The Role of Homework in Cognitive Behavior

Therapy of Depression
Michael E. Thase
University of Pittsburgh Medical Center

Judith A. Callan
Western Psychiatric Institute and Clinic

Homework is particularly important in the cognitive– behavioral treatment

of depression because the pervasive nature of the characteristic cognitive,
affective, and motivational disturbances undercut the impact of didactic and
supportive verbal interventions. Despite the importance of homework, a
relatively small number of studies have quantified the causal relationship
between homework completion and symptomatic outcome. Most of these
studies have limited power to detect small-to-moderate effect sizes and rely
on retrospective or incomplete measurements of homework that do not
distinguish between the quantity and quality of the assigned tasks. Neverthe-
less, there is relatively consistent evidence from correlational studies to
conclude that homework adherence is associated with significantly better
outcomes. These findings point to new questions for research (i.e., does
ongoing use of homework decrease the likelihood of relapse following
termination of time-limited therapy?) and have implications for clinical
practice. Examples of homework assignments are provided and strategies to
improve homework adherence are discussed.

The notion that an enterprise that involves learning new strategies to

overcome longstanding problems might be enhanced by extra practice outside
of formal didactic sessions is probably as old as teaching or training itself. It

Michael E. Thase, Department of Psychiatry, University of Pittsburgh Medical Center,

Western Psychiatric Institute and Clinic; and Judith A. Callan, Western Psychiatric Institute
and Clinic, Pittsburgh, PA.
Supported by MH-30915 (Mental Health Intervention Research Center) and MH-58356
from the NIMH.
Correspondence concerning this article should be addressed to Michael E. Thase,
Western Psychiatric Institute and Clinic, 3811 O’Hara Street, Pittsburgh, PA 15213-2593.
E-mail: thaseme@msx.upmc.edu

162
Journal of Psychotherapy Integration Copyright 2006 by the American Psychological Association
2006, Vol. 16, No. 2, 162–177 1053-0479/06/$12.00 DOI: 10.1037/1053-0479.16.2.162
Special Issue: Homework and CBT of Depression 163

was the shift toward viewing psychotherapy as a form of teaching or training

that is still relatively novel (see, e.g., Shelton & Levy, 1979). Nevertheless,
Burns and Spangler (2000) found examples of more pragmatic suggestions for
homework assignments in papers about psychotherapy that date back more
than 60 years. The basic rationale for the use of homework is that the therapy
“hour” (whatever the length) represents a only a fraction of time in a person
spends living with depression and such assignments provide a means to en-
hance mastery of newly learned coping strategies, facilitate generalization of
skills to novel situations, increase self-efficacy, and ultimately reduce vulner-
ability to relapse (Detweiler & Whisman, 1999).
Homework is particularly relevant to treatment of depression because the
characteristic motivational, cognitive, and behavioral disturbances “conspire”
to minimize the impact of verbal interventions such as psychoeducation,
advice, and development of new insights. Depressed people tend to be less
active, energetic, and enthusiastic, as well as more pessimistic and less able to
perform complex tasks. Procrastination and avoidance dominate waking life,
memory tends to be focused on negative events, and pessimistic predictions
(e.g., “That sounds good but it won’t work for me!” or “Why bother? It won’t
make a difference”) undercut willingness to try new strategies. Even when a
new way of coping is attempted, the tendency to evaluate outcomes reinforces
helpless/hopeless cognitions about problem solving. Indeed, one might spec-
ulate that without successful use of homework, depressed people may gain
little from cognitive behavior therapy (CBT) over and above the nonspecific
factors that underpin all helping relationships.
Homework is not a monolithic entity, however, and in broadest use the
term represents an amalgam of the skill of the therapist, the nature (and
appropriateness) of the particular tasks, and the quantity and quality of the
work completed. Therapeutic skill, in turn, subsumes the abilities to persuade
a depressed person to attempt homework assignments, to incorporate home-
work into the regular flow of sessions, to reinforce successive approximations,
to deal with nonadherence effectively, and to select progressively more com-
plex assignments that “match” the needs and readiness of individual patients.
One of the major theoretical issues pertaining to homework concerns
causal attributions: Do people improve because they do homework versus
are people who are likely to improve more likely to do homework? A 3rd
possibility is that yet another variable, such as motivation or self-efficacy,
mediates both improvement and homework performance. As noted by
several researchers (Burns & Spangler, 2000; Detweiler & Whisman, 1999;
Kazantzsis, Ronan, & Deane, 2001), definitive resolution of this issue
requires a level of detail of investigation that has not yet been undertaken.
The results of the research reviewed below, however, do suggest that the
1st hypothesis is viable and worthy of further study.
164 Thase and Callan

STUDIES EXAMINING THE EFFECT OF HOMEWORK AND

OUTCOME

We identified 13 studies that examined the impact of homework on

CBT response in depression, as well as recent narrative (Detweiler &
Whisman, 1999) and meta-analytic (Kazantzis, Deane, & Ronan, 2000)
reviews. The 2 studies that utilized random assignment to homework versus
no homework conditions will be reviewed 1st, followed by the 11 studies
that used correlational methods. Lastly, the quantitative review of Ka-
zantzis and colleagues (2000) will be summarized.

