Show Summary Details
Page of

(p. 13) Cognitive–behavioural assessment 

(p. 13) Cognitive–behavioural assessment
(p. 13) Cognitive–behavioural assessment

Joan Kirk

Page of

PRINTED FROM OXFORD CLINICAL PSYCHOLOGY ONLINE ( © Oxford University Press, 2020. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Clinical Psychology Online for personal use (for details see Privacy Policy and Legal Notice).

Subscriber: null; date: 04 June 2020


Cognitive—behavioural assessment is based on simple principles and has clearly defined aims. These can be readily understood by therapists new to this approach, although they may need two or more assessment sessions with their first patients in order to achieve the aims of the assessment. These are to have agreed a formulation of the target problems with the patient, and to have sufficiently detailed information about factors maintaining the problem to be able to design and present a treatment plan. In addition, the therapist should have begun to educate the patient about the psychological model.

The first, and perhaps central principle of cognitive–behavioural assessment is that the ways in which an individual behaves are determined by immediate situations, and the individual’s interpretations of them. This therefore becomes the major focus of the assessment, with an emphasis on specific problems rather than global entities.

The characteristics of therapists which are believed to be important in other kinds of therapy are likely to be just as relevant in cognitive–behavioural treatment. The patient needs to feel safe to disclose important and often distressing information. This will be facilitated if there is a warm and trusting atmosphere, no risk of censure, and if the therapist is empathic and clearly committed to helping the patient overcome current difficulties.

Goals of cognitive–behavioural assessment

Cognitive–behavioural formulation of problems

Cognitive behaviour therapy has its basis in the experimental method, so the early sessions are used to devise an initial hypothesis (formulation) and treatment plan. The formulation is tested out in subsequent homework and treatment sessions, and modified if necessary.

Although most of the assessment takes place in the initial sessions, the assessment process continues throughout treatment. Therapists sometimes make the error of thinking that if they classify a problem (for example ‘height phobia’), this will designate the treatment (for example graded exposure). Clinicians have become increasingly aware that diagnostic (p. 14) categories give broad indications about what treatment might be useful, but that this is only a preliminary step which must be supplemented by more detailed information. What is the person doing, overtly or covertly, which he or she would like to change? What are the precipitants (situational, mental, or internal), for the problem behaviour, and in what settings does the problem occur? What are the consequences of the problem behaviour? In particular, what seems to maintain the behaviour, either in the short-term or long-term? What changes might be made in any of these to produce changes in the problem behaviour?

Most of this chapter describes how to derive a formulation and treatment plan. However, before focusing on this, there are two other functions of the assessment which will be discussed first. These concern the use of the behavioural interview to inform the patient about the cognitive–behavioural model and approach to treatment; and also the therapeutic qualities of the assessment.

Educating the patient about the cognitive–behavioural approach

The patient should be informed during the assessment that the cognitive–behavioural approach is largely self-help, and that the therapist aims to help the patient develop skills to overcome not only the current problems, but also any similar ones in the future. The therapist should emphasize the role of homework assignments, pointing out that the major part of therapy takes place in everyday life, with the patient putting into practice what has been discussed in treatment sessions. The collaborative nature of the therapeutic relationship should be discussed; the patient is expected to participate actively by collecting information, giving feedback on the effectiveness of techniques, and making suggestions about new strategies.

Information about die structure of treatment should also be given at this stage; for example, how many treatment sessions there will be, how long each will last, and where treatment will take place.

A cognitive–behavioural assessment also has a general educational role and focuses the patient on internal and external variables which may not have been seen as relevant to the problem. The patient is asked about situations, physiological states, cognitions, interpersonal factors, as well as overt behaviour, and about how each of these groups of variables relates to the problem. This questioning will be discussed in detail later in the chapter. Drawing attention to such functional relationships is part of teaching the patient about the psychological model. During the early stage of treatment, this helps to increase the agreement between the therapist’s and patient’s expectations of treatment: if they remain too dissimilar, the patient may decide not to pursue treatment.

Initiating the therapeutic process

The assessment interview has an important role in beginning the process of therapy. Patients frequendy present with an undifferentiated array of (p. 15) difficulties. As the therapist helps to clarify and differentiate between problems, so the difficulties are frequently reduced to manageable proportions, and the patient begins to believe that change is possible. For example, a patient who presented with a series of problems including weepiness and low mood, loss of enjoyment and interest, tiredness, sleep disturbance, self-dislike, hopelessness, was relieved to learn that these were all common symptoms of one problem (i.e. depression), for which there were well-established treatment approaches. In contrast, some patients erroneously assume that their difficulties reflect a single problem; for example, a patient believed that she had one major problem—a basic lack of control—but was relieved when it became apparent during the assessment that instead she had interrelated, separate problems, including binge-eating, alcohol abuse, debt, low mood, and poor interpersonal relationships, which could each be tackled separately.

The assessment emphasizes the possibility of change, by helping the patient to think of what may be achieved, rather than dwelling continually on problems. It also sets reasonable limits on what might be achieved through treatment; for example, it is unreasonable for an agoraphobic patient to aim never to experience unpleasant emotions, but it should be possible to go to a supermarket in comfort.

The assessment also allows the patient to see that variations in the intensity of distress are predictable in terms of internal and external events, and are not just arbitrarily imposed by fate. It is implicit that if the variations are predictable, they may also be controllable. Patients may not readily pick up functional relationships between symptoms and such events. For example a patient said, ‘Yes, I’ve really had a dreadful week. I was premenstrual, then I had the awful row with my brother about the anniversary of my mother’s death, then I had to start going to work without the car because it broke down, and to cap it all, I’ve been feeling awful and all my worries about the symptoms have started coming back. I don’t know what’s brought it all on again.’ Questioning helped her to see that the increase in her symptoms was not unpredicted but could readily be explained by the associated events and her interpretation of what the ‘relapse’ signalled.

The therapist should offer non-judgemental sympathy and concern about the patient’s problems and distress; this may provide enormous relief, especially if the patient has felt embarrassed, guilty, or hopeless, as is often the case.

Finally, an important function of the assessment is to establish whether there is anything which needs dealing with as an emergency. For example, if the patient is depressed, suicidal intent must be assessed; if someone is complaining of difficulties in managing children, the possibility of physical abuse must be explored.

In summary, the main goal of the cognitive–behavioural assessment is to agree a formulation and treatment plan with the patient. In addition, it (p. 16) allows the therapist to educate the patient about the treatment approach, and to begin the process of change. It also allows emergency factors to be assessed.

Modes of assessment

Although the major part of the behavioural assessment takes the form of a behavioural interview, this is only one of the modes of assessment which may be relevant in any given case. When assessing problems, it is useful to differentiate between four different classes of response— behavioural, physiological, cognitive, and emotional. Different assessment procedures give information about different response systems, and so it may be useful to assess a problem in more than one way, to allow a more accurate picture of change following treatment. This is particularly true if there is likely to be a lack of synchrony between the different measures (Rachman and Hodgson 1974). For example, a patient may change behaviourally, but still feel distressed and experience physiological changes when in fearful situations. Thus, assessing only the patient’s self-report of distress would mask progress and it would be usefully supplemented by a behavioural test (see p. 49) in which the patient is asked to carry out problem behaviours. Table 2.1 summarizes the major modes of assessment which may be considered when assessing problems.

Table 2.1 Modes of assessment

  • Behavioural interview

  • Self-monitoring

  • Self-report (questionnaires, global rating scales)

  • Information from other people

    • interviews with key individuals

    • monitoring by key others

  • Direct observation of behaviour in clinical settings

    • role-play

    • behavioural tests

  • Behavioural by-products

  • Physiological measures

The major part of the behavioural assessment takes the form of an interview, but this is supplemented by information collected and recorded (‘self-monitored’) by the patient after the interview. Much of a behavioural interview will be directed at defining problems in the kind of detail necessary for subsequent self-monitoring. The principles of measurement which are relevant to self-monitoring also apply to the other aspects of assessment summarized in Table 2.1. Therefore, the principles of measurement relevant to self-monitoring and the other assessment modes (p. 17) will be presented here, before the behavioural interview and other modes of assessment are discussed.

Measurement in cognitive–behavioural assessment and treatment

The application of the experimental method to problems of individual patients, as advocated by Shapiro (1961b), is central to the cognitive–behavioural approach: a formulation is used to make predictions about the effects of particular interventions (treatment techniques, etc.), and these are then tested out in treatment. Thus, therapy with any one patient can be seen as a single-case experiment, and much of treatment revolves round measures taken both during treatment sessions, and between sessions. Such quantification can conveniently be restricted to self-monitoring and questionnaires in most patients, with direct observations or physiological recording seldom required. It may require ingenuity to find individualized measures which adequately reflect the patient’s problem, although standard measures are readily available for many situations.

Advantages of measurement

  1. 1. Retrospective estimates by patients of the frequency of behaviours are notoriously unreliable (Barlow, Hayes, and Nelson 1984). Direct recording allows a more accurate description of the problem in terms of frequency, intensity, etc.

