Training and Supervision in
THE ACQUISITION OF COMPETENCE
IN COGNITIVE THERAPY
When, some years ago, the author took part in a national project to define
competencies for the different models of psychological therapy, it was apparent
that cognitive therapy practitioners had a very particular take on the acquisition
A large group of therapists of different orientations gathered in
London, first together and then separately in different orientation groups, to
come back together at the end of the day. By the end of the day, the cognitive
therapy group had pretty well completed a first draft describing the different
skills and competencies associated with the different stages of their model. When
all the groups had reassembled, however, it quickly became clear that hardly any
of the other groups had got so far – one group was still discussing who was the
client and who was the therapist!
This story signals both the strengths and dangers of cognitive therapy. It is
obviously good to have a set of skills and competencies that can be clearly
described, discussed and learnt. It may, however, be problematic if the parsimo-
nious ‘simplicity’ of the cognitive model is learnt in a way that lacks therapeutic
artistry and subtlety. This piece will suggest that, just as the cognitive model has
developed a more rounded and subtle model of therapeutic intervention, so must
its methods of training and supervision. The challenge for trainers and supervi-
sors is to help trainees and practitioners to develop sophistication in both scien-
tific and artistic competence in cognitive therapy, without losing the benefit of
the simple clarity of the model. The common core content of training and key
points about the training process will be described, including new training
models incorporating an opportunity for a personal experience of therapy,
in the form of self-practice and self-reflection. Descriptions of the different
elements of and approaches to supervision will follow.
The subjects of training and supervision in cognitive therapy would each
warrant coverage as a separate volume. This piece will therefore focus on two
particularly common issues of contemporary training and supervision. The first
issue is reflected in a common experience of cognitive therapy training, such as
what occurs when trainees may have to let go of at least some previous ideas and
skills – let go of one trapeze – before they have got a reliable set of new, cogni-
tive skills and concepts – to catch the other trapeze. Clutching at fresh air is an
experience that invariably leads to trepidation. As cognitive therapists become
more aware of the interpersonal processes in therapy, so cognitive therapy
trainers can use the same skills to reflect the interpersonal processes of training.
The second issue is the debate that concerns the degree of systemisation that
should be used in implementing cognitive therapy. One side of the coin is the
clarity and help that a therapist can gain from following a relatively prescribed
format of therapy, such as that described in a ‘manual’ or ‘protocol’. The other
side of this coin could be that over-adherence could easily lead to a mechanistic
format of therapy that is imposed upon the client.
TRAINING IN COGNITIVE THERAPY
From its inception, cognitive therapy has always been well received in Britain
(Beck, 2004). Various well-known British trainers, such as Melanie Fennell and
Adrian Wells, trained with Beck in Philadelphia and incorporated much of his
training structure and many of his methods into the training course at Oxford.
The Oxford course has also subsequently influenced other courses in Britain and
abroad. As the demand has grown for accredited therapists and accredited train-
ing, the Oxford and other training traditions, including that of behaviour
therapy courses, have informed the course content which has come to be
recommended for cognitive therapy by the British Association for Behavioural
and Cognitive Psychotherapies (BABCP; see www.babcp.org.uk)
CONTENT OF TRAINING
Aaron Beck notes that, by the mid-1970s, cognitive therapy had ‘warranted
admission to the arena of controversy’ (1976: 7) alongside other approaches to
psychological therapy.Beck has frequently argued that this admission was war-
ranted by the fact that cognitive therapy could claim to be:
a system of psychotherapy which provides (1) a comprehensive theory of psycho-
pathology that drives the structure of psychotherapy, (2) a body of knowledge
and empirical findings which supports the theory, and (3) research findings that
demonstrate its effectiveness. (Weishaar, 1993: 47)
These three factors may be regarded as the bedrock features of cognitive therapy
and the bedrock for training. The strength of the model derives further from the
fact that there is a well-established set of principles that operationalise the above
points in a down-to-earth and often parsimonious way. The principles are set out
in Table 1.
These principles help to ensure that the therapeutic model is focused on both
short-term interventions and pragmatic problem-solving and that, at the same
time, through informed and collaborative therapeutic relationships, it is able to
switch into other modes and to take on other types of issues. Principle 10, sug-
gesting as it does that a map of the client’s problems be drawn up, also ensures
that research-based conceptual models can be shared with the client and used to
guide therapy, and thereby underpin the whole enterprise.
The teaching content of cognitive therapy is generally divided into theory and
practice elements (see Table 2) and application elements (Table 3) Most training
courses are likely to include these elements but, as we have noted many times
throughout Cognitive therapy, second edition (Sanders and Wills, 2005), the
applications of cognitive therapy have now expanded to the point where partic-
ular courses may seek to include some and exclude others, perhaps depending
on the needs of the trainees and/or training establishment.
Many courses in cognitive therapy use measures of therapist competence, such
as The Cognitive Therapy Scale (CTS) to assess whether therapists are doing
what they are supposed to do. The CTS (Young and Beck, 1988) was revised in
2000 (James et al., 2000) with 12 items measuring:
1. Agenda setting and adherence to the agenda;
2. Providing and eliciting feedback;
4. Pacing and efficient use of time;
5. Interpersonal effectiveness: genuineness, warmth, empathy, trust;
6. Eliciting appropriate emotional expression;
7. Eliciting key cognitions;
8. Eliciting and planning behaviours;
9. Guided discovery;
10. Conceptual integration;
11. Application of methods of change: use of cognitive and behavioural
Each category is given a six-point rating, from 0 (absence of feature or highly
inappropriate performance) to 6 (excellent performance, even in the face of
patient difficulties). The rating scale specifies exactly what the therapist should
be doing within each category, thereby giving highly specific instruction and
Table 1 Principles of Cognitive Therapy (based on Beck & Emery,
1985; Beck, 1995)
1. Cognitive therapy is short term, at least initially.
2. Cognitive therapy uses problem-solving.
3. Cognitive therapy is structured.
4. Cognitive therapy is collaborative.
5. Cognitive therapy is active and directional.
6. Cognitive therapy is based on a sound therapeutic relationship.
7. Cognitive therapy uses regular homework assignments.
8. Cognitive therapy uses guided discovery and Socratic dialogue.
9. Cognitive therapy is educational.
10. Cognitive therapy is based on cognitive conceptualisation.
Table 2 Theory and practice content of cognitive therapy training
Principles of cognitive therapy.
