For trainers

Learning the Conversational Model

Before the model was even described, Robert Hobson had already pioneered the use of audiotape and videotape feedback for teaching and supervision. He established the routine use of recordings in supervision (in those days in the form of audiotapes and less commonly videotapes) and he was one of the pioneers in the use of videotape as a teaching medium to model how therapists work. From the outset, learning from feedback was intrinsic to this way of working. Hobson developed a repertoire of ways of introducing new therapists to effective practice, and his systematic use of feedback from sessions is still as relevant today.

Using the films in training

The Cases films can be used as standalone examples of therapy with discussion about how to formulate the problems presented and discussion about what approach might be most appropriate for the particular client.

They can also be used as examples to try out different tools that are in the Learner section on How to use the Resources

So, for example, you might encourage someone you are training to try transcribing a section, then looking for key words and phrases, then identify the interventions being used and then a tentative formulation.

Working in groups is ideal, but the films are designed for use individually as well.

They were done as live role plays and are not intended to “perfect therapies”. So there are bound to be areas where you can discuss alternative approaches that may have been appropriate.

The “Minute Particulars” series draws from the research described below- again they can be used in a group or alone. The films deal with three levels of skill (Introductory, Intermediate and Advanced) starting with simple examples where the narrator gives suggestions of what to say. These examples can be used to familiarise the group with the process of actually saying out loud what you would have said in the therapist’s place.

In the later examples in each section and especially in the later films there is more room for discussion of options.

There is a symbol that comes up when a response is prompted- the prompt is just a few seconds but you can pause for fuller discussion.

If you are meeting a group regularly you can mix the films with role plays as described below.

Backing from research

The process of developing the Conversational Model in a research context is fully described in the original reports (Maguire et al., 1984, and Goldberg et al., 1984), but the research is relevant here primarily in terms of the eventual development of teaching methods. The research team had been influenced by the work of Ivey et al (1968) in the use of microskills teaching of counselling and therapy. 

The micro-skills method involves the definition, modelling and practice of desirable therapeutic behaviours until there is mastery of each component skill. However such teaching is time intensive and could benefit from a degree of simplification. With this in mind, an aim of the research was to automate the earlier parts of skill acquisition as far as possible and to concentrate teacher time on more complex issues such as timing and choice of intervention after the trainee had learned and practised the basic components.

After piloting several approaches, a set of three videotapes was compiled from extracts of the actual therapies carried out in the validation phase of developing the model. This had involved five therapists with, and five without, prior experience of being trained by Robert Hobson. The researchers identified aspects of the model that were congruent with his teaching that were also observed in realistic clinical use by those trained by Hobson as well as by Robert Hobson himself.

The first tape was a compendium of these basic skills:

• how to start the interview
• using statements rather than questions
• using a negotiating style
• using personal pronouns (“I” and “We”)
• staying with the “here and now”
• identifying verbal, vocal, non-verbal and internal cues
• expressing hypotheses (understanding, linking and explanatory)
• use of metaphor

These skills have been further elaborated into the form of the Manual described in Chapters 4 to 9, but at the time it was very unusual to use a microskills approach for a psychodynamic or relational approach.

The second tape was a longer example with a narrative explaining how the different aspects could work together as a whole with a linked commentary. In the third tape, each concept was exemplified and followed by practice excerpts following a standard format. A short vignette demonstrating the relevant type of intervention from a therapy session was shown followed by examples of increasing difficulty. After showing each practice example there was a ten-second pause when the trainee was encouraged to say out loud how he or she would have responded using the relevant type of intervention. If the trainee became stuck the teacher could pause the tape to allow discussion and further practice, but otherwise the tape would continue to allow the trainee to listen to how the therapist actually responded. The tape used typical rather than ideal examples to give the trainee confidence in using the model at a basic level.

The whole package was accompanied by a set of teaching notes explaining the principles of the model and how to use the tapes. After watching the videotapes the trainees formed seminar groups with three trainees and a teacher. The trainees played recordings of their own cases with the group paying close attention to how different aspects of the model could be applied in practice. This was not only an efficient use of teacher time but also established a peer group with members facilitating the progress of other members.