Randomized Studies

Kornblith, Rehm, O’Hara, and Lamparski (1983) studied the effect of regular
homework assignments on the efficacy of group self-control therapy in 22 women
with major depression. They found no evidence that homework improved the
outcome of this form of Group CBT. However, patients in the groups utilizing
homework completed only about one half of their assignments. Moreover, some
patients in the “no homework” groups devised and carried out their own tasks. As
such, the independent variable manipulation was relatively weak. This is particu-
larly important because with only 11 patients in each condition, the study only had
the statistical power to identify large effects, such as a between-groups difference
of ⬎8 points on the Beck Depression Inventory (BDI; Beck, Ward, Mendelson,
Mock, & Erbaugh, 1961).
Neimeyer and Feixas (1990) also conducted a randomized study to eval-
uate the impact of homework on the outcome of Group CBT. Outpatients
with major depressive disorder (MDD) were randomly assigned to CBT
groups that either did (n ⫽ 32) or did not (n ⫽ 31) use homework assignments.
Treatment consisted of 10 90-min weekly sessions, and homework assignments
were standardized. The patients in both types of groups improved significantly
on the BDI and Hamilton Depression Rating Scale (HAM-D; Hamilton,
1960). Time ⫻ Treatment interactions were not statistically significant, which
indicates that homework did not systematically improve therapeutic outcome.
When pretreatment severity was taken into account, however, patients in the
groups utilizing homework assignments did improve significantly more on the
HAM-D than those in the “no homework” groups. The parallel analysis of the
BDI revealed a trend did not reach statistical significance. These finding,
admittedly post hoc, suggest that the use of homework facilitated recovery
only when patients had more severe depressive symptoms. Said another way,
homework was not necessary for improvement when patients with milder
depressions were treated with Group CBT.
Special Issue: Homework and CBT of Depression 165

Correlational Studies

Fennell and Teasdale (1987) examined the correlates of early symp-

tomatic improvement among 34 depressed primary care outpatients with
major depressive disorder. Patients were randomly assigned to receive
either Beck’s model of CBT in addition to pharmacologic treatments as
usual (TAU) or TAU alone. Overall, whereas patients in the TAU condi-
tion obtained minimal symptomatic benefit, the addition of CBT was highly
effective (Teasdale, Fennell, Hibbert, & Amies, 1984). Improvement dur-
ing the first 3 weeks of CBT was significantly correlated with the subse-
quent response; there was no such early predictive effect in the group that
received TAU alone. Such early improvement was accounted for by 2
homework-related events: 1) a positive initial evaluation of the “Coping
With Depression” booklet (typically the patient’s 1st homework assign-
ment), and 2) completion of homework assignments between the 2nd and
3rd weeks of therapy.
Persons, Burns, and Perloff (1988) studied variables associated with
CBT response in 70 independent practice patients treated for depressive
disorders. They found that patients with high initial BDI scores and those
that met criteria for the endogenous depression subtype had poorer out-
comes. Moreover, the therapists’ ratings of patients’ homework adherence
predicted lower final BDI scores. Like Neimeyer and Feixus (1990), they
found that the impact of homework adherence was greatest among patients
with high pretreatment BDI scores. In other words, although it was gen-
erally helpful to complete homework assignments, homework had the
greatest impact for patients with more severe depressions. In fact, among
the more severely depressed subgroup, patients who were rated high on
homework adherence improved nearly 3 times as much as those who did
not (i.e., 73% vs. 23% reduction in BDI scores).
DeRubeis and Feeley (1990) studied the correlates of improvement of
25 adult outpatients with MDD treated with individual CBT. They identi-
fied 1 factor representing “concrete” or symptom-focused therapeutic
methods that was associated with symptom reduction early in the course of
therapy, but not with improvement later in therapy. Homework was 1 of
the interventions included in this factor.
Burns and Nolen-Hoeksema (1991) examined depressed outpatients’
reported use of coping strategies in relation to their willingness to complete
homework assignments. The study involved 125 outpatients with MDD
who were treated with at least 12 weeks of individual CBT. They found that
patients’ reported use of coping strategies prior to treatment did not
predict either subsequent homework adherence or response at week 12.
Moreover, self-reported willingness to try new coping strategies was only
166 Thase and Callan