  2. 2. Measurements during treatment sessions, as well as between sessions, allow the patient and therapist to modify treatment if necessary. For example, ‘thought satiation’ (where the patient focuses on a distressing thought for a lengthy period) was being used with an obsessional patient who rated her distress every three minutes during treatment sessions, and also three times daily. Her rating sheet is shown in Figs. 2.1 and 2.2. Ratings of distress about the thoughts consistently decreased during treatment sessions, but daily ratings of tension indicated that this was increasing from day to day. Given the ‘effectiveness’ of the procedure during sessions, this deterioration would have been missed without the daily ratings.

  3. 3. Measurement can have therapeutic effects, providing the patient with consistent and accurate information about progress. For example, an agoraphobic patient reported, ‘It’s been dreadful, I’ve not been out at all, I couldn’t do anything.’ Examination of her daily homework diaries of trips from home (see Fig. 2.3) allowed her to see that although she had slipped back from the previous week, she was going out significantly more and feeling less anxious than a month previously. The session also (p. 18) (p. 19) coincided with her menstrual period, and a review of other premenstrual weeks in the diaries indicated that she tended to be more anxious and less active at such times.

    Patients low in confidence and self-esteem tend to give themselves little credit for progress. For example, a patient who had successfully overcome her fear of staying in the house by herself dismissed this by saying, ‘Well, that was never a problem anyway; it’s handling the children I really have difficulty with.’ Reference to early record sheets revealed that at the beginning of treatment she had been very distressed if alone in the house, and avoided it as far as possible. This demonstrated how successful she had been in solving what had been a major problem (see Fig. 2.4).

    Regular measurements also ensure that the therapist and patient remain focused on the agreed treatment goals.

  4. 4. Measurement allows the therapist to establish whether the treatment has been delivered as intended; for example, an exposure programme (see Chapter 4) may not work because the patient is not exposing appropriately. Many therapists check on homework informally; for example:

    Therapist ‘Are you going out regularly as we agreed?’

    Patient ‘Oh yes, I’m going out from time to time.’

    Th. ‘How often is “from time to time”?’

    Pt ‘Oh, whenever I need to.’

    Th. ‘And how often is that?’

    Pt ‘On most days.’

    However, even this amount of questioning did not provide the precise information which a daily diary would provide. In this case the patient had actually been out three times in a week and this low frequency of exposure explained why little progress was being made.

Fig. 2.1 Self-rating scale for distress during sessions of thought satiation with an obsessional patient

Fig. 2.1 Self-rating scale for distress during sessions of thought satiation with an obsessional patient

Fig. 2.2 Three-dmes-daily tension-rating scale for an obsessional patient

Fig. 2.2 Three-dmes-daily tension-rating scale for an obsessional patient

Fig. 2.3 Daily diary of trips from home by an agoraphobic woman

Fig. 2.3 Daily diary of trips from home by an agoraphobic woman

Fig. 2.4 Diary of a patient anxious about staying alone

Fig. 2.4 Diary of a patient anxious about staying alone

(p. 20) In summary, measurement has a central role in cognitive–behavioural assessment and treatment, and may occur in different assessment modes. The behavioural interview is generally the starting point for assessment, and this will now be discussed in detail.

Behavioural interviewing

Initial stage

Most patients will have little idea what to expect of the assessment interview. It helps to put the patient at ease and to begin to establish rapport if it is clear that the therapist has read the relevant referral letters, and if the stage is set for the remainder of the session. After the therapist and patient have introduced themselves, the session might begin,

‘Dr… wrote to me about the problems you would like help with. I gather that you are feeling tense and anxious most of the time, and that you are concerned about your drinking. Apart from that, I do not have a lot of detail. I would like you to tell me briefly how you see the problems just now. Then we will discuss for a quarter of an hour or so how the problem developed. And then we will go over how things are now in more detail. So can you begin by telling me briefly what you see as the main problems?’

It is useful simply to listen to what the patient has to say about the problems. He or she has probably spent a great deal of time thinking about them. On the other hand, it is easy for the patient to misinterpret what is required, and be ready to launch into an historical account of the problem, which is not useful at this stage. The therapist should listen carefully and communicate caring concern to the patient. Comments such as, ‘This is obviously very difficult/upsetting for you’ help to engage the patient, and to establish rapport. Summarizing and paraphrasing what the patient has said, and reflecting back feelings, allows the therapist to communicate that the problems are understood. For example, after listening to a lengthy description of a patient’s problems, the therapist said, If I am right, you are saying that you try very hard to please other people, and to put their well-being before your own, but it sounds as though this makes you feel very worked up at times. Is that right?’ The patient can be encouraged to expand in relevant areas if increased interest is shown, both non-verbally with nods and eye contact, as well as verbally through comments and questions. However, therapists should ensure that their initial preconceptions about patients’ problems do not excessively influence either their questions, or their interpretations of the patients’ replies.

Some patients find it difficult to describe their problems, or give only vague descriptions. It may then be helpful to ask questions such as, ‘Can you describe to me what happened last time you were upset?’, ‘When was (p. 21) that?’, ‘What was the first thing you noticed?’, ‘In what way has your life changed since you developed these problems?’, ‘What does the problem prevent you doing?’, ‘What have you had to give up because of the problem?’ The use of paraphrasing may then help the patient expand on relevant aspects.

At this stage, which lasts five or 10 minutes, only a general outline of the problems is required. The therapist picks up clues about possible antecedents and maintaining factors for use later in the interview, but only notes them at this stage. It is useful to provide the patient with a summary of the problems, and to obtain feedback on its accuracy. The therapist could say, for example, ‘You seem to be saying that your major problem is palpitations, and your worry about them. But in addition to that, you are worrying about your daughter’s current relationship, as well as your husband’s attitude to it. Have I got that right? Are there any other problems we have missed out?’

When there is more than one presented problem, the therapist and patient should work out together which problem should be the initial focus of intervention (see p. 416).

The assessment then moves on to look at how each problem began and subsequently developed, before a more detailed analysis of the current situation is made. Each identified problem is analysed in turn, covering the steps summarized in Table 2.2.

Table 2.2 Stages in the Behavioural Interview

Brief description of problems


  • precipitants

  • time course

  • predisposing factors

Description of problem behaviour:

Cognitive–behavioural assessment

  • What?

  • When?

  • Where?

  • How often?

  • With whom?

  • How distressing?

  • How disruptive?

Contexts and modulating variables:

  • situational

  • behavioural

  • cognitive

  • affective

  • interpersonal

  • physiological

Maintaining factors:

  • situational

  • behavioural

  • cognitive

  • affective

  • interpersonal

  • physiological


Coping resources and other assets

Psychiatric and medical history

Previous treatment:

  • response

  • current medication

Beliefs about problem


Mood/mental state

Psychosocial situation:

  • family

  • psychosexual relationships

  • accommodation

  • occupation

  • social relationships

  • hobbies/interests

Preliminary formulation

Development of problem

This part of the assessment is considerably briefer than in other kinds of psychotherapeutic assessment, since historical information is only collected if it is of direct relevance to the development of the presenting problem, and the understanding of current maintaining factors.


There may have been a very clear onset for a problem; for example, a driving phobia may develop immediately after a car accident. However, even in such apparently straightforward cases, the therapist will require further information in order to understand the problem and how it is maintained. For example, the driving phobia may be maintained by thoughts about crashing, avoidance, and, perhaps, thoughts about disfigurement remaining from the initial crash.

For many patients, the problem will have developed gradually, with a succession of events contributing to the patient’s recognition that there is a problem. These events may be directly related to what is eventually identified as the problem; for example, a patient may have left three jobs before recognizing that he has difficulty dealing with authority figures at work. The patient on the other hand, may realize that there is a problem which is getting worse, but be unclear about how it started, or why it is (p. 22) (p. 23) deteriorating. In such cases, there may be stressful life events or major changes associated with the onset of the problem and changes in its intensity. It may be useful to run through a list of some typical life events, for example, death or illness in family or friends, break-up in a relationship, moving house, job change, etc. There will be areas of particular relevance for any given problem; for example, loss would be especially relevant for a depressed patient, physical illness in a relative or friend for a patient with panic attacks.


The way the problem has developed since its onset should be established. The problem may, for example, have persisted steadily, or it may have deteriorated or fluctuated. It is useful to establish why the patient presented for help at this particular time, as this may reflect other difficulties. For example, a woman with increasing social anxiety only presented for help when she changed jobs and could not face explaining to her colleagues that fears of contamination would prevent her using the staff canteen. If there have been fluctuations in the severity of the problem, and particularly if it is long-standing, then it is helpful to plot an event-time chart, with variations in problem severity down one side of a time-line, and life changes down the other side. An example is given in Fig. 2.5. As in other parts of the interview, the predictability of the problem is emphasized, to help the patient understand why variations in its severity have occurred. As well as life events, changes in mood should be enquired about, as should any ‘treatment’ interventions, either formal or informal (through, for example, self-help groups, voluntary bodies, churches, etc.).

Fig. 2.5 Time-event chart for a patient with anxiety symptoms

Fig. 2.5 Time-event chart for a patient with anxiety symptoms

Some patients will want to spend an excessive time describing the development of problems. This may be because of inaccurate expectations about the interview, or because they have spent hours pondering about their problems and wish to share their thoughts. It may be necessary to remind such patients that the major focus of treatment is on immediate circumstances. The therapist could say,

‘We need to spend most of our time on what is happening now, because that is what we are going to try and change. Although we need an oudine of how the problem developed, we need to spend most of our time on what we can change.’

