Copyright is owned by the Author of the thesis.  Permission is given for 
a copy to be downloaded by an individual for the purpose of research and 
private study only.  The thesis may not be reproduced elsewhere without 
the permission of the Author. 
 



Client Attitudes 

Towards Homework 

In Cognitive Therapy 

Yvonne H. James 

2001 



ACKNOWLEDGMENTS 

I would like to thank my supervisors, Paul Merrick and Nik Kazantzis for their guidance, 

support, and advise throughout all stages of this research. I would especially like to thank 

Nik for his generosity in sharing his resources on homework, and for his enthusiasm for the 

subject. 

I would like to also thank staff and clients at the Waitemata Health Cognitive Therapy 

Centre, without whose help this research could not have been done. 

Finally, I would like to thank my family for their patience and support. 



ABSTRACT 

This study examined client attitudes towards homework in cognitive therapy in relation to 

client diagnosis and symptom severity, and in relation to homework compliance. 

Participants attending a cognitive therapy outpatient facility completed two self-report 

attitude measures at each therapy session over a three-month period. The results showed a 

relationship between attitudes and symptom severity. Participants with depression exhibited 

more avoidance and less mastery and pleasure associated with homework completion, 

while participants with anxiety exhibited increased mastery and pleasure and decreased 

avoidance. Negative attitudes (avoidance, difficulties, and obstacles) were associated with 

non-compliance. Pleasure was significantly associated with quality of homework 

completion. In summary, there was a positive relationship between attitudes towards 

homework, compliance, and levels of symptom severity. 

11 



ACKNOWLEDGMENTS 

ABSTRACT 

TABLE OF CONTENTS 

LIST OFT ABLES 

LIST OF APPENDICES 

TABLE OF CONTENTS 

Chapter 1: Cognitive Therapy: The Context For Homework 

Introduction 

Principles of Cognitive Therapy 

Role of Cognition 

Summary 

Chapter 2: Cognitive Therapy: Research and Specific Uses 

Introduction 

Empirical support for cognitive therapy 

Mood Disorders 

Anxiety Disorders 

Other Disorders 

Summary 

Chapter 3: Homework and Cognitive Therapy 

Introduction 

Definition 

Homework Terminology 

Practical Support for Homework 

ii 

iii 

vi 

vii 

1 

1 

2 

4 

7 

8 

8 

8 

9 

11 

13 

14 

15 

15 

15 

16 

17 

111 



Theoretical Support for Homework 

Empirical Support for Homework 

Administration of Homework 

Summary 

Chapter 4: Research Aims 

Introduction 

Hypotheses 

Summary 

Chapter 5: Measuring Attitudes Towards Homework 

Introduction 

Literature on attitudes towards homework 

Present Measures 

Practical Considerations 

Measure 1: Attitudes Towards Homework Scale 

Response Scale and Subscales 

Measure 2: Homework Performance Measure 

Summary 

Chapter 6: Method 

Introduction 

Ethical Approval 

Sample 

Intake assessment 

Procedure 

Analysis 

18 

19 

25 

31 

32 

32 

32 

34 

35 

35 

35 

37 

38 

39 

40 

43 

45 

46 

46 

46 

46 

51 

63 

64 

IV 



Data Analysis procedures 

Summary 

Chapter 7: Results 

Introduction 

Subscales in the Attitudes Towards Homework Measure 

Disorders and Attitudes Towards Homework 

Chapter 8: Discussion 

Introduction 

Present Study 

Limitations 

Contribution to research 

Conclusion 

References 

Appendices 

65 

70 

71 

71 

71 

73 

78 

78 

78 

83 

84 

85 

87 

107 

v 



LIST OF TABLES 

Table 1 - Demographic Variables 

Table 2 - Participant Disorders (DSM-IV) 

Table 3 - Pre-Treatment Symptom Severity Scores 

Table 4 - Effect of BDI Score Data Transformation 

Table 5 - Attitude Subscale Scores 

Table 6 - BDI-11 Scores and Normative Data 

Table 7 - Mean STAI-Y-1 Scores and Normative Data 

Table 8 - Relationships with Mastery, Pleasure, Difficulty, and Attitude Subscale 

Scores 

48 

50 

52 

54 

67 

69 

70 

72 

Table 9 - Correlation of Attitude Subscales and Pre-Treatment Symptom Severity 74 

Table 10 - Attitude Differences with Mood and Anxiety Disorders 75 

Table 11 - Correlation of Homework Performance and Attitude Subscales 77 

VI 



LIST OF APPENDICES 

Appendix A - Beginning of Session Form 

Appendix B -Attitudes Towards Homework Scale 

Appendix C - Abbreviated Homework Performance Measure 

107 

108 

109 

Vll 



Chapter 1: Cognitive Therapy: The Context For Homework 

Introduction 

Cognitive therapy began in the early 1960s with A T. Beck's research on depression 

(Dattilio & Freeman 1992; A. T. Beck, 1964). He found that depressed clients consistently 

exhibited a negative bias in their cognitive processing. From this he developed a cognitive 

model of depression, which was later extended to other emotional disorders, such as 

anxiety disorders (A. T. Beck, 1963; 1964; 1967). 

The development of cognitive therapy was also influenced by a number of other early 

theories. These include Kelly's (1955) model of personal constructs and beliefs in 

behavioral change, Arnold' s (1960) cognitive theories of emotion, and Lazarus ' s (1966) 

work that linked the role of cognition with emotional and behavioral change. The theory of 

rational-emotive therapy (Ellis, 1962) was consistent with cognitive therapy theory in that 

individuals adopt patterns of cognition based on underlying assumptions. The manner with 

which this is dealt with therapeutically differs, however, with rational-emotive therapists 

persuading the client that dysfunctional thoughts are irrational, while cognitive therapists 

seek to test the validity of such cognitions using a collaborative approach. Other influences 

on cognitive therapy include Bandura' s (1977) social learning theory and cognitive 

modeling. This also coincided with a shift in behavior therapy towards a more cognitive 

approach (Dattilio & Freeman, 1992). Mahoney' s cognitive control of behavior (1974) and 

Meichenbaum's cognitive behavior modification (1977) were also important influences on 

cognitive therapy. 

1 



Cognitive therapy is based on the theory that a person's feelings and behaviors are largely 

determined by how the person structures experiences (A T. Beck & Weishaar, 1986). From 

this perspective a person's psychopathology is an exaggeration of a normal adaptive 

response (Dattilio & Freeman, 1992). For example, anxiety may be a useful warning 

response, which encourages a person to check out details of a situation, whereas excessive 

anxiety can cause a paralyzing inability to respond to a harmful situation. Cognitive 

therapy proposes that distortions in thought often maintain dysfunctional mood states. 

Changes in thoughts may bring about change in affect and in behavior. In this way thoughts 

are viewed as a key point of intervention in cognitive therapy (Dattilio & Padesky, 1990). 

Principles of Cognitive Therapy 

Cognitive therapy examines client problems in cognitive terms. The theory of cognitive 

therapy proposes that patterns of cognition help maintain emotions and behaviors. Clients 

are frequently screened at the beginning of therapy to obtain measures of functioning, 

including baseline measures of symptom severity. For example, a common baseline 

measure of depressive symptoms is the Beck Depression Inventory (BDI; A T. Beck, 

Ward, Mendelson, Mock, & Erbaugh, 1961; A T. Beck, Steer, & Garbin, 1988). 

Sometimes clients are assessed for how they might function in therapy. For example, the 

Suitability For Short-Term Cognitive Therapy Rating Scale (SSCT; Safran, Segal, Vallis, 

& Shaw 1990a) assesses among other things, accessibility of automatic thoughts, awareness 

and differentiation of emotions, compatibility with cognitive material, acceptance of 

personal responsibility, and alliance potential. 

2 



Therapy starts by addressing current issues in order to bring some immediate relief through 

reduction in symptom severity. While therapy is present-focused, it may also involve the 

examination of childhood experiences that have helped shape beliefs and thoughts, and how 

these may have made the client vulnerable to certain problems by linking into current 

precipitating factors (A T. Beck & Emery, 1986). By identifying precipitating factors and 

client responses to them, clients are able to challenge and potentially change dysfunctional 

thought, emotional or behavioral patterns in their lives (J. S. Beck, 1995). 

Cognitive therapy takes an educative approach (A T. Beck & Emery, 1986), and teaches 

clients skills to enable them to resolve present and future problems. Clients learn to become 

more aware of negative thoughts and beliefs, to identify, evaluate, and respond to them. 

Therapy is focused and goal-oriented. A variety of techniques are used to change thinking, 

motivation and behavior, which are linked to the specific problem area and are tailored to 

clients ' individual needs . 

Cognitive therapy is based on a sound therapeutic alliance. This requires good therapist 

skills, and a collaborative approach between client and therapist involving active client 

participation in therapy. Part of the collaborative process in therapy includes setting the 

agenda for the therapy session, reviewing and setting homework. An agenda is set at the 

beginning of each session, which provides the structure for that session. Homework is 

central to therapy. It is a means of ensuring that the work of therapy is applied throughout 

the time between therapy sessions. It also gives clients a chance to apply what has been 

learned to everyday situations, as well as to the problem situation. The review of 

homework assigned from the previous session generally comes early in the session, and the 

3 



assignment of new homework comes towards the end of the session (J. S. Beck, 1995). 

Data, for example daily activity schedules, or other material the client has collected as part 

of homework, often provides the material to be dealt with in the main portion of the 

session. 

Cognitive therapy is brief and time-limited, with generally fewer than 25 sessions. Often 

initial sessions are closer together with later sessions more spaced as the level of clients' 

adaptive skills increases. Periodic booster sessions are often used at the end of therapy to 

maintain gains and to prevent relapse of the presenting problem. (A T. Beck & Emery, 

1986; J. S. Beck, 1995). 

Role of Cognition 

Cognitive therapy proposes that an individual's previous expenences influence the 

formation of attitudes, rules and assumptions. These attitudes, rules and assumptions 

influence the thoughts that automatically arise in response to particular situations. 

Cognition may be understood as a three-level structure comprising automatic thoughts, 

intermediate beliefs, and core beliefs (Padesky, & Mooney, 1990). These can be thought of 

as three concentric circles with the automatic thoughts being the outer, most visible and 

accessible level, followed by intermediate beliefs, with core beliefs being at the deepest 

level and not generally being part of an individual's conscious awareness. Automatic 

thoughts are unplanned and situation specific. They arise spontaneously in response to a 

situation, and may involve verbal or visual images or both. 

4 



Automatic thoughts are the most accessible to the client. Bringing automatic thoughts to 

immediate awareness is one of the first tasks of a client in cognitive therapy. Intermediate 

beliefs are rules, attitudes and assumptions, which guide a person 's life and organize how 

situations are perceived. Evidence about these intermediate beliefs can be gathered and 

examined to decide the extent to which they are true. This can be done in the therapy 

session, or assigned as homework. Core beliefs or schemas comprise the deepest level of 

cognition. Core beliefs are unconditional, inflexible and absolute beliefs of the self: the 

world, and of others (Padesky, & Mooney, 1990). They are usually formed in childhood 

and are not usually part of an individual's awareness. Core beliefs may be assessed by 

noting an individual's typical reactions to events or circumstances over the course of 

therapy, and can sometimes be recorded by the client as automatic thoughts on thought 

record sheets (J. S. Beck, 1995). A critical incident or activating event may trigger a person 

into activating past maladaptive beliefs, generating current maladaptive automatic thoughts, 

thereby determining the way an individual thinks, feels, and reacts to a current situation. 

While thoughts are regarded as central to the way an individual's world is structured, the 

cognitive therapy model can be conceptualized as an inter-related structure of environment, 

cognition, affect, physical response and behavior (Greenberger & Padesky, 1995). The 

environment includes the individual 's living situation, history and culture. Cognition 

includes thoughts, images, and memories . Emotions include moods and feelings. Behavior 

includes actions and what a person actually does in response to the other components of the 

model. Physical response includes any biological or physical sensation a person may 

experience. Change at any point of the model will impact on the entire situation 

5 



(Greenberger & Padesky, 1995). Specific psychological problems are related to specific 

cognitive content. For example those with depressive disorders display a negative view of 

the self, the future and experience. Those with panic disorder show a catastrophic 

misinterpretation of bodily sensations and mental experiences. Those with obsessive­

compulsive disorders sense a repeated warning or doubting about safety and use repetitive 

acts to ward off such a threat (A. T. Beck & Rush, as cited in Kaplan & Saddock, 1998). 

Errors of cognition, also known as infonnation processing errors, are common to most 

people to varying degrees, and are theorized to maintain dysfunctional states. One of the 

tasks of therapy is to identify and modify such errors of cognition. The following examples 

(A.T. Beck, Rush, Shaw, & Emery, 1979) are among the most common cognitive errors. 

Over-generalizing occurs when what is true in one case is attributed to other similar cases. 

The therapeutic intervention is to oppose the faulty logic of this way of thinking and 

establish greater discrimination of the similarities and differences of various situations. 

Selective abstraction occurs where an individual focuses only on negative life events. This 

may be examined by keeping a journal, which includes the good things and successes the 

client tended to forget. Catastrophizing occurs when an individual always thinks that the 

worst eventuality will happen and is challenged by calculating the probabilities of what 

might happen, and recording evidence of what did, and did not happen. Dichotomous 

thinking occurs when everything is seen as either one extreme or the other. This may be re­

evaluated by using a continuum to chart the true impact of events. 

Other cognitive errors include: magnifying or minimizing, or over-valuing the importance 

of a situation or certain information; emotional reasoning, or thinking something must be 

6 



true because it "feels" true; using shoulds and musts (imperatives) involving self-messages 

that things "should" or "should not" be a certain way; personalizing (blaming) or holding 

oneself responsible for something not in one's personal control; tunnel vision or seeing only 

the negative aspects of a situation; and lastly disqualifj;ing or discounting the positive by 

self-messages that positive experiences, deeds or qualities do not count. 

(J. S. Beck, 1995). 

Summary 

Clients in cognitive therapy are urged to examine cognitions related to their problem. They 

learn to develop alternative and more helpful cognitions and behavioral responses. Much of 

the therapeutic work is done by checking hypotheses generated from the problem area, 

often as homework assignments . 

7 



Chapter 2: Cognitive Therapy: Research and Specific Uses 

Introduction 

There is a growing demand for therapies, which are brief, effective and therefore cost­

efficient. For example, in New Zealand, university training of psychologists follows a 

predominantly scientist-practitioner model (Kaz.antzis & Deane, 1999), and cognitive 

behavioral approaches are widely used as a training model. Homework is an essential part 

of cognitive therapy, providing activities that support the therapy session and providing 

therapeutic practice for the time between sessions. It is also an integral part of the 

collaborative process between client and therapist. Homework is widely held responsible 

for the long-term maintenance of therapeutic gains (e.g., Kovacs, Rush, Beck, & Hollon, 

1981 ). 

Empirical support for cognitive therapy 

Cognitive therapy has been found to be effective m outcome studies for unipolar 

depression, generalised anxiety disorder, panic disorder, bulimia, social phobia, and 

obsessive-compulsive disorder, with controlled studies now also examining cognitive 

therapy for personality disorders, psychosis, and post-traumatic stress disorder, among 

other disorders (Blackbum & Twaddle, 1996). An article by DeRubeis and Crits-Cristoph 

(1998) examining the research on psychological treatments for adult disorders, found that 

cognitive and behavioral approaches dominated the list of efficacious treatments, especially 

for anxiety disorders. To be listed as efficacious, treatments had to lead to a reduction or 

remission of the disorder at a higher rate than that expected over time with no treatment, or 

had to outperform an alternative active treatment. 

8 



Mood Disorders 

Cognitive therapy was determined an efficacious treatment for mood disorders, with some 

studies showing cognitive therapy to have a superior outcome, some showing cognitive 

therapy equal in effectiveness to pharmacotherapy, and some studies showing a 

combination of cognitive therapy and medication as the most useful. One of the earliest 

controlled outcome studies was that of Rush, A. T. Beck, Kovacs, and Hollon (1977) who 

examined the comparative efficacy of cognitive therapy and imipramine. Since then, some 

studies on depression have found that cognitive therapy is at least as effective as 

antidepressant medication in outpatient treatment of depression, and m some cases, 

superior. For example, Dobson (1989) in a meta-analysis of28 studies found that cognitive 

therapy clients on average had a better outcome than 70% of pharmacotherapy clients. 

Other studies for Major Depressive Disorder have found that cognitive therapy showed 

effects equal to those of pharmacotherapy (Hollon et al. , 1992; Murphy, Simons, Wetzel & 

Lustman, 1984; Rush et al. , 1977). Some studies have found a combination of cognitive 

therapy and pharmacotherapy more effective than either treatment alone for depressed 

outpatients, for example, Blackbum, Bishop, Glen, Whalley, and Christie, 1981 ; Hollon et 

al. , 1992; Murphy et al. , 1984). There is also evidence of superior post-treatment gains 

where cognitive therapy is available for depressed in-patients along with medication 

(Miller, Norman, & Keitner, 1989). 

The distinction becomes more pronounced in favor of cognitive therapy when the 

percentage of recurrence of depression in comparison to the use of medication alone is 

examined. For example, Evans et al., (1992) found at 24 months post-treatment, clients on 

9 



tricyclic antidepressants who did not continue medication had a 50% recurrence of 

depression, while those who continued medication had a 32% recurrence rate. Those who 

had cognitive therapy had a 21 % recurrence, with those having cognitive therapy and 

tricyclic antidepressants having the lowest recurrence rate at 15%. Blackburn, Eunson, and 

Bishop, (1986), in a 24-month post-treatment study, found a similar trend of results, with 

cognitive therapy showing even more favourably with a 78% recurrence of depression in 

the medication only group, with a drop to 23% recurrence in the cognitive therapy 

condition, and 21 % in the cognitive therapy with medication group. Other studies which 

have also shown cognitive therapy to have a symptom relapse prevention effect include 

~~~iffEvans et al., (1992); Kovacs et al., (1981); Simons, Murphy, Levine & Wetzel, 

There continues to be some disagreement over the use of cognitive therapy with severely 

depressed clients. In a National Institute of Mental Health study, Elkin et al (1989), found 

that medication was significantly superior to placebo or clinical management, but that 

cognitive therapy was not significantly superior to the placebo and clinical management 

conditions. From this study it was suggested that medication should be used for severe 

depression, which in the Elkin et al., (1989) study was defined as 20 or more on the 17-item 

Hamilton Rating Scale for Depression (HRSD; Hamilton 1967 ). Other studies, however, 

have found no negative relationship between severity of depression and cognitive therapy, 

including McLean and Taylor (1992); Hollon et al, (1992); and Thase , Simons, Cahalance, 

McGreary, & Harden., (1991 ). A study by Persons, Bums & Perloff (1988) showed a much 

larger effect of homework compliance on outcome for clients with high initial BDI scores. 

Those with scores over 20 who were compliant gained on average a 77% reduction in BDI 

10 



scores, while those who did not do homework showed little improvement. A recent 

correlational study by Burns and Spangler (2000) found similar results. Bums and Spangler 

suggest that the effect of homework appears to be sufficiently great as to lead to nearly 

complete elimination of symptoms in clients with mild to moderate depression. 

Anxiety Disorders 

Cognitive therapy is an efficacious treatment for a range of anxiety disorders. Cognitive 

and behavioral treatments have been used for clients with Generalized Anxiety Disorder 

with a success rate of 59% at follow-up three to six months later (Butler, Cullington, 

Hibbert, Klimes, & Gelder, 1987; Butler, Fennel, Robson & Gelder, 1991 ; Power et al., 

1990), making cognitive behavioral treatment with applied relaxation a specific treatment 

for this disorder (DeRubeis & Crits-Cristoph, 1998). A recent study by Durham et al. , 

(1999) found cognitive therapy to be clearly superior to analytic psychotherapy for 

Generalized Anxiety Disorder, as well as cognitive therapy being associated with a 

significant reduction in medication use. Clients in the cognitive therapy condition were also 

more positive about their therapy with approximately two thirds in the intensive cognitive 

therapy condition achieving clinically significant improvements over a six-month period. 

With Social Phobia, exposure therapy was an effective treatment with moderate success 

maintained at follow-up (Hope, Heimberg, & Bruch, 1995; Mattick, Peters, & Clarke, 

1989; Turner, Beidel, & Jacob, 1994 ). There were mixed results with a meta-analysis by 

Feske and Chambless (1995) reporting that a combined approach of exposure with 

cognitive therapy was not superior to exposure, in contrast to Butler, Cullington, Munby, 

11 



Amies, & Gelder, (1984), which showed that exposure with cognitive therapy 

outperformed exposure alone. 

Cognitive therapy has been found efficacious for Panic Disorder (A T. Beck, 1988; Clark 

et al, 1994), with cognitive therapy superior to other treatments at post-treatment (J. G. 

Beck, Stanley, Baldwin, Deagle, & Averill, 1994; Klosko, Barlow, Tassinari, & Cerny, 

1990; Williams & Falbo, 1996). At follow-up, Clark et al ., (1994) reported 80% of clients 

were panic-free post-treatment with gains maintained at 15-month follow-up, with Ost and 

Westling (1995) reporting panic-free rates of 65% for applied relaxation, and 74% for 

cognitive therapy at post-treatment. Effects continued increasing, so that at one-year 

follow-up the result was 82% for applied relaxation and 89% for cognitive therapy. Gould, 

Otto, and Pollack (1995), in their meta-analysis of treatment outcome for Panic Disorder 

found that cognitive behavioral treatment was at least as effective as pharmacological 

treatments, with the advantages of lower rates of attrition than pharmacological treatment. 

For Obsessive-Compulsive Disorder, exposure with response prevention therapy is 

effective (Foa, Steketee, Grayson, Turner & Latimer, 1986), with cognitive therapy 

showing even more effective results (A T. Beck & Emery, 1986; Salkovskis, 1985). For 

Agoraphobia, exposure and systematic desensitization are effective treatments (Gelder et 

al., 1973; Chambless, Foa, Groves, & Goldstein, 1979; McDonald et al., 1979). For Post­

Traumatic Stress Disorder systematic exposure to traumatic stimuli is effective (Cooper & 

Clum, 1989). With rape victims exposure (at 56%) was superior to supportive counseling 

(at 33%) (Foa, Rothbaum, Riggs, & Murdock, 1991). 

12 



Other Disorders 

Cognitive therapy is widely used in the treatment of eating disorders including Bulimia 

Nervosa, Anorexia Nervosa, and Obesity. Most empirical support for cognitive therapy for 

eating disorders has been established for bulimia. For example, it was found that the use of 

cognitive therapy in the treatment of bulimia resulted in a substantial reduction in bulimic 

behaviors as well as a change in the cognitive distortions associated with bulimia 

(Lewandowski, Gebing, Antony, & O'Brien, 1997). Cognitive factors have been considered 

to maintain the condition (Fairburn, 1985; Fairburn & Cooper, 1989). Cognitive behavioral 

therapy is at least as effective as other therapies, with Craighead and Agras (1991) in an 

analysis of 10 studies of bulimia, finding a mean reduction in bulimic behavior of79%. 

Cognitive therapy is also widely used in the treatment of pain. Morley, Eccleston and 

Williams in examining pain excluding headache ( 1999) found that cognitive behavioral 

treatments resulted in clients experiencing less pain, showing less behavioral expression of 

pain, and showing increased levels of cognitive coping with their pain. In a study of 

headache clients, Bogaards and ter Kuile (1 994) found that cognitive therapy, relaxation, or 

biofeedback alone or in combination with relaxation, were superior to placebo therapy or a 

no treatment condition. Cognitive behavioral therapy is an effective, time-limited treatment 

for anger problems (R. Beck & Fernandez, 1998). 

Cognitive therapy is now being used in more ' difficult' areas of therapy. The area of 

personality disorders has been the recent focus of several studies with Vallis, Howes, and 

Standage (2000) suggesting that while there is evidence to suggest clients with personality 

13 



disorders do not benefit as much from cognitive therapy as those without personality 

disorders, a modified form of cognitive therapy may be useful for these clients. One study 

by Persons, Bums and Perloff (1988) found that clients with personality disorders were 

more likely than others to drop out of therapy early, possibly because of the interpersonal 

difficulties, which are central to personality disorders . Other studies have also found that 

cognitive therapy is less effective for clients who have difficulty accepting the rationale for 

cognitive therapy, have difficulty with homework exercises, and have difficulties with 

alliance (e.g. Fennell & Teasdale, 1987; Persons et al, 1988; Edelman, 1995). Vallis et al., 

(2000) also suggest that any clients with alliance difficulties would find any short-term 

therapy difficult. For clients with personality disorders there are conceptual models 

available for modifies treatment, including ways of dealing with interpersonal issues m 

therapy (e.g. Young, 1990; Linehan, 1993). 

Summary 

Cognitive therapy is an effective treatment for many disorders with an established record 

for the effective treatment of mood and anxiety disorders. These disorders dominate client 

diagnoses at the Waitemata Health Cognitive Therapy Centre, from which the sample was 

drawn for the present study. The participants in the present study had among them the 

following disorders (primary and secondary diagnoses) : Major Depressive Disorder, both 

single episode and recurrent, with post-natal onset; Dysthymia; Bipolar Disorder 

(unspecified); Panic Disorder; Panic Disorder with Agoraphobia; Anxiety Disorder 

(unspecified); Generalized Anxiety Disorder; and Social Phobia. 

14 



Chapter 3: Homework and Cognitive Therapy 

Introduction 

The use of homework to disconfirm patterns of dysfunctional thoughts and behaviors and to 

develop more functional responses is central to cognitive therapy (Neimeyer & Feixas, 

1990). While the ideas of A T. Beck et al., (1979) on the benefits, uses and administration 

of homework are still valid and relevant, there is now an increasing amount of research and 

refinement of these concepts to support ideas previously based on clinical observation and 

behavioral theory. 

Definition 

Homework constitutes activities that contribute towards therapy, which are completed by 

clients outside the therapy session (Kazantzis, 2000a). These activities support the therapy 

session and provide therapeutic practice for the time between sessions. Homework is an 

integral part of the collaborative process between client and therapist, adding value to the 

therapeutic relationship through the discussion involved in its assignment and review. 

Homework is a means of disconfirming negative thoughts and beliefs and shifting the focus 

of therapy from subjective, abstract ideas, to something real, substantial and detailed (A T. 

Beck et al., 1979). It is also widely held to be responsible for the long-term maintenance of 

therapeutic gains (e.g. Kovacs et al., 1981). Homework may cover a range of activities 

from observation, reading material related to therapy, monitoring thoughts and activities 

and recording them, to physically doing things or interacting with others. 

15 



Homework Terminology 

Doing homework has the effect of extending therapy into daily practice between sessions, 

hence the commonly used alternative wording between session tasks (e.g., Neimeyer & 

Feixas, 1990), or between session practice (e.g., Barlow, O'Brien , & Last 1984), for 

homework. Other less frequently used names for homework include extratherapy 

assignments (Kornblith, Rehm, O'Hara, & Lamparski, 1983), show that I can tasks (STIC: 

Ronan & Deane, 1998), in vivo behavioral practice assignments (Ingram & Salzberg, 

1990), self-help assignments (Bums, 1989), and home-based practice (Schmidt & 

Woolaway-Bickel, 2000). Whisman (1999) suggests that the choice of word used may 

influence the client's perception of homework, and for some with unpleasant past academic 

experiences the word homework may be aversive, while for others, the word homework 

may be clear and the work easily implemented. 

An increasing emphasis is now placed on the systematic administration of homework. This 

involves specific details of the homework task including a written 'prescription' which the 

client takes away, and includes the therapist keeping regular checks and noting levels of 

homework compliance (Levy & Shelton, 1990; Shelton & Levy, 1981 ; Kazantzis & Deane, 

1999). Compliance, or performance refers to the degree of homework completion, and 

comprises quality of homework completion and quantity of homework completion. The 

effect of homework depends on the degree of homework compliance attained by the client. 

16 



Practical Support for Homework 

Client Factors 

Homework allows the client to put skills learned in session into daily practice and enables 

the client to work towards stated goals without the therapist being present. Homework 

provides the opportunity for the client to monitor thoughts, feelings, and behaviors and 

experiment with these. This may involve testing thoughts and beliefs or modifying ways of 

thinking and behaving. It also provides the opportunity to practice new cognitive and 

behavioral skills learned in session (J. S. Beck, 1995). 

Successful homework completion is associated with feelings of increased self-efficacy as 

the client is able to take part in activities that are likely to enhance successful treatment 

outcome. Increasing self-efficacy is matched by a decreased sense of helplessness and 

increased motivation (Bandura, 1977). By increasing clients' activity levels through 

homework, a more positive affect is developed, especially as clients develop skills that can 

be generalized to other areas of their lives. Clients also gain skills for modifying any future 

dysfunctional cognitions (Primakoff, Epstein, & Covi, 1986). It is thought that regular and 

consistent completion of homework throughout therapy helps the client maintain gains 

made in therapy after the end of therapy, as clients have had practice in self-monitoring 

new skills and behaviors (A T.Beck et al., 1979). The underlying concept of homework 

assignments is that cognitive and behavioural changes in everyday life need to occur before 

the underlying belief structures of the client can be altered (Persons, 1989). There is also 

the assumption that homework compliance leads to a better outcome for the client 

(Detweiler & Whisman, 1999). 

17 



Therapist Factors 

Homework can also be used to gather information about the client, for example, by using 

activity schedules. The information from the activity schedule serves to review the client's 

week, to provide information for the current therapy session, and to provide a focus for the 

session, thereby avoiding any side issues (A T. Beck et al ., 1979). Homework is also useful 

for problem areas that cannot readily be dealt with in session, such as sexual dysfunction 

(Holzworth-Munroe, Jacobson, DeKlyen, & Whisman, 1989). Homework provides 

effective and efficient use of therapy, as clients are themselves responsible for a lot of the 

work of therapy, much of which can be done as homework, making more efficient use of 

therapist time. For example, clients undergoing relaxation and exposure techniques, once 

they are sure of the task, are able to do their own relaxation and graded exposure under the 

guidance of the therapist, but not in the therapist' s time. 

Theoretical Support for Homework 

Generality 

Much of the theoretical support for homework in therapy comes from the behavioral 

principle of generality. According to this principle, a trained behavior transfers from the 

training situation to the natural environment. Generality applies when this training then 

leads to a new behavior, which has not been specifically trained for, or when the trained 

behavior is maintained over time in the natural environment (Martin & Pear, 1999). Three 

important principles are needed for the therapeutic change to be effective. The change in 

behavior must show stimulus generalization from the training environment to the natural 

environment. The change in behavior should also be able to show, at least some of the time, 

18 



response generalization to new behaviors. Lastly, therapeutic change must also show 

maintenance over time in the natural environment (Martin & Pear, 1999). 

In cognitive therapy, homework is the vehicle for generalization. Skills learned in therapy 

sessions can be generalized to specific goal areas of a client's daily living, and then can 

generalize further to new areas, and the skills are then also available for use in future 

situations. Having available skills learned in therapy, allows the client to respond to a 

variety of situations with a range of more adaptive behaviors. 

Compensatory skills model of change 

The compensatory skills model of change in cognitive therapy extends this concept with the 

change from generalization being matched by a change in positive affect and cognitions ( 

Persons, 1993 ). A client in therapy learns skills to reduce negative symptoms by realigning 

negative automatic thoughts. Through homework practice these negative thoughts are 

replaced by more functional automatic thoughts and consequent behaviors. For example, a 

person who previously felt uncomfortable around 'authority figures' may, through 

therapeutic change, feel more confident and less anxious around such people. Consequently 

the person may feel and act more positively and in interactions with people deemed to be 

'in authority', achieve successful outcomes. 

Empirical Support for Homework 

What disorders is homework used for? 

In a survey of 500 studies involving homework from 1973 to 1980, homework was used, 

for example, for anxiety, non-assertiveness, depression, insomnia, obsessions and 

19 



compulsions, obesity, physical illness and rehabilitation, sexual dysfunction, stuttering, 

social skills, and smoking (Shelton & Levy 1981 ). This I ist has grown to encompass a 

diverse range of clinical conditions. Increasingly refined manuals are now being developed 

for specific conditions, including homework specific to that condition, for example, 

childhood problems (Ronan & Deane, 1998), Generalized Anxiety Disorder (Barlow, Esler 

& Vitali, 1998), Post-Traumatic Stress Disorder (Vaughan & Tarrier, 1992), and Social 

Phobia (Marks, 1995). There are now also practitioner guides available, offering the 

clinician a variety of ready-made assignments and advice for a range of presenting 

problems, for example Shelton and Levy (1981 ), and Schulthesis, 1998). 

Cognitive therapy is now used for an increasing number of disorders, including those not 

previously thought treatable with cognitive therapy, such as severe depress ion, bipolar 

disorder, and personality disorders. As treatment becomes available for disorders, so too, do 

forms of homework suited to the particular diagnosis. Modified forms of cognitive therapy 

are increasingly found of value to these clients, for example clients with delusions or 

hallucinations are noted to have difficulty with written tasks, but it has been suggested that 

those with less severe symptoms can carry out behavioral and practical tasks and report 

verbally on them (Nelson, 1997). While personality disorders are difficult to treat with any 

form of short-term psychotherapy because of alliance potential difficulties and difficulties 

with client security operations (Vallis et al., 2000), there are now available published 

conceptual manuals for treating personality disorders which address these issues (Linehan. 

1993; Rothstein & Vallis, 1991). 

20 



Increasingly cognitive therapy is being used for more severe cases of anxiety and 

depression. There have now been several studies suggesting that severely depressed clients 

are able to complete homework as well as those with mild symptoms, both in quality and 

quantity of completion (Edelman & Chambless, 1995; Startup & Edmonds, 1994). This is 

important when taken with the earlier study of Persons et al. (1988), which showed a mean 

reduction in BDI scores of 16.6 points for clients who regularly did homework, and only 

2.4 points for those who did not. Bums and Spangler (2000), also note homework 

compliance as being strongly correlated with reduction in depression scores. 

As homework is central to cognitive therapy, homework is increasingly used both in the 

general sense of daily records of thoughts and activities to more specific homework 

activities designed to meet the specific needs of specific individuals with specific 

conditions. For example Conoley, Padula, Payton, & Daniels (1994) in examining types of 

homework, found that to increase the effectiveness of homework, homework assignments 

should be matched to the presenting problem. The greater the match the greater the 

likelihood was that clients completed homework. Homework assignments may serve 

different needs for different client groups (Leucht & Tan, 1996). Based on the work of 

Fennel and Teasdale (1987) they give as an example that for clients with depressive 

symptoms, cognitive therapy is useful for some in helping clarify their situation and that 

this in itself may combat feelings of hopelessness, while for others, skills-building to deal 

with problems may be what is needed. 

21 



Who uses homework in therapy? 

There is increasing interest in the role of homework in cognitive therapy. According to one 

analysis of outcome studies, 68% of psychotherapists surveyed from 1973 to 1980 reported 

using homework (Shelton & Levy, 1981 ). A more recent study showed 98% of 221 New 

Zealand practicing psychologists surveyed, reporting using homework assignments as part 

of a treatment package (Kazantzis & Deane, 1999). 

Do clients do their homework? 

Clearly therapists assign homework, but for homework to be effective there has to be 

compliance. This is a growing area of research and involves issues such as systematic 

administration of homework as a means of enhancing compliance. A problem with 

compliance research is the degree of control extended over clients, that is, it is difficult to 

monitor what clients do all of the time. Startup and Edmonds (1994) made the observation 

of "dismantling" methodology (Kazdin, 1980), (where a control group is given a treatment 

package, and the treatment group is given the same package plus homework), that while it 

is possible to accurately control whether or not therapists give homework, it is not possible 

to control whether or not clients do it. Komblith, Rehm, O'Hara, and Lamparski (1983) 

experienced this difficulty. They designed a study in which homework setting was 

manipulated with the result that inclusion of homework in therapy had no greater effect 

than the treatment without the homework condition. However, it was later found that some 

clients in the control condition formulated and implemented their own homework, while the 

average rate of compliance in the homework condition was only 55%. 

22 



Compliance is also enhanced by matching difficulty levels to client abilities. Setting 

homework with a relatively low level of difficulty, while still encouraging clients to move 

beyond their current abilities is advocated (Leucht & Tan, 1996). The Conoley et al. , (1994) 

study found a negative correlation between homework compliance and the level of 

difficulty of homework. Utilizing clients ' strengths, for example fitting homework to suit 

the activities at which clients are competent, or using sensory modalities which fit the 

client, increased the likelihood of homework completion. 

Getting clients into the homework habit early in therapy is also important. There is some 

evidence to support the idea that success in homework assignments given early in therapy is 

positively related to therapy outcome (Startup & Edmonds, 1994; Fennell & Teasdale, 

1987). Early sessions were chosen in the Startup and Edmonds (1994) study, so as not to 

confound results with positive results from the benefits of therapy as therapy progressed. 

The results suggested that it may be therapeutically useful to get clients early into the habit 

of homework as an additional focus on the task at hand, and to enhance feelings of self­

efficacy. Motivation to do homework has been linked to enjoyment, in that if a person is 

intrinsically motivated to do an activity, the person does it for the reward or enjoyment it 

provides. Wankel (1993), in a study involving individuals in exercise routines found 

enjoyment can be enhanced when clients set specific goals, when they perceive themselves 

to have choices, control, and some flexibility over the goal, when they can involve family 

and friends in the activity and when that activity is both challenging and rewarding. 

23 



Is homework compliance associated with outcome? Does homework add to therapy? 

Two further aspects which follow on from compliance are whether homework compliance 

is associated with outcome, and whether homework adds anything to therapy (Kazantzis, 

2000b ). A number of researchers have found that clients who comply with homework 

assignments show enhanced progress in therapy (e.g. Neimeyer & Feixas, 1990). The 

Neimeyer and Feixas study of clients with depression, using a dismantling methodology, 

showed that clients in the homework condition showed a significant reduction in depressive 

symptoms over the course of therapy. However, at six-month follow-up there were no 

differences in both groups . The suggestion from Neimeyer and Feixas was that long-term 

gains require successful consolidation of skills learned in therapy. Persons et al. , (1988), in 

a study of depressed clients found that clients who did homework showed three times as 

much symptom reduction as those who did no homework. 

While there have been many studies involving homework with a recent search showing 71 9 

studies from 1980 to 1999 (Kazantzis, 2000), very few have involved quantitative studies 

and results have been very mixed. For example, some studies have shown that homework 

produces improvement in symptoms as opposed to homework without the treatment 

condition (Harmon, Nelson, & Haye, 1980; Kazdin & Mascitelli, 1982). Kazantzis (2000b) 

used a statistical aggregation of findings of quantitative studies on homework and found 

that homework assignments produce positive effects on therapy (r = 0.36) and that 

homework compliance is associated with therapy outcome (r = 0.22), showing that clients 

had a better chance of improving in therapy if they did complete their homework. A study 

by Bums and Nolen-Hoeksema (1991), found that homework compliance made an additive 

and independent contribution to the recovery rates of clients with mood disorders. 

24 



Kazantzis (2000b) suggests that significant effects have not been reliably obtained from 

some studies, possibly because of a lack of design sensitivity, which has reduced the 

probability of detecting an effect. 

Administration of Homework 

Matching homework to the problem 

Homework tasks are specific to the particular diagnosis and the individual's presenting 

problems. For example, it may be desirable for a severely depressed client to do some 

housework as part of a basic self-care package, while a client with obsessive-compulsive 

symptoms who spends much of the day cleaning and tidying, may be encouraged to do less 

housework. For behaviorally-based problems, Shelton and Levy (1981) list four broad 

types of intervention: giving the client new information, having the client alter his or her 

physical environment, changing the existing reinforcement, and having the client learn a 

new skill or extinguish an old one. Thus an obese woman may be given reading material 

about obesity, she may be told to eat only at the kitchen table and not in front of the 

television, she may put aside money saved from buying junk food towards new clothing or 

other desired non-food items, and may learn, for example if frustration was a cause of 

eating, to get out of the house and walk around the block or some other distracting behavior 

instead of eating, until more adaptive means of dealing with the frustration are learned. 

Stages of homework 

Homework may vary according to the stage of therapy. Early homework is likely to involve 

filling out pretreatment questionnaires. This is relatively non-anxiety-provoking for the 

client and can help the therapist gain essential background information on which to build 

25 



therapy. Defining therapy goals may also be an early homework task. Other early 

assignments may include reading educational material, viewing videos, or other similar 

material. This may have the effect of enhancing early compliance. For example by 

providing some of these activities pretreatment, the therapist may be viewed as caring and 

organized, the client may enter therapy with enhanced positive expectations of therapy, and 

the client may have some knowledge of cognitive therapy and what to expect. Such early 

assignments also provide training in doing homework, and provide the expectation that 

homework completion is part of therapy (Shelton & Levy, I 981 ). Early compliance with 

homework has been shown to be predictive of outcome in cognitive therapy for depression 

at termination (Startup & Edmonds, 1994). 

Monitoring thoughts and behaviors is a common form of homework assignment in early 

therapy, often gaining complexity as therapy progresses. For example a simple ABC sheet 

(antecedent, behavior, and consequences) helps a client monitor the reality of sequences of 

behaviors and also provides material for subsequent therapy sessions and future homework. 

Such records will often fonn part of a standardized approach to particular diagnoses. For 

example, a depressed person may record negative automatic thoughts, which at a later stage 

of therapy may include alternative thoughts and consequent mood changes. Activity 

schedules are also common for depressed persons for homework as a means of motivating 

them from inaction into action. A person with an eating disorder may record mood and 

thoughts before eating, what was actually eaten, and thoughts and mood after eating. 

Later stages of homework tend to include increasingly closer approximations of the desired 

goal. A spider phobic person may start with homework assignments based on relaxation 

26 



and gradual exposure, such as looking at books on spiders, to later visualizing spiders in a 

relaxation setting. Later the person may view real spiders and may actually handle them if 

this is a desired goal. Later stages of homework may focus on consolidating gains made in 

therapy, and at the end of therapy, may center on relapse prevention measures (J. S. Beck, 

1995). 

The therapeutic relationship 

One of the first essentials of homework administration is a strong therapeutic relationship. 

Goldstein (1975) described this as a process of developing liking, respect, and trust. Being 

perceived as someone likeable (warm and friendly) credible (an expert professional) and 

similar to the client, are also relationship-enhancing qualities (Shelton & Levy, 1981 ). A 

strong therapeutic relationship ensures the increased likelihood of relevant material being 

obtained from the client, and the client's willingness to help set and follow through with 

therapeutic goals. Ideally both therapist and client work together in a collaborative 

approach to formulate homework tasks. The client's willingness to participate in homework 

is essential, as homework assignments can be seen as steps towards, or approximations of 

the therapeutic goals, and thus a crucial part towards attaining those goals (Shelton & Levy, 

1981). 

Ways to enhance compliance 

Systematic administration of homework has been suggested as a way of increasing clients ' 

motivation to engage and comply with homework activities (Shelton & Levy, 1981 ). When 

clients do not know how to complete a homework task, or do not believe homework will 

help, or if there are factors in clients' lives, which make compliance difficult, or do not 

27 



support compliance, then clients are more likely to be non-compliant with homework tasks 

(Shelton & Levy, 1990). Shelton & Levy (1981) further suggest that non-compliance may 

occur from the therapist setting unrealistic goals, a mismatch between the difficulty of the 

assignment and the client's ability to do the homework, a lack of commitment from the 

client, a lack of positive reinforcement in the client's environment, and a lack of familiarity 

with the expectation of doing homework. That is, the client may not be aware of his or her 

therapeutic responsibility to be actively involved in therapy, and may take, instead, a 

passive role. These factors need to be clarified with the client to maximize compliance. 

Persons (1989), suggests a concern with perfectionism, which may make homework goals 

seem unattainable, a fear of failure, and the need to please others, may be obstacles to doing 

homework. Edelman and Chambless (1995) suggest comorbid personality disorders, initial 

symptom severity, and fears or uncertainty at the commencement of treatment may be 

reasons for not doing homework. 

Steps can be taken to enhance compliance in self-report homework. The client should be 

motivated to comply. One way of doing this is to frame the homework task as an 

experiment, or way of testing issues or finding out something new about the client. Results 

may then be used in therapy to modify the client's intermediate beliefs and underlying 

assumptions. Recording procedures should be easy and if necessary the client should be 

taught how to use these. The client should be able to discriminate the behavior to be 

recorded. Recording should be done as close as possible to the behavior and the amount of 

recording should start at a low level and increase as necessary. Activities monitored should 

be checked and feedback given to the client (Shelton & Levy, 1981 ). 

28 



For successful homework completion, it is suggested that homework assignments be tailor­

made for the individual and be within the individual's ability to complete (J. S. Beck, 

1995). Homework should be set collaboratively, with the client's input and agreement on 

the task. The therapist should provide a rationale for the homework, and where possible 

begin the homework in session, for example, with role-play or rehearsal. The therapist 

should help the client set up systems for remembering to do the homework and anticipate 

and deal with any potential problems that may arise in the execution of homework (e.g. A 

T. Beck et al., 1979; Bums et al., 1985; Kanfer, 1980; Levy & Carter, 1976). If there is a 

possibility of a negative outcome, this too can be anticipated and prepared for in advance 

(J. S. Beck, 1995). Ideally tasks are fully described, giving such details time, place, 

duration, and frequency, and both client and therapist have a written record of the task or 

tasks. 

A study by Conoley et al., (1994) found three factors which supported homework 

compliance. The task should use the client's strengths, match the client's presentation of 

the problem, and be at a low level of difficulty. Level of difficulty was based on the time 

the homework required, the amount of anxiety inherent in the task, and the resources 

available to the client to complete the homework. Matching the client's presentation of the 

problem and the task, involved active collaboration in identifying the problem, how well 

the matching of the task was explained to the client, and the clarity with which this was 

done. Matching homework to the client's strengths was based on whether the client had 

previously verbalized a history of success with a particular behavior, or a knowledge or 

interest in the behavior, or a belief that the client could be successful with the behavior. 

This study was based on earlier research by Conoley, Conoley, Ivey and Scheel (1991) who 

29 



posited the model that acceptability of therapy recommendations is based on the client's 

perception of the match between the problem and the recommendation, what the client 

believed about the recommendation, and the relationship between client and therapist. 

According to this model, compliance with the recommendation was affected by the 

acceptability of the recommendation and the client's ability to carry out the 

recommendation. Recommendation maintenance was achieved through carrying out the 

recommendation, with this having a tolerable level of disruption to the client, and with this 

creating a sufficient level of change for the client (Conoley et al., 1994). 

J. S. Beck (1995) has suggested the following client personality problems associated with 

homework avoidance: the avoidant client who avoids challenging homework and 

homework that would evoke dysphoria; the anxious and overwhelmed client, who may 

avoid homework if too many tasks are given; the dependent client, who does not do 

homework as it is therapeutically beneficial, and mental health would mean ending the 

relationship with the therapist; and the overly compliant client, who agrees to homework in 

session, but fails to complete it. The therapist may need to be mindful of these 

characteristics when assigning homework. 

Structure of the therapy session 

In the first part of the therapy session the therapist reviews homework from the previous 

session, reinforcing successes and discussing problems. In the second part of the therapy 

session, among the other agenda items, preparation for the next session's homework begins. 

The therapist engages in compliance enhancement strategies to increase the likelihood that 

these assignments will be completed. It is important to assure the client that he or she is 

30 



able to carry out the assignment. The therapist may need to teach the homework task with 

instructions, modeling, guided rehearsal, and corrective and reinforcing feedback. At the 

end of the session there is a review of the homework assignment for the next session. The 

assignment is handed out in written form and a commitment is gained from the client to 

carry out the assignment (Shelton & Levy, 1981 ). 

Summary 

Homework is usually set towards the end of each session. It is reviewed at the beginning of 

the next session. Reviewing homework reinforces gains made, or provides an opportunity 

to discover how the task did not go as planned. Homework provides material for further 

therapy and further homework tasks. The concept of systematic administration of 

homework is an area of increasing interest, as now that it is known that doing homework 

adds to therapy and doing homework is related to therapeutic outcome (Kazantzis, 2000b) 

there is increasing emphasis on the detail of how to get clients to do their homework, and 

on which aspects of homework are most beneficial. The present study, in examining client 

attitudes towards homework, fits into this broad area of discovering what is likely to 

enhance homework compliance. 

31 



Chapter 4: Research Aims 

Introduction 

The present study had two main aims. The first aim was to examine the extent to which 

client symptomatology affected clients ' attitudes towards homework. The second aim was 

to examine client attitudes towards homework in relation to homework assignment 

compliance (quality and quantity of completion). A self-report questionnaire, the Attitudes 

Towards Homework Scale, was designed to obtain some preliminary information on client 

attitudes towards homework assignments. A self-report measure used in previous 

homework research, the Homework Performance Measure (Kazantzis & Deane, 1998), was 

modified and used to measure homework compliance. 

Hypotheses 

For the purposes of this study attitudes were defined as ways of thinking or behaving 

towards someone or something (Cowie, 1989), in this case homework. Attitudes lie on a 

continuum from negative to positive. Both measures used in this study have a Likert-type 

response scale, which matches this continuum. Attitudes may be influenced by internal and 

external factors . These include the individual's environment, culture, history, and context 

of thinking and behaving. It also includes the individual's cognitions, emotions, behavior, 

and physical reactions (Padesky & Mooney, 1990). The individual's present situation is 

informed by the individual's history and environment. All of these factors impact on the 

individual's attitudes or responses to specific situations. Thus attitudes are central to an 

individual's response system. 

32 



Other factors were also expected to affect attitudes and responses in a clinical population as 

a function of the symptoms of the disorder, and the severity of those symptoms. High levels 

of symptom severity are often associated with reduction in functional ability. For example, 

lack of motivation, lack of energy, inability to focus, and/or reduced ability to interact with 

other people, are some features of Major Depressive Disorder (Kaplan & Saddock, 1998). It 

was expected that the more symptom distress experienced by the client, the greater would 

be the perceived difficulty in homework completion. Clients with severe pre-treatment 

symptoms were expected to have negative attitudes towards homework Clients with 

different presenting disorders were expected to differ in their attitudes towards homework. 

For example, a severely depressed client may have difficulty finding the energy to do 

homework. A client with obsessive-compulsive symptoms, however, may do excessive 

homework to alleviate anxiety. It was therefore expected that client attitudes towards 

homework would vary as a function of the presenting problem, and that there would be a 

positive relationship between negative attitudes towards homework and high levels of pre­

treatment symptom severity (Hypothesis 1 ). 

All behavior is maintained by its consequences. From the behavioral principle of positive 

reinforcement, it was expected that if homework were viewed positively and had positive 

consequences, or possibly if it were viewed negatively but had positive consequences, then 

more homework would be done (Martin & Pear, 1991 ). Clients may enjoy homework tasks, 

possibly from a sense of mastery, or pleasure, or both. Such clients would be expected to 

show increased quantity and quality of homework completion. 

33 



It is also possible for clients to do anxiety-provoking or otherwise unpleasant homework, 

because the consequences of doing the homework are positive and give a sense of mastery. 

For example, a ritual bound person who manages not to perform the ritual, may experience 

extreme anxiety in not performing the ritual. This is likely to be followed by relief that first, 

formerly held catastrophic beliefs are not valid, and second, that long-term symptom relief 

is now possible. From this comes an increased sense of mastery. It was expected that there 

would be a positive relationship between high ratings of mastery and pleasure and 

performance, or of either mastery or pleasure separately, and performance. An inverse 

relationship was expected, however, between high ratings of difficulty and obstacles, and 

performance. In this study it was expected that clients who had more positive attitudes 

towards homework on the mastery and pleasure sub-scales, would be more likely to have 

greater quantity and quality of homework completion (Hypothesis 2). 

Summary 

The hypotheses for the present study are: hypothesis 1, that a positive relationship between 

negative attitudes towards homework and high levels of pre-treatment symptom severity is 

expected; and that attitudes towards homework may vary as a function of the presenting 

diagnosis; and hypothesis 2, that clients who have a more positive attitude towards 

homework are more likely to have greater homework compliance and greater quality of 

homework completion than those clients with a less positive attitude towards homework. 

34 



Chapter 5: Measuring Attitudes Towards Homework 

Introduction 

The present study was concerned with two main questions. The first was: "To what extent 

does client symptomatology affect attitudes towards homework?" The second question was: 

"What is the relationship between client attitudes towards homework and their level of 

homework performance?" 

Literature on attitudes towards homework 

A literature search revealed little information available on attitudes towards homework. 

Much of the information available comes from compliance studies in medical research. For 

example, in a self-report study of causes of illness, patients who believed that their illness 

was medically based were more likely to be compliant with treatment than patients who 

believed in a non-medical cause for their illness (Foulks , Persons, & Merkel, 1986). Other 

studies involving attitudes and mental health included three studies on panic disorder. 

Katemdahl (1999) examining clients with panic disorder, researched the role personality 

disorders played in comorbidity and treatment response in mental disorders and suggested 

that patients with panic disorder had certain personality characteristics associated with 

phobic avoidance. Alnaes and Torgersen (1990) found that patients with panic disorder had 

increased sensitivity, compliance, and insecurity. Hoffart and Hedley (1997) found greater 

levels of somatization and hypochondriasis associated with panic disorder. 

Other studies are more closely related to cognitive therapy. March (1997) proposed a model 

of internal cohabitation as a theory to explain patterns of resistance towards homework. 

35 



While a client may agree that a homework task is reasonable and that it would be beneficial 

to do it, the client later changes his or her mind. This theory looks at some issues of 

avoidance. One method found to enhance homework compliance comes from Fennel and 

Teasdale (I 987) in a study of depressed clients. They found that clients who gave a positive 

response to a pamphlet explaining a cognitive model of depression, showed faster change 

and greater long term benefits than clients who did not respond positively. Persons, Burns, 

Perloff, and Miranda, (1993) found that dysfunctional attitudes and symptoms of 

depression and anxiety as reported by clients have more validity and usefulness than a DSM 

Axis I diagnosis (DSM-IV, American Psychiatric Association, I 994). There have, however, 

been few attempts to gauge such attitudes. 

One measure which involves attitudes, however, is the Mastery Scale (Pearlin, Lieberman, 

Menaghan, & Mui Ian, 1981 ), where mastery is defined as the extent to which clients regard 

their life-chances as being under their own control in contrast to being fatalistically ruled. 

Mastery is also thought to be one of the internal resources that protect individuals from the 

effects of social strain, along with low self-denigration and high self-esteem (Pearl in et al. , 

1981}. The Mastery Scale is a seven-item self-report measure with a four-point 

agree/disagree format. It comprises items about having control over things that happen, 

being able to solve problems, being able to change important things, not feeling helpless 

about problems, not feeling pushed around, feeling that there is some degree of personal 

control over future events, and a feeling of being able to achieve most things one really 

wants to do. Scores range from 7 (low mastery), to 28 (high mastery). 

36 



The Possible Reasons For Not Doing Self-Help Assignments Scale (Bums & AT. Beck, in 

A T. Beck et al. , 1979) was developed to explore reasons why some clients habitually did 

not do homework. It is a 15-item, self- report. true/false measure comprising a list of 

reasons clients had given for not completing homework during the course of therapy. In a 

detailed introductory paragraph the importance and reasons for doing homework are 

explained, as well as the fact that some clients express reluctance to do homework, or a 

desire to put off doing homework. The suggestion is that clients complete the form and 

discuss their results with their therapist. Items involve broad areas of difficulty and 

avoidance, while others involve issues relating to the client's perception of the therapeutic 

relationship. The final question allows for a longer open-ended response under 'other 

reasons'. 

Present Measures 

As no existing measure assessed client attitudes towards homework a new measure was 

developed. This involved an examination of the theory of cognitive and behavioral therapy 

to find areas that might involve attitudes towards homework. The Mastery Scale and the 

Possible Reasons For Not Doing Self-Help Assignments Scale also provided useful areas of 

exploration for the present study. The Attitudes Towards Homework Scale was devised for 

this study to examine client attitudes towards homework. A modified version of an existing 

scale, the Homework Performance Measure ( Kazantzis & Deane, 1998) was used to 

examine homework performance. These two, separate, self-report measurements were 

administered to participants at the beginning of each therapy session. 

37 



Practical Considerations 

There were six practical considerations for this study. First, the attitudes measure was 

designed to be brief, to increase the likelihood of client participation and completion. 

Clients were expected to complete the measure at the beginning of each therapy session 

about the homework from the preceding session. Second, data collection was limited to a 

short time frame to meet academic requirements. Clients were included in the study for 

whatever sessions they attended over this time frame. The measure had to be suitable for 

both beginning clients and those who had had many therapy sessions 

A third practical consideration was that the measure was to be filled in by clients who had 

completed or not completed homework assignments, to whatever degree. While compliance 

was also measured as a consequence of attitudes towards homework, the primary focus was 

on attitudes . That is, attitudes were not dependent on homework completion. Fourth, the 

measure was to be completed by clients whatever their diagnosis. While it was expected 

that clients with particular diagnoses might have similar type attitudes towards homework, 

items were worded as neutrally as possible to avoid targeting a particular diagnosis. 

A fifth practical consideration was that the measure was to be completed by clients who 

had been assigned any type of homework. It is not unusual for clients to be assigned 

multiple homework tasks in any given therapy session. Homework may also vary in type 

and degree of difficulty from early therapy to later stages of therapy. The measure had to 

examine responses in general towards homework, rather than examining responses to 

38 



specific tasks. Finally, aspects of the therapeutic relationship, homework relevance, and 

different types of homework assigned were not part of this study as these topics were 

addressed elsewhere within the broader homework study. 

Measure 1: Attitudes Towards Homework Scale 

Initially a large collection of items was gathered following a search of attitude measures 

and research on homework in cognitive therapy. These items were narrowed down to 18 

items deemed to adequately cover the domain under investigation. Following the basic 

behavioral principles of reinforcement, it was expected that if homework were a positive 

activity it would be done, so the domain of benefits to the client was explored. Questions 

were discussed such as: "Do clients enjoy doing their homework?", "Is there some reward 

for them for doing their homework?", "Do they fell good about themselves for having done 

their homework?'', "Does doing homework add something beneficial to their lives?". Stated 

somewhat simplistically, it might be expected that clients enter therapy expecting to benefit 

from it. As the rationale for doing homework is explained early in therapy, and the benefits 

of doing homework are made clear, it might be expected that all clients would do 

homework and yet clearly that is not the case. 

The next domain under exploration, then, was why clients did not do homework. Some of 

the questions discussed here involved: ''Did the clients know what to do for homework?", 

''Did the task seem beyond their capabilities?", "Were there practical difficulties that made 

homework difficult?", or did it involve that very human attribute of not wanting to do 

something even though it ultimately might be of benefit. These were the types of questions 

asked, as it was thought · that these areas might affect how a person thought, felt and 

39 



behaved towards homework tasks. That is, that these domains might affect a person's 

attitudes towards homework. 

It was ensured that items asked one question only, and that all items within each subset 

tapped different aspects of that set as a means of increasing internal consistency. Items 

were also checked for ease of reading, clarity, face and content validity (Streiner & 

Norman, 1998). The final 18 items were then evenly split into nine positively worded and 

nine negatively worded items. Negatively worded items are less sound psychometrically 

and have lower validity co-efficients than positively worded ones (Holden, F ekken, & 

Jackson, 1985). For example, they are more difficult to understand than positively worded 

items and raise the possibility that respondents may give an answer opposite to that 

intended. The even split was maintained, however, to avoid the potentially more likely error 

of response bias (Streiner & Norman, 1998). 

The instructions for completion of the form were brief and asked three things . Clients were 

asked to mark the selected box with a tick. Clients were asked to complete all questions. 

Clients were asked for their response to homework from last the session. As this was an 

important point it was set in bold type, and was also noted under the title so that the title 

read "Attitudes Towards Homework" and in smaller type "from last session" (see Appendix 

B). 

Response Scale and Subscales 

The Attitudes Towards Homework Scale comprises an 18-item self-report scale with a five 

point Likert-type response weighting (Likert, 1952). A number of research studies have 

40 



shown using reliability co-efficients that reliability drops as fewer categories are used 

(Streiner & Norman, 1998). Research evidence suggests that most people are unable to 

discriminate much beyond seven levels at a time. While having a seven-point scale would 

maximize the precision of the response, it would also involve a substantial increase in time 

spent filling in forms for respondents. The loss of reliability from a 7- point to a 5-point 

scale is 12 percent, dropping to 35 percent for two categories (Streiner & Norman, 1998). 

The choice of a 5-item scale provided the best balance of maximizing the precision of the 

measure while reducing the time spent filling in forms for the respondents. The five-point 

scale was also used rather than a forced choice format as it allowed for degrees of response, 

as attitudes and behaviors tend to exist on a continuum (Streiner & Norman, 1998). For 

example, a Likert-type response was also used by A T. Beck, when assessing mastery and 

pleasure as a means of inducing clients to recognize small changes and partial successes in 

therapy (A T. Beck et al., 1979). 

The items in the measure were constructed to survey client attitudes towards homework in 

a broad sense. After gathering a large collection of items related to the domains to be 

covered and collecting a large sample of items, the subscales were finalized as: mastery, 

and pleasure; comprehension; and difficulty, obstacles, and avoidance. Experiences of 

mastery and pleasure were considered positive features, while difficulties, obstacles, and 

avoidance were considered negative features of doing homework. Comprehension could 

either enhance or detract from the experience of doing homework, depending on the degree 

of comprehension. 

41 



Mastery and pleasure items were generated from cognitive therapy theory on mastery and 

pleasure. Mastery refers to a sense of accomplishment when performing a specific task and 

feeling in control of one's life. Pleasure refers to pleasant feelings associated with an 

activity, or a sense of self-satisfaction. Mastery does not need to relate to completion or 

magnitude of a task, and mastery and pleasure may be independent of each other. That is, 

mastery without pleasure is a step forward and even mild levels of pleasure are clinically 

important in establishing a sense of optimism and boosting morale (A T. Beck et al. , 

1979). Items related to mastery included: "I feel a sense of achievement at having 

attempted the homework", and "I am more in control of my problems from having 

attempted the homework". Items assessing pleasure included: "I feel good about having 

attempted the homework", and "I enjoyed doing the homework". 

Comprehension items involved understanding the content (what), the method (how), and 

the reason (why) of the homework assignment. Items related to homework comprehension 

included: "I understood what I had to do for homework'', I understood how to do the 

homework", and "I understood why I was asked to do the homework". 

Difficulty relates to how onerous the homework task was for the client. Difficulty as used 

here has a global meaning rather than asking for specific details of difficulty. Paradoxically 

being specific about precise areas of difficulty would be a task in therapy for making a 

homework task manageable, which was previously too onerous for the client. Items related 

to homework difficulty included: "The homework was too difficult for me", and ''There 

was too much homework for me to do". 

42 



Obstacles examines practical difficulties for the client, including practical reasons why 

homework might not have been done. There are many potential obstacles to doing 

homework and this selection of items attempted to cover a wide range of them. 

Items related to homework obstacles included: "My life is too stressful for homework right 

now'', "I had no trouble remembering to do the homework", "I didn 't have an opportunity 

to do the homework", "Unexpected events made it difficult for me to do the homework", 

and "It was hard to find the energy to do the homework" . 

Avoidance items were selected as representing some of the issues in the Possible Reasons 

for Not Doing Self-Help Assignments Scale (A. T. Beck et al. , 1979). Items related to 

avoidance of homework included: "l don't like doing homework unless I can get it right", 

"I felt worse when thinking about the things I had to do for homework", "It was easy for 

me to find time for the homework", and "I had trouble getting started on the homework". 

Measure 2: Homework Performance Measure 

A measure of compliance was included in the present study as the effect of homework 

depends on compliance. Compliance as used here comprises quantity and quality of 

completion as separate variables. There are specific features of homework performance, 

which enhance maximum therapeutic benefit. Quantity of completion is an important 

factor (Primakoff, Epstein & Covi, 1986; Shelton & Levy, 1981 ), as is the quality of the 

homework done (Primakoff et al., 1986; Schmidt & Woolaway-Bickel, 2000). Homework 

difficulty is also important, as the assignments should be within the ability of the client to 

perform, yet be sufficiently challenging to increase client skills and confidence in their 

ability to move beyond the present level of functioning (AT. Beck et al., 1979; J. S. Beck, 

43 



1995). As early as 1986, Primakoff et al. had noted that research had not examined quality 

as opposed to quantity of homework completion. Quality of compliance is increasingly 

being examined as a factor related to outcome. For example Schmidt and Woolaway-Bickel 

(2000) found both quality and quantity of homework completion were positively correlated 

with outcome, but that quality of completion was the better predictor of positive outcome. 

The original measure (Kazantzis & Deane, 1998), asked clients to rate compliance, quality 

and difficulty, on each homework activity separately. The original measure was modified to 

ask clients to rate homework compliance, quality, and difficulty globally. There were two 

reasons for modifying the original measure. First, the extended protocol for use with the 

original measure was not being used by therapists in the present study. For example, 

therapists in a previous study asked clients to rate their homework performance on each of 

1 to 3 homework assignments administered on a prescription form for clients to take away 

as a reminder. The prescription form then served to guide clients' rating performance for 

each task separately. Second, the measure was modified to match the attitudes measure, 

which also asks clients for a global assessment. 

The modified homework performance measure was also designed to be brief to facilitate 

the likelihood of completion of the form. Clients were asked to complete the measure 

before each session regarding the homework from the previous session. Compliance was 

assessed by the item: "How much homework did you do?". Quality was assessed by the 

item: "How well did you do the homework?". Difficulty was assessed by the item: "How 

difficult was the homework?" (Kazantzis & Deane, 1998, see Appendix C). 

44 



Summary 

The present study used two measures to assess client attitudes towards homework: The 

Attitudes Towards Homework Scale, devised for this study, and a modified version of the 

Homework Performance Measure ( Kazantzis & Deane, 1998). Both were brief, self-report 

measures. Both were to be filled out by clients at the beginning of each therapy session 

about clients' experience of homework from the previous therapy session. 

45 



Chapter 6: Method 

Introduction 

This study examined client attitudes towards homework in cognitive therapy within a 

broader research program on homework conducted at the Waitemata Health Cognitive 

Therapy Centre. The research program was designed by Nikolaos Kazantzis at the School 

of Psychology, Massey University, Albany, in conjunction with Paul Merrick and other 

staff at the Cognitive Therapy Centre. The center is a Waitemata Health clinical facility, 

accepting referrals from general practitioners, Waitemata Health agencies, and a variety of 

other community groups. Clients referred to the center have a range of psychological 

problems. The center essentially accepts for treatment all adult clients, excluding those with 

psychoses, organic brain disorders, and substance abuse. Data for this study was collected 

over a three-month period from August to October, 2000. 

Ethical Approval 

No 2000/137 Homework Study: Auckland Ethics Committee 

No HEC 98/87: Massey University Ethics Committee 

Sample 

The present sample comprised 18 adult participants in various stages of cognitive therapy at 

the Waitemata Health Cognitive Therapy Centre. As shown in Table 1, the present sample 

comprised 12 females and 6 males, with an average age of 36 years, ranging from 27 to 48 

years. Of these, 15 were New Zealand European and 3 were non-New Zealand European. 

Eight participants were single, and 8 were married. One was separated, and 1 was divorced. 

46 



Twelve participants were referred through mental health agencies, with 6 referred by their 

general practitioners. 

47 



Table 1 - Demographic Variables 

Characteristic 

Age 

M 
Mdn 
mm1mum 
maximum 

Gender 

Male 
Female 

Referral Source 

35.77 
34.50 
27.00 
48.00 

Community Mental Health Centres 
General Practitioners 

Marital Status 

Single 
Separated 
Divorced 
Married 

Ethnicity 

NZ European 
Other 

Primary Diagnosis 

Mood disorder 
Anxiety disorder 

Note: Varying n due to missing data 

Participants 

n % 

6 
12 

12 
6 

8 
1 
1 
8 

15 
3 

9 
9 

33.33 
66.67 

66.67 
33.33 

44.44 
5.556 
5.556 
44.44 

83.33 
16.67 

50.00 
50.00 

48 



As shown in Table 2, participants were grouped by primary diagnosis. Diagnoses were split 

into mood disorders (n = 9) and anxiety disorders (n = 8) with one participant added to this 

group with hypochondriasis, which, while a somatoform disorder has fear-based symptoms. 

A secondary diagnosis for this client also indicated anxiety symptoms. As might be 

expected in such a clinical sample, there was a degree of comorbidity with two of the mood 

disorder participants also having a secondary diagnosis of anxiety symptoms and three of 

the anxiety disorder participants also having a secondary diagnosis of mood disorder 

symptoms. 

49 



Table 2 - Participant Disorders {DSM-IV) 

Characteristic Participants 

n O/o 

Mood Disorders 

Major Depressive Disorder 7 38.9 

- with post-natal onset 5.6 

Bipolar Disorder (unspecified) l 5.6 

Anxiety Disorders 

Panic Disorder with Agoraphobia 2 11 . l 

Generalised Anxiety Disorder 2 11.1 

Anxiety disorder (unspecified) 2 11.1 

Panic Disorder 5.6 

Social Phobia 5.6 

Somatofonn Disorders 

Hypochondrias is I 5.6 

The most frequent diagnosis for mood disorder in the sample was Major Depressive 

Disorder, with I participant with post-natal onset, but essentially the same features as 

Major Depressive Disorder. One participant was in a depressive phase of Bipolar Mood 

50 



Disorder. The anxiety disorder group was much more diverse. Three subjects had a 

secondary diagnosis of Dysthymia. In a clinical population, Dysthymia precedes major 

depressive disorder in 15 to 25 percent of cases (DSM-IV, 1994) indicating a considerable 

overlap of symptoms in the two groups. 

Intake assessment 

Prior to the commencement of therapy, participants had a screening interview where a 

diagnosis was made, and completed an intake assessment, which comprised a battery of 

measures designed to be sensitive to changes in cognitive therapy. The following measures 

for depression, anxiety and cognitive functioning were used: the Beck Depression 

Inventory (BDI; A T. Beck et al.,1961); the Hamilton Rating Scale for Depression 

(HR.SD; Hamilton, 1960); the State-Trait Anxiety Inventory - state version (ST AI-Y-1 ; 

Spielberger, Gorsuch, Lashene, Vagg, & Jacobs, 1983); the Beck Hopelessness Scale 

(BHS; A T. Beck, Weissman, Lester, & Trexler, 1974); the Dysfunctional Attitudes Scale 

(DAS; A T. Beck, 1976; Weissman, & A T. Beck, 1978); the Automatic Thoughts 

Questionnaire (ATQ, Hollon & Kendall, 1980); and the Suitability for Short Term 

Cognitive Therapy Rating Scale (SSCT; Safran et al. , l 990a; Safran, Segal, Shaw, & 

Vallis. 1990b)). Details of participants' pre-treatment scores on these tests are shown in 

table 3. 

51 



Table 3 - Pre-Treatment SymjXom Severity Scores 

n M Mdn SD range 
minimum maximum 

Beck 
Depression 17 24.76 26.00 12.86 7 51 
Inventory 

State-Trait 
Anxiety 15 46. 13 43 .00 13.83 26 73 
Inventory 

Automatic 
Thoughts 16 76.93 66.50 31.40 39 132 
Questionnaire 

Dysfunctional 
Attitudes 9 116.00 96.00 55.81 15 198 
Scale 

Beck 
Hopelessness II 10.09 11.00 5.61 2 17 
Scale 

Hamilton 
Rating Scale 14 12.78 12.00 5.42 3 22 
for Depression 

Note . Varying n due to missing data 

52 



In the present study, particular attention was given to the BDI and the ST AI-Y-1. These 

measures were used for the symptom severity analysis of hypotheses 1 and 2, as client 

primary diagnoses involved predominantly mood and anxiety disorders. The BDI is a 

measure of symptom severity of depression, while the STAI-Y-1 is a measure of current 

(state) symptoms of anxiety. Details of the other measures are also included, as they 

provided part of the intake assessment upon which diagnoses were based, these diagnoses 

further providing the split into the two groupings used to examine hypotheses 1 and 2. For a 

variety of reasons, not all participants completed all the pre-treatment screening measures 

(seen values on Table 3). 

Beck Depression Inventory 

In the intake assessment clients were administered the BDI-IA. For the purposes of this 

study the scores were converted to the BOT-IT rating scale based on an equipercentile 

equating method (see A. T. Beck, Steer & Brown, 1996, p.26). 

53 



Table 4 - Effect of BDI Score Data Transformation 

Statistic BDI-IA BDI-II 

M 21.64 24.76 

Mdn 22.00 26.00 

SD 11.39 12.86 

Variance 129.74 165.44 

Range 39.00 44.00 

Minimum 6.00 7.00 

Maximum 45.00 51.00 

The BDI is a unidimensional 21-item self-report instrument for assessing the severity of 

affective, cognitive, motivational, psychomotor and vegetative components of depression 

(Oei & Shuttlewood, 1997). The original measure was developed in 1961 (AT. Beck, et 

al., 1961) and was replaced by the BDI-IA (A. T. Beck et al., 1979). The original BDI was 

based on clinical observations and symptom descriptions reported by psychiatric clients. 

The BDI-11 is a substantial revision of the measure. It was developed to better match the 

criteria for depressive disorders in the American Psychiatric Association's Diagnostic and 

Statistical Manual of Mental Disorders - Fourth &i.ition (DSM-IV, 1994). 

The BDI is used for adults and adolescents over 13 years of age (A. T. Beck et al., 1996). 

Items are rated on a 4-point scale and total scores range from 0 to 63 . In the BDI-IA clients 

report how they have felt in the previous week including the day of assessment. For the 

54 



BDI-II the time frame is extended to match the DSM-IV time criteria for major depression 

of two weeks, including the day of assessment (Brantley et al., 2000). In both versions 

higher scores indicate greater levels of depression. For clinical screening purposes for 

major depression, scores are: 0 to 13 (minimal depression), 14 to 19 (mild depression), 20 

to 28 (moderate depression), and 29 to 63 (severe depression), (A T. Beck et al., 1996). An 

important distinction is that the BDI reflects the degree of depression and not the diagnosis 

of depression (A T. Beck et al. , 1996). The BDI was never intended to be a diagnostic 

screening device even though it is now widely used in this way. That is, the BDI measures 

a depression syndrome rather than the presence of a depressive disorder and does not 

include a full coverage of criteria listed for depressive disorders in the DSM-IV (Brantley et 

al. , 2000). 

The BDI is one of the most frequently used measures to assess depression and has strong 

psychometric properties. It has high indices of reliability across clinical and non-clinical 

populations (A T. Beck et al. , 1988). The BDI also correlates highly with other well­

established measures of depression including the HRSD, the Zung Self-rating Depression 

scale (Zung, 1965) and clinical ratings of depth of depression (Brantley et al. , 2000). The 

BDI has the added advantage of a positive relationship with the BHS showing the 

convergent validity of the BDI. The BDI also correlates more strongly with the HR.SD than 

with the Hamilton Rating Scale for Anxiety (HRSA; Hamilton, 1967) showing the 

discriminant validity of the BDI (Brantley et al., 2000). This is important because of the 

high degree of comorbidity of depression and anxiety disorders. 

55 



Normative data for the BDI-II were obtained from samples from four different psychiatric 

outpatient clinics and one college student group (A T. Beck et al., 1996). 

The participants in the current study produced a mean BDI score of 24.76 (Mdn = 26, SD= 

12.6). This is slightly higher than the normative data for the BDI, where the normative 

outpatient sample produced a mean BDI score of 22.45 (SD = 12. 75), but are still within 

the same grouping of scores for moderate depression of 20 to 28. The normative sample of 

college students produced a mean BDI score of 12.56 (SD= 9.93), (A T. Beck et al., 

1996). 

State-Trait Anxiety Inventory (STAI Form Y-1) 

The State-Trait Anxiety Inventory is a measure of anxiety, which differentiates between 

state anxiety and trait anxiety (Spielberger et al., 1 983 ). State anxiety is defined as a 

palpable reaction or process happening at a given time and at a given level of intensity, 

whereas trait anxiety is defined as a relatively stable individual difference in anxiety­

proneness (Spielberger et al., 1983). In the pretreatment assessment, participants in the 

present study were assessed on the state anxiety scale (ST AI Form Y-1 ), and not on the trait 

anxiety scale (ST AI Form Y-2). The state scale comprises 20 self-report statements that ask 

how the individual feels right now at the time of form completion. Each ST AI item has a 

weighted score of 1 to 4 with higher scores indicating higher levels of anxiety for anxiety­

present items. For anxiety absent items (state items 1, 2, 5, 8, 10, 11, 15, 16, 19, and 20) the 

scores are reversed. Scores for both state and trait anxiety can range from 20 to 80 

(Spielberger et al., 1983). There have been substantial revisions to the STAI since 1979 to 

provide clearer discrimination between anxiety and depression, and to replace items found 

to be psychometrically weak for younger, and less educated clients, and lower 

56 



socioeconomic groupings. An added improvement was the more even balance of anxiety 

present and anxiety absent items (Spielberger et al., 1983). 

The ST AI is widely used in research and in clinical practice. The state version is used to 

measure transitory anxiety, for example with dental or medical treatments, job applications, 

and exams, and in psychotherapy and behavior modification programs. The state measure 

has been found to be a sensitive indicator of change in transitory anxiety as experienced by 

clients in psychotherapy settings (Spielberger et al., 1983). The trait scale has been widely 

used in assessing clinical anxiety in medical, surgical, psychiatric and psychosomatic 

clients, and also for screening purposes where high levels of anxiety might be a safety 

issue, for example, in military recruitment. A potential problem with the ST AI is that it 

correlates highly with self-report measures of depression, suggesting some problems with 

discriminant validity (A. T. Beck et al., 1988). 

The ST AI-Y-1 measure used in pretreatment assessment in the present study has been 

normed on working adults, college student, high school students, and military recruits 

(Spielberger et al., 1983). Participants in the present study produced a mean score of 46.13 

(Mdn = 43.10, SD = 13.84). As might be expected, this is somewhat higher than the 

normative scores. The normative data for working adults produced mean scores for males 

of 35.72, and for females of 36.20; the normative data for college students produced mean 

scores for males of 36.47, and 38. 74 for females. The normative data for the military recruit 

grouping produced mean scores for males of 44.05, and 47.01 for females . The military 

recruit population was assessed shortly after starting a stressful training program. 

(Spielberger et al., 1983). 

57 



Hamilton Rating Scale for Depression 

The Hamilton Rating Scale for Depression was originally designed to provide a quantitative 

assessment of symptom severity for clients diagnosed with depression (Hamilton, 1960). 

While the BDI focuses more on the cognitive features of depression, the HRSD focuses 

more on the physical features of depression and anxiety (Cleary & Guy, 1977). The HRSD 

is designed to be administered by a trained clinician using a semi-structured clinical 

interview. Using a time frame of the previous week, each symptom is considered for 

severity and frequency . Ratings are based on any change from the client's 'usual self due 

to depression (Kobak & Reynolds, 1999). The HRSD consists of 21 items, each measuring 

a depressive symptom, although usually only 17 of these are used because of the 

uncommon nature of the remaining four items, such as 'depersonalisation '. 

The 17 items are rated on either a 5-point scale (0 to 4) with 0 (absent), I (doubtful), 2 

(mild to moderate), 3 (moderate to severe), with 4 generally reserved for extreme 

symptoms; or as a 3-point scale (0 to 2). This 3-point scale is used where quantification is 

difficult or impossible. Items are rated on this scale as 0 (absent), 1 (possible or mild), and 

2 (definite), (Kobak & Reynolds, 1999). There have been problems with inter-rater 

reliability, but an interview guide providing standardized instructions for administration has 

been shown to reduce this problem. The general population mean for the 17-item HRSD is 

6.25 (SD 4.24) (Grundy, Lambert, & Grundy, 1996). Participants in the present study 

produced a mean score of 12.78 (SD= 5.42), which, as might be expected, is higher than 

the normative data. 

58 



Beck Hopelessness Scale 

The Beck Hopelessness Scale is a 20-item self-rating scale, which measures the extent of 

negative attitudes about the future (i.e., pessimism) as perceived by adolescents and adults. 

Hopelessness as a psychological construct is measured by individuals' beliefs that nothing 

will work out well, that they will not succeed whatever they do, that they will not achieve 

their goals, and that their problems will not be solved. (A T. Beck & Steer, 1993). 

The BHS comprises 20 true/false general statements of how the future is viewed. Of these 

11 are keyed as true and 9 are keyed as false. The scores are totaled with a possible range of 

0 to 20. Scores are rated as 0 to 3 (minimal hopelessness), 4 to 8 (mild hopelessness), 9 to 

14 (moderate hopelessness), and 15 to 20 (severe hopelessness). The general population 

mean is 4.45 (Greene, 1981 ). The BHS correlates highly with level of depression and 

predicts suicidal intent more accurately than depression measures (Wetzel, Margulies, 

Davis & Karam, 1980; Kovacs, A T. Beck, & Weissman, 1975). It is recommended for 

use with clients aged 17 or older until more information is available on psychometric 

properties for younger adults (A T. Beck & Steer, 1993). 

Psychometric studies of the properties of the BHS have used psychiatrically diagnosed and 

normal populations. Much of the data has not been based, however, on samples that meet 

criteria for formal diagnosis (according to the DSM). Rather most of the studies have been 

directed towards the measure' s use as an indicator of suicidal behaviors. It is used 

especially as an indirect indicator of suicide risk both for depressed individuals and for 

those with a history of suicide attempts. Moderate and high levels of hopelessness in 

depressed clients are seen as indicative of some degree of active suicidal desire (Nekanda-

59 



Trepka, Bishop, & Blackbum, 1983 ). A study by Beck, Steer, Kovacs, and Garrison (1985) 

found a 90% predictive rate of eventual suicide at a cut-off rate of 9 or above in a 5 to 10 

year follow up of 165 patients who had been hospitalized with suicidal ideation. 

Participants in the present study provided a mean score of 10.09 (SD= 5.61), which again, 

as might be expected is considerably higher than the normative general population data at 

4.45 . 

Dysfunctional Attitude Scale 

The Dysfunctional Attitude Scale is a 40-item self-report scale, which measures 

dysfunctional attitudes characteristic of individuals cognitively vulnerable to depression 

(Olinger, Kniper & Shaw, 1987). It was originally a 100-item scale (A T. Beck, 1976), and 

was later revised into two 40-item forms (Weissman & A T. Beck, 1978). Using a 7-point 

scale, subjects indicate their degree of agreement or disagreement with the scale items. The 

range of possible scores is 40 to 280 wit