weakly correlated with homework adherence (r ⫽ .14, p ⫽ .06). A regres-

sion analysis was next performed to determine the additive effects of
homework adherence and use of coping strategies on outcome, controlling
for pretreatment depression severity. Homework adherence accounted for
a small (6.6%), but statistically significant amount of variance in final BDI
scores. In a subsequent report based on an expanded study group (n ⫽
185), Burns and Nolen-Hoeksema (1992) confirmed that therapists’ ratings
of homework completion had a modest, but statistically significant (r ⫽ .25,
p ⬍ .01) association with improvement in BDI scores.
Startup and Edmonds (1994) examined homework-response relation-
ships among 25 outpatients with MDD who were randomly assigned to
receive either 8 or 16 weekly sessions of individual CBT. Therapists
categorized the type of homework assignment given for each session and
rated adherence on a 7 point scale (1 being the assignments was not
completed up to 7 equaling the assignment was completed as agreed); the
mean adherence rating was 4.56 (SD ⫽ 2.2). They found that early home-
work adherence explained 13% of the outcome variance, and was signifi-
cant even after controlling for age, sex, and pretreatment severity. The
relationship between early homework adherence and BDI scores at the
3-month follow-up was not statistically significant, however.
Bryant, Simons, and Thase (1999) examined 3 aspects of homework in
26 MDD outpatients treated with a 16-week, 20 session course of individual
CBT. They assessed: 1) relationship of homework adherence to outcome;
2) relationship of therapist behavior to homework completion; and 3)
relationship of selected patient variables to homework adherence. The
investigators used the Assignment Compliance Rating Scale (ACRS; Pri-
makoff, Epstein, & Covi, 1986) to assess homework adherence. Scores on
this operationally defined scale ranged from 1 (the patient did not attempt
the assigned homework) to 6 (the patients did more of the homework than
was requested). The mean ACRS rating across sessions was 4.01 (SD ⫽
1.38). Mean ACRS scores were significantly correlated with improvement
in the HAM-D, but not the BDI. There was also a significant correlation
between the therapist’s review of the homework from the previous weeks
and the patient’s adherence at next session (r ⫽ .39, p ⫽ .002). Of note,
whereas patients attempted to complete 90% of homework assignments,
therapists forgot to incorporate homework assignments more than 20% of
the time.
Shaw et al. (1999) evaluated the role of fidelity of therapy in 53 outpatients
with MDD who were randomized to receive 16 weeks of therapy with indi-
vidual CBT as part of the National Institute of Mental Health Treatment of
Depression Collaborative Research Program (Elkin et al., 1989). This study is
unique in that independent experts reviewed tapes of therapy sessions and
Special Issue: Homework and CBT of Depression 167

scored various aspects of the therapists’ delivery of CBT. Overall, they found
that a single therapist competency factor accounted for approximately 19% of
the outcome variance. The competency factor included 2 relevant items: the
therapists’ ability to incorporate homework assignments into the agenda at the
beginning of sessions and to craft appropriate new assignments at the end of
each session. These findings are particularly noteworthy because it should be
recalled that the overall results for CBT were rather disappointing in this study
(i.e., CBT was not significantly more effective than a clinical management-pill
placebo condition on most of the primary analyses of outcome and was
significantly less effective than clinical management-active imipramine on
some analyses of patients with more severe depression) (Elkin et al., 1989;
1995; Gibbons et al., 1993). Thus, it is plausible that some therapists’ failure to
systematically attend to the importance of homework played a critical role in
CBT’s relatively poor showing in this influential study.
Burns and Spangler (2000) examined the causal relationships linking
homework compliance with improvements in depression using data from 2
groups of depressed outpatients (n ⫽ 122 and n ⫽ 399) who received CBT in
a independent practice specialty clinic. Structural equation modeling was used
to examine 2 causal paths: 1) the causal effect of homework completion on
improvement in depression; and 2) the causal effect of improvement in de-
pression on homework completion. They found that high pretreatment BDI
scores did not significantly reduce homework completion. By contrast, home-
work completion was associated with significantly greater subsequent im-
provement. For each “quantum” increase in homework completion, patients
achieved 4 more points of improvement on the BDI.
Homework adherence also was examined in relation to outcome in
preliminary (n ⫽ 45) and final (n ⫽ 150) reports on a study of outpatients
with MDD treated with 3 forms of cognitive and behavioral therapies
(Addis & Jacobson, 1996, 2000). Overall, both cognitive and behavioral
therapies were as effective as the full “package” (Jacobson et al., 1996).
Homework compliance was measured on a 7-point scale after each session.
In addition, the investigators studied patients’ acceptance of the rationale
for treatment provided by therapists. Although the relationship between
homework adherence and outcome was not significant in the preliminary
study (Addis & Jacobson, 1996), a significant (albeit weak) association was
observed in the final report. For example, adherence during the first 4
weeks of therapy was associated with both early (r ⫽ .23, p ⬍ .01) and
end-point (r ⫽ .17, p ⬍ .05) improvement. Homework adherence during
the middle phase of therapy also was associated with change at the mid-
therapy assessment (r ⫽ .18, p ⬍ .05) and with final outcome (r ⫽ .17, p ⬍
.05). Patients’ acceptance of the rationale for treatment also was signifi-
cantly and independently associated with outcome.
168 Thase and Callan

Meta-analysis

Kazantzis, Deane, and Ronan (2000) conducted a meta-analysis of the

effects of homework on response in 27 ambulatory studies of treatment of
depression, anxiety, or mixed patient populations. They found strong evi-
dence that interventions that included homework were more effective than
interventions that did not (weighted r ⫽ .36). The effects observed in
studies of depression tended to be larger than in studies of anxiety disor-
ders. They found that patient adherence with homework assignments also
was significantly associated with outcome (weighted r ⫽ .22).

