The Effect of Setting Homework on the Therapeutic Alliance in the CBT of Difficult Clients

Published: 2019/12/10 Number of words: 2982

Cognitive Behavioural Therapy (CBT) is currently the treatment of choice for a number of psychological difficulties ( CBT is a directive, goal-oriented, problem-solving psychological therapy and as such can initially seem at odds with clinicians prising client-centeredness, openness, non-directedness and not-knowing. One of the most fundamental aspects of CBT is the setting of homework between therapy sessions in order to consolidate therapy work taking place during sessions. Because the therapeutic alliance (TA; Hotvath & Bedi, 2002) is harder to establish and maintain with more troubled clients, the impact of setting homework can be more visible in such a context. In this short essay I will be drawing on my practice with eating disorders, at the inpatient level, to briefly examine the impact of setting homework with the clients. I will argue that it can have both positive and negative effects on the TA and that CBT can in fact be a helpful choice for helping difficult clients.

The Therapeutic Alliance

Norcross has eloquently articulated the importance of the therapeutic relationship:

(S)ome clinicians conceptualize the relationship as a precondition of change, others as the fertile soil that permits change, and still others as the central mechanism of change itself (Norcross, 2010, p.114).

It is well established that the therapeutic relationship is a central feature of clinical work in many psychological therapies (e.g. Woolfe et al., 2009) and that the therapeutic alliance is the most important aspect of the therapeutic relationship (Norcross, 2010). According to the classic transtheoretical formulation by Bordin (1979), the alliance consists of goals, tasks and the bond of the therapeutic couple. At the outset, cognitive therapy conceptualised psychotherapy as a process of “(O)ffer(ing) the patient effective techniques for overcoming his blindspots, his blurred perceptions, and his self-deceptions” (Beck, 1976 p. 213). The therapist’s tasks in CBT include framing the treatment as a joint effort, working with the client to identify the goals of the therapy, encouraging and eliciting feedback from the client, and monitoring and managing the alliance. The client’s tasks involve work on discovering cognitions within the session and homework. Beck conceptualised a successful therapeutic relationship as a necessary component for the unfolding of cognitive therapy. He stressed that therapist’s qualities such as warmth, empathy and positive regard (Rogers, 1957) were essential in order for the techniques of CBT to be effective. One of the central principles of CBT is collaborative empiricism where the therapeutic couple works together in an equal and transparent manner to address the client’s difficulties. However, as the use of CBT extended to the treatment of more complex problems and particularly personality disorders (Beck, Freeman & Davis, 1990) it has become clearer that for those clients the basics often do not work (J. Beck, 2005). Setting homework can be quite a complex process for these patients and its creative use can help as much as it can hinder the therapeutic alliance.

Homework in CBT

CBT is different from other therapies in that it postulates that is not sufficient for clients to talk about their problems, but that they need to work on ways to solve them during sessions and try to implement solutions between sessions (J. Beck, 2005). Homework is a fundamental task that client and therapist agree on from the beginning of therapy. During sessions the therapeutic couple attempts to identify the client’s cognitions so that they may be understood and accepted or challenged. Homework may be used to assist in the discovery of cognitions by using diaries monitoring the client’s daily activities, or thought records aimed at identifying the “hot thoughts” that emerge during crises. Alternatively, homework can consist of behavioural experiments collaboratively designed to test the client’s predictions and to eventually generate alternative thoughts (Bennett-Levy et al., 2004). The use of homework is likely to be agreed on during the first meeting between the couple. It is therefore the task of the therapist, in collaboration with the client, to assign homework that is relevant to the client’s goals and appropriate to the client’s current needs. The completion of homework constitutes the client’s major contribution to the working alliance. Non-compliance with homework can be the result of inappropriately choosing the problem to focus on, or of the client’s faulty cognitions interfering with this process.

The positive impact of homework

I believe that the way homework positively impacts the TA is by making the client’s contribution to the alliance explicit and the progress of therapy potentially more observable. Compliance or non-compliance can therefore be an objective indicator of the success of the ongoing therapy. Non-compliance with homework does not necessarily constitute a problem insofar as the reasons for it are understood and worked through. In fact, the setting of homework may allow an on-line exploration of the client’s faulty cognitions, particularly as they apply to the exchanges between the therapeutic couple. The client may gain a sense of being understood, and potentially valuable information about the client’s maladaptive thought patterns may emerge that might otherwise have gone unnoticed.