The teaching package was evaluated with trainees who had had no previous experience as therapists. The evaluation assessed performance in specified areas at three points: before any training; between the video teaching and the seminar group; and finally after the complete package.

Summary of results

The tapes alone had a profound impact on therapist behaviour. Initially they had a style that was characterised as “interrogative” with very high rates of closed questions, and a didactic somewhat authoritarian manner. After training, this shifted to a more mutual or “negotiated” style. The later interviews typically showed negotiation and a degree of tentativeness, used personal pronouns such as “I” or “We”. The therapists also adopted a conversational structure encouraging exploration in the “here and now”, and used other key behaviours such as “understanding hypotheses”. The role-played sessions also had an overall shape with attention to opening the session through an introduction by name, statement of the length and purpose of the interview and an initial focus on immediate feelings in the session. Similarly the sessions after training gave more attention to closing the session with a tentative contract spelling out what areas might require further exploration.

The impact of the teaching tapes was so great that the subsequent group teaching had very little “headroom” for further change. However, the trainees commented at a follow-up evaluation two years later (Moss et al, 1991) that the group teaching had increased their confidence to practice the model and was valued by them.

The original teaching material for the research had been made with extracts from research tapes of real patient interviews from the earlier part of the study and so the tape was then remade with role-played examples for public use. A replication study (Margison & Moss, 1994) was then carried out with a group of trainees from non-teaching hospitals and with teachers with variable levels of prior experience. This style of replication was chosen in response to criticisms of the earlier project that the method was only applicable in teaching settings with particularly carefully chosen trainees, excluding trainees from ethnic minorities, and with expert teachers committed to this model. The replication study showed that the method of teaching was equally effective with the non-selected trainees and the teaching was equally applicable to non-UK graduates and was taught effectively by less experienced teachers.

At two-year follow up (Moss et al., 1991) the trainees were still able to adopt this style of therapeutic interview when requested. The follow up showed little if any decrement in performance of the model under the same test conditions as previously. When interviewed about the effects of the training, they reported a positive impact on their practice although the study did not examine actual examples of their work to confirm this.

The teaching was also replicated in another site with nurse trainees (Paxton et al., 1988). They used the training method in a two-day workshop format. The tapes have also been used extensively in the United Kingdom and elsewhere as part of basic therapeutic skills training for other professional groups including clinical psychologists, trainee counsellors and mental health workers.

In summary the early phase of development of the model in terms of teaching showed:

• the model can be described clearly and rated with a measure that is reliable
• A simple micro-skills teaching package using inexperienced teachers had a dramatic effect on changing trainees behaviour from an interrogatory style to a negotiated style.
• supplementary group teaching increased the trainees’ level of confidence
• after two years there had been little if any change in the trainees’ ability to use these skills (although the measures were on simulated patients rather than on their day to day practice)
• the original teaching material was re-made with simulated interviews of high plausibility and the method has subsequently been used widely and replicated with nurse trainees
• a factor analytic study of the pattern of skills acquired was consistent with the underlying model

The research focused on the early stage of skills acquisition and to develop the model beyond the initial training or research contexts, further work was carried out using supplementary teaching methods to prepare practitioners for the range and complexity of problems found in everyday use as described below.

A further study (Guthrie et al, 2006) showed that counsellors working in primary care could be trained efficiently to deliver PI therapy using this approach. As described in Chapter 3, twenty counsellors received a 12 week training course in PI therapy including supervised practice of cases from their primary care settings. The client outcomes were good with 50% showing clinically significant and reliable change.

Performance was assessed using videotaped sessions with simulated patients at 3 points in time: before training, after an intensive first week of training, and at the end of 12 weeks of supervision. Counsellors’ adherence to the model was assessed in relation to three patient scenarios: chronic depression, somatisation, and suicidality. Validity of the simulated sessions was verified by reference to counsellor behavior with actual patients using audiotaped sessions. After training, counsellors’ adherence to PIT increased without affecting their basic counselling skills. Patients with chronic depression and those with somatisation demonstrated improvement in specific items; there was, however, less evidence of model application among suicidal patients, where concerns about managing risk were predominant. Ratings of audiotaped sessions with actual patients confirmed that counsellors were able to apply the model in a real-life situation. The counsellors also reported a positive experience with enhanced skills after training.