MODERATORS OF HOMEWORK ADHERENCE

Nonadherence to homework assignments has been broadly explained

as an interrelationship among task, therapist, and patient variables (Det-
weiler & Whisman, 1999). The major task determinants of homework
completion are the perceived and objective difficulty of the task. One study
of a mixed patient population identified the length of homework assign-
ments as an important aspect of task difficulty (Conoley, Padula, Payton, &
Daniels, 1994). Research in behavioral medicine suggests that task diffi-
culty appears to be attenuated by the use of written assignments [(e.g., the
“behavioral prescription pad” of Cox, Tisdelle, and Culbert (1988)].
Therapist variables have not yet been studied extensively, although it
is clear that individual differences in the enthusiasm for using homework
assignments are an important component of competency (e.g., Shaw et al.,
1999). Bryant, Simons, and Thase (1999) likewise found that therapists who
make regular use of homework assignments from session to session elicit
the greatest homework adherence. In the study of Startup and Edmonds
(1994), patient’s perceptions of the therapist’s skill in describing homework
assignments similarly were not correlated with homework adherence.
However, this latter finding is almost certainly an artifact of attenuated
range or problems with measurement: it is implausible that indifferent or
insincere therapist presentations of homework assignments would be un-
related to adherence. Nevertheless, a global rating of therapist empathy
was independently associated with treatment response and did not influ-
ence homework adherence in one larger study (Burns & Nolen-Hoeksema,
1992).
Adherence to homework assignments has not been linked to patient
factors such as symptom severity, personality pathology, or self-reported
preference for active coping strategies in studies of CBT of depression (see,
e.g., the review by Detweiler & Whisman, 1999). Persons (1989) suggested
Special Issue: Homework and CBT of Depression 169

3 observations from her clinical practice that increased the likelihood of

nonadherence: perfectionism, fear of failure, and fear of displeasing the
therapist. Although these qualities “ring true” to our own clinical experi-
ences, they have not been tested using operationally defined ratings and
prospective assessment. Fennell and Teasdale (1987) did observe that
demoralization (“depression about depression”) was associated with a
more rapid response to CBT of depression, which in turn was linked to
early homework adherence. Conversely, Addis and Jacobson (1996) found
that patients who attributed their depressions to existential or interper-
sonal causalities were less likely to complete homework assignments than
patients who endorsed other causal factors.

METHODOLOGIC ISSUES AND FUTURE DIRECTIONS FOR

RESEARCH

Even a cursory review of this literature leads to the conclusion that

homework has not received “its due” in research on CBT of depression.
Despite its centrality to cognitive and behavioral models of change, and
surprisingly consistent evidence of a significant association between com-
pletion of assignments and treatment response, homework is typically
studied as an afterthought, utilizing retrospective ratings of uncertain
reliability and validity (see, e.g., Detweiler & Whisman, 1999; Kazantzis,
Ronan & Deane, 2001). Retrospective ratings are particularly vulnerable
to “halo” effects and patients who are responding to treatment may over-
estimate the amount of homework that they complete (see Smith, Leffing-
well, & Ptacek, 1999). Most studies have employed correlational designs,
which cannot determine causality, even when sophisticated mathematical
models are utilized (Kazantzis, Ronan & Deane, 2001). Lastly, aside from
the large studies of Burns and colleagues (Burns & Nolen-Hoeksema, 1991,
1992; Burns & Spangler, 2000; Persons, Burns & Parloff, 1988), most
studies are simply too small to have adequate statistical power to test
hypotheses about moderating effects (Kazantzis, 2000).
A new generation of research is now needed. Important methodologic
considerations include ensuring that therapists actually assign and review
homework tasks, prospective collection and measurement of the quantity
and quality of homework assignments, and frequent measurement of out-
comes so that temporal relationships can be examined. Lastly, although the
ecological validity of research designs in which patients are randomly
assigned to receive CBT with or without the use of homework can be
debated, there are other fertile avenues for research. For example, do
individualized homework assignments actually enhance progress or could
170 Thase and Callan

similar results be gleaned by using standardized tasks? When homework

assignments are not being completed, is it possible to intervene more
intensively and would this increase the chances for a favorable outcome?
Does continued use of self help interventions actually reduce the risk of
relapse after termination of therapy? Research that answers these ques-
tions will only strengthen the empirical basis of CBT.

CLINICAL CONSIDERATIONS

Although other therapeutic skills may be perceived as more valuable

or personally rewarding, the skillful use of homework is a reliable
“marker” of a cognitive-behavior therapist. Indeed, the ability to assign
homework and to incorporate those activities into subsequent sessions is
one of the few replicated indicators of the fidelity of therapy. Patients’
adherence to homework assignments, conversely, is a reliable (albeit mod-
est) predictor of therapeutic success. This pattern is evident within the 1st
few weeks of treatment. We therefore suggest that clinicians who do not
regularly utilize homework assignments should: 1) consider the possibility
that they are not providing the most effective form of therapy for their
patients and 2) should reconsider the appropriateness of calling themselves
cognitive– behavioral therapists!
Strategies to make better use of homework assignments have been
suggested by Burns and Auerbach (1992) and Broder (2000). Broder
(2000) emphasized goal setting as a logical companion to homework. As
part of this approach, a stepwise approach to problem solving is empha-
sized, with homework assignments framed as necessary steps or “sub-
goals.” Broder also suggested that the use of preprinted homework forms
and other prepared materials (e.g., audiotapes) may facilitate homework
adherence.
Burns and Auerbach (1992) suggested that the “art” of therapy in-
cluded being able to foster a shared expectancy that homework is an
essential component of therapy and to match the difficulty of homework
assignments with the patient’s readiness. In this respect, the therapist needs
to ensure that the importance of homework or self-help is emphasized as
part of the “socializing” the patient into the CBT model of therapy. They
outline the basic types of homework assignments (i.e., bibliotherapy, self-
monitoring, thought recording, cognitive restructuring exercises, and inter-
personal assignments) in relation to the progress of symptomatic improve-
ment (see Table 1).
Although it is encouraging that up to 90% of depressed patients
attempt to complete homework assignments (Bryant, Simons & Thase,
Special Issue: Homework and CBT of Depression 171