It may also be worthwhile pointing out that a problem may have developed for reasons which have become irrelevant and that entirely different factors may now be maintaining it. For example, a man developed erectile failure at a time when he doubted his previous wife’s affection for him; although he was now in a loving relationship, anxiety about his poor sexual performance maintained the erectile problem.

It is worth noting that the word ‘why’ is avoided as far as possible, as it tends to produce the response, ‘I don’t know’, or lengthy expositions (p. 24) about the origins of the problem in terms of lay psychology. Questions such as, ‘What was difficult for you about that situation?’ or, ‘How did you stay calm in that situation?’, on the other hand, produce more detailed information about factors currently maintaining the problem.

Predisposing factors

Information is sought about anything in the background which made it more likely that the patient would develop the target problem. More specific information about background factors relevant for specific disorders is available in the subsequent chapters. For example, if someone (p. 25) were depressed, then the patient would be questioned about family history of depression and childhood separation; an anxious patient would be questioned about emotional lability; a woman with orgasmic dysfunction would be questioned about sexual attitudes of her parents. However, depressed and anxious patients would not be asked about sexual attitudes, even though it could be argued that this might give a further understanding of the patient as a whole. Generally, the therapist only seeks information which makes it more likely that the target problem can be changed.

Behavioural analysis

This stage, during which problems are reviewed in detail, comprises the major part of the interview. The aim is to discover how the problem is currently maintained, in what way it is interfering with the patient’s life, and whether the problem is serving any useful purpose for the patient. There are two commonly used approaches to this.

Each problem can be analysed in terms of what O’Leary and Wilson (1975) termed the A–B–Cs—the Antecedents, Behaviours and beliefs, and Consequences. Each of these factors may increase or decrease the probability that the behaviour will occur. For example, a common antecedent to smoking cigarettes is sitting with a cup of coffee at the end of a meal; if the antecedents are altered (immediately moving away from the table at the end of the meal, drinking tea instead of coffee), then the likelihood of smoking is reduced. On the other hand, changing behaviour by deliberately smoking cigarettes too rapidly can help bring smoking under control. Finally, if there are positive consequences, for example, money saved by not smoking put aside for a specific activity, smoking is less likely in the future. For any given problem, changes may be possible in any or all of the antecedents, behaviours, or consequences; the assessment aims to identify what might be maintaining the problem and what can be changed.

A similar but more straightforward way of carrying out a behavioural analysis is to describe the contexts in which the problems arise, to look at the factors which modulate the intensity of the problems, and to assess the consequences, including avoidance, of them. This scheme will be adopted here as it is less complex but allows an adequate analysis of most problems.

Detailed description of problem

As a first step, it is useful to ask the patient for a detailed description of a recent example of the problem. This gives more specific information than is obtained from a general description, and provides clues about maintaining factors. If patients have difficulty in describing a recent incident, it may be helpful for them to close their eyes and imagine the scene, as (p. 26) though it were being played on television. The description of the problem should include internal events such as thoughts, feelings, and physical symptoms, as well as overt behaviours.

A patient who presented with worries about bowel function was asked:

Patient ‘Let’s go over things in more detail. You say you worry about going to the lavatory. When was the last time you really worried about it?’ ‘This morning before breakfast.’

Therapist ‘Could you talk me through it, letting me know what happened, how you felt, what you did, what thoughts were going through your head and so on. What was the first thing which happened?’

It may be helpful for the patient to talk freely for a few minutes, but this should be followed up with questions, until the therapist has a clear idea of what happened in the specific example and the sequence in which it occurred. The patient in die above example focused on how he felt, without describing what he did, and so the therapist questioned him about other aspects of the problem. The patient replied:

Therapist ‘I just felt terrible. I can’t think about anything else at these times, and I get really uptight. My muscles were all tense, I got hot and sweaty, I got terrible butterflies, and my stomach felt all upset. But I knew that if I went to the loo I would tighten up and be unable to do anything.’ ‘You say you felt all tense, hot and sweaty and so on. Were there any other physical sensations this morning?’

Patient ‘I often feel lightheaded, but I know I won’t faint.’

Th. ‘And this morning, what did you do when you were feeling like this?’

Pt ‘Oh, I paced around my room, but I daren’t go outside, or into the kitchen in case I saw someone and they asked me what was wrong.’ Th. ‘Would it be unpleasant if someone asked what was wrong?’

Pt ‘Well, it would be so embarrassing. I could hardly tell them, could I?’

Th. ‘I can see you’d find that difficult. How long did you stay in your room until you felt a bit easier, until you settled down?’

Pt ‘About twenty minutes, then I could go out of the room.’

Th. ‘And then what did you do?’

Attention then turns to a wider description of the problem, with the therapist continually trying to elicit specific detail rather than generalities. For each problem the therapist should have a picture of: what the problem is (when, where, how often, and with whom it occurs); how distressing it is; and how disruptive it is.

Contexts and modulating variables

As outlined in Chapter 1, an assumption of behavioural theory is that abnormal behaviour has been learned, and that such behaviours can be triggered by external or internal cues which have become associated with the problem behaviour. Thus, a woman with bulimia nervosa found that (p. 27) she had many more urges to binge when she was in areas of town where there were food shops; a woman with agoraphobia noted that she felt highly anxious in particular shops.

A detailed assessment of contextual triggers is required because treatment plans often include manipulation of the contexts in which problems occur; for example, the woman with bulimia could initially reduce her bingeing by planning routes which circumvented food shops. In addition, treatment frequently involves variations in the modulating variables associated with particular cues. The range of possible triggers is almost infinite: for example, an obsessional patient may constantly ritualize at home but never at work, an agoraphobic patient may be anxiety-free if in a town where she is unknown, a compulsive gambler may only gamble when angry.

The patient may not be aware of the contexts in which the problem occurs, nor of the modulating variables. It is generally necessary for further information to be collected, either through self-monitoring or a behavioural test. For example, a patient was distressed by his high frequency of casual homosexual contacts, but was unclear about what prompted him to seek them out. Daily monitoring helped him to see that the behaviour was related to boredom, irritability, and tension, and only occasionally to sexual frustration. A sample from his diary is shown in Fig. 2.6.

Fig. 2.6 Self-monitoring of casual homosexual contact

Fig. 2.6 Self-monitoring of casual homosexual contact

When considering contexts in which problems arise, six broad areas should be covered, as listed in Table 2.2; while it is not necessary to go through these in order, the interview should be sufficiently structured for (p. 28) each to be discussed: a major goal is to shift from a global, all-or-nothing view of the problem, to one in which the patient may begin to see it as predictable.

Situational cues

Problems are often worse in some situations than others. For example, a patient with bowel problems gave the following account:

Therapist ‘Right, that’s a pretty clear picture of how things are when they are bad. Now let’s look at the sorts of things which make the problem more likely to occur.’

Patient ‘Actually, it’s there all the time, I never stop thinking about it.’

Th. ‘Yes, I’m sure it feels like that. But you’ve mentioned that at least some of the time you can put it to the back of your mind and get on with things. What is likely to set it off in the morning?’

Pt ‘Well, if I wake up and can go to the lavatory straightaway then I’m OK. But if someone is around on the corridor, and I can’t go, then I start to get uptight.’

Th. ‘So, it’s worse if people are around?’

Pt ‘Oh yes, much worse.’

Th. ‘And which people make it worse, anybody or are some people easier?’

Pt i’m not too bad at home with my family, because they know there’s a problem and leave me alone. It’s people on the corridor I particularly dread.’

Th. ‘But what about total strangers in college, say?’

Pt ‘Oh yes, they’re difficult, nearly as bad as those in the flat.’

Th. ‘Right, so having people around makes a difference. What else affects the situation? You are implying it is easier at home than at college. Is that just because of the people, or is it other things about the situation?’

Pt it’s certainly worse if it’s quiet.’

Th. ‘What’s so bad about quietness?’

Behavioural cues

Symptoms may be precipitated by a wide range of behaviours. For example, an obsessional woman was not troubled by the presence of knives in her kitchen, but became very distressed whenever she used them; an obsessional man found going through doorways or over steps very difficult; a man concerned about his health found that any mildly strenuous activity resulted in anxious thoughts.

Cognitive factors

Patients may believe that problems occur unpredictably because they pay little attention to what thoughts are going through their minds at the time and immediately beforehand. At the initial interview it may be difficult for the patient to identify the relevant thoughts, or to focus on thoughts at the appropriate level of specificity. This may be because the thoughts were not attended to; or it may be that, when the (p. 29) patient is not upset, the thoughts are discounted as ridiculous and exaggerated. The identification of dysfunctional thoughts, which is an important step in the treatment of many problems, is discussed in detail in Chapters 3 and 6. At the assessment stage, questioning about thoughts may introduce the patient to their role in precipitating problems. For example, one patient said:

Therapist “But sometimes I just start worrying about it for no reason at all. I can just be sitting in front of my typewriter, and suddenly feel awful.’ ‘Can you think of a time like that?’

Patient ‘Yes, it happened last week.’

Th. it’s very unusual for symptoms to come out of the blue. Let’s try and imagine the scene in as much detail as we can, and see if we can piece together what set it off.’