These teaching methods were also used in evaluating post-qualifying training in psychotherapy for social workers addressing PI behaviours among 14 core psychotherapeutic competences, This research showed the course to be effective overall (Firth et al, 1999). The training methods for both short courses for psychiatrists, nurses, counsellors, psychiatric social workers and clinical psychologists, (eg, Paxton et al, 1988, Paley et al, 2003a, Paley et al, 2003b, Moss et al, 1991, Margison, et al,1994) and as part of longer training (Firth et al, 1994) appear to be effective across professional groups.

Role plays are another way of hearing feedback about performance as a therapist. They have been used widely in developing therapist skills and awareness for this model of psychotherapy and as an adjunct to supervision of clinical experience. However, trainees are often worried that they will not be able to cope with “difficult” situations that occur outside the immediate training context. So, from an early stage we drew on the work of Kagan (1980) to develop a way of addressing these therapist fears.

Kagan suggests that therapists typically “feign clinical naiveté” with a supervisor to avoid possible humiliation about their basic fears. He described these as fear of being engulfed or engulfing (seducing) the other; or being attacked / attacking the other. These are seen as the basic fears in any interpersonal interaction, but they are amplified in the close observation characteristic in learning psychotherapy. We used role-plays to set up a series of escalating difficulties to desensitise the trainees to their fears and to help them to discuss the difficulties they experienced in an open way. We developed four stages of increasing complexity and challenge for the trainee therapists.

In our adaptation of Kagan’s methods, a standard format is used for all the role-plays (Margison, 1991). A group, typically of about four therapists in training and a facilitator meet regularly. Two chairs are kept for the members currently playing therapist and client. The “client” is asked to prepare a couple of sentences summarising the basic information about age, gender, and main problems. Sometimes the role-play will be based on an initial session but role players are also surprisingly adept at starting as though several sessions into an ongoing therapy. Either the client or the therapist (or occasionally both) are given cue cards which have a brief instruction explaining what should be done. Typically this is a brief statement about what is the “agenda”, such as assuming that the client has to be told that the therapist is going to be away for three weeks holiday. The teacher stops the role-play after a two or three minutes in the early examples, or somewhat longer in the later, more complex examples. After the role-play a standard sequence is followed with the two role players expressing their feelings about the interaction before opening up the discussion to the whole group. It is crucial in this teaching method to establish an atmosphere that is conducive to learning. Other group members are encouraged to describe what they saw, heard or felt. Also they are asked to comment about how they might have felt as the client or responded as the therapist. Criticism of the “therapist” is specifically discouraged.

Commonly trainees have had bad experiences of role-plays where they have felt humiliated or exposed. So, it is important for the teacher to establish an atmosphere that allows exploration in depth for the role players, but with a safe atmosphere. In a group that has been working together for only a few weeks it is usually possible for the members to be quite open about their anxieties as therapists. A reliable structure of the teaching session prevents inappropriate drift towards the training group becoming a quasi-therapeutic group.

The training begins with low key examples of how sensitive the therapeutic situation is to therapist behaviour. So, for example, a trainee may be given a prompt card, which instructs the therapist to be “over-reassuring”, “lean too close to the patient”, “avoid eye contact” or similar undesirable behaviours. The therapist can use these early examples to become sensitive to nuances in the interpersonal situation and also become less anxious about the role-play itself. This is preparation for the next phase when the sessions focus on how to make practical arrangements about breaks, reviews of progress, session times. These particular role-plays have a dual purpose: The experience allows the therapist to be able to structure a therapy with confidence whilst developing sensitivity to the undercurrents of the therapy.

A common therapeutic error amongst beginners is their tendency to equate listening with passivity and as a result they may give insufficient structure. Whilst learning these basic techniques of structure and contracts they also become more sensitive to the emotional undercurrents involved in negotiating apparently practical issues. For example, any issue related to timing or frequency may be linked with separation issues or feeling neglected. Issues regarding review of progress are often tinged with feelings of being criticised or, in extreme cases humiliation and narcissistic wounds. This second phase of the role-play training prepares the therapist for the inevitable multiple levels involved in any therapeutic conversation.