Table 1. Types of CBT Homework

Assignment Example
Self monitoring Rating moods in relation to activities
Bibliotherapy Assigned readings such as “Coping
With Depression” (Beck &
Greenberg, 1974) or Feeling Good
(Burns, 1980).
Behavioral activation Scheduling periods of exercise or
participation in formerly
pleasurable tasks to offset
inactivity and increase access to
reinforcement.
Behavioral symptom management strategies Practicing use of techniques such as
progressive muscle relaxation,
thought stopping, or sleep hygiene
to improve specific symptoms.
Cognitive assignments Planned practice using the Daily
Record of Dysfunction Thoughts,
“Examine the Evidence,” or “Pros
and Cons” techniques.
Schema assessment Review of cognitive assignments to
identify common these, writing a
brief autobiography or filling out
the Dysfunctional Attitude Scale.
Note. CBT ⫽ Cognitive Behavior Therapy.

1999, 10)% flatly refuse and some larger proportion approach homework
half-heartedly. What can be done to facilitate adherence? First, it is im-
portant to keep in mind that behavior is both motivated and maintain by
consequences, and it is useful to understand the phenomenology of non-
adherence. Burns and Auerbach (1992) described the use of a “self-help
memo” to proactively address some of the most common reasons for
nonadherence. The exercise includes a listing of 25 reasons not to complete
homework assignments, which can be scored on a scale of personal rele-
vance ranging from “not at all” to “a lot” (see Table 2). As the nonadher-

Table 2. Commons Reasons For Not Doing

Homework
1. All or none thinking
2. Fear of disapproval
3. Hopelessness
4. Unexpressed anger
5. Coercion sensitivity
6. Depressive realism
7. Conceptual mismatch
8. Entitlement
9. Fear of change
10. Shame
Note. Adapted from Burns, Adams, and Anastopoulos
(1985).
172 Thase and Callan

ent patient is unlikely to complete this as a homework assignment, it is

necessary to use a portion of the therapy hour to collect these ratings. The
items that receive high scores can then be examined collaboratively for
evidence of cognitive distortion and addressed with the standard “5 col-
umn” intervention for dysfunctional thoughts. Paradoxical intention and
role-playing, especially using role reversal, are illustrated as creative ap-
proaches to help engage the recalcitrant patient in the process of doing
homework. Flexibility also is emphasized, including discarding the term
“homework” when patients report that earlier academic endeavors have
left an unsavory “taste” for the concept. The terms “self-help exercise” and
“practice sessions” may better convey the more independent and adult-like
nature of this type homework.
Second, the patient’s difficulty completing particular homework assign-
ments should be viewed as an indication of the therapist may have failed to
clearly describe the task, selected an assignment that was too demanding,
did not clearly link the assignment to the material addressed during the
sessions, or misperceived the patient’s willingness to endorse the merits of
the particular task. As with other interchanges between the patient and
therapist, collaboration is made explicit by asking for feedback and ensur-
ing that the assignment is perceived to be relevant and “do-able.”
Additional strategies that we have found helpful to increase the like-
lihood of homework adherence include asking the patient to set-aside or
schedule a specific time on a daily activity sheet to work on homework
assignments, to plan for interim check-ins (by telephone or e-mail) to
review homework activities, and to request the assistance of a supportive
significant other. We routinely ask patients to keep a loose leaf or spiral
bound notebook to provide an enduring record of their homework activ-
ities. The importance of assignments also can be conveyed by devoting a
significant portion of each session to completing tasks that have gone
undone and by explicitly incorporating that work within the session’s
agenda of new material.