Pt ‘I was sitting in my office, typing away, and suddenly I felt dreadful.’

Th. ‘Do you often feel bad in your office?’

Pt ‘No, as I said, I usually feel OK when I’m at work.’

Th. ‘So, can you remeber what were you actually doing on this occasion?’

Pt ‘I was typing an annual review.’

Th. ‘So, you’re sitting in your chair, typing the annual review, was anyone else around?’

Pt ‘No, nobody was in, and I’d been left with a hugh pile of stuff to get through.’

Th. ‘And can you think what actually went through your mind at that point?’

Pt ‘I thought, I shall never get this lot done, I’ll be tired out, and I shall be in a state for this evening.’

Other patients are fully aware that their thoughts play a major role in the presentation of symptoms, and may self-critically point out that they talk themselves into feeling bad. They sometimes feel hopeless and out of control about having got themselves into this mess, with all the responsibility for getting out of it thus resting fully on their shoulders. This offers an opportunity for pointing out an area of control.

Therapist ‘You mean that if you start thinking “I’m going to feel bad”, this brings on the symptoms?’

Patient ‘Yes, it’s ridiculous, I just bring it on myself. Anybody would think I enjoy feeling like this.’

Th. ‘I am sure you don’t. But you’re in a very strong position, having realized that the way you think affects how you feel.’

Affective states

A variety of mood states may affect the problem. Depression and anxiety are the most obvious areas, but other states like irritability, cheerfulness, excitement may be relevant. If, for example, a patient feels more anxious in phobic situations when she is irritable, she could look at what makes her irritable and whether this can be changed. It would also be useful to discuss whether she attributes her physical (p. 30) ‘symptoms’ to anxiety, when some could be attributed to irritability. A woman with a social phobia described this effect:

Therapist ‘Do other sorts of mood make any difference? What about if you’re excited?’

Patient ‘I never feel excited, at least not for any length of time. If I feel excited about something, it reminds me that I shall probably be anxious when I get there, and that makes me nervous.’

Th. ‘Do you mean that the feelings you get when you are excited are similar to the ones you get when you feel anxious?’

Interpersonal factors

Social factors are relevant in the majority of problems presented. These include problems such as assertiveness or social anxiety in which interpersonal factors are central, through to a wide range of problems in which social variables are implicated but more peripherally. For example, a girl with an eating disorder was asked:

Therapist ‘Does it make any difference if you’re with someone?’

Patient ‘Oh, I can’t bear to eat if anyone is around.’

Th. ‘What is it that you dislike?’

Pt ‘Well, I can feel them watching me all the time to see if I’m eating normally, and I start thinking that they think I’m greedy as soon as I put the first mouthful in. Though I know that’s ridiculous.’

Th. ‘Is there anything else you dislike about it?’

Pt ‘Well, if I’m honest, I guess it means that I can’t go ahead and really binge, even if I want to.’

The behaviour of family members and other key individuals may have a marked effect on the problem; for example, criticism by another person frequently exacerbates obsessional rituals; the presence of a child may facilitate coping responses in a phobic patient. The beliefs of family members are also relevant, depending on whether they see the problem as physical, incurable, moral weakness, or whatever.

Physiological factors

These can be relatively specific to the problem; for example, a palpitation may precipitate severe anxiety symptoms in a patient concerned about cardiac functioning. On the other hand, there are more general factors, like tiredness, phase of menstrual cycle, caffeine intake, which may affect either general arousal level, or may affect the problem directly. Furthermore, some behaviour may only occur in specific physiological states; for example, after alcohol consumption.

As with many of the cues given in the above examples it is not sufficient simply to establish the antecedents for a behaviour; the patient’s interpretation of the cues, whether these are situational, behavioural, physiological, or interpersonal, are central to the assessment.

(p. 31) Maintaining factors

Having built up a reasonable picture of the conditions under which the problem is most likely to occur, the next step is to look at what is maintaining the problem. The major focus is on the immediate consequences of the problem behaviour. In simple terms, as described in Chapter 1, behaviour which is followed by unpleasant circumstances is less likely to recur, behaviour followed by pleasant events is more likely to recur in the future. The most potent consequences of the problem are the patient’s thoughts and other reactions to the problem, as these frequently set up a series of vicious circles which maintain the problem. Longer-terms events are usually less relevant, and indeed may appear to contradict the basic operant principle; for example, an obsessional patient may continue with time-consuming rituals even though in the long term this is jeopardizing both employment prospects and family harmony.

Immediate consequences

As with the triggers, these reactions can be classified into six broad groupings, as shown in Table 2.2.

For example, a woman with increased frequency of micturition found that bladder sensations were triggered by a great many situations and behaviours, including arriving at work, beginning a journey by bus or car, eating a meal, bending over, carrying anything. She then thought, ‘I must go to the loo’, and if this were not possible, her reactions included behavioural responses like crossing her legs, sitting still; thoughts like, ‘If I do not go immediately, there’s going to be a terrible mess’, and a whole series of thoughts about how uncomfortable she felt; affective changes, mostly in anxiety level; and a wide range of physiological symptoms including headache and stomach-ache, as well as more non-specific symptoms of anxiety. All of these reactions kept her focused on bladder sensations, which increased as a result; this then increased the reactions, and so on in a vicious circle. There were also interpersonal consequences; for example, her aunt might say, ‘Are you not going to go to the loo now that you’ve eaten that sandwich?’ As soon as she was able to empty her bladder, all of these reactions disappeared, thus reinforcing her belief that she had an abnormal bladder which she needed to empty frequently, and thus reinforcing her high frequency of micturition.

The patient will frequently give clues about maintaining factors during earlier parts of the assessment, but these need to be supplemented with detailed and specific questions. For example, a woman was feeling anxious much of the time because of a difficult family situation. She was increasingly worried about her ability to cope at work as a teacher.

Therapist ‘You say that if you are anxious in the classroom then the children get out of control. What happens?’

Patient ‘Well, they seem to sense that I’m feeling bad, and they play up. It can deteriorate into a riot within seconds.’

(p. 32) Th. ‘Can you think of an example of when that happened, as quickly as that?’

Pt ‘Well, it didn’t happen within seconds, but one day last week it built up very quickly.’

Th. ‘Can you tell me about that?’

Pt ‘Well, I was feeling really bad, light-headed and tense, and they got more and more out of control.’

Th. ‘What were you doing with them?’

Pt ‘They were painting, and two or three started flicking paint, then it all built up. I should have noticed when it started really.’

Th. ‘What were you doing instead?’

Pt ‘I guess I was too busy thinking about how I was feeling.’

Th. ‘Can you remember what thoughts were going through your head?’

Pt ‘Oh yes, I spend all my time thinking—“I can’t carry on like this, there’ll be a riot if I don’t pull myself together.”’

Th. ‘What happened that day? Did it build up to a riot?’

Pt ‘No, it was playtime so I could go off to the staff room to cool down.’

The therapist then gave the patient a summary, so that she could begin to see which factors might be changed, and could provide feedback on the accuracy of the summary:

Therapist ‘So what you seem to be saying is that if you are tense and worried in the classroom, this affects your concentration so that you’re not so good at nipping trouble in the bud; if any trouble erupts, then you worry about it really getting out of control, get even more tense, and so it builds up.’

Patient ‘That right. I can’t do anything.’

Th. ‘It sounds as though you just wait for relief, until you can get out of the classroom. And I suppose that is just strengthening your belief that you really cannot cope.’

Pt ‘Well, it’s true, I can’t.’

Th. ‘So we’ll need to work out what you can do that would build up your self-confidence about coping at work.’

This summary has raised the possibility of altering a number of maintaining factors—nipping trouble in the bud, worrying thoughts about riots, increased tension—so that the situation can be improved without ‘escaping’, and so that the patient can again feel confident in the classroom.

The response of relatives and friends to the problem is obviously relevant, and it is usually necessary to ask in detail about particular situations. General descriptions of behaviour, like ‘supportive’, give little information; nor do phrases like, ‘He never says anything’. A socially phobic woman gave much more useful information when specifically questioned:

Therapist ‘How did your husband respond when you went red?’

Patient ‘Oh, he never says anything really, he’s not much help.’

Th. ‘Can you remember what he actually said that day in the cafe?’

Pt ‘He said he wasn’t going to sit there if I was going to make a fool of myself.’


The avoidance of or relief from distressing emotion is frequently an immediate effect of a problem behaviour, and is often the most potent maintaining factor (see Chapter 1). Many treatment plans will include steps to overcome it, and the therapist therefore requires an extensive description of what is avoided. The discussion of passive avoidance can be introduced with a general question like, ‘What things have you stopped doing, or what places have you stopped going to, because of this?’ Active avoidance can be addressed with a question like, ‘Are there things which you have started doing, or are doing differently, because of the problem?’ or ‘What could you stop doing if the problem went away?’ This should be supplemented by questions about home life, relationship with partner and children, work, social life, hobbies and interests, each of which may be affected by avoidance. If the problem is chronic, the patient may have become unaware of the extent of the avoidance, and questions such as, ‘How would your life be different if you did not have this problem?’, ‘In an ideal world, if the problem disappeared, what would you be doing that you can’t do now?’ are helpful. In addition, reports in the literature can guide specific questioning; for example, patients with worries about cardiac function may avoid any mildly strenuous activity like going upstairs; patients with eating disorders may avoid a wide range of fattening foods on ‘health’ grounds.