Cognitive model of psychological problems.
Conceptualisation behavioural therapy.
Interpersonal and therapeutic processes.
feedback. It is very valuable in this way, giving detailed guidance about the form
of therapy. We also know that therapist competence, along with therapeutic
alliance, is related to outcome and may well be one of the key common factors
across the psychotherapies (Trepka et al., 2004).
LEARNING PROCESSES IN TRAINING
Although the content of training may look straightforward, learning cognitive
therapy, either as a beginning trainee, or in ongoing professional development,
is a more complex process. Because cognitive therapy is often understood as
incorporating ‘psycho-education’ – teaching clients about the nature of psycho-
logical problems and about the way out of them – as part of its working meth-
ods, there are inevitable parallels between the educational process in therapy and
the education process in training. Although there is a distinct content that needs
to be imparted, attention also needs to be paid to the processes that trainees
follow when learning cognitive therapy. Such processes are described in adult
education theory, such as the Lewin-Kolb cycle and Schon’s concept of reflective
practice (see Bennett-Levy, 2002; Bennett-Levy et al., 2004). In brief, the Lewin-
Kolb cycle suggests that there are different stages in the cycle of learning and that
there is a characteristic sequence of stages, as shown in Figure 1. Learning theo-
rists such as Honey and Mumford (1992) have suggested that, although this is a
common sequence, people vary in how they move through it: some people –
‘reflectors’ – need lots of reflection time, whereas others – ‘pragmatists’ – need
more action-based experiences.
In educational terms, this means that trainers and tutors probably need to spend
time thinking about the balance between different types of learning activities and
the links between them, such as following a slab of theory with some kind of prac-
tical exercise. For example, in a session looking at some of the principles in
Table 1, it can be useful to ‘operationalise’ one of the principles in a practice appli-
cation. Trainees could be asked to take one of the principles, such as agenda setting
or homework, and practise a possible line of rationale-giving that one might
use to introduce it to a client. An additional dimension might be added to this
by suggesting that trainees select principles with which they have some difficulty.
Using Uncertainty and the
Given the cognitive model of understanding belief and attitude change, we
should not expect that all trainees will immediately and fully accept all of the
Table 3 Application elements of cognitive therapy
Types of anxiety – panic, social and simple phobias, agoraphobia, OCD, etc.
Post-traumatic stress disorder.
Stress and low self-esteem.
Family and couple problems.
training content, particularly the principles (Table 1) Indeed we should be sur-
prised if they did. One way of conceptualising the principles is to view them not
as blind facts, but as beliefs about therapy (Wills, 2005), with ‘belief’ in this con-
text following the definition of Alford and Beck as ‘ideas that are worthy of
being tested’ (1997: 85). Thus, by introducing principles as testable hypotheses,
such as ‘therapy works better when it is structured’, we are introducing trainees
to the ideas of questioning, finding out, and hypothesis testing, which resonate
throughout cognitive therapy.
What Padesky (1996) calls ‘therapist beliefs’ – the beliefs of individual ther-
apists about what they are doing in therapy – may also come into play. For
example, the therapist may believe that ‘clients will not like it when I ask them
to do homework’. Therapist beliefs may be in the form of ‘learning heuristics’
or rules governing the way in which the trainee will implement the therapy
principles. In a longitudinal study of cognitive therapy trainees, Wills (2005).
found that many trainees had an initial learning heuristic that told them that
‘I must follow the steps of cognitive therapy very exactly, otherwise it will not
work.’ This heuristic often led to rigid and ineffective practice. A more effec-
tive and personal style only developed when trainees were able to change this
heuristic to something more like, ‘I can try to do this my way and see if it
works’, thus introducing an experimental and discovery mode into the learn-
ing process. Counsellors learning cognitive therapy experienced the most dif-
ficulty when they were trying to implement the structure of therapy, especially
in setting an agenda (Wills, 2005). This seemed to relate to a conjunction
between a previous immersion in person-centred counselling and sets of ‘ther-
apist beliefs’ that defined structuring sessions as being ‘directive’ and disre-
spectful of the client. Sometimes these beliefs would be accompanied by
catastrophic predictions about the client’s likely reaction to such ‘disrespect’.
Again, such beliefs are open to be tested, by trying out the difference between
bringing little structure or more structure to the sessions.
Other cognitive therapy skills seem, in their own right, to be particularly hard
to learn to do really well. Trainees often report difficulties in learning with
Socratic dialogue, finding it difficult not to ‘lead’ clients towards certain answers
that seem obvious.
Figure 1 The Experiential learning cycle (Bennett-Levy et al., 2004; adapted
from Lewin, 1952; Kolb, 1984).
Sometimes the client is saying something like, ‘I’m such a failure, so useless’, or
whatever, and you just feel like saying to them, ‘oh, for God’s sake, what about
the two wonderful kids you have brought up?’ Not very Socratic, eh? (Wills,
research interviews, 2005)
This may be because at this moment, the therapist has moved from a ‘guided dis-
covery’ perspective to a ‘changing minds’ perspective (Padesky, 1993) Padesky
makes the point that one of the characteristics of good guided discovery may be
that the therapist does not always know where the dialogue is heading and is
genuinely open to wherever it may lead. If we translate that thought into the
training situation, it may be that there are times when the trainer has to find that
same interest in the ‘wanderings’ of the trainee as he or she lurches towards the
‘aha’ moment of learning.