CASE EXAMPLE

Mr. K., a 47-year-old factory worker, was referred for CBT by his
family physician after several unsuccessful trials of antidepressant medica-
tion. He met criteria for MDD, single episode (nonpsychotic), severe
(BDI ⫽ 35). The depressive episode was about 9 months in duration. Mr.
K. was having great difficulties in his workplace: He had been demoted
about 1 year previously (3 months prior to the onset of the depressive
episode), and he considered his employer’s actions to be grossly unfair. The
Special Issue: Homework and CBT of Depression 173

demotion was precipitated by complaints from Mr. K.’s supervisees, who

found him critical, demanding, and insensitive. Mr. K. also had been
“written up” for using vulgarities and making several racial slurs. He had
appealed the decision through his union and lost. Although the new job
was not demanding and the demotion resulted in only a modest reduction
in pay, he felt humiliated because he had held that position some 18 years
earlier. Mr. K. also felt trapped: He wanted to resign but knew that he
could not find another job with comparable pay and benefits. Mr. K. also
recognized that he was the wage earner for his family (which included a
wife and 3 sons), and that he was two thirds of the way to a well-
compensated early retirement plan.
Mr. K.’s primary symptoms were cognitive and emotional. He spent a
great deal of time ruminating about the unfairness of his employer’s actions
and his sense of betrayal by the union officials. Mr. K. had thoughts like: “I
can never count on anyone” and “No matter how hard I try things won’t
get better.” He felt diffuse anger and irritability, as well as great sadness
and pessimism about the future, and was on the verge of tears most of the
time. Mr. K. had suicidal and homicidal thoughts. He had not lost his
appetite and had gained perhaps 10 pounds. Mr. K. was less active and
stopped doing previously preferred hobbies. His most distressing physical
symptom was marked fatigue, which he stated that could be felt in “his
bones.” He slept about 1 more hour each day than normal. Although he
had abused alcohol and marijuana in his teen and 20s, he remained
abstinent.
Mr. K.’s history and symptomatic presentation were fully appropriate
for treatment with CBT. Nevertheless, he was skeptical about the chances
for therapy. At the end of the initial session, he reported feeling comfort-
able with the developing rapport with the therapist but semijokingly said
that the intervention he really needed was getting his old job back. The
initial homework assignment, to read the Beck and Greenberg (1974)
pamphlet “Coping With Depression,” resulted in a few derisive comments
scrawled on the margin. Mr. K. stated that the therapy approach was
“Mickey Mouse” and did not think that it was appropriate to try to “talk
away” his grief over having been so mistreated. The activity monitoring
scheduled enclosed in the pamphlet was blank, despite the apparent agree-
ment that this was a reasonable and relevant task.
After engaging Mr. K.’s help to set the agenda for the 2nd session, the
therapist spent 5 minutes helping Mr. K. to fill in the blanks on the
schedule. It was apparent that when Mr. K. was at work, his mood was
mostly sullen and angry, and when he was at home, he felt sad and
withdrew from his family. For 5 straight nights he sat alone in his den,
reliving conversations from the workplace that reinforced his sense of
mistreatment and loss of esteem. The therapist pointed out how the activity
174 Thase and Callan

schedule could help target particular behaviors that increase or decrease

dysphoria.
The next item on the agenda was looking for a relationship between
Mr. K.’s thoughts and feelings about the pamphlet and his pessimism about
the therapy. This discussion took the balance of the hour. Mr. K. grudg-
ingly acknowledged that his low appraisal of the therapy may have been
flip and that there was no evidence that the therapist intended to dismiss or
trivialize his grief. The reference to Mickey Mouse was used to illustrate
how a cognitive distortion (name-calling) could undercut a positive expe-
rience. At the end of the session, Mr. K. again agreed to use the activity
schedule to help keep track of his moods and to begin to keep a diary of his
negative predictions about the future.
At the 3rd session, Mr. K. had again not completed an activity sched-
ule, nor had he made any effort to begin a thought diary. He stated that he
had agreed to do the assignment to get the therapist “off his back.” The
therapist apologized for mistaking Mr. K.’s acquiescence for agreement
and obtained Mr. K.’s consent to again spend session time to complete the
missing assignment. A spiral bound notebook was provided and the initial
“hot thoughts” were recorded. Later that session, the behavior-feeling-
thinking relationship during his “down time” at home was targeted. Mr. K.
agreed that he was physically capable of doing more things that used to
give him pleasure, but again expressed reservations about the wisdom of
the focus and suggested that it smacked of “not dealing with his problems.”
Mr. K. did identify a small woodworking project that would require about
4 hours of work and agreed to consider devoting 1 hour per night to the
project. The therapist also attempted paradoxical intention by insisting that
the patient not even attempt to complete an activity schedule during the
intervening 4 days. Mr. K.’s mood lifted briefly, and he chuckled “You
don’t think that I’m going to fall for the old reverse psychology, do you
doc?”
At the next session and over the next 8 weeks, the pattern of home-
work nonadherence continued. Mr. K. was unwilling to attempt even the
briefest homework activities, remarking that he would not give in to the
“tyranny” of homework. Such resistance was puzzling because there was
progress within sessions on identifying automatic negative thoughts and
developing rational alternatives. Mr. K. accepted that he had been unwill-
ing to admit that he had made mistakes as a supervisor and that, while the
demotion felt unjust, there was evidence that the employees he supervised
had valid complaints. He accepted a “coping card” that summarized his
rationale responses to thoughts about not being able to stand 10 more years
working in a job that he did not want. However, he reported that he would
not be “caught dead” looking at the card when he was at work. On one
Special Issue: Homework and CBT of Depression 175