Patients can be quite skilled at subtle avoidance even when they appear to be exposing themselves to difficult situations. For example, a socially phobic woman continued to go out socially but had developed an almost total avoidance of eye-contact; this allowed her to avoid the disapproval and boredom she feared she would see if she looked at people interacting with her. Avoidance may be extensive even though the problem appears to be circumscribed. For example, a woman with a phobia about vomiting had described how it affected her work as a nursery nurse where she felt unable to deal with ill children. Specific questioning revealed a much wider pattern of avoidance.

Therapist ‘You say that if you see dirty people, or drunks, you get worried in case they are sick. Does this affect you socially? Are there places or people you avoid because of this?’

Patient ‘I suppose there are. Hundreds of them. I won’t go to pubs when they are busy… or go out late in the evening when people may have drunk a lot… I don’t go to parties in case people get drunk.’

Th. ‘Are there other social situations like that?’

Pt ‘I’m not so bothered about dinner parties if I know the people, but I’m not keen on restaurants in case people eat or drink too much. I won’t (p. 34) even fly because people might drink to settle their nerves before they fly and be sick in the departure lounge—I once saw that happen.’

Th. That was an unfortunate coincidence that you were there. It can’t happen often. What about other public transport. Is that affected?’

Pt ‘Well, I don’t go on trips with the nursery in case the children are travel sick. And I wouldn’t go on a long distance coach trip either— not because I’m travel sick but in case somebody else was sick.’

Th. ‘Can you think of other situations like that?’

Pt ‘It sounds ridiculous, but I stopped going to the squash club because somebody was sick once. They’d probably just got an upset tummy—if anybody says they feel ill, I do anything to get away.’

Th. ‘Are there any other steps you take, in case anyone is sick?’

Pt ‘Yes, I will never go out without wet tissues so that I could clean myself up if I was affected by it. And I carry barley sugar sweets to give to the children in case they feel sick. I’m very careful with my diet as well, particularly if I am going out.’

Further questioning about friends’ illnesses and ‘dirty’ people was supplemented by a homework exercise, when she made a list of how life would be different if there was no problem. This was to give her more opportunity to identify avoided situations.

Long-term consequences

The pattern of avoidance described by a patient may raise the question of whether the presented problem is part of a wider difficulty. However, there is no supposition of underlying needs as in concepts such as ‘secondary gain’. For example, it could be postulated that the vomit phobia described above allowed the woman to avoid social interactions. This could be pursued in the assessment interview, but the issue may not be fully resolved until further into treatment—successful resolution of a specific difficulty may reveal a wider problem, or failure to proceed may indicate associated difficulties. If there were an associated problem, it could be approached within the cognitive–behavioural framework described above without invoking underlying conflicts. Dealing with associated problems in this way side-steps the issue of whether the patient is really motivated to improve (p. 36), and evidence of a related problem can be taken at its face value with no implications of ambivalence about change.

After the detailed description of maintaining factors, it is worthwhile asking a broad question like, ‘Are there any more general ways in which your life would change if you no longer had this problem?’ This may simply point to a general increase in self-esteem for example; on the other hand, it may indicate longer-term changes, for example within a marriage, or in terms of independence, which may need building into the formulation.

(p. 35) Coping resources and other assets

People differ in their methods of coping with problems and distress, and in the extent to which they rely on themselves rather than on other people. This ranges from the individual’s familiarity with specific strategies, like relaxing the shoulders when tense, through to more general assets, like being able to communicate distress to others.

Initially, questioning can be directed at coping with the target problem. For example, the therapist can ask, I would be interested to know what you do which helps you control the problem, even if it only has a small effect. Can you tell me what things you’ve found helpful?’ It may be useful to give an example: ‘You mentioned that you can sometimes prevent things building up if you go out of the room and count ten before coming back. Are there other things like that which you have found useful?’ It is also helpful to discuss how the patient has dealt with other difficult situations, partly to highlight any demonstrable capacity for coping, and partly to determine which skills were used. For example, the therapist could ask, ‘Let’s think of another time in your life when you have had to deal with a difficult situation, something which was distressing. Can you think of an example?’

From this the interviewer can move to a wider discussion of the patient’s assets, skills, and strengths. These include environmental features (for example, a supportive spouse, a satisfying job, having a car available for homework tasks); skills which may generally facilitate change (good record-keeping, highly skilled bee-keeping as a source of self-esteem); and strengths, such as a sense of humour, willingness to tolerate discomfort, persistence, interpersonal warmth, which make it more likely that treatment suggestions will be carried through.

Previous psychiatric and medical history and treatment

A description should be obtained of previous history, particularly of similar episodes. The patient’s response to previous treatment is particularly important. This is partly because it may predict current response to treatment, and, in the case of poor outcome, might give information about pitfalls to avoid. In addition, the patient may have developed beliefs about the nature of the problem (e.g. ‘It responded to tablets so it must be physical’) or its possible outcome (e.g. it didn’t even respond to tablets so it must be incurable’) on the basis of previous response. Current medication should be noted, particularly psychotropic, but also other medication (e.g. hormone treatments) which may affect psychological functioning.

Beliefs about the problem and treatment

Patients are unlikely to engage in treatment if the approach the therapist offers is not congruent with their beliefs about the nature of the problem. (p. 36) Although the therapist seeks to structure the interview so that information necessary for a treatment plan is obtained, and so that patients can be educated about the cognitive–behavioural approach, nevertheless opportunities should be given throughout the interview for them to educate the therapist about their perception of the problems. Patients should be asked, for example, whether they think their problems are likely to change and what they think would be likely to help them. It is helpful to ask whether anyone close to them has had similar problems, and what the outcome was.

The impact of the patient’s beliefs on assessment and treatment need not be subtle. One woman with long-standing obsessional problems was unable to give a coherent description of her difficulties, or even to sit down during the interview. It would have been entirely useless to pursue a ‘standardized’ assessment interview, until it had been established that she believed that she was being reassessed for the lobotomy she had been offered 20 years previously. Other beliefs have indirect subtle effects; patients can provide detailed information about their problems, but reveal nothing about their general beliefs about them. For example, a woman had described a cancer phobia. It transpired that an aunt with hypochondriacal ideas had died after a long period of institutionalization in a psychiatric hospital; the patient believed that ultimately she would suffer the same fate, although vaguely hoping that treatment could stave off the evil day.

Patients with physical symptoms frequently believe that they have a physical condition which will only be helped by physical treatment, and this is also true for some depressed patients. While some beliefs require immediate intervention, others can be dealt with in subsequent sessions. Many change spontaneously during treatment; for example, a belief about the hopelessness of a condition may begin to change as soon as any improvement occurs. Ways of questioning and challenging beliefs are described in detail in Chapters 3 and 6.

It is preferable for the patient’s beliefs to be elicited during the initial interviews, but occasionally the patient may be unwilling to disclose them at this stage, or even be unaware of them. As assessment continues throughout treatment, further discussion of beliefs may arise when there is a block in progress.

Engagement in treatment

Most cognitive–behavioural treatments demand a high level of commitment on the patient’s part, and many treatments fail because the patient does not apply the agreed procedures. It would be useful to identify those people who are most likely to fulfil their side of the treatment bargain, but attempts to predict success in treatment by measuring a unidimensional or unitary ‘desire for change’ have been disappointing (Bellack (p. 37) and Schwartz, 1976). Instead it is helpful to discuss with the patient some of the components which make up a desire for change, correct any misconceptions, and together make an informed decision about whether it would be worthwhile continuing with treatment. First, the level of distress or inconvenience associated with the problem must be compared with the distress and inconvenience likely to result from the treatment. This balance is likely to change during treatment, and hence should be periodically reassessed. The patient’s beliefs about the problem and treatment (as above) should be explored, and erroneous beliefs corrected. This could involve providing new information (for example, about probable outcome); or by getting the patient to question the validity of beliefs (for example, that the right tablet would take the problem away). Finally, it is useful to discuss whether changes on a broad front, in addition to those in the target problem, would, on balance, be positive. For example, a man is unlikely to energetically pursue treatment suggestions if he knows that his wife will leave home once she is sure he can cope.

Failure to progress

Even after an initial discussion of this kind, the patient may make some progress and then discontinue homework assignments. This kind of block can be associated with the homework itself, or with the patient’s perceptions of the homework. Since similar principles apply whether engagement is an issue at the assessment or subsequent stages, the general principles will be reviewed here.

The homework task

The patient and therapist should focus on what prevented completion of the homework. Was the homework merely suggested rather than explicidy planned? Was it too vague? Was it recalled accurately? Homework tasks should ideally be written down by therapist and patient. Had the therapist routinely reviewed homework on previous occasions? If homework is not reviewed, usually at the start of each session, patients come to perceive it as unimportant. Was the rationale for the homework understood? The patient should be asked to summarize the homework for the therapist, thus identifying gaps and misunderstandings. Were there practical difficulties which interfered with the homework? (e.g. diary forms were not accessible, patient had insufficient funds to complete the task).