In order to learn cognitive therapy, particularly for those who have several years
or decades of experience in other therapeutic models, trainees might have to go
through a considerable period of confusion and discomfort while their therapy
‘beliefs’ and learning heuristics adapt to their new situation. The period of confu-
sion has a striking similarity to the phenomenon of aporia, described in Chapter 5
of Cognitive therapy, second edition (Sanders and Wills, 2005) as a necessary part
of cognitive change, facilitated through Socratic dialogue and guided discovery.
Such confusion is a defining experience of training, as described in the seminal
paper by Worthless et al. (2002). There may be trainees who have moved effort-
lessly towards competence but we would struggle to recall them. However, the
cognitive dissonance generated by such confusion may be a necessary motivator
for movement towards insight. If they wish to help trainees through this period of
confusion, trainers in cognitive therapy may wish to ensure that their teaching
methods are as Socratic as they would want their trainees to be with their clients, as
strongly suggested by Christine Padesky and others (Padesky, 1996). Training in cog-
nitive therapy cannot therefore just be about ‘imparting knowledge’ in a context-free
way but should, as far as possible, mirror the principles of cognitive therapy itself.
The concept of ‘training heuristic’ used in Cognitive therapy, second edition
can be applied to trainees’ search for meaning and competence, clearly heard in
this reflective account of a trainee after finishing training:
I am normally a confident person and I back myself but for a short while there, in
the middle of training, I lost that trust. I thought that I had to apply this model as
a 1, 2, 3, 4, 5 … And it really wasn’t working … but it became like a safety
blanket … 1, 2, 3 … And then I thought, well, this is no good … and at that same
time, I talked to the tutor and he said, ‘A., you could just try doing it’ and I thought,
yes, come on A., just do it … And I realized that I could do it 1, 2, 3, 4 and it would
still be wrong … So I just did it, I did it in my own way, I suppose, and it worked,
and, to my amazement, the tutors thought it worked too! (Wills, 2005).
This ‘silent’ dialogue going on inside the trainee’s head is very reminiscent of the
way Beck (1976) describes finding out about the two streams of thinking going
on within some of his patients. This discovery led on to the concept of ‘negative
automatic thoughts’ and to the basis of the theory of cognitive therapy. Just as
these silent ways of negative thinking could be related to feelings of dysphoria
and anxiety in patients, it is easy to see how negative training heuristics can be
associated with such feelings in trainees. A finding that surprised me in my
research (Wills, 2005) was the pervasive anxiety that seems to surround all
aspects of training, captured in Worthless et al.’s paper (2002) Sometimes it is
helpful to link what one might call ‘negative thoughts about training’ and anxi-
ety with trainees, although, in our experience, they may be guarded in discussing
them until after being awarded their certificates of competence! The training
arena in which these thoughts are most likely to come up is within case discus-
sions or supervision groups. Being aware of negative training thoughts and feel-
ings can help a supervisor to respond more sensitively and helpfully to the needs
of trainees. Such thoughts and feelings do not of course entirely disappear after
the completion of training and so can also be helpfully addressed in ongoing
Personal Therapy and Cognitive
In many counselling and psychotherapy traditions, personal therapy is seen as
an essential requirement for training and practising as a therapist (Wilkins,
2006) and therapists rate this as an important aspect of training (Macran and
Shapiro, 1998; Orlinsky et al., 2001). Although it is accepted wisdom that per-
sonal therapy is useful, there has not been a great deal of research on its effect
on subsequent therapeutic practice, nor on what learning mechanisms can
explain its effectiveness (Bennett-Levy, 2002). In a review, Macran and Shapiro
(1998) found that personal therapy improved empathy, warmth and genuineness
and, given the importance of these factors in therapy, such improvements in
themselves are likely to have a good impact on outcome.
By contrast, in the world of cognitive and behavioural therapies, personal
therapy is not currently seen as an essential, or widely accepted, aspect of train-
ing (EABCT, 2001). The British Association for Behavioural and Cognitive
Psychotherapies (BABCP, 2000), while not insisting on personal therapy for
accreditation, does stipulate: ‘therapists must ensure that they can identify and
manage appropriately their personal involvement in the process of cognitive
and/or behaviour therapy’ and ‘therapists must have developed an ability to
recognise when they should seek other professional advice’.
Counsellors and therapists from other traditions have reported being some-
what mystified by such lack of interest in individual therapy. What they have
gained from their own therapy – an awareness and understanding of their own
issues and how these might impact on the therapeutic relationship, the experi-
ence of being a client, a place to take difficulties that might arise during training
or as a result of client work, material with which to conceptualize difficulties in
therapy – are all seen as equally valid within cognitive therapy but have not,
until recently, been explicitly focused on. In addition, we know that reflection is
an important aspect of learning, and personal therapy provides the opportunity
to reflect on both professional and personal issues. Part of the reason for the lack
of focus on personal therapy may be historical: many of the ‘old school’ practi-
tioners of cognitive therapy came from behavioural traditions and clinical
psychology, where personal therapy has not always been seen as essential or even
relevant. Those seeing cognitive therapy as an educational model, involving the
application of appropriate techniques, again might see therapy for therapists as
Given the strong belief in the benefits of personal therapy for therapists, it
may be rather baffling that there is little evidence of its overall effectiveness
(Roth and Fonagy, 1996). This may be because of the way it has sometimes been
implemented as a compulsory element in some training courses. We think that
this may not be a good enough reason to undertake therapy, partly because it is
important for the ‘client’ to be able to decide when the time is exactly right for
this commitment. We are also aware that undertaking therapy in this way has
led to some quite adverse and even abusively exploitative outcomes for some of
the trainees and therapists involved.
We are glad to report the need for personal experience of therapy is increas-
ingly recognised within cognitive therapy and that our colleague, James Bennett-
Levy, has devised a method for facilitating such an experience in a way that
overcomes some of the difficulties described earlier by formatting it as a safer
and more educational experience. Working in Australia and more recently in
Oxford, he has developed a training method called Self Practice/Self Reflection,
where trainees undergo cognitive therapy with a training partner, and reflect in
writing on the process of each session, thinking through the implications of the
therapy experience for themselves, for their clients and for cognitive therapy.