occasion Mr. K. interpreted the therapist’s statement of fact about the

efficacy of homework to be an example of “the carrot and the stick”
approach and playfully reminded the therapist that mathematical averages
did not pertain to individuals. He permitted his wife to purchase the book
Feeling Good (Burns, 1980), but he refused to read a single page. Mr. K.
rejected the therapist’s offer to schedule brief telephone consultations to
assist with completion of self-help activities. He also declined to use an
audiotape recorder to do self-help activities verbally rather than with pen
and paper.
Of note, during Session 8, Mr. K experienced a marked positive mood
shift when reviewing the “Reasons Not to do Homework” handout from
the Burns (1980) book. Mr. K. reported that the shift was triggered by the
comforting thought that “Other people must refuse to do homework, too.”
Later that session, when discussing the potential advantages of not doing
homework, Mr. K. volunteered that he had despised school, had been
involved in epic struggles with his physically abusive father over the
completion of schoolwork, and had chosen trade school instead of conven-
tional high school so that he would “never again” have to do homework.
“Sensitivity to coercion” was identified as the relevant core belief. Al-
though Mr. K. was able to identify a number of differences between the
present circumstance and the past, he stated that he was unwilling to “go
back on his word.” An attempt to relabel these exercises with some less
personally offensive name, such as practice sessions or self-help exercises,
was flatly rejected as semantics.
After 8 weeks of therapy, Mr. K’s BDI was 22. He acknowledged
feeling “somewhat better” and attributed the improvement to learning to
better cope with his negative thinking. He continued to express concern
about the aims of therapy and wondered aloud if therapy might turn him
into a “Stepford Employee” (i.e., an emotionless automaton). Mr. K.
continued to refuse to try any homework. Nevertheless, he attended
every session and worked with increasing enthusiasm during sessions.
Mr. K. did express pride that he was making progress by doing therapy
“his way.”
During the subsequent 8 weeks of therapy, slow improvement contin-
ued, and Mr. K.’s BDI scores at Sessions 14 and 16 were 13 and 14. Mr. K.
stated that he had come to terms with the fact that he had been “screwed”
by his employer, but that he needed to stay on the job because his family
depended upon him as the wage earner. He was able to appropriately rank
the magnitude of his employer’s “betrayal” in relation to other injustices
that had befallen humankind. At the final session of therapy, Mr. K. also
mentioned that he frequently did cognitive exercises “in his mind” and
planned to continue this particular unnamed form of covert homework.
176 Thase and Callan

SUMMARY

Homework is an important component of CBT, and homework adher-

ence increases the chances of successful treatment of depression. Although
generally relegated to “side show” status as both a topic of research (when
compared to studies of therapy efficacy or process), and as a favored
strategy of therapists, there is reason to believe that greater attention
increasing the quantity and quality of homework may increase the efficacy
of CBT and enhance long-term prophylaxis against relapse and recurrent
depressive episodes.

REFERENCES

Addis, M. E., & Jacobson, N. S. (1996). Reason-giving and the process and outcome of
cognitive-behavioral psychotherapies. Journal of Consulting and Clinical Psychology, 64,
1417–1424.
Addis, M. E., & Jacobson, N. S. (2000). A closer look at the treatment rationale and
homework compliance in cognitive-behavioral therapy for depression. Cognitive Ther-
apy and Research, 24, 313–326.
Beck, A. T., & Greenberg, R. L. (1974). Coping with Depression. New York: Institute for
Rational Living.
Beck, A. T., Rush, A. J., Shaw, B. F., & Emery, G. (1979). Cognitive therapy of depression: A
treatment manual. New York: Guilford Press.
Beck, A. T., Ward, C. H., Mendelson, M., Mock, J., & Erbaugh, J. (1961). An inventory for
measuring depression. Archives of General Psychiatry, 4, 561–571.
Broder, M. S. (2000). Making optimal use of homework to enhance your therapeutic effec-
tiveness. Journal of Rational-Emotive & Cognitive-Behavior Therapy, 18, 3–18.
Bryant, M. J., Simons, A. D., & Thase, M. E. (1999). Therapist skill and patient variables in
homework compliance; controlling an uncontrolled variable in cognitive therapy out-
come research. Cognitive Therapy and Research, 23, 381–399.
Burns, D. (1980). Feeling good: The new mood therapy. New York: William Morrow.
Burns, D. D., & Auerbach, A. H. (1992). Does homework compliance enhance recovery from
depression? Psychiatric Annals, 22, 464 – 469.
Burns, D. D., & Nolen-Hoeksema, S. (1991). Coping styles, homework compliance, and the
effectiveness of cognitive-behavioral therapy. Journal of Consulting and Clinical Psy-
chology, 59, 305–311.
Burns, D. D., & Nolen-Hoeksema, S. (1992). Therapeutic empathy and recovery from
depression in cognitive behavioral therapy: A structural equation model. Journal of
Consulting and Clinical Psychology, 60, 441– 449.
Burns, D. D., & Spangler, D. L. (2000). Does psychotherapy homework lead to improvements
in depression in cognitive-behavioral therapy or does improvement lead to increased
homework compliance. Journal of Consulting and Clinical Psychology, 68, 46 –56.
Conoley, C., Padula, M. A., Payton, D. S., & Daniels, J. A. (1994). Predictors of client
implementation of counselor recommendations: Match with problem, difficulty level,
and building on client strengths. Journal of Counseling Psychology, 41, 3–7.
Cox, D. J., Tisdelle, D. A., & Culbert, J. P. (1988). Increasing adherence to behavioral
homework assignments. Journal of Behavioral Medicine, 11, 519 –522.
DeRubeis, R. J., & Feeley, M. (1990). Determinants of change in cognitive therapy for
depression. Cognitive Therapy Research, 14, 469 – 482.
Detweiler, J. B., & Whisman, M. A. (1999). The role of homework assignments in cognitive
Special Issue: Homework and CBT of Depression 177