Patient’s beliefs about homework

If the homework tasks were set up appropriately, then non-compliance indicates broadly that the patient does not except that the homework tasks will help achieve the goals for treatment—either that they are irrelevant, or that other factors (e.g. incompetence or hopelessness) will prevent progress. This may be because the task is irrelevant, perhaps revealing new facets of the problem; or (p. 38) because the patient fears the outcome of the homework; or because the patient has misperceived the relationship between the homework and the goal. For example, previous homework may have been aimed at die acquisition of a new skill, but the patient may have been dispirited that practice did not result in any discernible improvement in the problem, and hence be reluctant to spend further time on it. Further discussion of the role of skill acquisition in overcoming problems may make the homework more relevant.

On the other hand, non-compliance may tap into patients’ more fundamental beliefs about either themselves or their problems. This can be followed up using the approach for eliciting cognitions described in Chapters 3 and 6. For example, the patient could be asked to imagine in detail what it would be like attempting a homework assignment, and to report any thoughts which went through his or her mind.

Psychosocial situation

Information is sought about the current situation as summarized in Table 2.2, without a detailed personal and family history. American texts emphasize the use of test batteries including inventories for demographic and background data covering family, religious, sexual, health, and educational histories (Cautela and Upper 1976). However, evidence for their usefulness is not strong, and there is less willingness in Britain to complete multiple questionnaires, and so there has been relatively little use of them.

Many patients expect to give lengthy descriptions of their past lives, and while they should be gently dissuaded from doing this, care must be taken not to communicate that they are not allowed to bring up difficult or embarrassing material.

Preliminary formulation

By this stage in the interview, the therapist should be in a position to give the patient a preliminary formulation of the problem. This would include a brief description of the current problem, an explanation of how the problem developed (including predisposing factors and strengths as well as immediate precipitants), and a summary of maintaining factors. As the treatment plan will be based on the formulation, it is important that the patient is asked for feedback on its accuracy.

For example, a 28-year-old woman presented with a five-year history of a bird phobia. Towards the end of the assessment interview, the therapist presented a preliminary formulation, in which liberal use was made of questions rather than statements in order to facilitate feedback. It is important that the patient is not overloaded with information, and is given ample opportunity to comment as the formulation is presented—a useful rule is for the therapist to talk in sentences, rather than presenting lengthy uninterrupted paragraphs.

(p. 39) The therapist began by giving a brief summary of the problem, emphasizing the symptoms experienced by the patient when she was confronted by a bird (or other feathery object) and her increasing avoidance of places where birds might be encountered. The discussion then moved on to the development of the problem.

Therapist: ‘So, as I understand it, your first memories of unpleasant experiences with birds date back a long way, like the ones with the birds at the seaside when you were small, and you were terrified when you saw a gigantic bird in the pantomime “Sinbad the Sailor”. There were no more frightening incidents until your teens when you saw the film of “The Birds”. You also say you have always been a nervous, excitable sort of person, who reacts very strongly to things, and have been rather anxious at times of pressure, is that a reasonable summary?’…

‘Let’s go on to the time when your fear of birds really became extreme, and see if we can understand that. You say you had moved to your new house very recendy, and although you’d been married for a couple of years, this was the first time you’d been away from your mother in the sense that you couldn’t call on her for help, as you had previously. You had moved to the country, although you felt rather uncertain about the move and you had no friends in the village. So all of that had made you generally fairly tense, is that right?’…

‘And then the problem became extreme when you came down into the living- room one morning and found a jackdaw flapping around; you became very frightened, rushed out and closed the door, and didn’t go back until your husband came home and got rid of it for you. Now, I think if you had found the bird at any other time, it would have frightened you, or startled you, but you were generally tense anyway because of the other things we’ve mentioned. This extra fright pushed your anxiety up to a very unpleasant level, and you associated all of that anxiety with the bird. This would probably have settled down, but the next day when you went out, you saw a line of ducks walking past your gate. One of these ducks was flapping its wings, reminding you of the jackdaw. As die ducks approached you became very anxious and rushed back inside. This strengthened the association in your mind between birds and anxiety. Your body had learned to respond with fear each time you saw a bird, or even thought of going places where there might be birds. So what we will aim to do in treatment is to let you learn other ways of responding to birds, to weaken the association between birds and fear. Does that sound reasonable?’…

‘If we think about what has kept the problem going, I think it may become clear what sort of steps we need to take to help you with this. I think there are two important factors. One is the way in which you have gradually avoided more and more bird situations—although it may be common sense to let the fear settle down, avoidance is one of the most important factors in strengthening fears of this kind. What happens is that each time you see a bird, or think about going anywhere where there might be birds, you get anxious; as you avoid the situation, your anxiety decreases, and you never get the chance to find out that nothing dreadful happens when you are near birds, or that you can cope with your anxiety feelings. This confirms your beliefs that being with birds is bound to make you anxious, and strengthens the association between birds and fear. Can you see any ways that we could begin to overcome that?’…

(p. 40) ‘The other important element is what happens when you are actually confronted by birds. You have given a very clear picture of your immediate response when you see a bird—like your heart pounding, feeling cold; you get anxious thoughts like, “What if it flies into me?” and an unpleasant image of a large black bird flapping its wings right in your face; you begin to cry and cling to whoever is with you; and you feel anxious. You describe this amount of fear as unpleasant but just about tolerable, and it would probably die down quite quickly when nothing dreadful happened; but because you know how unpleasant the anxiety symptoms have been in the past, you react to these initial symptoms and the whole thing builds up very rapidly. For example, you notice your heart pounding and feeling light-headed, and you get the thought “I’m feeling all anxious, I shall pass out”, “The bird can sense I am anxious”, “I must get out of here”. These thoughts rapidly increase your anxiety level, so your physical symptoms increase.’

‘There are a whole set of vicious circles of this kind, which increase your fear when you are in an actual “bird situation”. This means that another thing we need to do is to break into these vicious circles, and help you learn ways to reduce your anxiety symptoms when they begin to build up, rather than to increase them. For example, you said that when you become anxious in the house when feathers come out of pillows, you distract yourself by thinking of what jobs you have to do next. Are there other things that you do like that, which settle you down when you are feeling slightly anxious?, for example, thinking about the bird being trapped and scared itself, rather than thinking of it as a predator?’

The therapist then asked the patient to summarize the main points of the formulation—that the problem developed at a time when the patient was in a state of high arousal; that it was a learned response; that it was maintained by avoidance, and a series of vicious circles. Further discussion of the treatment plan followed on from the summary.

The presentation of the formulation generally highlights the need for further information, which can be obtained from either self-monitoring or one of the other sources, described in the remainder of this chapter. Alternatively, the preliminary formulation can be held over until self-monitoring data is available; this allows the therapist to prepare the formulation between sessions, and to incorporate the information from self-monitoring. In either case the formulation is a working hypothesis, which can be altered at any stage during treatment on the basis of new information. While changes are more likely at the beginning of treatment, treatment blocks may arise at any stage, and may alter the weighting of various factors in the formulation. For example, the presenting problem of a 30-year-old woman was her excessive fluid intake, amounting to 12–15 pints daily. The preliminary formulation emphasized her misinterpretation of a wide range of bodily signs (tiredness, tension, dry mouth, having eaten, feeling hot, headache, as well as thirst) as cues for her needing to drink. Initial intervention focused on her developing different ways of responding to this range of cues, when it then became apparent that another major factor was her misinterpretation of bladder signals. (p. 41) The revised formulation was discussed with her, and new interventions planned.

Self-monitoring often provides the essential information around which the formulation can be tested, as well as allowing progress to be assessed. The latter is easier if there are agreed treatment goals. The advantages of setting goals, and the ways of doing this, will now be described.

Goal setting

Goal setting involves agreeing with the patient detailed, specific goals for each of the problem areas which are going to be worked on, as well as setting up intermediate sub-goals. Many of the principles involved in goal setting overlap with those for devising measurement, as the measures are usually related to goals.

Advantages of setting goals

There are many points in favour of setting goals at the assessment stage. First, it helps to make explicit what the patient can expect from treatment; for example, it is unreasonable to expect never to have an argument with your spouse. It can pin-point areas of miscommunication between therapist and patient, and may help the patient decide whether to continue with therapy. For example, a patient who was interested in learning about the historical antecedents for her panic attacks did not feel that a treatment aimed at reducing the frequency of her panics addressed the appropriate issue.

Setting goals also emphasizes the possibility of change, and begins to focus the patient on future possibilities rather than simply on symptoms and problems. It also reinforces the notion that the patient is an active member of the therapeutic relationship, and that full involvement is required: the patient will not be ‘done’ to.

Defined goals help to impose structure on treatment. This allows the presenting problems to be addressed, with less risk of diverting into a series of crisis interventions. It also prepares the patient for discharge, making explicit that therapy will be terminated when goals are achieved; or that therapy will be discontinued if there is little progress towards them. This is not to say that goals cannot be re-negotiated during treatment, but that this should be done explicitly, together with the patient, thus reducing the risk that patient and therapist are pursuing different agendas.

Finally, setting goals provides the opportunity for an evaluation of outcome related directly to the individual’s presented problems.