Bennett-Levy found that SP/SR impacted on therapy in a number of ways. The
trainees reported a ‘deeper sense of knowing’ of cognitive practices. They gained
a deeper understanding of therapy, understood themselves better and demon-
strated improvement in cognitive therapy skills (Bennett-Levy, 2001; Bennett-
Levy et al., 2003). They also noted a re-emphasis on therapeutic relationship
skills: ‘the experience of being “in the client’s shoes” demonstrated starkly some
of the anxiety and difficulties in making changes, even as high functioning indi-
viduals; and served to emphasize how valuable empathy, understanding, respect,
tolerance and guidance of the therapist is’ (Bennett-Levy et al., 2003: 150) The
study found that SP/SR helped trainees to develop self-reflection, enabling them
to reflect both during and after sessions. We know that client perception of empa-
thy is correlated with positive outcome; and if SP/SR leads therapists to be more
empathic, as judged by our clients, then it is likely to lead to better outcomes. A
review of SP/SR in CBT training (Laireiter and Willutzki, 2003) confirms
Bennett-Levy’s findings, showing that trainees report substantial personal and
professional gains from using SP/SR. They report improved self-insight and self-
awareness, and a better understanding of the role of the therapist and the process
of therapeutic change, as well as a better understanding of and skills in CBT
methods. As summarised by Laireiter and Willutzki: ‘Although empirical evi-
dence is not extensive at present, it supports the notion that most of these goals
may be attained by a combination of person- and practice-related self-reflection
together with self-practice of CBT methods’ (2003: 28)
Personal therapy or SP/SR may take a while to filter through to become incor-
porated within cognitive therapy training, but is certainly becoming recognised
as a means of improving understanding of ourselves and of personal aspects of
our clinical practice.
LEARNING THE STRUCTURE: THERAPEUTIC
ARTISTRY VERSUS STANDARD THERAPY
FOR STANDARD PROBLEMS
… nor should standard, protocol-based practice stifle innovation, but should
serve as a springboard to it, since there are many situations in the psychothera-
pies where there is no research evidence or even clinical consensus to indicate
the best approach. (Parry, 1996)
Once the basics are learned, a competent cognitive therapist develops an artful
ability to conceptualize interlocking problems, make intervention choices, and
solve problems in an efficient and effective manner to facilitate client learning
and change. These ideals are challenging to achieve. (Padesky, 1996: 269)
Many therapy procedures are now well written up into ‘manual’ and ‘protocol’
formats, and one approach to training can include introducing trainees to the
use of these materials. On the one hand, cognitive therapy does promote the
importance of protocols, since research has suggested that therapists who do not
follow the structure are not as effective as those that do (Schulte et al., 1992;
Wilson, 1997) On the other hand, however, developing an individual conceptuali-
sation for each client we see, and following therapy accordingly, is recommended
(Butler, 1998; Jacobson et al., 1999) Jacqueline Persons (1989) argues that a fail-
ure to use individual conceptualisation means that our treatment would be
overly dictated by symptom identification. She uses the analogy of treating fever.
If we were to only treat fever according to the symptoms, we would not be able
to distinguish between fever caused by malaria and that caused by pneumonia,
and this would mean that we would be likely to make treatment mistakes.
Learning to balance a structured, standard approach in therapy with working
with individual clients can be an extremely challenging part of training, particu-
larly for those therapists used to a looser, client-centred and client-directed way
of working, who fear that structure will lose the essence of working with the
person rather than with symptoms. However, to reject standard therapy on this
basis may risk losing the therapy’s effectiveness or at least diluting its value. The
therapeutic procedures of cognitive therapy have gained a lot from being clearly
delineated and structured, and there are strong arguments for therapists to
follow them closely. For many trainees, using a standard protocol may come as
a relief, offering clear guidelines as to what to do in therapy. But even this is not
without its problems.
Henry et al. (1993) report, admittedly in the context of training in psycho-
dynamic therapy, that training using protocols and/or manuals is likely to
result in therapist hostility towards clients – presumably when the clients do
not conform to the client behaviour required by the protocol – and thus to
adverse results. The problem raised by Henry and colleagues is one that is fre-
quently reported in training. Trainees, coping with the aporia crisis of leaving
one form of practice behind without having fully developed a new form of
practice, see the process of mastering the protocol as a way out of this anxi-
ety. They therefore frequently report going through a stage where they over-
apply the protocol or the model:
I seemed to spend more time thinking about applying the model than thinking
about the person in front of me. This felt very uncomfortable to me. In retrospect,
I think that I got into a way of thinking about how to do cognitive therapy that was
almost neurotically structured … (cognitive therapy trainee, quoted in Wills, 2005)
More optimistically, however, trainees in the Wills (2005) research reported that
as time went on they were able to reconceptualise the situation with a learning
heuristic that was more helpful to them. The solution for these trainees seemed
to lie in developing the paradoxical ability to follow the protocol in their own
way and in a way that was responsive to the needs of their individual clients.
In some ways the protocol versus individual therapy argument has become more
polarised than it needs to be, and successful therapy involves a balance between
‘science’, or following the structure, or ‘art’, working with the individual. Artistry
and science are clearly evident in Aaron Beck’s therapeutic abilities and his ideas. It
is perhaps significant that, for cognitive therapy, he has been an advocate for both
modes of operation. This fact in itself may give us hope that the two approaches are
not necessarily mutually exclusive. Faced with a bewildering mass of issues and
symptoms, a therapist can find help in both the individual conceptualisation and in
the treatment methods and rationales of the manuals.
There are, also, different understandings of what structure means in practice.