The third group of role-plays focuses on how easily the therapeutic alliance can be damaged and persecutory dynamics established. For example, one role-play simply asks both participants not to take anything the other says at face value. As this is enacted the trainee becomes aware of the risk of invalidating the other’s experience. This helps the trainee to understand the processes described by Meares and Hobson (1977) as a “persecutory spiral”. By enacting the patient role as well it is possible for the trainee to realise just how sensitive the person in that position will be to minor discrepancies, ambiguities or inconsistencies in the therapist. These examples also give practice in how to repair ruptures in the therapeutic alliance.

Many beginning therapists in service contexts have had no experience of their own personal therapy, which is a prerequisite for most trainees wishing to practice primarily in psychotherapy. Whilst role-plays are clearly not capable of providing an equivalent experience, they can increase the trainee therapist’s sensitivity to interpersonal nuances which otherwise are learned only through therapeutic failures with actual clients.

In the fourth stage of the role-play training, a series of role-plays deliberately simulates what Kagan described as “therapeutic nightmares” such as being told of intense suicidal or violent impulses, sexual feelings towards the therapist and antagonistic and hurtful comments about the therapist. Kagan had developed a number of “stimulus vignettes” on videotape, which the trainee could watch and reflect on their emotional response. Our method draws on that technique, but uses a role-play situation to enact the challenging experience. The therapists have to think on their feet and respond in “real time”, but in a safe environment with no risk to clients. Most trainees find these sessions a demanding but positive experience.

Examples of role play cards

Role Plays Level One: Examples

• Avoid all eye contact with your client/ therapist
• Speak only technical jargon
• Ask only direct factual questions

Role Plays Level Two: Examples

• You are taking an unplanned three week vacation after the next session
• Discuss your need to change the session time

Role Plays Level Three: Examples

• Do not accept what your client / therapist says at face value
• Comment in an emotionally neutral tone about some aspect of your client / therapist’s appearance or clothing

Role Plays Level Four: Examples

• Discuss with your therapist that you fear you might lose control and smash things / harm yourself
• Discuss feeling embarrassed but that you need to tell your therapist how attracted you are to him / her
• Tell your therapist that his / her habit of pausing before answering is driving you crazy

Follow up role play training

After the formalised sequence of role-play situations it is possible to extend this model of training into actual clinical work. For example, in a supervision group a trainee can role play his or her client, or can ask another group member to role-play the client whilst the trainee stays in the therapist role. The advantage of this method is that it allows the therapist to become familiar with “what if?” scenarios in dealing with particularly challenging material.

We evaluated the effectiveness of this role play training in acquiring therapy skills (Palmieri et al, 2007) and the research suggested that role play training was effective in learning new skills. A 15-item role-play competence measure was developed. Ratings by three judges of 34 role plays from psychodynamic interpersonal therapy training showed good inter-rater reliability (.73–.79) and internal reliability (.84–.96). Validity was supported as scores were statistically significantly associated with the length of psychotherapy training experience. Most participants achieved satisfactory ratings on the key skills of this model of therapy after the training.

These methods can be supplemented by the use of peer teaching groups first described by Kagan (1980) when the members rotate roles between the roles of commentator/timekeeper, facilitator, “patient” and “therapist”. These techniques are a particularly powerful way of getting across the idea of different therapeutic perspectives. When teaching resources are scarce it is possible to have several groups of more experienced trainees running simultaneously with techers moving between the groups to add an additional perspective.

We have used these sessions regularly with therapists at different levels of experience and it is possible to develop new role-plays relevant to any specific topic such as working with somatic distress or managing self-destructive behaviour depending on the experience and clinical setting of the participants. Participants include established PI therapists who take part in the sessions to maintain their skills as part of continuing professional development.

With established therapists who want to progress to the role of supervisor we have used the same structure to role play the difficulties that arise in a supervisory setting where members role play supervisor and therapist and reflect on experiences with different styles of supervision.