Individual case formulation (ICF) has a long history within psychological therapy (see Chapters 1 and 2 by Bruch in this volume). This chapter discusses case formulation as a process that draws upon scientific knowledge and experimental reasoning but, as I hope to show, this is only part of a mix that includes knowledge of culture and practical wisdom. In other words, case formulation (and therapy in general) cannot simply be a matter of applying science in a manner comparable with engineering. Scientific methodology is a component of a therapist’s expertise in the sense that it contributes to a systematic analysis of problems and provides guiding principles that draw upon well-validated psychological processes.

Within the general field of applied psychology, ICF is distinctive because it takes an idiographic approach to the task of applying expertise. It differs from the mainstream nomothetic view that expertise is based on (a) theoretical models that have been developed to explain a type of problem (e.g. depression, anxiety and obsessions) and (b) techniques and methods that have been shown in controlled trials to ‘work’ in ameliorating a certain kind of problem. According to the nomothetic approach, a therapist’s expertise consists in knowing which theoretical model to use to understand a problem and knowing what the evidence shows in terms of the techniques likely to resolve it. In current jargon, it is ‘evidence-based’.

The idiographic approach to applying knowledge in formulation places more emphasis on a therapist’s ability to build a ‘unique theory’ to make sense of a client’s problem. The difference is only one of emphasis because both nomothetic and idiographic approaches draw upon ‘bottom–up’ and ‘top–down’ reasoning. A ‘bottom–up’ process can be defined as one that is driven by data or observations, while a ‘top–down’ process involves searching for certain kinds of data within the presenting problem, guided by theoretical principles or a general model. The reason for preferring an emphasis on bottom–up processes in ICF is scepticism about the assumption that any problem falls neatly into a type for which there is an adequate theoretical model or set of treatment guidelines. The ICF therapist is more likely to draw upon a variety of theoretical principles and knowledge resources. Consequently, in order to formulate systematically, she or he has to be more reliant on disciplined procedures for collecting and evaluating the observations that are made. The final case formulation may not resemble anything that has been produced in any previous case analysis.

As already noted, nomothetic and idiographic approaches lie on a continuum. It would be foolish for an ICF therapist to ignore evidence that has accumulated in outcome studies even if it relates to rather rigidly defined problem categories. Moreover, theoretical models that have been produced to account for specific ‘disorders’ (i.e. problems diagnosed as expressions of a psychiatric disorder) have been developed on the basis of research evidence and are always worth considering. Contrariwise, it is also necessarily true that therapists who follow a nomothetic approach have to be sensitive to the individual circumstances of their clients. A manual that follows a disorder-specific theoretical model cannot be allowed to become a bed of Procrustes to which a client is fitted regardless of their unique needs.

A criticism often meted out to the ICF idiographic approach is that it is ‘subjective’, meaning that it is biased by untested assumptions or simply that it is too ‘intuitive’. In this chapter, I will stress that the process of case formulation needs to be systematic and aware of its potential biases. No one, I believe, wishes to advocate ‘on the hoof’ inspired guesswork. The model of applied science put forward by detractors of ICF is also open to criticism. It has followed the argument of Paul Meehl (1954) for a more ‘objective’ actuarial approach in applied psychology. He showed that a clinician’s intuitive judgement about what to do is often inferior to using a set of objective or operational procedures when the outcomes (e.g. a diagnosis or a therapeutic outcome) could be specified in advance. Once the desired outcome is known, the predictive accuracy of the objective measure or procedure can be gradually improved over time, based on empirical research and ‘hard data’. This, broadly speaking, has been the mainstream approach for adding to knowledge about ‘what works for whom’ in the field of psychological therapy.

This approach to applied science works well when input and output (e.g. type of problem and desired outcome) can be specified in a way that satisfies all concerned. In the area of social epidemiology, a great deal has been learned about the phenomenology of common problems by systematically studying the relationship between well-defined variables (e.g. a history of childhood sexual abuse and a later diagnosis of schizophrenia). All of this knowledge can contribute to a therapist’s expertise. However, clients’ problems and the goals they aspire to reach are largely defined by clients, not by the experts treating them. Problems are typically obstacles, compulsive habits or deficits of some kind that interfere with the attainment of an aspiration. They are rarely, as in medicine, problems that the patient wishes to have diagnosed as a treatable disorder. This means that nomothetic knowledge of the kind that technique X is the best treatment for disorder Y is less relevant in psychological therapy.

Meehl’s approach is, of course, valuable in many technical fields. As technology develops, there is less need to rely on the judgement of an expert. The early pioneers of flight had to assess all the variables that kept the plane in the air, whereas modern pilots can leave most of the decisions to computers. Rather like one of the pioneers of flight, an early conception of an expert in applied psychology was a person trained as a scientist who heroically investigated and solved whatever problem was thrown at them. This was understandable given that the old paradigms of medicine and psychoanalysis were obviously failing, and psychologists took it upon themselves to make a fresh start using empirical and experimental methods.

One model of the expert in ICF remains that of the so-called scientist–practitioner (Raimy, 1950). Do we still have need of them? Subject to the cautions expressed earlier about fitting clients to models and methods rather than vice versa and with regard to research and development, the answer is obviously yes, just as we still need test pilots to see how a newly designed plane performs. I believe that the answer is also a qualified yes in the sense that every client presents a practitioner with a novel and unique problem to solve. A therapist has to be a general problem-solver, and part of this expertise is grounded in an experimental style of reasoning originally developed for scientific purposes.

It should not be controversial to assert that there are novel, unique and impossible to categorize aspects of any individual client’s problem. There is often no well-validated model or technique to turn to, even if one wished to follow that path. Consequently, it is necessary to fall back on principles of problem generation and behaviour modification in general, rather than categories of problem and specific techniques. An analogous situation might be presented to a geologist who attempts to explain the combination of natural forces that have led to a particular rock formation. By virtue of historical accident, the formation is unique. The equivalent of Meehl’s equations might not be available.

In fact, when attempting to understand a client’s problem, the situation is not one of prediction (by filling in the values of an established formula) but of making coherent sense of all the prevailing aspects of a client’s life taken together. A knowledge of general principles should, of course, give rise to particular hypotheses – which in a sense are predictions – but they have to be tested by collecting further data from the client. A prediction could take the form that if a client continues to do X, an undesirable consequence, Y, will follow. A number of such predictions could be made. However, it will still be necessary to weigh up these predictions and put them in the balance. Perhaps, doing Z would lead to even worse consequences, and the client is limited to a choice between X and Z. The point I am making is that resolving a problem involves pragmatic choices and practical wisdom. Perhaps, it has not occurred to the client that doing W is worthy of consideration. There is always an element of choice in how a problem is solved because a problem is a manifestation of unique cultural–historical circumstances. Problems do not always have obvious solutions. The sense that has to be made of any problem also has to be owned by the client and accepted as valid. However, clients also seek out therapists on the basis of their acknowledged expertise. A therapist must make sense of a set of observations, notwithstanding the fact that this creative leap of the imagination must be tempered by knowledge of what is possible and feasible. Included in this set of limits is the influence of unbending universal psychological processes. For instance, are a client’s choices limited by their capacity to learn?

The creative element is held in considerable suspicion by some writers on case formulation. Haynes and O’Brien (2000: p. 59) express a bias against ‘subjectivity’ when they state: ‘There are few empirically-based guidelines for making data-based clinical judgments and almost no guidelines for designing intervention programs. The dearth of guidelines has encouraged the use of purely intuitive approaches to clinical judgment.’ Haynes and O’Brien are certainly in favour of an individual approach to formulation but they seem unwilling to go the further step of accepting that the basis of clinical judgement cannot be purely quantitative. There simply are limits to the use of algorithms.

Scientific Reasoning in Case Formulation

An approach to studying individual phenomena that emphasizes the reasoning processes of the scientist has an illustrious history. For instance, there is still a great deal to be learned from the writings of Claude Bernard (1813–1878), a famous French physiologist, who wrote about his experimental method (Bernard, 1957). Of course, Bernard was interested in establishing the validity of universal generalizations – a nomothetic enterprise – but he experimented on single animals. His subject was the internal environment of the body, an area of study that was complicated enough, but psychological therapists are faced with a far more unwieldy set of variables. These interact dynamically from week to week and include a number of unpredictable environmental events as well as a client’s reflexive awareness of any changes that are occurring within their own life and sense of well-being. And unlike Bernard’s animal subjects, clients react to the methods used by therapists to investigate their problem. For instance, a client may ‘know’ what a therapist is up to when following a certain line of questioning. Case formulation is necessarily a collaborative process. A therapist is not attempting to establish universal generalizations (as in Bernard’s case), but there are still certain lessons we can take from him; these are lessons connected with the way causal inferences are made from observations.

1. He feels it important to separate the act of observation from the act of reasoning. ‘The observer listens to nature: the experimenter questions and forces her to unveil herself’ (Bernard, 1957: p. 6). This separation is suggested on the basis of a methodological caution: there are, of course, two phases in the experimental method. The aim of the experimenter is to induce new observations and, having done so, he or she reverts to observing without prejudice. Bernard also notes that this separation is hard to achieve because the initial facts may be viewed selectively according to preconceptions and the new ‘induced’ facts may not be perceived with an open-minded attitude that allows the information they convey to truly answer the question posed by the experimenter.
   
2. The observed ‘facts’ must be accurate to begin with. The observer might be unaware of the reasoning that she/he brings to the facts, and one aim of the experimental method is to bring these to light. Reasoning deals with two facts – comparing or relating the observed fact and the induced fact.
   
3. Bernard therefore stresses that ‘Pile up facts or observations as we may, we shall be none the wiser’ (1957: p. 16). He refers to the induced facts as controls in the reasoning process.
   
4. Inducing a new fact always involves an idea (or a hypothesis) that comes from intuition. Bernard, with his background knowledge, was not, of course, pulling rabbits out of a hat. But even he admitted that it was sometimes necessary to go fishing in ‘groping experiments’ (1957: p. 21).
   
5. While stressing that observation should be unbiased, the stage of reasoning involves an active, questioning mind according to ‘various hypotheses that suggest themselves’ (p. 22). However, the experimenter then reverts back to observing the induced facts without preconception. He or she must be ready to abandon the idea that inspired the experiment. The experimenter is always sceptical, and hypotheses are always provisional.
   
6. Bernard was suspicious of generalizations that gathered in a range of facts. ‘Scientific generalizations must proceed from particular facts to principles, and principles are the more stable as they rest on deeper details.’
   
7. An experiment is a supposition about the causes of perceived activity, and then measures are taken ‘to make facts appear’ (p. 32), which logically inform the experimenter that the supposition is confirmed or negated. The supposition must of course be of a kind that is potentially verifiable or refutable. Bernard does not think that there are any rules for thinking up a good idea. In some cases, this is just a rare creative talent, but hopefully, among therapists, experience and training can make a contribution as well. Ultimately, it is good method, not creativity alone, that promotes lasting scientific developments.
   
8. Bernard was well aware of the dangers of his method. Nevertheless, for him, ‘… it is better to know nothing than to keep in mind fixed ideas based on theories whose confirmation we constantly seek, neglecting meanwhile everything that fails to agree with them’ (p. 37).

Bernard seems very much aware of what we would now term biases in information processing (Kahneman & Tversky, 1973; Ross, 1976; Turk & Salovey, 1988). He also stresses the process of thinking creatively about a collection of related observations and bringing them together in a synthesis. This is now referred to as ‘abductive reasoning’ (Evers & Wu, 2006). Before expanding on the processes involved in bringing the facts together in a coherent and productive formulation, I will examine how Bernard’s style of single-case experimentation was taken forward in psychology.

Single-Case Methodology

One example of this approach comes from the group of researchers associated with M. B. Shapiro at the Institute of Psychiatry in London. Victor Meyer was part of this group, and he later developed his own model of formulation at the Middlesex Hospital (see Bruch, Chapters 1 and 2). In the late 1950s and early 1960s, therapists began to see themselves as both scientists and practitioners and used single-case experimental methods to investigate a client’s problem with the aim of developing a theoretically based intervention. The general approach is described by Shapiro (1961) who underlined the importance of (a) devising measures that are geared to the unique nature of a client’s problem (b) establishing experimental control over a phenomenon of interest, meaning that a therapist comes to understand what produces it or removes it, and (c) ensuring that relationships observed between variables found in a single client can be replicated again in the same client or in other clients with a similar problem. The underlying rationale was to work out the parameters and determinants of a problem with the aim of discovering an individually tailored solution.

Another very important development in single-case investigation took place in North America, influenced by B. F. Skinner’s methods and concepts. The origins of this method were not therapeutic although they came to be applied in clinical settings (e.g. Lindsley, 1959). Over time, a sub-field of research methodology developed, now known as single-case experimental design.

Although single-case methodology of this type is still alive and well, most therapists find a commitment to this experimental method too demanding in routine practice environments. It proves time-consuming to carry out thoroughly, even assuming that the parameters of a problem can be identified with the level of precision required. Nevertheless, the early examples of this approach created a new emphasis on causal relationships and exerted a strong influence in a less rigorous form, shaping the style of work of some of the early pioneers of behaviour therapy. For instance, in a functional analysis of a client’s problem, data are collected in a ‘baseline period’ by an observer or by the client herself or himself between sessions. A change of conditions is then introduced, and data collection continues. The conditions may then be changed back to baseline. With the help of these experimental designs, inferences about potential causal relationships (i.e. what produces or removes a behaviour) can be taken to a highly sophisticated level (see Barlow, Nock, & Hersen, 2008).

The spirit of hypothesis generation and testing, in a more clinician-friendly form, was carried forward by Meyer and also by Joseph Wolpe and Ira Turkat. Wolpe describes his interviewing approach in a book edited with Turkat, who had worked previously with Meyer (Wolpe & Turkat, 1985). These early pioneers focused on the concrete details of a problem, and their interview method involved a process of testing hypotheses about its underlying causal determinants. They demonstrated that even without the benefit of precise measurement and the experimental manipulation of conditions, a practitioner could reach sound conclusions with implications for selecting a successful intervention. An account of the interviewing style developed by Meyer and Wolpe now follows.

The Hypothesis-Testing Interview

This style of interview is very much in keeping with Bernard’s experimental philosophy. Taking our cue from Bernard, we can ask what is a ‘fact’ and what is an ‘induced fact’? What kinds of knowledge does the therapist draw upon? At what level of complexity (theoretical principles, comprehensiveness or scope, integration or coherence) should a hypothesis be framed? If we regard a case formulation as a kind of synthesis that results from the hypothesis-testing interview, how should it be expressed – in words, symbols, diagrams, etc.?

Related posts:

The Development of Case Formulation Approaches The Therapeutic Relationship as a Critical Intervention in a Case of Complex PTSD and OCD Cognitive-Behavioural Formulation and the Scientist-Practitioner The UCL Case Formulation Model Case Formulation and the Therapeutic Relationship Generalized Anxiety Disorder

Stay updated, free articles. Join our Telegram channel

Join