One notion would hold that inside the therapist’s head there is a very clear and
structured scheme of rules on how to proceed and this is matched by another
consisting of clear and structured therapeutic behaviours. Another notion is that
there is a clear and precise structure of rules inside the therapist’s head but that
the ‘art’ of therapy depends on using the rules to steer between different poten-
tial actions. Alford and Beck use the term in both these different ways: ‘The ther-
apist needs a clear structure … a clear rationale on which to base a way of
proceeding … without that it will be difficult to maintain a scientific stance in
practice’ (1997: 72); ‘Structure is a defining characteristic of cognitive therapy
and a reliable way to discriminate between it and other therapeutic modes … [it
includes] … the presence of an agenda, setting homework and asking the client
for feedback at the end of the session’ (1997: 76)
The first usage may be deemed as a kind of structured way of thinking about
and understanding a situation, a way of conceptualising or formulating a client’s
history, for example. The latter we might better call ‘steps that structure action’.
While a scheme of such steps can act as a facilitator of clear work with clients, it
may also be open to the danger of resulting in mechanical work. This is why
trainees also need to be ‘wired to the cognitive way’ (i.e. having a thorough way
of using cognitive conceptualisations), as Young (in Weishaar, 1993: 38.) puts it.
He goes on to say: ‘Technical cognitive therapists … [are] … not really attuned to
the interpersonal component of the therapy, just with delivering a technique’
(1993: 39) What we are talking about here may be the art of learning to sail a
yacht by catching the wind rather than by resorting to the outboard motor.
In some ways, cognitive therapy may itself have to accept at least part of the
blame for the state of things regarding its implementation. As a discipline, it has
been over-reliant on highly quantitative outcome research and has not encour-
aged much qualitative process research, with the notable exception of the
in-depth, qualitative work of Bennett-Levy (2002) A strongly quantitative bias
may have been necessary in the opening phase of the history of cognitive
therapy but development of qualitative research of cognitive therapy is now long
overdue (Grant et al., 2004) As we will argue again later, we may need to develop
different research modes – probably those of the more qualitative, process-oriented
style – to get at more subtle questions such as, ‘when is it good to depart
from the structure/protocol and when is it necessary to stay with it?’ These are
questions that crop up time after time when an open learning approach is used
in cognitive therapy training. The question can also become a major focus of
SUPERVISION IN COGNITIVE THERAPY
Supervision in cognitive therapy probably occupies less of a ‘hinterland’ than it
does in some of the other major models of psychotherapy and counselling.
Although it has always been a part of the actual practice of the model, it has
not been much written of or theorised about, perhaps once again showing the
practical and pragmatic nature of the model. That it has always been there,
however, can be seen in the fact that Beck strongly argued for the presence of
cognitive therapy supervisors in the major research trial conducted by the US
National Institute for Mental Health in the 1980s (Weishaar, 1993)
There are signs now that cognitive and cognitive-behavioural therapists are
beginning to think much more about the nature of supervision (Bennett-Levy,
2002; O’Carroll, 1999; Padesky, 1996; Townend et al., 2002) There are two
main forums in which cognitive therapy supervision takes place: supervision as
part of a training course (more likely to be group supervision) and supervision
as part of ongoing development after training (more likely to be individual
supervision) Both authors have practised and experienced individual and group
supervision in both of these contexts. Supervision after training may be particu-
larly important in helping to prevent practice ‘reverting to the mean’ after train-
ing. It is easy to understand how, when the frequently intensive atmosphere of
the training situation is left behind, therapeutic practice can, first, stop develop-
ing and, then, revert back to previously established models of practice (MacKay
et al., 2000) Post-training environments, despite what quality assurance and mis-
sion statements may say, may not nurture innovative practice, even when it is
empirically supported (Wills, 2005) There is also a significant problem whereby
highly trained therapists may be quite quickly promoted into posts where they
do less, or sometimes no, direct therapy work (Ashworth et al., 1999) In addi-
tion, supervision can be informal peer supervision, as in a discussion between
colleagues, or self-supervision, where we allow time to reflect on our work
(Padesky, 1996) Christine Padesky’s (1996) excellent chapter on supervision in
cognitive therapy gives clear guidelines about how to bring cognitive therapy
principles into supervision. For example, she recommends that supervision
reflect the structure of cognitive therapy by using pre-planned questions as a
focus for each supervision session, with themes such as:
• Is there a cognitive model for understanding and treating this client problem
and is this being followed?
• Does the therapist have the knowledge and skills to implement therapy
• Is the therapy following expected patterns?
• What might be interfering?
CONTENT AND PROCESS IN COGNITIVE
Cognitive therapy is a vast and growing area of application, and in comparison to
the early models, where depression was the only area to learn about, now there is
much extra that one could learn about. Thankfully, this does not necessarily mean
that we have to know in detail the content of every intervention for every prob-
lem. Many of the different applications operate to the same sort of principles and
offer similar types of interventions using similar skills. The main variation, as indi-
cated earlier, would be on the development of a conceptualisation for particular
problems, mindful that our clients are individual people, whom we understand
and aim to help in individual ways. From a training and supervision point of view,
the variety in cognitive therapy is really more about ‘being able to find out about’
rather than ‘knowing about’. Curiosity is a quality that is good for trainees and
supervisees to foster, backed up by the ability to think in a clear and scientific way.
Therefore, knowledge about the content of therapy for particular client diffi-
culties will probably always play quite a significant role in cognitive therapy
supervision. Being able to use diagnostic system criteria to get in the right ‘ball-
park’ with clients is often something that more senior practitioners can help
trainees with. However, we would stress the use of such criteria as ‘ballpark esti-
mates’ rather than rock-solid facts. Help with understanding symptoms may be
less relevant to more ‘seasoned’ supervisees – a good example of how super-
visory needs vary over the development of a career. Townend et al. (2002) found
in their survey of cognitive behavioural practitioners that ‘conceptualisation’
was the issue most raised in CBT supervision. Conceptualisation can contribute
to the review of either the content or process of the therapy.