For more troubled clients who are disconnected from their thoughts and feelings, and operate primarily at the level of observable behaviour where actions speak louder than words, the act of giving homework may be experienced as concrete evidence of the therapist’s caring. It can be seen as the therapist “giving” something to the client, a gesture of concern and interest, and the use of homework can strengthen the relationship. Almost all anorexic patients I have worked with operate at this level and pose formidable challenges to the alliance. For these and other patients operating at a borderline level of personality organisation (Fonagy et al., 2002), the neutrality of the therapist in humanistic or psychodynamic approaches may be experienced as disinterest or criticism. Severe levels of disturbance can challenge the couple’s capacity to reflect on these negative cognitions. As a first intervention, CBT can be quite helpful in preparing the ground for more challenging therapy later on. Homework can strengthen the sense of collaboration and bond between the two participants.

I want to offer an example of this principle by demonstrating how the therapist’s undertaking of homework himself was essential to prove that talk of collaboration and mutuality was more than just talk. Miss AB, initially very mistrustful of the clinic staff, challenged me to do some homework myself. I believe that she felt demeaned by the request of an assignment and also believed that she would be the only person that must do work. She humorously asked me to do something that she perceived as embarrassing to me. Once we discussed my homework assignment, she seemed to have some evidence that I was also doing difficult work and that embarrassment could be understood and worked through.

The setting of homework can also structure the time in the client’s life in between sessions. This can be directly observable in the treatment of depression which often involves clients that are paralysed by their negativity. Working towards “behavioural activation”, homework per se can be a highly therapeutic intervention as its completion increases the client’s physical and mental mobility (Beck & Alford, 2008). Another example can be seen in the structuring of the anxiety provoking openness of the client’s life. Miss A, a chronically unwell 30 year old, who I discovered may have had Asperger’s syndrome in addition to her anorexia. Her experience at that point was that she needed goals to work towards, moving from one to the next without interruptions. “Being” was at this stage intolerable and left her feeling depressed and desperate, a common experience for people with an autistic diagnosis. She found homework to be a scaffolding experience during her inpatient visit that protected her from feeling lost in the gaps between focused activities. She also gained a sense of agency by being able to use the tools that we had agreed on to arrive at discoveries that she proudly presented in our sessions.

As mentioned before, the rationale for homework in CBT is that the work of therapy is insufficient unless the client can be guided to apply his work outside sessions. It is likely that certain clients can be helped tremendously in building their sense of agency and confidence by succeeding in practical tasks during the week. This can lead to a positive feedback loop where success strengthens the bond of the couple and this improves the collaborative atmosphere that helps to appropriately set and execute homework. Miss B, like many clients with AN, felt very critical of herself and her capacity to be impactful. This has led to a routinised existence governed by profound rigidity. We started by setting very modest tasks, such as switching the order between brushing her teeth and having a shower in the morning. As we reflected on her growing capacity to succeed in assignments, Miss B felt grateful for being encouraged to try and was able to slowly move to attempting harder changes in her life on the ward. I think that for Miss B homework was a necessary intervention and indeed just talking at that point alone would not have sufficed.

The negative impact of homework

For the same reasons that homework can be helpful, it can also jeopardise the therapeutic alliance. Beck (Beck et al., 1990) stressed that cognitive therapy is a collaborative venture and that it is important to be mindful of how the collaboration can be destroyed by the activities of patients with personality disorders. Homework non-compliance can be used, I think, as a very direct way of sabotaging the work, in so far as homework signifies explicit collaboration. This can be due to the faulty cognitions some clients have about close relationships and their role in these. The setting of homework is a very concrete exchange and can have a powerful impact on the therapy.

To begin with, the setting of homework sets expectations. This is much more explicit than, say, silences in psychodynamic therapy that clients may find hard to tolerate because of an unconscious belief that the therapist expects them to give something by talking. Miss C did not complete the homework we agreed on and, after a few meetings, disengaged from therapy stating that it was useless and did nothing for her. Miss C reported having had a very controlling mother and she expressed rage at her mother’s suffocating behaviour. On reflection it is highly likely that she had experienced my perceived expectations as a form of control. This underscores the possibility that clients may be affected by a request for homework, but not allow opportunities to work through the problem because the act of bringing it up in therapy is felt to be weakness. Although, there may have been an agreement that the client undertake this task, it does not follow that the client is necessarily aware of the meaning of agreeing until it is experienced. The task of the therapist, her contribution to the working alliance, does not end with the assignment, but continues in the management of the aftermath.

Another issue, closely related to the first, is that in other forms of therapy the notion of the client doing work in between therapy (their contribution to the therapeutic alliance) differs. As clinicians we hope that clients will hold what happened in session in mind and that they will reflect on it during the week. We hope to hear references to the previous session, how it disturbed them, or helped them and so forth. It may be quite some time before clients can bring in the work they did in the week in relation to therapy, or indeed before they do any work outside therapy at all. It may be that working through being able to have a relationship with the therapist is a priority for them. However, setting homework does not allow the space and time for this to occur, as concrete work is demanded almost immediately. Miss D was a very young anorexic woman. She consistently told me she had forgotten to do her homework, because she had spent all of her time thinking about her weight. I attempted to understand what the problem was, but to no avail. Miss D was able to complete her therapy and she brought increasingly more of herself in sessions. She was a very quiet person and required great gentleness and care to connect with. She reported finding the therapy helpful in our last session. I was left wondering whether for this client setting homework was working against rather than for us, communicating to her that I could not respect the fact that she did not wish to allow the therapy to “continue” during the week. For her coming and being with me was what was helpful at the moment.

Finally I want to offer a more detailed instance where the setting of homework too closely resembled a historical experience in the client’s life. Miss E and I had agreed that at the end of every session she would go away with a number of tasks or experiments that might allow us to get insight into her thinking. Miss E only partially completed her homework on every occasion. Miss E had been excluded at school due to her violent and disruptive behaviour. Additionally, she had had severe academic difficulties. It was not surprising to hear her tell me that what I was asking her to do was, well, like homework. In this case I validated her view by acknowledging that I could see how she had drawn this parallel. I thought that it might make things between us difficult and we needed to tease out the differences between homework then and now. In line with socialising her to the broad CBT model in use, I explained the rationale and clarified that she was not being assessed and that my offer of help was not related to her performance. Miss E explained that she could not understand some parts of the homework. Again the risk that this situation would prove to her that she was inadequate and “always got things wrong” was present and the threat to the therapeutic relationship imminent. After exploration did not seem be possible I invited her to try together in session. Interestingly, on most occasions Miss E would express her confusion about something in the homework and then quickly work through this on her own while I acknowledged that we could be confused together about it. I think she was testing us with incomplete homework, trying to tease out what sort of person I would be; with a lot of hard work I believe it became possible to experience me as collaborator rather than a teacher or parent.


In this essay I offered some instances where the setting of homework offered vital aid to difficult clients, as well as instances where it was less rather than more helpful. Although, homework is fundamental in CBT it is not synonymous with it. It may be the case that assignment of homework as a particular task may be premature or need to be renegotiated or withheld for some time until the client is ready for it. I have given examples from very disturbed clients to illustrate this thesis. It appears to me that the more disturbed clients are the more of a high risk-high gain approach homework is. However, it seems that there must be a sufficient minimum of a therapeutic alliance between therapist and client before homework is used. As CBT has been applied to more complex difficulties such as personality disorders, there has been a growing appreciation by prominent clinicians and writers that extensive problem solving and alliance strengthening might be needed before homework can be successfully set (J. Beck, 2005). I believe then that homework is a very powerful intervention when it comes to difficult clients and it requires significant care and supervision before and throughout its use.


Beck, A. T. (1976). Cognitive Therapy and the Emotional Disorders. London: Penguin.

Beck, A. T., & Alford, B. A. (2008). Depression: Causes and Treatment. Philadelphia, PA: University of Pennsylvania Press.

Beck, A. T., Freeman, A., & Davis, D. D. (1990). Cognitive Therapy of Personality Disorders. London: Guilford Press.

Beck, J. S. (2005). Cognitive Therapy for Challenging Problems: What to Do When the Basics Don’t Work. London: Guildford Press.

Bennett-Levy, J., Butler, G. Fennell, M., Hackmann, A., Mueller, M., & Westbrook, D. (2004). Oxford Guide to Behavioural Experiments in Cognitive Therapy. Oxford: Oxford University Press.

Fonagy, P., Gergely G, Jurist E, & Target M. (2002). Affect Regulation, Mentalization, and the Development of the Self. New York: Other Press.

Horvath, A. O. & Bedi, R. P. (2002). The Alliance. In J. Norcross (Ed.), Psychotherapy Relationships That Work: Therapist Contributions and Responsiveness to Patients (pp. 37-70). New York: Oxford University Press.

Rogers, C. R. (1957). The necessary and sufficient conditions of therapeutic personality change. Journal of Consulting Psychology, 21, pp. 95–103.

Cite this page

Choose cite format:
Online Chat Messenger Email
+44 800 520 0055