Unified Psychotherapy and Clinical Science, 2, 120.
Constantino, M. J., Westra, H. A., & Antony, M. M. (2015, June). Mediators of treatment effects in a randomized controlled trial testing the
efficacy of augmenting CBT with motivational interviewing for severe
generalized anxiety disorder. Paper presented at the meeting of the
Society for Psychotherapy Research, Philadelphia, PA.
Craske, M. G., & Barlow, D. H. (2006). Mastery of your anxiety and panic:
Therapist guide (4th ed.). New York, NY: Oxford University Press.
Federici, A., Rowa, K., & Antony, M. M. (2010). Adjusting treatment for
partial or nonresponse to contemporary cognitive-behavioral therapy. In
D. McKay, J. S. Abramowitz, & S. Taylor (Eds.), Cognitive behavioral
therapy for refractory cases: Turning failure into success (pp. 1137).
Washington, DC: American Psychological Association. http://dx.doi
.org/10.1037/12070-002
First, M. B., Spitzer, R. L., Gibbon, M., & Williams, J. B. W. (1996).
Structured clinical interview for DSMIV Axis I disordersPatient ed.
(SCID-I/P, version 2.0). New York, NY: Biometrics Research Department, New York State Psychiatric Institute.
Flynn, H. A. (2011). Setting the stage for the integration of motivational
interviewing with cognitive behavioral therapy in the treatment of depression. Cognitive and Behavioral Practice, 18, 46 54. http://dx.doi
.org/10.1016/j.cbpra.2009.09.006
Hara, K. M., Westra, H. A., Aviram, A., Button, M. L., Constantino, M. J.,
& Antony, M. M. (2015). Therapist awareness of client resistance in
cognitive-behavioral therapy for generalized anxiety disorder. Cognitive
Behaviour Therapy, 44, 162174. http://dx.doi.org/10.1080/16506073
.2014.998705
Jensen, C. D., Cushing, C. C., Aylward, B. S., Craig, J. T., Sorell, D. M.,
& Steele, R. G. (2011). Effectiveness of motivational interviewing
interventions for adolescent substance use behavior change: A metaanalytic review. Journal of Consulting and Clinical Psychology, 79,
433 440. http://dx.doi.org/10.1037/a0023992
Kazantzis, N., Lampropoulos, G. K., & Deane, F. P. (2005). A national
survey of practicing psychologists use and attitudes toward homework
in psychotherapy. Journal of Consulting and Clinical Psychology, 73,
742748. http://dx.doi.org/10.1037/0022-006X.73.4.742
Kazantzis, N., & Shinkfield, G. (2007). Conceptualizing patient barriers to
nonadherence with homework assignments. Cognitive and Behavioral
Practice, 14, 317324. http://dx.doi.org/10.1016/j.cbpra.2006.08.003
Martin, T., Moyers, T. B., Houck, J. M., Christopher, P. J., & Miller, W. R.
(2005). Motivational Interviewing Sequential Code for Observing Process Exchanges (MI-SCOPE) coders manual. Retrieved from http://
casaa.unm.edu/download/scope.pdf
Meyer, T. J., Miller, M. L., Metzger, R. L., & Borkovec, T. D. (1990).
Development and validation of the Penn State Worry Questionnaire.
Behaviour Research and Therapy, 28, 487 495. http://dx.doi.org/10
.1016/0005-7967(90)90135-6
Miller, W. R., Benefield, R. G., & Tonigan, J. S. (1993). Enhancing
motivation for change in problem drinking: A controlled comparison of
two therapist styles. Journal of Consulting and Clinical Psychology, 61,
455 461. http://dx.doi.org/10.1037/0022-006X.61.3.455
Miller, W. R., & Rollnick, S. (2002). Motivational interviewing: Preparing
people for change (2nd ed.). New York, NY: Guilford Press.
Miller, W. R., & Rollnick, S. (2009). Ten things that motivational interviewing is not. Behavioural and Cognitive Psychotherapy, 37, 129 140.
http://dx.doi.org/10.1017/S1352465809005128
Moyers, T. B., Martin, T., Houck, J. M., Christopher, P. J., & Tonigan, J. S.
(2009). From in-session behaviors to drinking outcomes: A causal chain
for motivational interviewing. Journal of Consulting and Clinical Psychology, 77, 11131124. http://dx.doi.org/10.1037/a0017189

793

Moyers, T. B., Martin, T., Manuel, J. K., Hendrickson, S. M. L., & Miller,
W. R. (2005). Assessing competence in the use of motivational interviewing. Journal of Substance Abuse Treatment, 28, 19 26. http://dx
.doi.org/10.1016/j.jsat.2004.11.001
Moyers, T. B., Martin, T., Manuel, J. K., Miller, W. R., & Ernst, D. (2010).
Revised Global Scales: Motivational Interviewing Treatment Integrity
3.3.3 (MITI 3.1.1). Retrieved from the University of New Mexico Center
on Alcoholism, Substance Abuse and Addictions (CASAA) website:
http://casaa.unm.edu
National Institute of Clinical Excellence (NICE). (2011). Generalized
anxiety disorder and panic disorder in adults: Management in primary,
secondary and community care. Leicester, UK: British Psychological
Society. Retrieved from http://www.ncbi.nlm.nih.gov/books/
NBK83459/
Newman, C. F. (2002). A cognitive perspective on resistance in psychotherapy. Journal of Clinical Psychology, 58, 165174. http://dx.doi.org/
10.1002/jclp.1140
Patterson, G. R., & Forgatch, M. S. (1985). Therapist behavior as a
determinant for client noncompliance: A paradox for the behavior modifier. Journal of Consulting and Clinical Psychology, 53, 846 851.
http://dx.doi.org/10.1037/0022-006X.53.6.846
Rogers, C. R. (1956). Client-centered theory. Journal of Counseling Psychology, 3, 115120. http://dx.doi.org/10.1037/h0046548
Schmidt, N. B., & Woolaway-Bickel, K. (2000). The effects of treatment
compliance on outcome in cognitive-behavioral therapy for panic disorder: Quality versus quantity. Journal of Consulting and Clinical Psychology, 68, 1318. http://dx.doi.org/10.1037/0022-006X.68.1.13
Shaw, B. F., Elkin, I., Yamaguchi, J., Olmsted, M., Vallis, T. M., Dobson,
K. S., . . . Imber, S. D. (1999). Therapist competence ratings in relation
to clinical outcome in cognitive therapy of depression. Journal of
Consulting and Clinical Psychology, 67, 837 846. http://dx.doi.org/10
.1037/0022-006X.67.6.837
Stiles, W. B., Honos-Webb, L., & Surko, M. (1998). Responsiveness in
psychotherapy. Clinical Psychology: Science and Practice, 5, 439 458.
http://dx.doi.org/10.1111/j.1468-2850.1998.tb00166.x
Szkodny, L. E., Newman, M. G., & Goldfried, M. R. (2014). Clinical
experiences in conducting empirically supported treatments for generalized anxiety disorder. Behavior Therapy, 45, 720. http://dx.doi.org/
10.1016/j.beth.2013.09.009
Tracey, T. J., & Ray, P. B. (1984). Stages of successful time-limited
counseling: An interactional examination. Journal of Counseling Psychology, 31, 1327. http://dx.doi.org/10.1037/0022-0167.31.1.13
Waltz, J., Addis, M. E., Koerner, K., & Jacobson, N. S. (1993). Testing the
integrity of a psychotherapy protocol: Assessment of adherence and
competence. Journal of Consulting and Clinical Psychology, 61, 620
630. http://dx.doi.org/10.1037/0022-006X.61.4.620
Westen, D., & Morrison, K. (2001). A multidimensional meta-analysis of
treatments for depression, panic, and generalized anxiety disorder: An
empirical examination of the status of empirically supported therapies.
Journal of Consulting and Clinical Psychology, 69, 875 899.
Westra, H. A. (2012). Motivational interviewing in the treatment of anxiety. New York, NY: Guilford Press.
Westra, H. A., Aviram, A., Kertes, A., Ahmed, M., & Connors, L. (2009).
Manual for rating interpersonal resistance. Unpublished coding manual,
York University, Toronto, Ontario, Canada.
Westra, H. A., Constantino, M. J., & Antony, M. M. (in press). Integrating
motivational interviewing with cognitive-behavioral therapy for severe
generalized anxiety disorder: An allegiance-controlled randomized clinical trial. Journal of Consulting and Clinical Psychology.
Wolf, A. W., & Goldfried, M. R. (2014). Clinical experiences in using
cognitive behavior therapy to treat panic disorder. Behavior Therapy,
45, 36 46. http://dx.doi.org/10.1016/j.beth.2013.10.002
Young, J., & Beck, A. T. (1980). Cognitive therapy scale rating manual.
Unpublished manuscript, University of Pennsylvania, Philadelphia, PA.

794

AVIRAM, WESTRA, CONSTANTINO, AND ANTONY

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Zickgraf, H. F., Chambless, D. L., McCarthy, K. S., Gallop, R., Sharpless,

B. A., Milrod, B. L., & Barber, J. P. (2015). Interpersonal factors are
associated with lower therapist adherence in cognitive behavioral therapy for panic disorder. Clinical Psychology & Psychotherapy. Advance
online publication. http://dx.doi.org/10.1002/cpp.1955
Zinbarg, R. E., Craske, M. G., & Barlow, D. H. (2006). Mastery of your
anxiety and worry: Therapist guide (2nd ed.). New York, NY: Oxford

University Press. http://dx.doi.org/10.1093/med:psych/9780195300024

.001.0001

Received September 12, 2015

Revision received December 31, 2015
Accepted February 22, 2016

Members of Underrepresented Groups:

Reviewers for Journal Manuscripts Wanted
If you are interested in reviewing manuscripts for APA journals, the APA Publications and
Communications Board would like to invite your participation. Manuscript reviewers are vital to the
publications process. As a reviewer, you would gain valuable experience in publishing. The P&C
Board is particularly interested in encouraging members of underrepresented groups to participate
more in this process.
If you are interested in reviewing manuscripts, please write APA Journals at Reviewers@apa.org.
Please note the following important points:
To be selected as a reviewer, you must have published articles in peer-reviewed journals. The
experience of publishing provides a reviewer with the basis for preparing a thorough, objective
review.
To be selected, it is critical to be a regular reader of the five to six empirical journals that are most
central to the area or journal for which you would like to review. Current knowledge of recently
published research provides a reviewer with the knowledge base to evaluate a new submission
within the context of existing research.
To select the appropriate reviewers for each manuscript, the editor needs detailed information.
Please include with your letter your vita. In the letter, please identify which APA journal(s) you
are interested in, and describe your area of expertise. Be as specific as possible. For example,
social psychology is not sufficientyou would need to specify social cognition or attitude
change as well.
Reviewing a manuscript takes time (1 4 hours per manuscript reviewed). If you are selected to
review a manuscript, be prepared to invest the necessary time to evaluate the manuscript
thoroughly.
APA now has an online video course that provides guidance in reviewing manuscripts. To learn
more about the course and to access the video, visit http://www.apa.org/pubs/authors/reviewmanuscript-ce-video.aspx.