It is also inevitable that issues concerning therapeutic process will often be
raised in supervision. Interestingly, I kept a supervision case book of issues raised
in supervision over a period of two years and this record shows that over two-
thirds of all supervision discussions were more focused on therapeutic process
than on therapeutic content. This was made all the more remarkable by the fact
that the greatest proportion of the supervision discussions had taken place
within the context of a training course. Perhaps the reason for this is that many
cognitive therapy techniques are relatively simple to execute and are often
backed by very clear rationales. This is not by any means saying that cognitive
therapy is easy to do but that its clarity means that most of the difficulties occur
in relation to its application, rather than to its execution.
Process issues in cognitive therapy supervision often seem to carry the flavour
of ‘I’m doing great cognitive therapy but the client isn’t!’ Some examples of
typical discussions are presented in Figure 2.
We described how such cognitive-interpersonal events can be understood and
worked with in cognitive therapy in Cognitive therapy, second edition. The
‘seven-eyed’ model developed by Hawkins and Shohet (2000) can be very useful
for identifying and working with such ‘interpersonal markers’ as those noted
earlier (see Figure 3)
The first three process modes of the Hawkins and Shohet model relate mainly
to fairly conscious material content of the therapy. The other process modes
relate mainly to less conscious processes in both the therapy and the supervision.
Processes 1 and 2 of the Hawkins and Shohet model focus on the content of the
therapy sessions and the strategies being used in them respectively.
In the case of client M. (see Figure 2), the client had been diagnosed as being
dysthymic and was indeed very emotionally flat and ‘hard to engage’. One of her
criticisms of the therapy, however, concerned filling out an inventory to measure
her depression. The therapist uses many of these inventories and to save resources
and the rainforest had started getting clients to fill them out in pencil and then
erasing their marks for reuse later. M. had objected to this and had asked for a
fresh sheet each time. This was the first time that any client had made this point
and the therapist had begun to wonder if the client might have an entitlement
schema (Young and Klosko, 1994) The supervisor helped him think this through
and how this might link to the choice of which strategies to use, for example,
whether he should agree to this request or not. What would be the implications
of acceding or not acceding to the request for the therapeutic relationship, the third
process mode of the Hawkins and Shohet model? The therapist spontaneously
began to consider, ‘I wonder if I like this client? What is she doing to me? Am
I a bit attracted to her?’ This is the therapist process focus – process mode 4
of the model. The supervisor was now aware of a reaction in himself:
‘I have to make this therapist stand his ground here. I can’t let him be pushed
around by this client. This has been a weak side revealed in his work before.’
This is process mode 5 of the model: focusing on the supervisory relationship
itself. Finally, the supervisor decided to step back from this latter reaction and
ask himself, ‘Am I getting drawn into this? Is this perhaps driven by my stuff
about my strong need to patrol my own boundaries? I could just go to a more
I get the impression that this client doesn’t suffer fools gladly and I wonder if she
sees me as a fool!
M. is an extremely bright but depressed student client. She is very committed to doing
CT therapy and indeed sees it as the answer to all her problems. The supervisee had
only seen this client three times but on each occasion she had made a criticism of the
way the therapy was being conducted and each time had asked for specific changes in
the way it was being done. The supervisee made the above comment.
I described one aspect of the client’s behaviour as ‘over the top’ and he stormed
out of the room.
F. was a depressed middle-aged man who was finding it difficult to sustain relationships
with women. The female supervisee was exploring with the client his thoughts and
behaviours as he ‘showered his girlfriend with gifts’, and tentatively wondered how the
girlfriend might experience it. The client reacted badly – though he did call back and
some discussion followed. The supervisee was able to reflect on how the client was
affecting her and how she might use such ‘cognitive interpersonal’ information better in
Figure 2 Process Issues in Cognitive Therapy Supervision
neutral position and help the therapist review the pros and cons of acceding or
not acceding to the request. He’d probably learn more from that.’ This is super-
vision process mode 6: focusing on the supervisor’s process.
Since devising the six process modes of the original model published in 1989,
Hawkins and Shohet (2000) have added a seventh process mode in their second edi-
tion. The seventh process mode concerns how the therapy and the supervision relate
to the wider context of the therapy and supervision. The case of client F. (see Figure
2) could profit by analysis from the perspective of the same six modes that have
been applied to client M., above. In this case, however, the seventh process mode
may also become relevant. Clients who storm out of sessions may choose to go into
‘complaint mode’. Most organisations now have well worked-out, formalised ways
of dealing with complaints and these may relate to wider quality assurance systems.
Such systems and procedures may strike us as bureaucratic and/or ‘covering one’s
backside’ but do at least offer some protection to potentially vulnerable human ser-
vice workers. Single-handed practitioners may not expect to get caught up in such
situations and may have very little in the way of ‘organisational’ coping systems in
place. Good supervision would not only help the therapist reflect on the practice
issues, as per Figure 2, but also on the systems and potential personal liability issues
in such incidents. This might include severely practical aspects concerning whether
the incident had been recorded and whether any witnesses were available. In gen-
eral, supervision in therapy and counselling has been good and proactive in dealing
with psychological process issues, and passive and poor in dealing with system
codes and ethics
Figure 3 The seven-eyed model of supervision
process issues. Townend et al. (2002), for instance, report that, in their survey of
supervision, therapist safety was never discussed.
A further systems issue that has arisen in the nascent literature on supervision
for cognitive therapists is that of ‘dual relationships’ (Townend et al., 2002) In
the NHS, for example, where there are strong management systems, conflict
may arise when one person is trying to hold the roles of both ‘managerial super-
vision’ and ‘developmental supervision’. Townend and colleagues report that
supervisees did not generally report this as a problem but suggest caution in this
regard. They further suggest that a written contract – covering such matters as
expectations, methods and procedures – can act as a useful protection against
contamination effects between the two roles.
Townend et al.’s (2002) survey reports that many cognitive practitioners are
influenced by the supervision models offered by Padesky (1996) and Beck and
Liese (1997) These models are based on the desirability of congruence between
the way therapy is conducted and the way supervision is conducted. This point
covers issues such as congruent structure, setting agendas, goals and using feed-
back to guide the process, and congruent methods, such as conceptualisations,
role-plays and use of tapes. In this way, the supervision session mirrors cognitive
therapy itself, and is a powerful means of experiencing the process while work-
ing on the supervision issues themselves.