How to set goals

  1. 1. Whenever possible, goals should be stated in positive terms, so that it is explicit what the patient is moving towards rather than away from. For example, a patient could aim to ‘clean the kitchen calmly within one (p. 42) hour’ rather than ‘have no handwashing rituals during housework’. Gam- brill (1977) refers to the ‘dead man’s solution’ as that which could be achieved by a dead man; for example, a dead man would have no panic attacks, no urges to binge, no sleepless nights. She suggests that such solutions be avoided. As it is often difficult to turn the patient away from symptoms, and towards positive goals, it may be useful to say something like:

    ‘It’s as though you’ve been wearing glasses which are very good at focusing on symptoms and problems. I want you to start wearing glasses which pick up evidence that you’re coping, evidence of success. So it’s useful if we are clear what success would be like.’

    Specific questioning may help to focus the patient on positive targets. For example, a patient said that she wanted to ‘stop being irritable all the time’, and she was asked, ‘What would you do that was different if you were not irritable?’ Barlow et al (1984) suggest asking the patient to make three wishes, or to describe a typical ideal day. It is also helpful to pick up positive goals as they are mentioned throughout the interview (for example, ‘I wish I could just invite friends round for supper, like I used to’), and remind the patient of these if necessary.

  2. 2. Goals should be specific and detailed. Patients are often aware in general terms of how they would like to be. For example, when the therapist asked what she would like to achieve through coming to the hospital, one patient replied:

    Therapist ‘I’d just like to be normal, like everybody else.’ ‘Being normal means different things to different people. If you felt normal, how would you be different from how you are now?’

    Patient ‘I would be more like I used to be.’

    Th. ‘What would tell you that you were more like you used to be? What would you be doing that you’re not doing now?’

    The patient was eventually able to list the following goals for herself: to go out shopping alone in supermarkets; to stay overnight in the house alone; to initiate contacts with friends; to have friends round to her house. Questioning then helped to specify where/when/how often particular behaviours would occur if each goal was met. If possible, goals should be phrased so that more than one person could agree if the goal was achieved, as this is likely to increase the reliability of measures related to goal achievement.


Self-monitoring is the most widely used adjunct to behavioural interviewing, and is almost invariably used both at the initial assessment stage and to monitor subsequent change. The introduction of self-monitoring at (p. 43) the beginning of treatment emphasizes the self-help, collaborative nature of treatment. It is flexible, can be applied to a wide range of overt and covert problems, and can give information about many aspects of problems. Barlow et al. (1984) point out that there are two stages in self-monitoring: first, the individual has to note that the behaviour, thought, emotion, event has happened; and secondly has to record that it has happened. These stages should be borne in mind when self-monitoring is initially devised and care taken to obtain accurate measurement.

Accuracy of self-monitoring

The accuracy of self-monitored information is increased if a few general rules concerning measurement are followed. Thus, only appropriate and meaningful information should be requested, without overburdening the patient. The importance of self-monitoring should be emphasized, it being made clear that subsequent treatment sessions will focus on the material. Explicit rather than tacit agreement to monitor should be obtained.

The accuracy of self-monitoring increases if the patient is aware that its accuracy will be assessed (Lipinski and Nelson 1974). This is often difficult clinically, but can be achieved most directly if the events can intermittently be monitored by someone else; for example, time spent handwashing by an obsessional patient could also be monitored by a spouse.

How to self-monitor

A measurement procedure should be relevant to the question asked, should measure what it is supposed to measure (valid), and should provide a reasonably consistent account of how things actually are (reliable). There are various ways of increasing relevance, validity, and reliability.

Specific, clearly defined targets

It is difficult to achieve reliable measurement of vague concepts like ‘self-confidence’. The feature or event to be measured should be defined in detail, as far as possible in ways which would allow different observers to agree about its occurrence. For example, a patient who was lacking in self-confidence was asked,

Patient ‘How would you know if your self-confidence improved? What would you be doing that you are not doing now?’ ‘I wouldn’t be panicky like I am now, and I’d be doing all sorts of things I can’t do now because I don’t have the confidence.’

Therapist ‘Can you give me some examples?’

Pt ‘Well, I would invite some of my neighbours in for coffee—they must think I’m very funny because I never do. I would begin to show my pictures at local Exhibitions. I could drive myself and do the shopping on my own—I passed my test years ago.’

(p. 44) Further discussion elicited a list of indices of ‘self-confidence’ which were amenable to self-monitoring. The criterion that measurements should refer to events which are observable may be difficult for internal states but it may be possible to measure the external effects of an internal state. For example, it would be easy to disagree about ‘Mr G was angry’, but easy to agree about a more detailed ‘Mr G shouted’, ‘Mr G kicked door/furniture’. This could be supplemented by Mr G counting the number of angry thoughts, and giving self-ratings of anger.

In general, the instructions about what to record should include requests for information about the frequency, intensity, and duration of the targeted problem where these are relevant.

Aids to recording

The therapist should provide the patient with a form or recording device which allows easy record keeping. Patients usually cannot draw up record forms for themselves until they have become skilled at recording. The patient should be clear about what and how to record information. This is best achieved by going through a worked example with the patient.

Meaningful and sensitive measures

The most meaningful measures are often different from the most sensitive ones. For example, a girl was being trained in assertiveness skills as a way of increasing her self-esteem. The most meaningful measures were self- ratings of behaviours associated in her mind with self-esteem (e.g. being able to initiate social contacts), and questionnaires related to self-esteem (Rosenberg 1965). However, these indices would be insensitive to small daily changes during therapy and could only be used, say, monthly. In order to look more immediately at whether changes were occurring, more sensitive measures (such as the number of times each day that she said ‘sorry’) were also used.

Simplicity of measures

It is usually helpful to use multiple measures for each problem since there is no single ‘true’ measure of a problem which will adequately reflect all aspects (see p. 8). However, patients should not be bombarded with demands for information. It takes time to acquire skill in record-keeping, and it is better to keep it simple, particularly at the beginning of treatment. Data should only be collected if the patient and the therapist are clear about what it will be used for. Patients are much less likely to keep records if they appear irrelevant.

Timing of measurement

Recordings should be made as soon as possible after an event (behaviour, thought, or feeling) has occurred. If the patient stores up examples and (p. 45) records them all at the end of the day, then some examples will be forgotten and others distorted. This will be particularly true if, for instance, the patient is depressed, low in self-esteem, or anxious, and is recording examples of achievement or coping. It is important, therefore, that the means of recording is easy to carry and use—a notebook, for example.

Types of self-monitored information

There are many different kinds of data which can be monitored. Specific examples are given in each of the subsequent chapters, but a broad description follows to allow the reader to design the most relevant self-monitoring for specific problems.

Frequency count

If there is a relevant and meaningful aspect of the problem which can be counted, then this will provide the most accurate information. This has wide applicability and it is worthwhile trying to find discrete aspects of a problem to count; for example, number of visits to family doctor per week, number of self-critical thoughts, number of hairs pulled out, number of panic attacks, number of arguments with spouse. The data can be recorded on diaries, or as frequency logs on cards; but for problems with high frequencies, it is easier to use a mechanical counter (for example, a golf counter or knitting counter).

Duration of problem

It may be appropriate to measure the duration of the target event or behaviour. Examples are: how long an agoraphobic patient spent away from home, time spent handwashing, time spent studying, and how long it took a patient to settle down after an episode of overbreathing. This information can be recorded in a diary, unless a stopwatch with an elapsed time indicator is available: this stores cumulatively how much time elapses whenever a switch is in the ‘on’ position.


These are used when information is required about a patient’s affective or subjective state, and are frequently obtained in addition to the frequency and duration measures described above. They are less reliable than more direct measures, and ‘anchor points’ may change as the patient improves unless great care is taken to specify what the points on the scale mean. For example, the meaning of ‘mildly distressing’ on a five-point scale from ‘not at all’ to ‘extremely’ distressing may change as the patient gradually has less frequent highly distressing experiences.

Ratings are more reliable if they are made at the time the problem occurs. If the problem or event occurs discretely and infrequently, the (p. 46) patient can be asked to rate each time it occurs; an example would be the intensity of ‘urge to check’ in an obsessional patient. If the problem occurs continuously or very frequently, then the patient may be asked to record for a fixed period of the day, either chosen because it has particular significance for the problem (for example, recording distress of ‘fatness’ thoughts in the hour following a meal), or because it is felt to be representative of the day. It may be necessary to arrange cues to remind the person to record; for example, there are portable timing devices which buzz when a recording is required, either at a fixed or varying time interval. Although less reliable, it may be more useful to ask the patient to make a rating of a subjective state averaged, for example, over a day, or hourly, or three times per day. Accuracy may be improved if the patient is also asked to rate the worst feeling during the day, to distinguish this from the remainder of the day.

Rating scales differ in their form, and range from visual analogue scales where a standard length line is provided and a mark can be made at any point along it (see Fig. 2.7 for an example), to numerical scales with a set of separate and distinct response categories, one of which must be marked (see Fig. 2.1 for an example).

Fig. 2.7 Visual analogue scale for therapist ratings of amount of eye-contact shown by a patient

Fig. 2.7 Visual analogue scale for therapist ratings of amount of eye-contact shown by a patient


These are widely used, and often include frequency counts, duration measures, and self-ratings, but in addition include information about the circumstances in which the event occurred. It is important to specify closely what information is required; otherwise a vast amount may be recorded in a form which it may be impossible to assimilate, and with little check on whether the same material was attended to on different occasions (Figs. 2.3, 2.4, and 2.6 provide examples of diaries in addition to the numerous examples in individual chapters).