therapy for depression: Potential methods for enhancing adherence. Clinical Psychology:
Science and Practice, 6, 267–282.
Elkin, I., Gibbons, R. D., Shea, M. T., Sotsky, S. M., Watkins, J. T., Pilkonis, P. A., et al.
(2005). Initial severity and differential treatment outcome in the National Institute of
Mental Health Treatment of Depression Collaborative Research Program. Journal of
Consulting and Clinical Psychiatry, 63, 841– 847.
Elkin, I., Shea, M. T., Watkins, J. T., Imber, S. D., Sotsky, S. M., Collins, J. F., et al. (1989).
National Institute of Mental Health Treatment of Depression Collaborative Research
Program: General effectiveness and treatments. Archives of General Psychiatry, 46,
971–982.
Fennell, M. J. V., & Teasdale, J. D. (1987). Cognitive therapy for depression: Individual
differences and the process of change. Cognitive Therapy and Research, 11, 253–271.
Gibbons, R. D., Hedeker, D., Elkin, I., Waternaux, C., Kraemer, H. C., Greenhouse, J. B., et
al. (1993). Some conceptual and statistical issues in analysis of longitudinal psychiatric
data. Archives of General Psychiatry, 50, 739 –750.
Hamilton, M. (1960). A rating scale for depression. Journal of Neurology, Neurosurgery, and
Psychiatry, 23, 56 – 62.
Jacobson, N. S., Dobson, K. S., Truax, P. A., Addis, M. E., Koerner, K., Gollan, J. K., et al.
(1996). A component analysis of cognitive-behavioral treatment for depression. Journal
of Consulting and Clinical Psychology, 64, 295–304.
Kazantzis, N. (2000). Power to detect homework effects in psychotherapy outcome research.
Journal of Consulting and Clinical Psychology, 68, 166 –170.
Kazantzis, N., & Deane, F. P. (1999). Psychologists’ use of homework assignments in clinical
practice. Professional Psychology: Research and Practice, 30, 581–585.
Kazantzis, N., Deane, F. P., & Ronan, K. R. (2000). Homework assignments in cognitive and
behavioral therapy: A meta-analysis. Clinical Psychology: Science and Practice, 7, 189 –202.
Kazantzis, N., Ronan, K. R., & Deane, F. P. (2001). Concluding causation from correlation:
Comment on Burns and Spangler (2000). Journal of Consulting and Clinical Psychology,
69, 1079 –1083.
Kornblith, S. J., Rehm, L. P., O’Hara, M. W., & Lamparski, D. M. (1983). The contribution
of self-reinforcement training and behavioral assignments to the efficacy of self-control
therapy for depression. Cognitive Therapy and Research, 6, 499 –528.
Neimeyer, R. A., & Feixas, G. (1990). The role of homework and skill acquisition in the
outcome of group cognitive therapy for depression. Behavior Therapy, 21, 281–292.
Persons, J. B. (1989). Cognitive therapy in practice: A case formulation approach. New York:
Norton and Company.
Persons, J. B., Burns, D. D., & Perloff, J. M. (1988). Predictors of dropout and outcome in
cognitive therapy for depression in a private practice setting. Cognitive Therapy and
Research, 12, 557–575.
Primakoff, L., Epstein, N., & Covi, L. (1986). Homework compliance: An uncontrolled
variable in cognitive therapy outcome research. Behavior Therapy, 17, 433– 446.
Shaw, B. F., Elkin, I., Yamaguchi, J., Olmsted, M., Vallis, T. M., Dobson, K. S., et al. (1999).
Therapist competence ratings in relation to clinical outcome in cognitive therapy of
depression. Journal of Consulting and Clinical Psychology, 67, 837– 846.
Shelton, J. L., & Levy, R. (1979). Home practice activities and compliance: Two sources of
error variance in behavioral research. Journal of Applied Behavior Analysis, 12, 324.
Smith, R. E., Leffingwell, T. R., & Ptacek, J. T. (1999). Can people remember how they
coped? Factors associated with discordance between same-day and retrospective reports.
Journal of Personality and Social Psychology, 76, 1050 –1061.
Startup, M., & Edmonds, J. (1994). Compliance with homework assignments in cognitive-
behavioral psychotherapy for depression: Relation to outcome and methods of enhance-
ment. Cognitive Therapy and Research, 18, 567–579.
Teasdale, J. D., Fennell, M. J. V., Hibbert, G. A., & Amies, P. L. (1984). Cognitive therapy for
major depressive disorder in primary care. British Journal of Psychiatry, 144, 400 – 406.