CONCLUSION: HELPING THERAPISTS
TO ESTABLISH THEIR OWN STYLES
Cognitive therapy walks a narrow line between scientific precision and thera-
peutic artistry. As the cognitive model has developed, it has become abundantly
clear that the therapy itself and the training and supervision modes informed by
this model are usually accompanied by subtle interpersonal processes. Such
processes were not entirely expected in the early days of cognitive therapy.
Interpersonal processes receive much of their subtlety from their idiosyncratic
nature, a factor relatively easily integrated into a cognitive approach, based as it
is on uncovering individual appraisal and meaning.
One of the challenges facing trainee cognitive therapists, and indeed those
longer in the tooth, is in balancing structured, standard therapy with working
with our individual clients. There is no doubt that the therapeutic procedures of
cognitive therapy have gained much from being clearly delineated and structured.
There are therefore strong arguments for therapists to follow them quite closely.
These arguments also stand in favour of supervisors helping trainees and super-
visees to follow them quite closely. The devil, however, is quite often in the detail.
How closely should cognitive therapists stick to the protocol and what licence can
they give to their therapeutic artistry?
We wonder if the debate between these two approaches has not become rather
polarised into an either/or mode. It seems extremely unlikely that one therapist
would implement even highly specified treatments in exactly the same way as
another therapist or that either of them would work in exactly the same way
with different clients. Our own experience tells that even when a client seems to
be very much ‘with the programme’, there are hundreds of little decisions
specific to that client and that therapy that still have to be made. Equally, even
when client and therapist are not following a manual, it is also extremely
unlikely that the therapy is not following some kind of sequence known to the
therapist from his or her reading of the literature and from previous experience
with clients. It is likely then, that there is far more of a continuum between high
and low structure than is sometimes implied in the current debate.
Even when following an established sequence of therapy, clinicians themselves
consider that their clinical judgement is a vital part of the equation for success
(Butler, 1998; Nelmes, 2001) Huey argues that therapy would be likely to be
enhanced by ‘effective clinical reasoning, i.e., teaching future practitioners how
to, rather than what to, think about complex psychopathology’ (1999: 2)
It would be helpful now to have more research on the kind of small decisions
that sometimes can keep things on track but at other times allow for successful
diversion. Such research would give guidance on what one might called the ‘nat-
ural history’ of the therapy process. The forum of supervision, both during and
after training, seems to be the most likely location where such data would natu-
rally accumulate. Within psychotherapy generally, supervision has been more
prevalent and prominent than it has been in the cognitive and behavioural field
and yet has also been somewhat under-researched. A recent review of research
on supervision in counselling and psychotherapy (Wheeler, 2003) has, however,
revealed a new focus on outcome and it is now time for cognitive therapists to
add its experience to this developing and important field.
Bennett-Levy, J. (2006) Therapist skills: their acquisition and refinement.
Behavioural and Cognitive Psychotherapy, 34 (1): 57–78.
Goldfried, M.R. (ed.) (2000) How therapists change: personal and profes-
sional reflections. Washington, DC: American Psychological Association.
Padesky, C.A. (1996) Developing cognitive therapist competency:
teaching and supervision models. In P.M. Salkovskis (ed.), The frontiers
of cognitive therapy (pp. 266–92) New York: Guilford Press.
1 The cognitive therapists also tended to dress more conventionally. Goldfried
(2000) claims that, as a behaviourist travelling to the humanistic Mecca at Esalen
in the 1970s, he used a two-sided case so that the two appropriate dress styles
could be kept separate. The Esalen side was smaller, clothing being, at that time
in Esalen, optional!
Alford, B.A. & Beck, A.T. (1997) The integrative power of cognitive therapy. New York:
Ashworth, P., Williams, C. & Blackburn, I.-M. (1999) What becomes of cognitive therapy
trainees? A survey of trainees’ opinions and current clinical practice after postgraduate
cognitive therapy training. Behavioural and Cognitive Psychotherapy, 27 (3): 267–77.
BABCP (2000) Minimum training standards for the practice of CBT. London: BABCP.
Beck, A.T. (1976) Cognitive therapy and the emotional disorders. New York: International
Beck, A.T. (2004) Origin, evolution and current state of cognitive therapy: The inside
story. Keynote address, Congress of the European Association of Behavioural and
Cognitive Therapies, Manchester, September.
Beck, A.T. & Emery, G. with Greenberg, R.L. (1985) Anxiety disorders and phobias: a
cognitive perspective. New York: Basic Books.
Beck, A.T. & Liese, B.S. (1997) Back to basics: fundamental cognitive therapy skills for
keeping drug-dependent individuals in treatment. National Institute on Drug Abuse
Research Monograph Series, 165: 411–30.
Beck, J. (1995) Cognitive therapy: basics and beyond. New York: Guilford Press.
Bennett-Levy, J. (2001) The value of self-practice of cognitive therapy techniques and self-
reflection in the training of cognitive therapists. Behavioural and Cognitive
Psychotherapy, 29: 203–20.
Bennett-Levy, J. (2002) Navel gazing or valuable training strategy? Self-practice of
therapy techniques, self-reflection, and the development of therapist expertise. In
J. Henry (ed.), First European positive psychology conference proceedings. Leicester:
British Psychological Society.
Bennett-Levy, J., Lee, N., Travers, K., Pohlman, S. & Hamernik, E. (2003) Cognitive ther-
apy from the inside: enhancing therapist skills through practising what we preach.
Behavioural and Cognitive Psychotherapy, 31: 143–58.
Bennett-Levy, J., Butler, G., Fennell, M., Hackmann, A., Mueller, M. & Westbrook, D.
(2004) The Oxford guide to behavioural experiments in cognitive therapy. Oxford:
Oxford University Press.