Reactivity of self-monitoring

When the patient begins to record the occurrence of an event, its frequency changes (Barlow et al. 1984). This phenomenon is called the reactivity of self-monitoring, and it occurs whether or not the monitoring is accurate. It may occur because the monitoring interrupts an automatic chain (p. 47) of behaviour, and allows the person to decide whether to continue; for example, the chain ‘sight of stranger getting out a cigarette—urge to smoke—reaching in pocket—getting out cigarette’ may be interrupted if the person has to rate ‘urge to smoke’ before getting out a cigarette. Clinically, it can be useful as the changes are almost always in the therapeutic direction. It is more problematic, however, when the data from self-monitoring are being used to establish, for example, a baseline.

In summary, self-monitoring has a central role in cognitive–behavioural assessment and treatment. If the patient does not self-monitor in spite of adequate care in setting it up, this can be treated as any other kind of non-compliance (see p. 37).

Self-report questionnaires

There is a somewhat arbitrary distinction between self-monitoring and self-report, but the latter refers to more retrospective and global information than self-monitoring. The most frequent source is questionnaires, which have the advantage that normative data are often available against which the patient can be compared. It is worth emphasizing again that self-report provides different but not necessarily inferior information to more direct measures. For example, there may not be a perfect correlation between physiological recordings of cardiac function, and a patient’s self-report of palpitations. While the physiological data may be important, the patient’s perception of cardiac function is equally relevant.

It is only worthwhile using questionnaires which have demonstrable psychometric soundness. Content validity is particularly important and refers to the extent to which the questionnaire adequately measures the relevant area. This should have been determined by the authors of the questionnaire on an empirical rather than logical basis, and validational data should be easily accessible.

Subsequent chapters give information about relevant questionnaires, but a wide range of examples is available in Cautela and Upper (1976) and Bellack and Hersen (1988).

Information from other people

Additional information may be obtained from other people, in each of the areas discussed so far. Thus, key people may be interviewed, may monitor information as it occurs in vivo, or may provide more global retrospective information. The relevant people include the therapist, relatives or other key individuals for the patient, or staff interacting with the patient.

Interviews with key individuals

The main aims of such interviews are identical with those when interviewing the patient. These are to derive and present a formulation of the (p. 48) problem; to educate the relatives or others about the nature of the problem and the psychological approach to treatment; and to engage them in treatment if this is relevant. More specifically, it is useful to establish during the interview what impact the problem has on the key individual; what that person’s beliefs are concerning the problem; and how the person responds to, or copes with the problem. Information may also be available about avoidance which had not been mentioned by the patient. This part of the assessment may be longer than the interview with the actual patient if the problem behaviour is more distressing to others than to the patient. For example, the husband of a woman with panic attacks believed that his wife was suffering from madness, and that the major aim of treatment was to keep this distressing fact from her as long as possible. The husband’s beliefs only became clear after a lengthy interview in which his pessimism about therapeutic outcome was discussed. Further examples of the central role of information from others are given in Chapter 9.

It is important to check whether the relative or other person wishes the therapist to keep any of the information confidentially (and to have made a similar check with the patient before interviewing the other person). If so, it is worthwhile discussing whether the request is based on unreasonable fears.

Monitoring by key others

This may be used to enhance the accuracy of self-monitoring, but it can also provide specific information about the impact of the patient’s problem on others. This will be particularly relevant where other people are heavily involved in the problem; for example, a spouse reassuring a hypochondriacal patient, or relationship problems (including those with children). The general principles for obtaining accurately monitored data are exactly the same as for self-recorded data, and it should be set up with similar attention to detail.

Observations by staff in therapeutic environments are discussed in detail in Chapter 9.

Direct observation of behaviour

It is often useful to have direct observation of a problem behaviour; for example, a patient may broadly outline handwashing rituals, but be unable to give a detailed description. Sometimes it is difficult to arrange the observation in naturalistic settings. One example is where the patient describes gross inadequacies in social skills, and it is unclear whether these represent deficits or anxiety about social performance.

Observation of naturally occurring behaviours

If the relevant behaviours occur with the therapist, then measures can be taken at the time, provided that the situation can be standardized. Such (p. 49) measures can include frequency counts, duration measures, and ratings. For example, with a patient who complained of abdominal distension, the therapist counted the number of burps per session; with an electively mute patient, the therapist counted the number of words per session. In such examples, the length of the session should either be constant or the frequency count calculated on the basis of a constant session length. With a depressed patient, ratings of self-criticism and sad mood were made for the patient’s response to the standard question, ‘How have things been this week?’ Another example of useful ratings were those made by a therapist of the amount of eye contact made by a socially withdrawn patient in each session, the visual analogue scale shown in Fig. 2.7 being used for this purpose.

If the target behaviours do not occur spontaneously in the clinical setting, it may be possible to contrive the situation so that the behaviour can be observed by the therapist. Two common examples are role-play and behavioural tests.


If the problem involves interactions with other people, then role-play with a stooge allows direct observation of the problem behaviour, and can be repeated pre- and post-treatment to assess change. Where possible, the role-play should be videotaped, and then rated on relevant dimensions by independent observers who have practised using the rating scale; this method has been used to assess the efficacy of social skills training (Trower, Bryant, and Argyle 1978). In another study, couples with marital problems were asked to discuss problem topics, and their interactions were videotaped and subsequently coded (Bornstein, Bach, Heider, and Ernst 1981).

It cannot be assumed, however, that there is a high correlation between performance in role-played situations and that occurring in everyday settings. Role-played performance is sensitive to situational variables; for example, assertiveness may vary according to whether a good friend or acquaintance makes a request and whether or not a reason is given for the request. When ratings of role-play performance are used to assess treatment changes, treatment should not have focused specifically on the role-played tasks. It will otherwise be impossible to determine whether the improvement has generalized beyond those specific tasks.

Behavioural tests

These allow direct observation of a wide range of problem behaviours, and many examples are given in subsequent chapters; for example, avoidance tests for phobic patients (p. 106) and behavioural tests with obsessional patients (p. 143). The measures obtained from such tests can include specific, objective measures (for example, length of time spent confronting a phobic object), as well as ratings by both the patient and (p. 50) therapist. For example, an obese patient shop-lifted much of the food required for her binges; a behavioural test in a supermarket focused on the chain of behaviours which led up to stealing, with the aim of breaking into this chain at multiple points. The measures included the amount she stole (e.g. when carrying different bags, wearing different clothes), and self-ratings of urge to steal at various points in the shop. Another example was the behavioural test for a patient with writer’s cramp, who was asked to write out a standard passage; the measures included the length of time taken and number of words completed, with ratings of discomfort by the patient, and therapist ratings of ease of pen-hold.

Behavioural by-products

These measures are indirect and do not focus on the problem behaviour itself. They have the advantage that they are objective and relatively free from observer bias. A common example is weight as a by-product of eating, used with patients with eating disorders. Other examples include the amount of money spent on food by patients who binge, the amount of soap used per week by obsessional patients, and the number of hairs pulled out by patients with trichotillomania.

Physiological measures

Physiological processes may be monitored indirectly; for example, a patient could self-monitor frequency of headaches, or a socially phobic patient could rate the amount perspired in social situations. Although there is an extensive literature on psychophysiological measurement, its use in routine clinical practice is limited by the cost and availability of equipment. However, psychophysiological changes may precede other changes, for example subjective and behavioural ones, and low-cost devices are increasingly available. Examples are given in subsequent chapters, particularly where the problems are largely somatic (p. 251).


The principal aim of a cognitive–behavioural assessment is to derive a formulation and treatment plan. Most of the information will be collected during interviews with the patient, and a preliminary formulation may be discussed following the initial interview. However, to complete the formulation it will almost always be necessary to obtain further information. Self-monitoring by the patient will generally be required and questionnaires may be relevant. In addition, information from relatives or others may be helpful. Direct observations of problem behaviour often highlight facets of the problem which it would otherwise be difficult to assess. Unless the behaviour occurs spontaneously in the clinical setting, it (p. 51) may be necessary to set up role-play tasks or behavioural tests. In some cases, physiological data may be collected, but these will often be indirect, for example ratings by the patient.

It may take two or even three sessions to complete this preliminary assessment and arrive at the formulation. Many studies have found that die majority of change occurs during the first few treatment sessions, and it would be unfortunate to minimize this by introducing inappropriate strategies before the therapist and patient had an adequate understanding of the problem. Having agreed together the nature of the problem and probable maintaining factors, the therapist and patient are in a position to make changes in the antecedents, consequences, or the behaviour itself, and to monitor the effects. The rest of the book focuses on the therapeutic approaches to specific problems.

Recommended reading

Barlow, D. H., Hayes, S. C., and Nelson, R. O. (1984). The scientist practitioner. Pergamon, Oxford.Find this resource:

    Bellack, A. S. and Hersen, M. (1988). Behavioural assessment: a practical handbook, (3rd edn). Pergamon, New York.Find this resource: