Professional Documents
Culture Documents
Tee Kazantzis 2011
Tee Kazantzis 2011
Tee Kazantzis 2011
net/publication/227585893
CITATIONS READS
49 6,167
2 authors, including:
Nikolaos Kazantzis
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
All content following this page was uploaded by Nikolaos Kazantzis on 08 December 2017.
Despite the consensus regarding the importance of col- 2008). However, there has been little theoretical analy-
laborative empiricism (CE) in Beck’s cognitive therapy, sis, and no definition or assessment of the construct.
absent are operational definitions, measures, or empiri- Although CE is frequently referred to as a distinctive
cal investigations of the construct. Existing research and defining characteristic of cognitive therapy, and as
has centered on constructs such as the working alliance
a core component of the therapeutic alliance (e.g.,
Beck, 1989; Bishop & Fish, 1999; Cross, 2007; Datti-
and has produced inconsistent findings. It is unclear
lio, 2000; Dudley & Kuyken, 2006; Friedberg & Clark,
whether CE is related to treatment outcomes, or
2006; Friedberg & Gorman, 2007; Kuyken, Padesky, &
whether it mediates other cognitive change processes.
Dudley, 2008; Merali & Lynch, 1997; Overholser,
This article argues that the core meaning of collabora-
1995; Rutter & Friedberg, 1999; Segal & Shaw, 1996;
tion in CE, ‘‘sharing the work,’’ is not captured by the Will, 1995), it remains unclear whether CE directly or
most frequently used construct of the therapeutic indirectly facilitates hypothesized mechanisms of cogni-
alliance in cognitive therapy research. A theory of CE, tive change and positive therapeutic outcomes. Existing
based on self-determination theory, is developed that discussion of CE in the literature has been limited to a
integrates the collaborative and empirical aspects brief outline of two components (collaboration and
of CE and addresses the motivational aspect of the empiricism). Absent are analyses of how these compo-
construct. nents interact, or a theory of how CE may benefit
Key words: cognitive, collaboration, collaborative clients in therapy.
empiricism, empiricism, relationship, therapy. [Clin This article will first review the current definition
Psychol Sci Prac 18: 47–61, 2011]
and assessment of CE in cognitive therapy. Second, the
article will provide a theoretical discussion to clarify
and elaborate on the nature and interaction between
A contradiction awaits the reader who reviews the lit- the elements of collaboration. Finally, the article will
erature on collaborative empiricism (CE). The past propose new components of CE that can form the
40 years have seen widespread agreement regarding the basis of a more fully articulated definition of the con-
central role of CE in cognitive therapy (Hollon & struct, which can also form the basis of an operational
Beck, 1979; Kuyken, Padesky, & Dudley, 2009; definition, and a new measure.
Madsen, 2009; Persons, 2008; Tryon & Misurell,
EXISTING DEFINITIONS OF COLLABORATIVE EMPIRICISM
Address correspondence to Nikolaos Kazantzis, School of The therapeutic collaboration was first outlined as a
Psychological Science, La Trobe University, Victoria 3086, means for identifying ‘‘raw data’’ and a means for
Australia. E-mail: N.Kazantzis@latrobe.edu.au. encouraging client identification, observation, and
2011 American Psychological Association. Published by Wiley Periodicals, Inc., on behalf of the American Psychological Association.
All rights reserved. For permissions, please email: permissionsuk@wiley.com 47
evaluation of introspective beliefs in Beck, Rush, Client: Gee, well, I don’t know…
Shaw, and Emery’s (1979) Cognitive Therapy for
Depression (see also Beck, 1967). These seminal works Perhaps they are tired … maybe, even they are
waiting for a lead from someone else? Is it possible
define CE as a basic strategy or form of interaction
they think I don’t make much of an effort?
underlying the relationship between client and thera-
pist in cognitive therapy (Hollon & Beck, 1979). Therapist: Those sound like some good possibilities.
Client and therapist actively collaborate to identify I really like how you are looking around the situa-
problem situations, and to design, implement, and tion from different perspectives. I wonder how
evaluate empirical tests of the client’s beliefs. A key could we figure out how much of the ‘‘not talking’’
objective is for CE to enable experimentation of is due to other people’s views of you … or any of
beliefs and behaviors outside therapy via the vehicle these other ideas you have generated?
of homework assignments (Beck, 1970, 1995; Kazant-
zis, MacEwan, & Dattilio, 2005). This collaborative Client: Hmmm. Maybe I could take more notice of
sharing of the work of experimental testing distin- what happens when I am socializing.…
guishes CE. For example, a therapist aiming to point
Therapist: Good, let’s figure out how to be specific
out errors in thought process through disputation
about how you will notice that.…
may be attempting to foster empiricism, without col-
laboration:
A true sharing of the work of identifying ideas to be
tested, and then devising means of testing them out,
Client: I don’t want to go to the dinner; people
clearly has a different presence in therapy. However,
never talk to me—they think I’m boring.
much of the literature following the early teaching in
CE is less than clear. While there is agreement that CE
Therapist: Sounds like you’re mindreading again.
is important for cognitive therapy (cf. Kuyken et al.,
When you’re at the dinner, notice whether people
2008, 2009; Persons, 2008; Tryon & Misurell, 2008),
actually seem bored, or whether you’re basing your
subsequent analysis and discussion have been largely
feeling on what other people think, or worse yet,
based on practical, commonsense grounds, rather than
what you think they’re thinking.
firm conceptual foundation. For instance, CE has been
The same lead from the client could be pursued by described as ‘‘embedded in every clinical action and
a therapist adherent to CE, as outlined in A. T. Beck’s decision’’ (Friedberg & McClure, 2002, p. 44), which
therapy, by taking the following approach: comprises the ‘‘cornerstone of CBT’’ and is ‘‘used
throughout the entire course of treatment’’ (Dattilio,
Client: I don’t want to go to the dinner; people
2000, p. 39). There is also confusion as to whether CE
never talk to me—they think I’m boring.
is a relationship attribute or technique, or both a pro-
Therapist: That sounds like a tricky situation for you cess and an approach (e.g., Gleeson & McGorry,
Would you like to talk about that some more—and 2004). In some cases, CE has been referred to as a
see what we can figure out here as a way to help? major therapeutic method and strategy in Beck’s ther-
apy (Abramson, Alloy, & Dykman, 1990; Bell, Grech,
Therapist: Ok, well if I am hearing you correctly, a Maiden, Halligan, & Ellis, 2005; Dattilio, 2000). In
concern that you have is that people think you are other cases, CE is referred to as a ‘‘philosophy which
boring, and this is connected to your conversations forms the foundation of cognitive therapy’’ (Wilkes,
with people. It might help us to find out under 1994, p. 309), as well as a ‘‘specific technique’’ and an
which situations this happens the most.… For the ‘‘atmosphere’’ that pervades the patient–therapist rela-
moment, I wonder what might be the reasons that tionship (Turkington & Siddle, 1998, p. 237). Thus, it
people (including you) might not talk much some- remains unclear whether CE can be considered a
times? requirement for the therapeutic relationship in Beck’s
by increasing the meaningfulness of new beliefs for the associated with clients’ satisfaction with that session
client. (Eisenthal, Koopman, & Lazare, 1983). Clients are
The application of SDT to CE raises several more likely to return for therapy after an intake inter-
hypotheses for empirical examination. SDT suggests view in which the therapist shares the problem formu-
that the collaborative empirical investigation of clients’ lation and negotiates mutual therapeutic goals,
intrinsic data increases the core SDT variables of compared to an interview where neither is done
autonomy, competence, and relatedness, thereby (Tracy, 1977). Similarly, therapist–client discussion of
increasing motivation to engage in behavior change. goals and expression of thoughts and feelings are associ-
SDT also clarifies the functioning of collaboration and ated with observer ratings of ‘‘goodness’’ of therapy
empiricism in CE in terms of these core variables. sessions (Hoyt, 1980; Hoyt, Xenakis, Marmar, &
Before outlining a research agenda informed by these Horowitz, 1983). Simply specifying goals is enough to
hypotheses, existing empirical work on collaboration increase clients’ ratings of therapy helpfulness (Gold-
and empiricism will be reviewed. stein, Cohen, Lewis, & Struening, 1988). Client-rated
agreement with therapist and client-rated experience of
Empirical Evaluation of Collaborative Empiricism goal consensus at session 2 has been shown to predict
There has been no empirical investigation of CE in cog- the reduction in symptoms on the Symptom Check
nitive therapy research trials. However, research in other List (SCL; Derogatis, 1974), including up to six months
contexts has found a relationship between aspects of col- later (Dormaar, Dijkman, & de Vries, 1989). Higher
laboration in CE and therapeutic outcomes (Boardman, response congruence, in which therapist verbal
Catley, Grobe, Little, & Ahluwalia, 2006; Bordin, 1994; response directly addresses the subject of the client’s
Clemence, Hilsenroth, Ackerman, Strassle, & Handler, immediately preceding statement, is associated with
2005; Colson et al., 1988; Connor-Greene, 1993; higher likelihood of the client returning for therapy
Creed & Kendall, 2005; Diamond, Liddle, Hogue, & after the initial session (Duehn & Proctor, 1974).
Dakof, 1999; Hatcher, 1999; Hatcher & Barends, 1996; Finally, topic determination, the proportion of topic
Le Bloc’h, de Roten, Drapeau, & Despland, 2006). initiations subsequently followed by the other partici-
One such aspect of collaboration in CE is the fram- pant in therapy (client or therapist), predicts the con-
ing of the client as central to the process of shared tinuation of therapy past session 3 (Tracey, 1986).
decision making. Sharing decisions with the client These aspects of collaboration do not address the core
regarding the process of therapy, and developing a clear meaning of collaboration in CE. Nevertheless, shared
rationale and explanation for the course of a session are decision making, sharing the formulation, negotiation
determination are aspects of collaboration in CE for Empirical work has shown a relationship between
which there is empirical support. aspects of collaboration and therapeutic outcomes. This
Not all studies have found a relationship between suggests that CE, which incorporates these aspects of
aspects of collaboration and outcome. Beyebach and collaboration, may also be related to outcomes. Never-
Carranza (1997) failed to replicate the relationship theless, results of this research have been mixed, and
between topic determination and engagement reported much of this work has focused on collaboration as a
by Tracey (1986). However, nonengaged clients in the client factor, rather than a property of the dyad. Col-
Beyebach and Carranza study used more domineering laboration in CE goes beyond the meaning of collabo-
language with therapists, interrupted more often, and ration in these studies. Thus, while research to date
engaged in more conflict. It is possible that these provides qualified support for a relationship between
behaviors disrupted collaboration, resulting in lower aspects of CE and outcomes, research focusing specifi-
continuation of therapy (Tryon & Winograd, 2002). cally on CE is needed. For this, a measure is needed
Collaborative involvement has also been used as a that incorporates the aspects of collaboration discussed
measure of collaboration in therapy process. Collabora- earlier and operationalizes the meanings of collabora-
tive involvement is the mutual involvement of client tion and empiricism specific to CE.
and therapist in a helping relationship (Tryon & Wino- As with collaboration, there has been no specific
grad, 2002). The term ‘‘mutual involvement’’ suggests investigation of the empirical component of CE in
that this construct has a meaning similar to collabora- cognitive therapy trials. However, there has been some
tion in CE. However, studies using the construct have evaluation of one form of empirical evaluation of cli-
operationalized collaborative involvement as the com- ents’ beliefs in cognitive therapy in the form of behav-
pliance or cooperation of the client to the strategies of ioral experiments (BEs; Beck et al., 1979; Clark, Beck,
the therapist. For example, Schmidt and Woolaway- & Alford, 1999; Dobson & Dobson, 2009). Behavioral
Bickel (2000) operationalized collaborative involvement experiments are widely used and recommended in cog-
as homework completion. They found that therapists’ nitive therapy as a component of standard treatment for
ratings of compliance with homework predicted out- a range of disorders (Beck, 1995; Bennett-Levy et al.,
comes, but clients’ ratings of compliance did not. This 2004; Padesky & Greenberger, 1995). It has been sug-
suggests a relationship between clients’ compliance and gested that the very ubiquity of BEs in successful ther-
outcomes (see review in Kazantzis, Whittington, & apy attests to their value (Bennett-Levy et al., 2004).
Dattilio, in press), but is a long way from ‘‘mutual However, despite their importance in cognitive therapy,
involvement’’ in the sense of a genuinely collaborative only one study has directly investigated the efficacy of
sharing of the work indicative of CE. BEs in belief change (Bennett-Levy, 2003). In this
In an interesting study, O’Malley, Suh, and Strupp study, 27 trainee cognitive therapists compared the
(1983) found that client involvement at session 3 pre- effectiveness of BEs and automatic thought records in
dicted outcome, whereas involvement at sessions 1 and changing their own cognitions. The authors reported
2 did not. They concluded that client involvement that BEs produced significantly higher cognitive and
may not be simply a quality that the client brings to behavioral change compared with thought records.
therapy, but may be facilitated by qualities of the thera- Several trainees credited the higher efficacy of BEs to
pist. Windholz and Silberschatz (1988) found that ther- their experiential aspect, noting that BEs led to new
apist ratings of client involvement were related to cognitions that were more powerfully believed than
outcomes, but not client or observer ratings. It is sug- those which were understood rationally but not experi-
gested here that involvement may also be a function of enced in a BE.
the interaction between client and therapist, rather than There is some empirical work supporting a mediat-
a quality of one or the other in isolation. If true, the ing effect of CE on therapeutic outcomes. While there
genuine collaboration inherent in CE could be has been no direct investigation of the relationship
expected to increase client involvement. between case conceptualization and outcomes in