Butler, G. (1998) Clinical formulation. In A.S. Bellack & M. Hersen (eds), Comprehensive
clinical psychology. Oxford: Pergamon, (pp. 1–24).
European Association of Behavioural and Cognitive Therapies (EABCT) (2001)
Minimum training standards. London: EABCT.
Goldfried, M.R. (ed.) (2000) How therapists change: personal and professional reflec-
tions. Washington, DC: American Psychological Association.
Grant, A., Mills, J., Mulhern, R. & Short, N. (2004) Cognitive behavioural therapy in
mental health care. London: SAGE.
Hawkins, P. & Shohet, R. (2000) Supervision in the helping professions: an individual,
group, and organisational approach. Buckingham: Open University Press.
Henry W.P., Strupp, H.H., Butler S.F., et al. (1993) Effects of training in time-limited
dynamic psychotherapy: Changes in therapist behavior. Journal of Consulting and
Clinical Psychology, 61: 434–40.
Honey, P. & Mumford, A. (1992) The manual of learning styles. Maidenhead: Peter
Huey, D. (1999) The potential utility of problem-based learning in the education of clinical
psychologists. British Association for Behavioural and Cognitive Psychotherapies Annual
Jacobson, N.S., Schmaling, K.B., Holz-Munroe, A., Katt, J.L., Wood, I.F. & Follette, V.M.
(1999) Research structured versus clinically flexible versions of social learning based
marital therapy. Behaviour Research and Therapy, 27: 173–80.
James, I.A., Blackburn, I.M. & Reichelt, F.K. (2000) Manual of the revised cognitive therapy
scale. Newcastle upon Tyne: Newcastle Cognitive & Behavioural Therapies Centre.
Kolb, D.A. (1984) Experiential learning: experience as the source of learning and devel-
opment. Englewood Cliffs, NJ: Prentice Hall.
Laireiter, A.R. & Willutzki, U. (2003) Self-reflection and self-practice in training of cognitive
behaviour therapy: an overview. Clinical Psychology and Psychotherapy, 10: 19–30.
Lewin, K. (1952) Field theory in social science: selected theoretical papers. Edited by
Dorwin Cartwright. London: Tavistock.
MacKay, H., West, W., Moorey, J., Guthrie, E. & Margison, F. (2000) Training in
psychodynamic-interpersonal therapy: primary care counsellors’ experiences of chang-
ing their practice. British Association for Counselling and Psychotherapy Research
Conference, Manchester, May.
Macran, S. & Shapiro, D. (1998) The role of personal therapy for therapists: a review.
British Journal of Medical Psychology, 71: 13–25.
Nelmes, M. (2001) Therapy protocols versus individual conceptualisation: the views of
CBT therapists. Unpublished MA thesis, University of Wales, Newport.
O’Carroll, P. (1999) Cognitive-behaviour therapy supervision: evidence-based? Paper
given at BABCP Annual Conference, Bristol, July.
Orlinsky, D.E., Botermans, J.F. & Rønnestad, M.H. (2001) Towards an empirically
grounded model of psychotherapy training: four thousand therapists rate influences on
their development. Australian Psychologist, 36: 139–48.
Padesky, C. (1995) Cognitive therapy of anxiety: key treatment principles and methods.
Audiotape. Newport Beach, CA: Center for Cognitive Therapy. www.padesky.com
Padesky, C. A. (1993) Socratic questioning: Changing minds or guided discovery? Congress
of the European Association of Behavioural and Cognitive Therapies. London,
Padesky, C.A. (1996) Developing cognitive therapist competency: teaching and supervision
models. In P.M. Salkovskis (ed.), Frontiers of cognitive therapy. New York: Guilford Press,
Parry, G. (1996) NHS psychotherapy services in England: review of strategic policy.
London: NHS Executive.
Persons, J.B. (1989) Cognitive therapy in practice: a case formulation approach.
New York: W.W. Norton.
Roth, A. & Fonagy, P. (1996) What works for whom? A critical review of psychotherapy
research. New York: Guilford Press.
Sanders, D. & Wills, F. (2005) Cognitive therapy (2nd edn). London: SAGE.
Schulte, D., Künzel, R., Pepping, G. & Schulte-Bahrenberg, T. (1992) Tailor made versus
standardized therapy of phobic patients. Behaviour Research and Therapy, 14: 67–92.
Townend, M., Iannetta, L.E. & Freeston, M.H. (2002) Clinical supervision in practice: a
survey of UK cognitive behavioural psychotherapists accredited by the BABCP.
Behavioural and Cognitive Psychotherapy, 30: 485–500.
Trepka, C., Rees, A., Shapiro, D.A. & Hardy, G.E. (2004) Therapist competence and out-
come of cognitive therapy of depression. Cognitive Therapy and Research, 28: 143–57.
Weishaar, M.E. (1993) Aaron T. Beck. London: SAGE.
Wheeler, S. (2003) Research on supervision of counsellors and psychotherapists:
a systematic scooping study. Rugby: British Association for Counselling and
Wilkins, P. (2006) Professional and personal development. In C. Feltham & I. Horton
(eds), Handbook of counselling and psychotherapy (2nd edn). London: SAGE,
Wills, F. (2005) The influence of trainees’ beliefs about therapy on the acquisition of com-
petence in CBT training. Paper presented at the British Association for Counselling and
Psychotherapy Research Conference.
Wilson, G.T. (1997) Treatment manuals in clinical practice. Behaviour Research and
Therapy, 34: 295–314.
Worthless, I.M., Competent, U.R. & Lemonde-Terrible, O. (2002) Cognitive therapy
training stress disorder: a cognitive perspective. Behavioural and Cognitive Psychotherapy,
30 (3): 365–74.
Young, J. & Beck, A.T. (1988) The cognitive therapy rating scale manual (rev. edn).
Philadelphia, PA: Center for Cognitive Therapy, University of Pennsylvania.
Young, J. & Klosko, J. (1994) Reinventing your life. New York: Plume.